

                                       
                 UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
                     SECTOR POLICIES AND PROGRAMS DIVISION
                 OFFICE OF AIR QUALITY PLANNING AND STANDARDS
                          OFFICE OF AIR AND RADIATION
           

Date:		June 27, 2017

To:		Plywood and Composite Wood Products (PCWP) ICR Project File

Subject: 	Summary of Comments and Responses for Comment Period Ending 
		November 7, 2016 Regarding the PCWP Information Collection Request
		Docket ID. No. EPA-HQ-OAR-2016-0243
	


I. Background

	On September 8, 2016 (81 FR 62125), EPA solicited comment on the first draft Information Collection Request (ICR) for the plywood and composite wood products (PCWP) industry. We received 11 comment letters during the comment period and three additional comment letters following the close of the comment period. A list of commenters and their docket IDs is presented in Table 1. We reviewed and considered each of these comments and, as a result, revised the ICR. Section III through VI of this document provides our responses to some of the general comments and concerns. Section VII contains tables summarizing specific comments on particular sections of the ICR and our responses. Any changes made to the ICR as a result of these comments are also reflected in the revised version of the ICR spreadsheet questionnaires.

                          Table 1. List of Commenters
                                   Docket ID
                            (EPA-HQ-OAR-2016-0243-)
Commenter Name
Commenter Affiliation
                                     0005
Russell Strader

Boise Cascade Company, Boise Cascade Wood Products LLC (BCWP)
                                     0006
Dana Lee Cole
Hardwood Federation
                                     0007
James R. Rabe
Masonite International Corporation
                                     0008
Robert W. Glowinski 
American Wood Council (AWC)
                                     0009
Will Telligman
Southeastern Lumber Manufacturers Association (SLMA)
                                     0010
Emma C. Cheuse and Nicholas Morales
Earthjustice on behalf of Sierra Club
                                     0011
Kevin Cheung

Western Wood Products Association (WWPA)
                                     0012
Stephen E. Woock
Weyerhaeuser Company
                                     0013
Kurt Bigbee
APA- The Engineered Wood Association
                                     0014
Vipin K. Varma
National Council for Air and Stream Improvement (NCASI)
                                     0015
Gary L. Heroux
Composite Panel Association (CPA)
                                  0016 (Late)
Traylor Champion
Georgia-Pacific LLC
                                  0017 (Late)

NCASI (supplementary comments on spreadsheet)
                                  0018 (Late)

NCASI (supplementary comments on instruction manual)


II. ICR Applicability

Comment: Multiple commenters suggested that the survey should be limited to operating major sources and should not include synthetic area sources. The commenters noted that synthetic area sources are not subject to the PCWP MACT standards. According to the commenters, some synthetic area sources have limited production rather than installing controls, and any controls installed are the same as controls installed at major sources. Eliminating synthetic area sources from the survey will not impair EPA's ability to conduct its analyses and rulemakings. One commenter noted that hardwood lumber producers are not likely to be major sources of HAP.

Response: In response to this comment, we reduced the scope of questionnaire for synthetic area sources to a 1-page response. We added the 1-page form as Appendix 1B of the survey instruction manual. Synthetic area sources are no longer required to complete the Phase I spreadsheet. The 1-page form requests synthetic area sources to provide the basis for their synthetic area source status, whether they accepted a production limit, made a process change, or installed technology to reduce emissions to below the major source threshold. Facilities are also requested to provide permit documentation (e.g., a construction or operating permit) reflecting synthetic area source status. Synthetic area sources that achieved area source status by application of HAP-reducing measures to PCWP process units would be asked to submit any PCWP process unit test reports used to demonstrate their synthetic area status. Process changes, control technology descriptions including process units controlled and type of control, and other relevant information specifying how HAP reductions were achieved are relevant for purposes of the RTR. Should a novel technology or approach for reducing HAP emissions be identified from a facility's response to the synthetic area source questions, the EPA may follow-up to better understand whether the approach could be applicable to further reduce emissions from other facilities in the PCWP industry.

We acknowledge that many lumber producers, and specifically hardwood lumber facilities, may not be major sources of HAP emissions. In attempt to minimize burden we have eliminated most lumber producers consistently listed as area or synthetic area sources in EPA data bases given the low probability that any lumber producers have implemented HAP emission reduction measures for lumber kilns. Softwood and hardwood lumber producers listed as major sources in EPA data bases have been retained in the draft ICR mailing list available in docket EPA-HQ-OAR-2016-0243. We request comment on this mailing list from PCWP producers, sawmills with lumber drying kilns, state and local air agencies, trade associations and other interested parties. We are particularly interested in definitive information to further refine the mailing list prior to mail-out of the ICR. Stakeholders are asked to review the list and provide updates if they are aware of any errors or omissions. Updates could include:
 identification of closed facilities that should be removed from the list,
 identification of true area source facilities that should be removed from the list,
 identification of inadvertently omitted facilities, 
 updates to major or synthetic area source status, as discussed below,  
 updates to parent company names, 
 updates to ICR recipient names (current "Mill Manager" by default),
 updates to the prevailing FRS ID for facilities listed with alternate FRS IDs,
 updates to mailing addresses, or
 updates to products produced.

Updates to the facility list can be submitted as comments to docket EPA-HQ-OAR-2016-0243.
   

III. Timing

Comment: Multiple commenters stated that more than 90 days are needed to provide quality responses to the proposed ICR. The commenters stated that extending the Phase I questionnaire from 90 to 180 days would reduce such burden. The commenters noted that EPA's estimated 250 hours would require environmental staff to dedicate more than 50 percent of one person's time to completing the survey in addition to their normal workload. These facilities typically do not operate with full-time, dedicated environmental staff and companies often share environmental staff among multiple PCWP facilities. This concern is even greater for small businesses. Outside consultants may not be the most efficient method of collecting data as knowledge of the plant and processes is necessary to accurately and efficiently complete the Phase I questions. The commenters noted that if the ICR is issued during the first half of the year, when numerous regulatory reports are due, it will be especially difficult to complete the survey in only 90 days.

One commenter stated that the requested 90-day response time to complete any testing in Phase II is inadequate. The commenter suggested that 180 days is reasonable to develop reliable data under Phase II.

To the contrary, another commenter argued that waiting to send initial letters until 2017, and then having 9 months before all data is returned in two phases will significantly delay rulemaking. The commenter stated that Phase I and II should be done concurrently because it is unlikely that data under Phase I will show that Phase II testing is not needed. The commenter also argued the scope of Phase II may need to be expanded to cover additional facilities, emission points, and HAPs, and should be done so with no additional delay. The commenter stated that not collecting data from enough facilities will hamper EPA's ability to ensure it has data from the relevant best-performing sources, and to ensure that it has sufficient data to assess the health risks this industry causes as the Act requires.

Response: In response to these comments on timing, we have extending the time for completion of Phase I from 90 to 120 days based on consideration of the revised burden estimate of 150 hours. On March 22, 2017, the EPA was ordered by the U.S. District Court (D.C. Circuit) to complete the PCWP RTR no later than June 30, 2020. This court-ordered deadline suit restricts our ability to provide additional time for respondents while ensuring sufficient time for the EPA to complete the required regulatory analyses. We expect to implement Phase I of the ICR in the second half of calendar year 2017 following completion of the process for and receipt of U.S. Office of Management and Budget (OMB) approval to conduct the ICR.

The EPA has decided to eliminate the Phase II emissions testing component of the ICR. The phased ICR approach originally planned was designed to allow us to gather and evaluate emissions data through Phase I before determining the scope of the necessary emissions testing. Our intent was to implement Phase II or portions of the Phase II testing only as needed to fill any data gaps remaining after completion of Phase I. The above noted court-ordered deadline does not allow enough time for completion of additional emissions testing. The EPA now intends to rely on the emissions data gathered through Phase I to complete the RTR. We disagree with the commenter suggesting that Phases I and II should be completed concurrently to avoid delays in rulemaking. Rushing into emissions test data collection that is not tailored to the Agency's specific data needs would exhaust time and resources and be more likely to delay rulemaking than our planned approach of focusing Agency and industry resources on the specific emissions testing needed to fill data gaps for purposes of establishing emissions standards. In addition to timing and resource burden considerations, there are technical reasons for evaluating the information submitted with Phase I prior to initiating Phase II testing. The Phase I information is needed to assess the configuration of process units to determine which units would be suitable for emissions testing. As explained in Section VI below, many of the process units included in the Phase II test plan are not configured for emissions testing because they are not enclosed or isolated from other process units. 

We acknowledge comments regarding the estimated burden hours and percent of company staff time required to complete the ICR. Extending the response time reduces the fraction of company staff time needed to respond to the ICR. We have also reduced the burden associated with responding to the ICR as discussed throughout this document. For example, we have eliminated the request for safety data sheets (SDS), continuous parameter monitoring system (CPMS), and continuous opacity monitoring system (COMS) data. As discussed further in section IV, we included available emission factors and provisional emissions calculations into the ICR spreadsheet to facilitate development of emissions estimates in cases where facilities do not have facility-specific data and opt to use the provisional calculations. Synthetic area sources would now submit a 1-page response instead of being required to complete the spreadsheet. We streamlined the request for tabulation of previous emissions test data, and have made numerous other clarifying changes throughout the survey spreadsheet that should reduce completion time. Our revised overall estimate of burden is an average of 150 hours per response as compared to our previous overall composite estimate of 250 hours. Due to the number of lumber mills included in our mailing list, we separated the burden estimates in the supporting statement to consider the different scope of responses to be submitted. The revised burden estimate for the current draft of the ICR for lumber mills is 59 hours, while the burden estimate for PCWP mills is 250 hours (for a composite overall estimate of 150 hours considering both lumber and PCWP mills). The actual burden for each facility will reflect the number of different types of process units operated at the facility. The EPA is not aware of any facility that would be required to complete all of the spreadsheet tabs. 

Allowing a longer response time should provide respondents an opportunity to use less contractor time, and thus, response costs. Further, providing a longer response period will assist technical organizations in their ability to advise their members and provide their members with technical support which will have the effect of reducing the survey response time further. 

We believe the ICR burden request is reasonable for multiple reasons. The EPA has not conducted an ICR for the PCWP industry since 1999. The information collected from 1999 is out of date and does not reflect the process unit and air pollution control configurations that are currently in use. It is important that the data used in this review of the PCWP rule is accurate because significant changes have occurred in the PCWP industry since NESHAP promulgation. A large number of facilities have both entered and exited the industry, and new process and pollution reduction measures are in place. The CAA only requires one residual risk review and technology reviews every 8 years. The accuracy of the one-time residual risk review will be significantly improved with collection of emission release point parameters and coordinates. The process and equipment configuration data supplied through this ICR will be used to perform the technology review, as well as to establish emission standards for more than 20 process units that were remanded for further action in the June 2007 D.C. Circuit court decision. Gathering updated information is not only prudent, but essential given the extent of regulatory actions to be completed in the forthcoming review of the PCWP rule. The supporting statement for the ICR provides a full description of the need and use for the ICR data as well as burden estimates.

IV. Burden

Comment: Multiple commenters stated that the ICR may be more burdensome and costly than projected. With regard to Phase I, the commenters stated expressed concern that the personnel with training needed to complete the ICR within the EPA's estimated number of hours may be unavailable. The commenters noted that completing the ICR will take the full-time attention of company environmental personnel and, in many cases, retained consultants. The burden will be especially acute with smaller companies with headquarter employees servicing multiple facilities.

To reduce burden the commenters suggested that EPA:
Refine the questions asked in the ICR to be as specific as possible. Questions and instructions need to be as clear as possible to understand exactly what information is needed for survey accuracy and submittal ease. The instructions and checklist should match.
Refine sources where emissions data is readily available and useful to evaluate emission standards. Miscellaneous sources have constraints that limit testing and where process data does not provide adequate insight into emissions with a variability in site-specific production and product specifications.
Use existing AWC/NCASI data to satisfy some of the data requests regarding:
Startup and shutdown of process and control equipment
Feasibility and cost of building enclosures for various sources (EPA's $12K per enclosure is low)
Costs of emission controls and emission reduction techniques
Environmental and energy dis-benefits of requiring additional emission controls such as oxidizers

Response: As explained in Section III, based on comments received, we have revised the ICR to reduce the burden associated with Phase I and we have eliminated Phase II. We have extended the response time to reduce the potential for the ICR to detract from the many duties of company environmental personnel and to reduce the need for retained consultants. Many of the smaller companies included in the ICR mailing list operate only one facility. We intend to offer a webinar on the ICR to assist all respondents with training, including smaller companies. We will also provide a phone and email hotline and develop a web site where answers to frequently asked questions will be posted.

We have revised the Phase I ICR spreadsheet and instruction document (including the checklist). Specific clarifications and burden-reducing changes are enumerated throughout this comment-response document, including Section III above and in our responses to comments in the tables at the end of this document.  

We welcome and will review data provided by industry trade groups and technical associations regarding startup/shutdown, costs, environmental and energy dis-benefits. The commenters have not indicated specific Phase I ICR questions that could be replaced with such information to further reduce the burden of Phase I. We have received information from the commenters regarding the feasibility of enclosing and isolating emissions from many of the miscellaneous process units, and we intend to use this information in our evaluation of whether process units are more suited for numeric emission limits requiring testing or work practices developed according to section 112(h). Regarding construction of enclosures, the commenters indicated that the EPA's estimate of $12,000 per enclosure is too low, but did not provide a revised cost estimate. We have not included the cost of enclosures in revised burden estimates because the Phase II test plan was eliminated. Nevertheless, a more-accurate enclosure cost would be valuable information as we consider the feasibility of testing as part of the forthcoming regulatory analysis.  

As noted in section III above, although the ICR imposes a burden on the PCWP industry, it is the most efficient way for information essential to completion of the RTR to be gathered. The affected industry will benefit from the EPA having comprehensive and realistic data with which to conduct the RTR which is court-ordered to be finalized by June 30, 2020. The data supplied will provide post-MACT and post-CARB process unit configuration and emissions data for inclusion in the residual risk analysis the EPA is required to conduct under CAA section 112(f)(2). Using out-of-date data that does not reflect the progress made in the industry to reduce HAP emissions could lead to overestimated risk levels that prompt additional rulemaking.  Similarly, updated information on PCWP facility practices, process unit and control configurations, and emissions will inform the EPA's technology review required under CAA section 112(d)(6). In addition, the information from Phase I will be used to address approximately 25 remanded process units for which the EPA currently has very little information. The EPA intends to evaluate emission standards (emission limits or work practices, as appropriate) under CAA sections 112(d)(2)-(3) and 112(h) and to perform the 112(f)(2) residual risk review for the remanded process units in one action. Without the Phase I data set, the EPA would have little useful information to conduct the required regulatory assessments of emissions, residual risk, environmental impacts, energy impacts, cost/economic impacts, and small business impacts. Our ability to perform these assessments accurately benefits the affected industry. Agency efforts to gather information from sources other than the affected industry would be time-consuming and costly, and would not yield information as up-to-date, comprehensive, and complete as could be obtained from affected facilities. The supporting statement for the ICR provides a full description of the need and use for the ICR data as well as burden estimates.

Comment: One commenter responded to the EPA's request for comment regarding the potential for utility of an emissions calculation tool. The commenter stated that an emissions tool pre-populated with emission factors may be helpful for facilities that may not have a robust inventory, but it depends on the quality of the emission factors (e.g., most of the AP-42 emission factors for wood products sources are pre-MACT and rated "use with caution"). The commenter suggested that use should be optional and the quality of any pre-populated emission factors should be provided. The commenter stated that they would like to review and comment on the factors included.

Response: The Primary HAP Emissions and Other HAP Emissions tabs were consolidated and transposed into a single HAP Emissions tab and columns comprising a "Provisional Calculation Tool" were built into the HAP Emissions tab. The Provisional Calculation Tool is an optional emission-factor calculation resource that facilities may choose to use for estimating emissions in the absence of more-accurate site-specific information. Instructions for using the Provisional Calculation Tool were added to the ICR instruction document. The columns and instructions pertaining to the Provisional Calculation Tool can be ignored by facilities not using the tool.
The provisional calculation columns include a description and reference for the emission factor used to calculate emissions to allow facilities to confirm that the emission factor is relevant for their operations in the absence of site-specific data. Emissions are pre-calculated for each release point based on the process unit throughput reported in the ICR, the relevant emission factor, a release point apportionment factor (for process units with multiple emission release points), and the combined HAP collection-control efficiency specified by the respondent. Actual emissions, allowable emissions, and maximum emissions estimates are pre-calculated for the respondent's optional use. Allowable emissions consider emissions allowed under the facility's compliance margin (e.g., the 5% additional emissions if a facility's control device performs at 95% reduction instead of the 90% reduction required under the PCWP rule). Maximum emissions are reported in lb/hr and were assumed to represent an hour of uncontrolled emissions for purposes of the acute risk assessment EPA performs. Respondents choosing to use pre-calculated, provisional emissions estimates can cut-and-paste special (as values) the provisional emission rates into the required estimated emissions columns in the HAP Emissions tab, and indicate that the provisional emissions were used by selecting 99-PCWP Provisional Calculation as the method for determining emissions.


The provisional calculations are primarily based on publicly-available wood products AP-42 emission factors to assist facilities that may be building an entirely new HAP emissions inventory using calculated emissions. Relevant emission factors were paired with the 400+ SCCs available for the PCWP industry. Many of the PCWP SCCs were newly created to cover the entire spectrum of PCWP process units and process variables. Approximately 60 SCCs that are not relevant for process unit types to be included in the ICR were removed from the SCC list. However, there are currently more available SCCs than there are AP-42 emission factors so it was necessary to apply closely related emission factors for many of the SCCs. After considering related emission factors, there are many SCCs with no emission factors are available. A memorandum in the docket titled, "Development of a Provisional Emissions Calculations Tool for Inclusion in the PCWP ICR," describes the methodology used to assign relevant emission factors. The EPA would appreciate additional stakeholder comment on the utility of the provisional calculations, whether they achieve the goal of reducing burden, and whether it would be more appropriate for facilities to develop their own emission estimates based on data or emission factors available to them as opposed to using AP-42 emission factors matched to related SCCs.


V. Magnitude of Emissions from Remanded Miscellaneous Units Compared to Regulated Sources

Comment: One commenter explained the vast majority of emissions data for miscellaneous sources have been generated as part of NCASI technical studies or limited screening-level studies. A review of this data (in the NCASI Wood Products Emission Factor Database) indicates HAP emissions from miscellaneous process units are a relatively small fraction of uncontrolled HAP emissions from presses and dryers with control requirements under the PCWP MACT rule. It is important to reiterate that most of these miscellaneous source emission data reflect pre-MACT conditions as they existed in the 1990s. Current miscellaneous process unit emissions (reflective of post-MACT and post-CARB data) would constitute an even smaller fraction when compared to emissions from presses and dryers. For instance, pre-MACT plywood press emissions were ~5% of the average uncontrolled emissions from the presses and dryers for which controls were required.

When pre-MACT emission rates associated with selected miscellaneous process units are compared against Production-Based Compliance Option (PBCO) emission rates specified in the PCWP MACT rule, the same conclusion emerges (i.e., emissions from miscellaneous process units are in many instances between 10  -  25% of PBCO emission rates). At an OSB plant using phenol formaldehyde plus methylene diphenyl diisocyanate (PF+MDI) resin, potential HAP emissions from blenders and formers are only 25% of controlled emissions from dryers and presses. Similarly, at an MDF plant using urea formaldehyde (UF) resin, HAP emissions from board coolers, formers and finishing operations are only 20-25% of controlled emissions from dryers and presses. This additional perspective is useful when considering the appropriateness of control technology-based standards for these sources (i.e., the potential for HAP emission reductions are much lower for this category of sources). More details on this comparative analysis are provided in Attachment No. 1 to comment EPA-HQ-OAR-2016-0243-0014.

Response: We acknowledge that the HAP emissions from the miscellaneous process units are relatively low compared to the currently regulated PCWP process units. The ICR will provide valuable information on the numbers of miscellaneous sources, emissions relative to other PCWP process units, and any applicable operational practices or innovative compliance approaches that could form the basis for emission standards reflecting the nature of these units commensurate with the feasibility of testing and controls. The ICR will also provide the information needed to estimate the costs, environmental, energy impacts associated with emission standards for the miscellaneous processes.


VI. Phase II Test Plan and Consideration of Equipment Infeasible to Isolate, Enclose, and Test

Comment: One commenter noted the proposed test plan provides an overview of required testing, approved sampling and analysis methods, and target pollutants for 19 emission sources. Five of these sources (softwood veneer dryers, rotary strand dryers, primary and secondary tube dryers, and green rotary dryers) are currently subject to emission standards under PCWP MACT I. EPA is proposing to require HAP metals (and PM) testing for these sources, presumably for use in residual risk modeling. The remaining fourteen sources in the proposed test plan would require organic HAP emissions and total hydrocarbon (THC) testing. 

The commenter stated that, first, the test plan presupposes that all 19 emission sources are equally amenable to emissions testing and that they can all be tested using the proposed procedures. Drawing from our extensive experience in emission testing using EPA-approved test methods and our familiarity with miscellaneous sources, it is our conclusion that the proposed contingency test plan may be infeasible to implement in many cases. Second, this test plan may not adequately consider the practical / technical limitations that dictate the testability (ability to isolate, enclose, and test) of some of these sources. Results from the NCASI Miscellaneous Source Survey relative to source collocation/interconnectivity indicate that the vast majority of vents or emission points associated with these process units, if tested, would not yield a result that represents emissions from an enclosed, 100% isolated process unit. In fact, EPA has used the term "emission point" to describe atmospheric venting locations for pneumatic dust collection systems associated with dry blending/forming and finishing sander, saw, and chipper process units. The use of this term would suggest EPA has an understanding of the challenge that lies ahead with regard to linking emission test results obtained from collocated sources ducted through common release points. Any testing program implemented without adequate consideration of these source characteristics and associated challenges would, at best, only provide non-representative emissions estimates for a group of collocated sources. More details on these challenges and options to improve the quality of data collected during the test plan are provided in Attachment 2 of comment EPA-HQ-OAR-2016-0243-0014.

The commenter noted that the EPA is proposing to initiate the Phase I survey to collect both facility-level and emissions information. If the emissions information collected during Phase I is inadequate to either assess residual risk and/or assess technological developments in practices, processes, or controls, EPA would require facilities to conduct emissions testing as outlined in the test plan. The commenter stated that many of the remanded miscellaneous sources included in the test plan currently have no reliable emissions information and are challenging sources for emissions testing. This is especially true if the goal is to develop source-specific emissions data for floor analyses and setting numerical emission standards. However, the Survey Instrument does not solicit information on these testing challenges. It presupposes that all these sources can be tested to generate reliable emissions information; i.e., they would fall into one of two categories: those with emissions information, OR those without. Therefore, if facilities do not report emissions information for a specific source category in the Survey, EPA would conclude there is lack of emissions information and subsequently mobilize the Phase II test plan. This conclusion would be erroneous and based on incomplete information.

The commenter identified only one instance (plywood presses) where a question related to the technical feasibility of enclosing the unit was posed. The NCASI Miscellaneous Source Survey solicited detailed information on process unit configurations with the explicit goal of identifying testability issues and testing challenges (e.g., source collocation, feasibility to enclose and isolate, etc.). Therefore, the commenter recommended that EPA adequately consider the testing challenges associated with these sources (also identified in the NCASI Miscellaneous Source survey) to narrow the list of sources included in the Phase II test plan and not mobilize the test plan solely based on the lack of emissions information.

In addition to the commenter above, other commenters expressed concern about the meaningfulness of any new data conducted under the proposed test plan, especially at the high estimated cost of testing. The commenters stated that any emissions data collected will not adequately characterize variability and there are serious implications of using a small data set to establish standards. 

According to the commenters, it is not feasible for EPA to prescribe or enforce a numeric emission standard for many of the sources included in the proposed test plan, and many of these sources will qualify for a design, equipment, work practice or operational standard in accordance with section 112(h). In addition, the commenters argued that the data that are collected will be very limited data or of poor quality and representativeness and any limit developed will not reflect the actual level of emissions from those sources nor will it reflect the achievable level of performance across all sources and operating conditions.

The commenters stated that the PCWP MACT remanded sources cannot be tested in a manner that assures the emissions are solely representative of the emission source or that assure emission limits established for those sources capture all the emissions attributable to that source. Even under circumstances where the tests could be reliable measures of the emissions from those units, it is not likely that the emissions tests could reliably take into account all the factors that affect the variability of emission rates at these units. It is not reasonable to attempt to test these units and establish emission limits based on the best performing units when emission performance is not based on the unit itself but rather the specific product being manufactured in that unit. It is also not reasonable to establish emission limits based on existing emissions information that does not take into consideration all these different factors. For these reasons the commenters strongly urged the EPA to forgo testing and setting emission limits for most of the remanded sources, and instead work with industry to develop reasonable work practice standards for these sources. 

The commenters concluded that, while existing data may not be appropriate to establish emission limits, it may be useful emission data to support EPA's work on the residual risk at these facilities. If EPA needs additional emissions information to support the residual risk evaluation, then additional testing for that purpose may be reasonable. Such a testing program could be designed to test worst case emission rates that would be useful for a conservative risk assessment even if that data was not useful for establishing emission limits from the various remanded sources.

Response: We acknowledge the challenges associated with isolating and capturing emissions from many of the miscellaneous remanded PCWP process units. After reviewing the detailed information submitted by commenters, and considering the time and burden associated with the Phase II test plan, we have decided to eliminate the Phase II of the ICR and rely on the information gathered with Phase I to complete the PCWP RTR. Phase I requests information on process unit configuration with respect to collection and controls, practices that reduce HAP emissions, and emissions test data from previously conducted emissions tests. The information collected in Phase I will enable us to evaluate the feasibility of testing remanded PCWP process units, the representativeness of any previously conducted testing, and the applicability of work practices that could be developed under section 112(h) for process unit types determined to be infeasible to test. Phase I includes survey questions to help us more fully characterize testability issues.


Comment: Commenters argued that the Phase II test plan should be limited to sources that will not qualify for a work practice standard approach. HAP emissions capture and measurement is not practicable at many of the listed sources and it will be challenging, if not impossible, to completely separate emissions from different types of sources that are located in close proximity to one another within the same building.

One commenter described the lack of reliable emissions information and inherent variability in HAP emissions from remanded miscellaneous sources. According to the commenter, the vast majority of emissions data for miscellaneous sources have been generated as part of NCASI technical studies or limited screening-level studies aimed at (a) better-understanding the complexities associated with such testing, (b) assessing the magnitude of HAP emissions from these sources and (c) understanding the potential for emissions variability. An analysis of existing research-level emissions data for miscellaneous sources in the NCASI database indicates that emissions from softwood plywood presses, board coolers, and sources that share dust collection systems like blenders/formers and saws/sanders, can be highly variable. More details on contributing factors are provided in Attachment 1 to comment 0014.

The commenter also described challenges associated with isolating, enclosing and testing miscellaneous sources in order to reliably quantify, and subsequently attribute, HAP emissions to specific process units. The commenter explained that NCASI conducted a comprehensive survey of a subset of the listed remanded miscellaneous source units as part of its Technical Studies program in 2014. The objectives of the survey were to (a) gather data on process characteristics and complexities, specifically the heterogeneity (and site-to-site variability) in source configurations, (b) identify instances where facilities had isolated, enclosed and quantified emissions from these units, (c) identify the need for work practices as opposed to emission limits where applicable and (d) gather data on operating practices that could be used for the development of work practice standards. The NCASI Miscellaneous Source Survey was comprehensive and included 12 particleboard mills, 11 MDF mills, 3 hardboard mills, 17 OSB mills, 19 softwood plywood mills, 1 softwood veneer mill, and 43 sawmills with a total of 188 lumber kilns. 

Approximately 1100 out of the 1250 miscellaneous process units surveyed were rated as being either very difficult or infeasible to enclose. Additionally, formers, blenders, saws, sanders and board coolers were frequently collocated in the same building with other process units in belonging to this group, and shared common dust-collection and pickup systems. These data would suggest it would be infeasible to enclose and isolate specific process units in order to quantify their emissions. Survey questions on the feasibility to access and carry out emissions testing on these sources indicated that the vast majority (>95%) were either very unsuitable or infeasible for testing. It is our assessment that the emissions variability observed in available emissions data can partly be attributed to the collocation of sources, the resulting interconnectivity in emissions and the contribution of fugitive emissions from collocated sources in air being drawn into a nearby source. More details and key conclusions are also summarized in Attachment 1 of comment 0014. 

Based on a review of the results from the NCASI Miscellaneous Source Survey and an analysis of available emissions data, the commenter concluded that any available emissions data would be inappropriate for use in source-specific floor analyses and setting of numeric emission standards, given the high prevalence of source collocation, varying equipment configurations, challenges in enclosing and testing emissions, and where quantified, the inherent variability in emissions.

Response: As explained in Section III and in the previous response, Phase II of the ICR has been eliminated. We have reviewed the results of the NCASI miscellaneous sources survey. We appreciate the work NCASI has done to gather information relative to the testability of miscellaneous sources, and we find this information useful for gauging the relative difficulty of isolating, enclosing, and testing the remanded process units. Commenters have noted that many of these process units have relatively low emissions. Phase I of the ICR does not duplicate, but will expand on the information collected from NCASI members in order establish national emission standards including work practices developed pursuant to section 112(h) for process units that are not feasible to test. Information on applicable work practices is needed before this determination can be made. We are interested in comments on potential work practices for emissions units that are not feasible to test, including whether numeric work practices based on resin HAP content or codified federal worker-safety thresholds would be viable options. Such comments would be helpful in further focusing the Phase I ICR questions.

Comment: One commenter suggested that emissions and fuel analysis data collected as part of the Boiler MACT ICR can inform EPA's estimates of HAP metal emissions from solid fuel direct-fired dryers.

The commenter noted that a technically justifiable alternative to HAP metals testing would be to leverage existing data for boilers in the Boiler MACT database (stokers/sloped grate units designed to burn kiln-dried biomass or suspension burners designed to burn biomass, etc.) to carry out the residual risk analysis. Another option would be to conduct an initial HAP metal risk assessment based on fuel analysis data (from the Boiler MACT database) to: (a) estimate the maximum potential for HAP metals emissions; and (b) applying PM removal efficiencies to arrive at controlled emissions. These two approaches may provide sufficient information for the intended purpose.

Response: We agree that use of fuel analysis data collected for the Boiler MACT for the types of biomass fired in PCWP dryers is a reasonable approach to approximating HAP metals emissions from PCWP dryers. A separate memorandum documents our analysis of the metals fuel analysis data from the Boiler MACT database and the provisional emission factors developed for use in preparing emissions estimates for risk modeling. These emission factors can be coupled with actual or estimated PM removal efficiency for the direct-fired dryer control device to estimate metal HAP emissions. 

Should the EPA find it necessary to establish emission limits for HAP metals, the EPA could rely the PM emissions data collected in Phase I of the ICR and any transferable PM metal HAP surrogacy conclusions from other applicable rules. We have also requested previous PM or HAP metal test reports from the PCWP industry for PCWP process units because PCWP-industry-specific data are preferable.  


VII. Detailed Comments on the Survey Instrument

Tables A through H present detailed comments on the PCWP ICR and EPA's responses in addition to the overarching comments and responses provided above. 

Column and row letters and numbers included in the tables below pertain to the columns and rows in the first draft of the ICR spreadsheet provided for comment on September 8, 2016. The letters and rows may differ slightly in the final spreadsheet due to addition or deletion of columns or rows as a result of public comments.

  
Plywood and Composite Wood Products ICR Comment Summary
A. Timing
No.
Commenter
Comment Summary
EPA Response/Action Taken
A1
AWC

Masonite

SLMA

CPA
Estimated time of 250 hours per facility within 90 days will impose significant burden on facilities and their environmental staff. Extending the Phase I questionnaire from 90 to 180 days would reduce such burden.
   Current time estimates (250 hours) would require environmental staff to dedicate ~50% of their time completing these efforts in addition to their normal workload. Concern is even greater for small businesses.
   Outside help may not be the most efficient method of collecting data as knowledge of the plant and processes is necessary to accurately and efficiently complete the Phase I questions.
   If the ICR is issued during the first half of the year, when numerous regulatory reports are due, it will be especially difficult to complete the survey in only 90 days.
See memo section III for response.
A2
CPA
The requested 90-day response time to complete any testing in Phase II is inadequate. 180 days is reasonable to develop reliable data.
See memo section III for response.
A3
Earthjustice
To avoid delays in rulemaking, it is important to complete the ICR as soon as possible. Waiting to send initial letters until 2017, and then having 9 months before all data is returned in two phases will significantly delay rulemaking. It is unlikely that data under Phase I will show that Phase II testing is not needed therefore, EPA should send out both surveys to be done concurrently.

For HAPs not already covered by the Phase II survey, EPA may need additional data and this ICR should cover that possibility such that no additional delay would be caused due to an expanded scope.
See memo section III for response.



B. Applicability
No.
Commenter
Comment Summary
EPA Response/Action Taken
B1

AWC

Hardwood Federation

Weyerhaeuser

Limit survey to major sources and do not include synthetic area sources, as the synthetic area sources are not subject to these MACT standards and information from these facilities could skew the data for major sources that are subject to the MACT requirement. Some synthetic area sources have limited production rather than installing controls, and those that did install controls have not installed any devices that are different than those installed at major sources. If they are included, questions should be limited to information about the control devices or production limit approach to achieve synthetic area status.

Based on information from two of Hardwood Federation's largest manufacturers with dry kiln capacity, hardwood facilities are not likely to be major sources of HAP.

Eliminating synthetic area sources from the survey will not impair EPA's ability to conduct its analyses and rulemakings. The control technologies employed are not substantively different from control technologies employed at the major sources subject to the NESHAP, and production limits provide no novel information about control technologies for RTR purposes. And because these sources are not required to meet the HAP reduction limits of the PCWP NESHAP, do not operate control devices according to the rule requirements, nor have to monitor and maintain records to demonstrate compliance with the NESHAP standard, the information would not be representative of the major source operations that do have to accomplish all those things.
See memo section II for response.

We removed the specific area source facilities identified by the Hardwood Federation from the facility list. As noted in section II, we request comment and definitive information to allow us to further refine the draft ICR mailing list which has been placed in docket EPA-HQ-OAR-2016-0243. 
B2
AWC

Weyerhaeuser
If base year is 2015, do not require the survey to be completed for facilities that were not in operation in 2015 and have not restarted (e.g., facilities that closed in 2016, facilities in the process of shutting down, facilities that have been idle for several years but still have an operating permit).
Change made. We updated the base year of the ICR to 2016 now that the ICR is expected to be implemented in 2017.

We also agree with the commenter that facilities that have permanently closed should not be required to submit a survey, even if they were operating during the survey year and an operating permit is still in place. As noted in section II of this memo, we are accepting comment on the draft ICR mailing list placed in docket EPA-HQ-OAR-2016-0243. Among other refinements to the list, commenters are asked to identify closed facilities that should not receive the survey. In addition, we have revised the 1-page form closed facilities would complete in lieu of the full ICR to allow respondents to certify that their facility: (1) is a true area source; (2) does not produce PCWP or operate a lumber kiln; (3) was closed in 2016, remains closed, and does not have an active permit, or (4) was closed in 2016, remains closed, but has an active operating permit. None of these facilities are required to complete the survey spreadsheet. However, permit documentation is requested to verify true area source status and a very brief explanation is requested for facilities that do not produce PCWP or are closed with no active permit. Facilities that are closed but maintain an active permit are requested to submit a copy of their active operating permit with the form in Appendix 1A. The EPA will use the permit to estimate permit-allowable emissions for inclusion in the residual risk modeling analysis. Actual emissions would be considered to be zero for the 2016 operating year.
B3
AWC

For sites that are co-located but operating under separate Title V permits, allow submittal of one survey per permitted facility, especially if each facility has separate environmental staff.
Change Made. We clarified the instructions to indicate that only one facility bearing a unique FRS ID is to be submitted per spreadsheet. It is likely that co-located facilities with separate title V permits will also have separate FRS IDs.



C. Burden Estimates
No.
Commenter
Comment Summary
EPA Response/Action Taken
C1
AWC

NCASI

CPA
ICR will be burdensome / more burdensome than projected
Phase I ICR cost/time estimate is low. 
   Lack of properly trained/shared personnel and time needed to complete the ICR within the EPA's estimated number of hours
   Current scope of survey is costly to facilities.  Completing the ICR will take the full-time attention of company environmental personnel and, in many cases, retained consultants. The burden will be especially acute with smaller companies with headquarter employees servicing multiple facilities.

To reduce burden it is suggested that EPA:
   Refine the questions asked in the ICR to be as specific as possible. Questions and instructions need to be as clear as possible to understand exactly what information is needed for survey accuracy and submittal ease. The instructions and checklist should match.
   Refine sources where emissions data is readily available and useful to evaluate emission standards. Miscellaneous sources have constraints that limit testing and where process data does not provide adequate insight into emissions with a variability in site-specific production and product specifications.
   Use existing AWC/NCASI data to satisfy some of the data requests regarding:
     Startup and shutdown of process and control equipment
     Feasibility and cost of building enclosures for various sources (EPA's $12K per enclosure is low)
     Costs of emission controls and emission reduction techniques
     Environmental and energy dis-benefits of requiring additional emission controls such as oxidizers
      
See memo section IV for response.


C2
AWC

NCASI

CPA
Phase II ICR cost/time estimate is low. Reasons included:
Test plan includes process units with significant technical limitations (according to NCASI documentation) where testing of sources is infeasible or would render the emissions data unusable as it would include contributions from various sources
Significant variability within products due to wood speciation, product specifications, and resin formulation.
Personnel costs, retained consultants, production disruptions, construction costs and testing would be extensive, complicated, and very expensive.

Consider for reducing burden
Consider less than 7 test runs. Request 4 runs from facilities where test data are available and sources meet Method 1 requirements, instead of 7 runs from a facility where no test data are available and sources may not meet all Method 1 location requirements without modification.
Evaluate whether there is a relationship between THC and organic HAP emissions or PM and metal HAP emissions so that only 3 runs of individual HAPs could be conducted and emissions extrapolated for 4 additional runs of only THC or PM emissions.
Use already determined MACT rules such as the industrial Boiler/Process Heater MACT where filterable PM is an appropriate surrogate for HAP metals for combustion units and therefore only requiring filterable PM testing instead of HAP metals as well. Alternatively, it is suggested that EPA use fuel testing and control device efficiency for estimating HAP metals emissions from direct-fired sources for its risk analysis.
Use available NCASI data for sources that indicate HAP emission are below detection levels (e.g., chippers, fiber washers). These sources will not contribute to the risk analysis and are work practice standard candidates. Review available information and should not require HAP testing for sources where available information indicates HAP emissions are below detection levels or where the risk analysis indicates they do not contribute above a de minimis amount to facility-wide risk.
Add required test methods to ERT to eliminate the separate PCWP MACT supplement spreadsheet submittal.
As discussed in Sections III and VI of this memo, Phase II of the ICR has been eliminated as a result of timing, burden, and testing feasibility considerations. Numerous specific comments relating to the test plan are listed in Table G at the end of this memo, including comments pertaining to feasibility of testing, the number of test runs, test method considerations, concurrent testing of THC with total HAP, concurrent metal HAP and PM testing, treatment of data below test method detection limits, and test data reporting. These comments are summarized, but are not responded to in detail as a result of our decision to forgo Phase II testing.


C3
SLMA
EPA has estimated that the average cost of responding to the Phase I ICR survey would be approximately $28,244 per facility. This financial burden is not justified, and EPA should reduce the scope of the survey. Lumber kilns are not included in the list of categories for Phase II testing; any testing of lumber kilns would further increase the financial burden.
As discussed elsewhere, we have made changes to reduce the burden associated with the ICR based on comments received. The ICR response for lumber producers is focused on lumber kilns so their burden is considerably less than for other PCWP producers. Phase II emissions testing has been eliminated and was not required for lumber kilns. 
C4
Earthjustice
EPA should issue this ICR, and should require the full scope of emission testing currently slated for Phase II as soon as possible, for a larger number of facilities, as well as expand the scope of emissions testing to be required. Testing of both currently regulated and unregulated HAPs is essential to ensure that the agency has current data to fulfill its CAA §112(d) and (f) rulemaking obligations. EPA should require emission testing of as many facilities as possible to ensure it gets the best available data to meet these requirements. Not collecting data from enough facilities will hamper EPA's ability to ensure it has data from the relevant best-performing sources, and to ensure that it has sufficient data to assess the health risks this industry causes as the Act requires. EPA should expand the number and types of facilities and emissions points for which the ICR will require testing.
See memo section III for response.


D. Phase I: General
No.
Commenter
Comment Summary
EPA Response/Action Taken
D1
AWC

Functionality of ICR spreadsheet and submittal:
   Questions and spreadsheets should be formatted to request only the information needed to collect emissions, assess risk, and identify technological innovations. This would reduce the number of columns in most tabs and provide ease of printing / portability.
   Compliance and Emissions Data Reporting Interface (CEDRI) has not been proven as an easier way to provide data and would be a burden on facilities if mandatory (could be optional). To ease reporting it is requested that the survey and attachments be submitted in the form of Microsoft Excel and other attachments via survey website or on a USB flash drive via certified or express mail.
In developing the survey, we made a concerted effort to limit the survey questions to only the information needed to gather process, controls, and emissions data needed to conduct the required analyses for the RTR and to address the remand. We have implemented most of the stakeholders' specific suggestions regarding the spreadsheet queries to further refine the spreadsheets, reducing the number of columns when possible. 
In designing the spreadsheet, we attempted to balance the need for printability/portability with the need to maximize utility of the data base to be constructed from the ICR responses. We reduced the number of columns as much as practical by removing questions and transposing the Primary HAP and Other HAP Emissions tabs into a single HAP Emissions tab. The instructions document has been edited to recommend use of the freeze/split pane and hide functions in the spreadsheet to aid in viewing information across the spreadsheet.  The spreadsheet has also been formatted for print. Finally, we intend to create a single pdf file showing the column headings and instructions for the final ICR in order to provide respondents who would like to have a printout with one file to print as opposed to printing every spreadsheet separately. 

Instructions for submittal of non-CBI responses through CEDRI have been incorporated into the second draft of the ICR. Uploading non-CBI ICR response materials into CEDRI is recommended, but we have retained the option for facilities to mail their completed spreadsheet on a disk or flash drive. See Section C and Appendix 6 of the revised instruction manual for a description of CEDRI upload procedures. Essentially, facilities would need to register for CEDRI, fill in a few key facility identification details on screen, and upload their non-CBI spreadsheet and other files into CEDRI instead of saving these files on a disk that is mailed. It is our understanding that many PCWP facilities are already registered for CEDRI.  The CEDRI system will conduct automated checks of the spreadsheets uploaded to ensure the data are formatted properly and meet basic criteria for translation to the federal Registry System (FRS). The EPA's preference is for facilities to use CEDRI because it reduces the Agency burden of handling and downloading files from numerous disks or flash drives. CBI data would not be submitted into CEDRI. All CBI would be saved on a disk and directed via mail to EPA's CBI office. 
D2
AWC
Emissions tool:
   An emissions tool pre-populated with emission factors may be helpful for facilities that may not have a robust inventory, but it depends on the quality of the emission factors (e.g., most of the AP-42 emission factors for wood products sources are pre-MACT and rated "use with caution").
   Use should be optional
   The quality of any pre-populated emission factors should be provided; AWC would like to review and comment on the factors included.
See the description of the provisional calculation tool in section IV. 

D3
AWC
Survey questions on operational parameters:
   Eliminate survey questions that relate to operational parameter data where there is no corresponding site-specific emissions data. Snapshots of operating variables will not inform EPA's analysis of risk or innovative technology.
   If the operational parameter data is important to the standard setting process for these sources, then explain the connection. If EPA's objective is to gather information on operational variability within or among facilities, then the range of operating parameters should be requested, not the typical or maximum parameter.
   The most informative question to ask for units that are not controlled under the current PCWP MACT is what practices are being used to minimize emissions.
Change made. We eliminated several parameter columns along with the request for a 24-hour snapshot continuous parameter monitoring system (CPMS) data in response to this and related comments. The request was originally included to help us better understand the monitoring recording methods and frequencies used by facilities for parameters of potential interest for work practices. Although many columns seeking parameter data have been removed, the ICR spreadsheet retains several questions on the practices being used to minimize emissions from uncontrolled process units, and specific questions relating to parameter ranges that are likely to be relevant to the potential for a process unit to generate HAP emissions. Operational parameters could form the basis for establishing work practice standards and help to distinguish among process unit types should subcategories of process units need to be considered.
D4
AWC
If facilities have recently submitted permit applications or received permits for modifications that involve changes to emissions or release parameters and would make 2015 operations and emissions non-representative of future conditions, EPA should provide an option for submitting post-project projected actual emissions and release parameters instead of 2015 actual emissions.
Change Made. For plants undergoing modifications, the EPA provided an option for submitting the ICR response based on post-project projected equipment information, actual emissions and release parameters instead of 2016 actual emissions. This has been clarified in the instructions document in the instructions regarding base year.
D5
Earthjustice


EPA has not plainly explained how emission factors might be used in reporting information under the ICR. EPA should ensure that only EPA-approved emission factors are used, if any, and should require clear documentation of what factors were used, so that the public can fully understand how the data submitted under the ICR were created.
As explained in section IV, the EPA has provided provisional calculations using publicly-available emission factors for optional use to facilitate development of HAP emissions estimates to be included in facility responses. The EPA has made clear that site-specific data are preferred. However, facilities that do not have emissions measurements for certain combinations of pollutants and process units are likely to estimate emissions using the most-representative emission factors available to them. The EPA's AP-42 emission factors, where they exist, will be only one source of emission factors facilities may use in the absence of site-specific measurement data. There have been both process improvements involving lower HAP adhesives and control technology developments since the AP-42 values were published so facilities may use site-specific data or updated emissions measurement or emission factors which reflect these technology changes when responding to the ICR. Requiring only "EPA-approved emission factors" is not in keeping with the technical principal/policy of using all the data before the agency. 
D6
Earthjustice
The ICR notes that EPA will have a "website to be developed" and an emission estimate calculator "to be developed" that facilities can use to respond to the ICR's request for emission estimate info. EPA should provide full information on that process so the public can understand the assumptions and analysis underlying it when it is released. Because it is not available at this time, it was not possible to submit a comment on this component of the ICR.
As explained in section IV, the EPA has provided provisional calculations using publicly-available emission factors for optional use to facilitate development of HAP emissions estimates to be included in facility responses. A memorandum documenting assignment of relevant emission factors for the various SCC codes has been placed in the docket for comment.

We also expect to develop a website to facilitate survey responses. The website development is more closely linked to a final ICR, and is unlikely to be finished prior to the final ICR. The website is a functional tool for facilitating survey response, not a policy tool so taking comment on it at this time is not useful.  
D7
Earthjustice
Request that EPA report the emission information in pounds not tons as many HAPs pose health threats at much lower levels of human exposure than 1 ton. This would give EPA and other stakeholders the most granular data available.
The residual risk modeling file format uses emissions data in tons per year. Decimals are used to reflect fractions of tons and to allow for conversion to pounds. 

E. Phase I: ICR Instruction Manual
No.
Commenter
Comment Summary
EPA Response/Action Taken
E1
AWC
Facility Emission Point Map
Requirement should be clarified to indicate that only HAP emission points are requested (e.g., not the wood yard or other material handling sources for which HAP emissions are not expected or estimated).
Change Made. The instructions were clarified to state that the emission points for process units listed in Appendix 2 for which HAP data are provided in the ReleasePt tab are required to be included in the emission point map. Wood yard and other material handling sources are omitted from Appendix 2.
E2
NCASI
Section D3 Process Data Tabs\Mill tabs, covering Facility identifiers
Relative to Columns B-G, some mills are going to find multiple numbers, especially the CEDRI ID. Provide guidance within the instructions document for mills with multiple ID numbers on which number to select for entry into the survey.
Change Made. Mills will use an ICR ID and FRS ID, which will be provided in letter from EPA. Mills will also be asked to provide their FRS ID. 

E3
NCASI
Release Point Detail Tab Instructions:
Under Stack parameters, PCWP Notes, page 19: Instructions cover baghouses that vent 360 degrees but do not address other baghouse vent configurations (side mount vents).
The footnote text for Stack Parameter Column L, page 19, has the following typo; "For fugitive are sources (codes 01 or 07)..." should be "For fugitive area sources (codes 01 or 07)..."
Page 21, Longitude/latitude coordinates instructions text has the following typo; "For fugitive are sources (codes 01 or 07)..." should be "For fugitive area sources (codes 01 or 07)..."
Change Made. The instructions were clarified to indicate emission release point type 03 (for horizontal vents) is available for baghouses with side mount vents.

Change Made. The text discussing fugitive area sources was updated to reflect a new emission release point type labeling approach. Fugitive (code 01) is no longer used. The revised emission release point type codes are as follows:
02  -  Vertical
03  -  Horizontal
04 - Goose Neck
05 - Vertical with Rain Cap
06 - Downward-facing Vent
07  -  Fugitive Area. Not used for new data. Reserved for historical data* 
08 - Fugitive Vent
09 - Fugitive Two-dimensional
10 - Fugitive Three-dimensional

*The fugitive area emission release point type 07 is being phased out. It is suggested that facilities reviewing historical NEI data for reference in compiling their ICR response replace their fugitive area release type with the most representative fugitive release type from 08 through 10.
E4
NCASI
The instructions provided in cell T8 of the ReleasePt Tab of the Excel survey indicate that:
"If this information is not known, use the Geographical Information Systems (GIS) maps that have been provided for your facility to determine where the release points are on the facility premises to the 6th decimal point."
There is no information about "provided GIS maps" in the instruction manual. EPA needs to provide more information about this so a mill can use this reference. If GIS maps have not been provided to a mill, then EPA needs to provide instructions on how to calculate the 6th decimal Lat/Long data coordinates from degrees, minute, second traditional coordinate measurement devices. EPA should also address how to factor in the inherent accuracy of GPS devices.
Change made. No GIS maps will be provided. The spreadsheet was revised accordingly. Mapping software/freeware can be used in conjunction with GPS devices to identify coordinates. A formula for converting from degrees, minutes, and seconds was added to the instructions manual.
E5
AWC
Where facilities do not provide site-specific release point details, EPA is expected to use default values in risk modeling. EPA should provide proposed default values as an attachment to the survey so these values can be reviewed and any appropriate alternatives suggested. Also, the default values used in EPA's risk analyses tend to be overly-conservative, so to avoid risk results that are not representative, EPA should follow up with facilities rather than using defaults especially when screening results show higher risks or anomalies.
Change made. We included information on default release parameters that would be used in the instructions document. We will follow up with facilities as needed to resolve anomalies. 

For stack heights and flow rates, we will consider values reported in ICR responses from similar sources across the industry as we develop and document defaults used in analyses. We are not providing these defaults in the ICR instructions because they will not be developed until after the ICR is complete, and then will be developed only if needed.
E6
NCASI
Permit Limit Tab Instructions:
Associated text on page 22, third paragraph: "You are asked to provide emission limits for HAP, VOC, and PM." To avoid confusion, please add the text that distinguishes particulate matter (M5) from filterable particulate matter (PM10 and PM2.5). Furthermore, it is unclear whether EPA is soliciting information on condensable PM. 
Change Made. We edited the PM section of the Permit tab to include separate columns for filterable PM (Method 5 or similar methods) and another "smallest" size fraction of PM including PM10, PM2.5, CPM, Primary PM10 (PM10 + CPM) or Primary PM2.5 (PM2.5 + CPM). Facilities are asked to provide the smallest size fraction included in their permit with or without CPM. We added a menu listing the relevant EPA test methods for the "smallest" PM fractions. Facilities are only asked to provide PM limits as they appear in their permits, not all size fractions of PM.
E7
NCASI
Resin Tab Instructions:
Amend existing instruction text on page 23, second paragraph:
"Describe HAP scavengers....by overwriting the word "specify" in the `Other' pull down menu choice."
To read: "Describe HAP scavengers....by selecting the "Other: {specify}" option within the pull down menu and then going to the formula bar to write in the type HAP scavenger by overwriting "{specify}"."
Change Made. Instructions text was revised for clarity.
E8
NCASI
Resin Tab Instructions:
EPA should provide a detailed discussion relative to the percent resin solids information they are requesting.  
Mills should keep in mind that the "%" value provided in some SDS sheets is not well defined. Although it is most likely the case that % value represents % by weight, it might be reasonable to verify that this assumption is correct.
EPA should be aware that resin suppliers are likely not required to know or provide the exact amount of formaldehyde and phenol content in the resin but they are probably required to certify that it is below a certain threshold. It is expected, therefore, that most mills will enter the "less than" <% content that is stated on the SDS sheet.  
Changes Made. We provided additional instructions to explain that we are asking for percent "by weight" to ensure we are receiving data on a consistent basis. We added instructions to request the less-than sign be added for the resin HAP percentages when reported as "less-than" in the SDS. We anticipate that these resin-HAP-content questions will be important for establishing resin-related work practices.  
E9
NCASI
EmTest Tab Instructions:
Page 27, first paragraph should say Appendices 2 and 3. (Also applies to other references to Appendix 3 in the text.)
Change Made. We reworded the instructions to refer to the process units listed in Appendix 2 instead of repeating the process units in the table in Appendix 3.  
E10
AWC
Emissions Testing and Monitoring Data:
   Facilities may not have tests dating all the way back to 2003; title V only requires retention of last 5 years. EPA should only require test reports that are readily available so the respondents do not have to retrieve test reports from the permitting authority.
   EPA should only require test reports that represent the current facility configurations.
   Limit the request to testing and monitoring conducted to demonstrate compliance with the PCWP MACT and testing and monitoring conducted on affected sources for HAP emissions.
   EPA should not require submission of opacity data. Opacity data are not relevant to EPA's RTR and opacity is not currently required as a monitoring parameter under the PCWP NESHAP.
Limited Change Made. We revised the date for submittal of HAP data to July 30, 2004, the original PCWP rule promulgation date. We also stated in the instructions that respondents may limit test data submitted to the 5 more recent test reports for a given process unit and pollutant if multiple tests have been conducted. The bulk of the previous HAP test data or surrogate THC data that exists is expected to be more than 5 years old because most facilities conducted emissions testing to demonstrate compliance with the PCWP rule in the 2007-2008 timeframe. Many facilities would have conducted an initial HAP or THC test only, with no repeat tests since 2007-2008. CAA section 114 provides the authority to collect previous or new emission test data. Given that Phase II of the ICR has been eliminated, the EPA will now rely on previous emission test reports to conduct the required CAA review. Upon consideration of this comment, we restricted our request for previous PM (and associated opacity) test reports to tests conducted within the past 5 years not to exceed 5 test reports showing compliance with the applicable permit limits because of the likelihood that PM tests have been conducted more often than HAP or THC tests.   

The EmTest tab asks respondents to note when the facility configuration has changed since testing. We clarified in the instructions that submittal of data for prior configurations is optional. We are not limiting our request for previous test data to current facility configurations because (1) prior facility configurations could provide insight into emissions from facilities with similar configurations, and (2) data from prior configurations would be useful when compared to data from current facility configurations. For example, data from before and after implementation of the CARB/TSCA rules would be valuable for quantifying the impacts of the CARB/TSCA rules on HAP emissions from affected facilities. Such before-and-after data provides useful insights into the changes in emissions associated with process changes. 

The request for previous test data includes the process units listed in Appendix 2. The request is not limited to the subset of units with HAP limits in the current rule, or to tests conducted to demonstrate compliance with the current PCWP HAP limits, because of the need for the EPA to address the remanded units. 

Change Made. The survey now includes a yes/no question in the EquipDetail tab regarding whether COMS are used. The request for Method 9 opacity data was retained in the EmTest tab but reduced in scope to request up to five Method 9 tests conducted within the last 5 years.
E11
Weyerhaeuser
Eliminate COMs data survey requirement
It will require a more substantial effort than EPA estimates to obtain and provide in the format that EPA plans in its draft survey instrument. Potential use of this information appears to be beyond what is necessary for the RTR since opacity data are not currently required under the PCWP NESHAP as a monitoring parameter for the specific PCWP NESHAP regulated HAP emissions. This data would not be associated with any HAP information that would facilitate its use in any potential evaluation of additional HAP limits under the remand.
Change Made. The request for COMS data was eliminated. The survey now includes a yes/no question in the EquipDetail tab regarding whether COMS are used. 



E12
Earthjustice
EPA lists for existing emission tests include only 2 speciated HAPs: methanol and formaldehyde. (Instructions p. 28-29) The rest of the list includes all total or combined HAP emissions values, or surrogates for such values. In addition to requesting tests for all of the pollutants and total pollutant values listed, EPA should also request any tests available for each of the HAPs known to be emitted and any other potential HAPs, including metals and other products of incomplete combustion.
Change Made. We have revised the request for existing test data to be more open-ended to ensure that we collect data for of any HAP known to be emitted. The menu in the EmTest tab now includes all six of the speciated compounds comprising "Total HAP" as defined under the PCWP rule. The terms "Total metals" and "Total non-PCWP HAP" were removed from the menu in the EmTest tab. Facilities are now requested to select common compounds or write in additional individual HAP compounds tested. Tests involving HAP groups measured by the same test method such as Method 29 HAP metals or Method 429 POM/PAH may be listed as "HAP metals" or "POM" on one row instead of listing every specific compound. The EPA will speciate compounds when extracting the data from the test reports received.  
E13
AWC
Parameter monitoring (Table 4):
   The request for continuous parameter monitoring data for several types of equipment should be eliminated because:
     There is no utility to gathering parameter data when there is no corresponding emissions data. It would be arbitrary for EPA to set a numeric work practice standard on temperature or moisture data collected in this ICR and limiting operational parameters without considering the associated emissions reduction and achievability of the standard in practice.
     The data will not inform EPA's risk and technology reviews in any way. Snapshots of operating variables will not inform EPA's analysis of risk or innovative technology.
     The data cannot be used to set any kind of standard
     Eliminating this request would provide a $172,215 reduction in burden according to EPA's supporting statement.
We eliminated the request for a 24-hour snapshot continuous parameter monitoring system (CPMS) data in response to this and related comments. The request was originally included to help us better understand the monitoring recording methods and frequencies used by facilities for parameters of potential interest for work practices. The ICR spreadsheet retains several questions on the practices being used to minimize emissions from uncontrolled process units, and specific questions relating to parameter ranges that are likely to be relevant to the potential for a process unit to generate HAP emissions. Operational parameters could form the basis for establishing work practice standards (if appropriate) or be used to develop and distinguish among process unit subcategories.
E14
AWC
Appendix 2:
The survey attachment that lists emission points is helpful to clarify that EPA is focusing on HAP emission sources only. Neither chippers nor any unit processing green wood should be included in the survey, as HAP emissions would not be expected from chippers and the limited NCASI data confirm that HAP emissions are non-detect from chippers. EPA should clarify that it is only requesting information on panel trim chippers, as this is the only type of chipper defined in the 2006 final rule as a PCWP process unit.
Change Made. Appendix 2 now more clearly references "panel trim chippers" instead of "chippers/flakers/hammermills" which could have been misinterpreted to include green end sources.  



F. Phase I: Spreadsheet
No.
Commenter
Comment Summary
EPA Response/Action Taken
F1
NCASI
Global Spreadsheet Issue:
   The following is an external reference (link) showing up in the Data-Connections-Edit Links Ribbon Bar: `PCWP_survey_REVISED_7-01-16_decolored.xlsx
Change Made.      


F2
AWC
Mill Tab:
   EPA should not ask for TRI, EIS, CEDRI, and NEI IDs  -  EPA has stated it is not their intent to use IDs to confirm data in those other systems, and all data submitted to those systems is being requested in the ICR.
   Columns AM and AN should be deleted and the survey only sent to major sources. Even if synthetic area sources are surveyed, column AN seems unnecessary  -  by definition a synthetic area source has accepted permit conditions that limit its emissions below major source thresholds.
   EPA has not specified the practical utility of asking about other federal regulations.
Change Made.  The ICR now only asks for ICR ID and FRS ID, both of which can be found in the ICR letter mailed to each facility. 

Change made. Columns AM and AN were deleted as requested because only major sources are required to complete the ICR spreadsheet. In lieu of the spreadsheet, true area sources would submit the form in Appendix 1A and synthetic area sources would complete the form in Appendix 1B of the instructions. 

The brief menu-selection questions requesting indication of rule applicability were retained. These questions aid our understanding of related rule requirements. Emissions of HAP outside the category are relevant for the facility-wide risk analysis. This allows us to more accurately consider cumulative regulatory burden.  
F3
NCASI
Prod Tab:
   Column G Nominal Production and Capacity Units of Measure  -  please provide "MLft/yr" as a production unit for thousand linear feet.
Change Made. 
F4
AWC
Prod Tab:
   EPA does not need the information in columns L, M, N (product sold internally, plywood core type, agricultural fiber).
   It is unclear why EPA is asking for additional 5-year utilization data in column K.
Change Made. The information requested in columns L, M, and N would be used in EPA's economic analysis of amendments to the PCWP rule. We rephrased the questions to eliminate the request for a specific percentage of product sold internally and agricultural fiber. The questions now allow users to indicate a range of plywood sold internally (NA, <50%, 50-100%) and to indicate the only the type of agricultural fiber, if used as furnish.       

Change Made. Column K was deleted.
F5
NCASI
EquipDetail Tab:
   Columns B-F  -  use "process unit" instead of "emission unit" to be consistent with the terminology used in 40 CFR Part 63, subpart A

Are survey respondents required to enter a complete duplicate set of Emission Unit IDs, Emission Unit Types, Descriptions, and Product entries if the product line makes more than one product? For example, if a particleboard mill makes UF and MF+MDI panels, does each product type need its own set of emission unit types covering from chipper through sander?
Change made. The term "emission unit" has been replaced with "process unit" throughout the ICR because "process unit" is defined in the PCWP rule.  

Survey respondents should not duplicate the process units listed in the EquipDetail tab. Data validation is included in the Process Unit ID column to prevent duplicate entries. The data in the EquipDetail tab will be used to count process units across the industry would be complicated by duplicate entries. Instructions have been added to the instruction document to clarify this point. 

Column H in EquipDetail allows entry of 2 products lines for the same press. For example, if "press 1"is used for two product lines, "UF" and "UF+MDI," then "UF, UF+MDI" can be specified in column H.
F6
NCASI
EquipDetail Tab:
Column D Emission Unit Type  -  Please clarify instructions in spreadsheet so that they match the options provided in the survey instructions document on page 16.
The options listed in the spreadsheet instructions are:
"Group together ducts/zones from an emission unit that are collected and controlled together such that emissions exit from a single release point." This text is redundant.
"However, if the emission unit has multiple emission release points, list each emission release point on a separate line and use a different emission release point ID# in column Q." This text appears to require the entry of emission release points as separate emission units. This is not correct because this column is to identify individual process units. There would only be one process unit even if that process unit had multiple emission release points. This is an example of how the use of the term "emission unit" is confusing. This text should be modified to state: "For a process unit that has multiple emission release points, list the process unit ID once and identify all associated emission release points in columns Q-V."
"For multiple emission units with a common release point...list the process units separately but be sure to use the same emission release point ID in columns Q-V." EPA needs to add text in bold "release" to spreadsheet instructions.
Change made. Instructions were revised.

F7
NCASI
EquipDetail Tab:
Column F Product  -  Instructions column heading indicates that the "Product" entry will be prepopulated from Prod Tab but cells contain drop down list.
Column G Product Line  -  Column instructions could use additional clarification. "Select the applicable product line from the drop down menu options provided and, if applicable, add any additional product lines separated by commas."
(F-G) Change made. The instructions were revised.

F8
NCASI
EquipDetail Tab:
Columns H-O Air Pollution Control Device Information:
The global instructions for filling out this area of the survey are confusingly dispersed within the individual column headings. EPA should compile the information in cells H8, I8, and J8 together within cell H7.
The separated "APCD list" options provided in cells L8, M8, and N8 should either be combined into one list and this entire list replaced/placed it in cells H8, J8, L8, and N8 or simply have survey respondents refer to the instruction manual and put a note stating that in the instructions for those columns.
(H-O) Change made. The APCD Type Codes are provided in the instructions document. The vertical lines in cells H7 through O7 were removed to clarify that the instructions provided apply to all APCD IDs and Types.


F9
NCASI
EquipDetail Tab:
Column P Air Pollution Control System  -  EPA can pre-populate the entries in this column by concatenating columns H, J, L, and N.
Concatenation would insert extra unwanted characters into the field (e.g., the "/" used to separate controls). The drop-down menu is simple and contains the most common control combinations, and is not encumbered by controls in parallel. An example of parallel controls was added to the instructions row in case such configurations exist.  
 
F10
NCASI
EquipDetail Tab:
Columns S-V  -  Formatting is not correct in rows 21-23. The black fill does not format as it does with the other rows.
 (S-V) Change made. Black conditional formatting was updated. All tabs now have data entry beginning on row 24 to facilitate extraction of spreadsheet data into the EPA's data base.

F11
NCASI
EquipDetail Tab:
EPA has added text to provide the option of grouping similar emission units of "lesser importance in terms of total facility HAP" into a combined group that can be entered into one row. If this option is selected, then the similar emission units are grouped together in Column E "Emission Unit Description" along with the number of units. This new option appears to conflict with the requirements stipulated under the Multiple emission units ducted to a single release point requirement. EPA should clarify any discrepancies between these two requirements.
Change made. The commenter is correct regarding the apparent conflicting instructions. We have eliminated the option to group multiple similar emission units of "lesser importance in terms of total facility HAP" because it is important for us to be able to count emission units nationwide.  
F12
AWC
EquipDetail Tab:
The tab should be adjusted to accommodate alternate operating scenarios, or the instructions adjusted to indicate only the primary operating scenario should be reported.
The heading of column AB should state that the compliance option should be the one associated with the most recent test (to match the instructions).
Change made. The instruction document was updated to indicate only the primary operating scenario should be reported, and that the Comments column could be used to describe any alternate operating scenario pertinent to EPA's understanding of equipment operation.

Change made.
F13
NCASI
ReleasePt Tab
Column I Exit Gas Temperature  -  Stack gas temperature is a stack gas parameter so this column should be moved to the Stack Gas Parameter section.
Column J Release Height  -  This parameter should be moved to Column I.
Column K Stack Diameter  -  This column should be included within the Release parameters and should replace Column J. There is also no discussion of equivalent stack diameter or how to report the stack diameter of non-traditional emission point configurations, such as a side mounted vent on a baghouse.
Column L Measured average flow rate of gas stream (ACFM) -  EPA is asking for acfm then states that if you do not have acfm, then enter dscfm.  
If EPA is going to allow the option of entering either acfm or dscfm then there needs to be an IF STATEMENT in Column P to address the auto-calculation formula in that column.
The suggested conservative default parameters of 72 oF for ambient temperature and 29.92 in. Hg for ambient pressure for every source without measured stack gas parameters may not be an appropriate approach for some process unit emission points.
EPA should provide a template that can be populated with stack test summary data. All that would be required is that the summary table within a stack test report be referenced into a spreadsheet template formatted to meet the needs of this ReleasePt tab. The template can then be pasted into the corresponding input sheet within this ICR. The ReleasePt tab would then simply reference the data within the input sheet. 
Column M Gas Flow Basis  -  move to Column K as the first question EPA asks relative to Stack gas parameters.
Column P Exit gas flow rate  -  Unclear whether this column is intended to be an extracted value from Column L. If a responder enters dscfm into Column L, there is a potential that incorrect flow rate results will be propagated.
(I and J) The exit gas temperature and release height are relevant for stack and fugitive sources. No change in placement is needed.

(K) Entry of the actual stack diameter is only required for stack release point types so placement of this column was not changed. The "stack" emission release point types available for selection in the "Emission Release Point Type" column include:
02 - Vertical
03 - Horizontal
04 - Goose Neck
05 - Vertical with Rain Cap
06 - Downward-facing Vent
Facilities can select from among the most representative of these emission release point types for non-traditional configurations (e.g., a side mounted vent could be labeled as 03-Horizontal).

Defaults will be applied for the diameter when fugitive emission release point types are selected:
07 - Fugitive Area [Reserved]
08 - Fugitive Vent
09 - Fugitive Two-dimensional
10 - Fugitive Three-dimensional

(L and P) Change made. Column L was updated to require ACFM without reference to DSCFM. The corresponding conversion formula noting defaults of 72ºF and 29.92 in. Hg. was deleted from the instructions.  

No stack test summary template was designed. Conditional formatting and validations were added to the ReleasePt tab to guide the user though population of the correct columns for each emission release point type. 

(M) Change made. Column M (gas flow basis) was moved to precede the ACFM column. 

F14
AWC
ReleasePt Tab
Some facilities may have modeled fugitive emissions from the ends of continuous lumber kilns as volume sources in permit modeling exercises, so EPA should include the option to characterize a source as a fugitive volume source in the ICR.
Change made. An emission release point type "10 - Fugitive Three-dimensional" was added to the spreadsheet for facilities that wish to characterize continuous kilns as volume sources.
F15
NCASI
Permit Tab:
Columns G, J, and M Non-PCWP NESHAP HAP permit limits  -  EPA should refine this list to only include PCWP sources to improve the efficiency of data entry. Furthermore, the list of "Other HAP" contains the 6 HAP compounds (acetaldehyde, acrolein, formaldehyde, methanol, phenol, and propionaldehyde). Updating these would avoid incorrect data entry.
Columns Q - S VOC Limits  -  Since VOC data are being requested, by default this indicates that methane is not to be included in the reported values. EPA should add non-methane options in the drop down menu.
Columns V & W PM Limits  -  Menu options in Column V do not match the menu options in Column W. For example, the "PM compliance test" menu in Column W includes Method 5 but "PM" is not provided as a menu option in Column V. Furthermore, EPA has provided PM compliance test options M201, M201A, and M202 but has not provided "filterable PM10", "filterable PM2.5", "CPM", "direct PM10", "direct PM2.5" menu options in Column V.
(G, J, and M) Change made. The instructions were rewritten to clarify the intent of these columns. The pull-down menu is provided for convenience but may be overwritten. The menu has been amended to remove the 6 PCWP HAP.

(Q-S) Change made. Non-methane VOC was added to the menu. 

(V, W): Change made. We edited the PM section of the Permit tab to include separate columns for filterable PM (Method 5 or similar methods) and another "smallest" size fraction of PM including PM10, PM2.5, CPM, Primary PM10 (PM10 + CPM) or Primary PM2.5 (PM2.5 + CPM). We added a menu listing the relevant EPA test methods for the different PM fractions including a choice of M201A/M202 for the primary PM limits. "Primary" PM10 and PM2.5 are also known as "direct" PM10 and PM2.5.  Reference to "TSP" was removed.  

F16
AWC
Permit Tab:
The list of all non-MACT limits and tests required by the permit is not relevant and creates extra burden for respondents. Opacity limits are not relevant to PCWP MACT compliance.
Agree with EPA's instructions that state-only air toxics limits based on modeling analyses should not be reported.
The heading of column E should state that the compliance option should be the one associated with the most recent test (to match the instructions).
Change made. Column E was reworded. The questions in column P and Z regarding non-PCWP compliance tests were deleted. Questions pertaining to permitted opacity limits were retained in lieu of the request for COMS data which was deleted in response to comments E10-E11.
F17
NCASI
Resin Tab:
Columns B  -  D Pre-populated Data  -  In cell B8, replace "Pre-populated Data" with "Pre-populated data will appear if "Yes" was selected in Column O of `Prod' tab"
(B-D) Change made to refer to correct column in the Prod tab.

F18
NCASI
Resin Tab:
Column F Principal resin type  -  There are two issues with `Principal resin type' field:
 Instructions for "Principal resin type" 
Currently states: "Describe HAP scavengers by overwriting "specify" the `Other' pull down." Change to read something like: "Describe HAP scavengers by selecting the "Other: {specify}" option within the pull down menu and then go to the formula bar to overwrite "{specify}" with the HAP scavenger type."
Also add "Other: {specify}: for HAP scavengers" at the end of the PF to PRF list of options shown
 EPA appears to be asking that both the type of resin and HAP scavenger be entered for each principal resin which requires two rows of data since both the resin type and HAP scavenger are entered into the same Column F. Column B (FRS Site ID), C (Product Line), and D (Product) are protected fields that are prepopulated so duplicate entries are not allowed. EPA needs to clarify how to input both the principal resin type and Other: HAP scavenger type for each resin type.
The best way to resolve this issue would be to add another column between column F and G for the entry of any HAP scavengers associated with the principal resin. If EPA agrees with this, then EPA will have to adjust the instructions in both the column headers and Instruction document.
(1) Change made. The word "principal" was eliminated from column F and instructions were added as requested to clarify that key resins, additives, or scavengers can all be listed in column F.

(2) Change made. Instructions were added to clarify that multiple rows can be entered for a given product line to accommodate products with multiple resins or additives to be reported.
F19
NCASI
Resin Tab:
Columns I  -  O  -  EPA could provide more details on the information being requested relative to resin solids. Provide a definition of resin solids and example calculation in the instruction document for the following:
Column I  -  "Resin percent solids application by weight of resinated wood material"; does "resinated wood material" refer to the uncured resinated furnish exiting the blender and being sent to the former or the final panel after pressing? EPA should ask for average values.
Column J  -  "total tons of resin solids used each year": This request covers the amount of resin used to make the product per year? That would depend on how much product was made which is reflecting market conditions. Since the parameters impacting market conditions are variable and not steady-state, how relevant is this information going to be? If this information can be used to estimate production throughputs, this could be CBI.
Column K  -  "average percent solids by weight in the resin": This should be an average value
Column L through O  -  percent by weight of formaldehyde, methanol, phenol, MDI, or other HAP >= 0.1% by weight: most SDSs will only have information on formaldehyde, MDI, and maybe phenol for PF resins. MDI resin is supplied in various forms and products. There is 100% pure MDI (MDI) and polymeric MDI (pMDI) which contains a percentage of MDI. %Methanol will be unknown. Methanol content will not be on the SDS sheets for UF or PF resins because the hazard level and toxicity are related to formaldehyde. Historically, resin manufacturers have not tracked the methanol content in resins nor has the amount of methanol in resins been required to be controlled. Column M (methanol) should only be an option to report since this information will not be readily attainable for most facilities and is expected to be highly variable.
(I) Change made. This question was deleted given the various ways it could be interpreted and considering the potential uncertainty in measurement of resin solids mass in relation to furnish at various moisture contents and points in the process. As noted below, we have retained a more important question of total annual resin solids used.

(J) The total annual solids usage (tons/yr) was retained to aid in mass-balance calculation of resin-related HAP emissions (i.e., resin solids tons/yr x % HAP by weight). 

(K) Change made. The word "average" was added to the resin percent solids column.

(L-O) We have retained the questions in columns L through O as is because of the relative importance of gathering resin HAP content information that could potentially be used in establishing work practices for infeasible-to-test process units downstream of resin application. As explained in the instructions document, "Unknown (UK)" is a valid answer when facilities do not have the requested information, cannot obtain the information without extraordinary effort, and cannot provide a reasonable estimate. Facilities may enter "UK" if they have no information on methanol. The option to respond with "UK" was added to the resin tab instructions in the instructions document.  

F20
NCASI
Resin Tab:
Column Q "...lower HAP resins..."  -  EPA should consider adding "Customer demand" or "market demand" as a menu option.
(Q) Change made. Customer and market demand were added to the menu.
F21
NCASI
Resins Tab:
Column E Product Trade Name and Catalogue Number  -  EPA should change the reference to "MSDS" the new system "SDS."
(E) Change made.

F22
AWC
Resin Tab:
EPA should ask for copies of the MSDS or the information from MSDS but not both.
Change made.  We eliminated the request to submit MSDS.


F23
AWC
All the Process tabs:
   EPA asks for operational work practices that reduce HAP emissions on some but not all of the process tabs. A column for providing this information should be provided for every process type.
   The columns requesting operating parameter data on all of the process tabs (e.g., temperatures, moistures, cycle times, batch lengths) should be removed as these will not inform the RTR.
   Survey questions that are speculative should be eliminated or made optional. For example, on several tabs, EPA asks each facility to speculate on all the operational and safety factors that might make it infeasible to enclose a particular process unit. The survey should ask facilities to submit available information, not to create new information.
We disagree that all operating parameter data in the tabs should be removed. Questions are generally limited to parameters that relate to HAP emissions (e.g., temperature and inlet/outlet wood moisture).  Parameter information will be useful in evaluation of work practices (as appropriate) or subcategories for process units that are infeasible to enclose and test. However, as discussed elsewhere, we have reduced the number of columns requesting parameter data and we have eliminated our request for CPMS data to reduce respondent burden. 

Regarding open-ended questions that the commenter refers to as "speculative," the instruction document explains that facilities that do not have the requested information, cannot obtain the information without extraordinary effort, and cannot provide a reasonable estimate, may enter "UK" for unknown. Regarding the example noted by the commenter, collecting operational and safety information is expected to help in the development of work practice standards where they are appropriate so we retained the question.
F24
NCASI
VeneerDry Tab:
Column H 2015 Dryer Throughput (MSF/yr 3/8")  -  EPA is requesting dryer throughput in units of MSF/yr 3/8-in basis but is allowing numerical values to be entered that are on alternative basis, such as MSF/yr. If an alternate basis is used, there needs to be a provision to make note of this in the comments column. This is not the correct method for data input. To avoid confusion and data errors, EPA needs to add another column to the right of Column H and allow the user to select the associated production rate units from a drop down menu.
Column AD What percentage of cooling zone...  -  Instructions in cell AD8 are not completely shown due to the row height. 
Cell AF7  -  typo: "Veneer Redryers and Rredrying in Conventional Dryers".  "Rredrying" should be "Redrying"
Changes made. A column was added to allow for selection of MSF/yr or MSF/yr 3/8" and the typos were corrected. 
F25
NCASI
RotaryDry Tab:
Column F drop down menu includes "Radio frequency" dryer option. Remove this section to avoid entry errors.
Change made. 


F26
NCASI
TubeDry Tab:
Column F drop down menu includes "Radio frequency" dryer option. Remove this section to avoid entry errors.
Change made. 
F27
NCASI
ConvDry Tab:
Column F drop down menu includes "Radio frequency" dryer option. Remove this section to avoid entry errors.
Change made. 
F28
NCASI
Lumber Kiln Tab:
Column E  -  Drop down menu includes "Radio frequency" dryer option. Remove this selection to avoid entry errors.
Column K header  -  Change text to indicate: "Describe the method by which the change in the lumber moisture content is monitored".
Columns L  -  R  -  Cell L8 is the header for the "Batch Kilns" information section, EPA should note here that all data is for 2015 operations.
Column L  -  EPA should add a column in the Batch Kiln section. Column L should be the new column for "Design Kiln Capacity, MBF per charge". Change header of existing Column L "Batch kiln number of cycles per year" to "Number of kiln charges dried per year." Now one can match up these two entries with Column G entry.
(E) Change made.  Radio-frequency was dropped from the firing method column. "Electric Dehumidification Kiln" was added as a menu choice to ensure a characterization is available in the table for this type of kiln most commonly used for hardwoods.

(K) Change made.

(L-R) Change made.

(L) Change made.
F29
NCASI
Lumber Kiln Tab:
Columns M through R for Batch Kiln operating parameters:  The survey is asking for min and max drying times. To get any applicable outcome from the min/max drying time ICR collection segment, EPA is going to have to attempt to differentiate between the factors that go into variable drying times. To clarify or improve this section of the survey, EPA should re-work Columns M-R by incorporating the following components:
Impacts from seasonal variability: This is probably the only parameter that has a consistent impact on drying times because composition and heat input are variable parameters. The initial wood moisture content may vary by season.
Impacts from composition of kiln charge: Are the kiln charges primarily composed of uniform or variable dimensioned lumber?  
Impacts from Heat Input
Are there other considerations impacting kiln drying cycle times that should be considered
{The commenter provided background on lumber drying and specific examples of survey questions that could be added (see p. 8-9 of supplemental comment).}
We acknowledge there are many factors that impact the length of the kiln drying cycle. We reviewed the numerous additional survey questions suggested by the commenter that illustrate the variability in kiln cycle length but elected not to include the questions in the survey. Because kiln cycle time is so variable and site specific it is unlikely that any work practice could be defined in terms of kiln batch cycle length. We believe the questions in the survey pertaining to kiln design, lumber outlet moisture content, and maximum target temperature are sufficient to inform potential work practice standards. 
F30
NCASI
DFDryFuel Tab:
Columns B (FRS Site ID), C (Emission Unit Type), and D (Dryer Emission Unit ID)  -  These columns are not being pre-populated from the EquipDetail tab except for `Fiberboard mat dryer'. This could be due to reference errors within the VLOOKUP formulas within the column range AM:BJ of the `Dryer Fuel Calcs' tab.
Change made.
F31
NCASI
Press Tab:
Columns J and K Panel thickness  -  Instructions should be merged between these two columns.
Column P Unloading method  -  This column is missing a drop down menu.
Column S Pressing cycle length  -  EPA is asking for press cycle time to be entered in this column but instruction also wants: "List thickness basis used if standard thickness is not produced". Provide example of a non-standard thickness entry.
Column W Production Rate  -  Same comment as Column S, provide example if correct.
Column AD Describe method used... - What does "...and provide capture efficiency measure" mean as asked for in the instructions. Is the user supposed to select the method used to determine TTE capture efficiency and then add to it the measured number? Clarify and/or provide example.
Column AE Are there constant human... - typo in the instructions, "(e.g. yes, exposureby worker on press elevator)".
(J-K) Change made.

(P) Change made.

(S) Change made. Example added.

(W) Change made. Example added.

(AD) Change made. Phrase deleted.

(AE) Change made.


F32
NCASI
BC Tab:
Column D Emission Unit ID  -  There are no pre-populating formulas provided in cells D13 and D18.
Change made. The formulas were checked. All tabs now have data entry beginning on row 24 to facilitate extraction of spreadsheet data into the EPA's data base.
F33
NCASI
EWP Press:
Column I Press throughput... - Instructions are cut off due to row height. Redundant units in drop down menu; Mft/year and MLF/yr are the same units. Delete one of these options.
Changes made.
F34
NCASI
FB-HB Tab:
Column N Firing Method  -  Cell N16 might already filled out as "Direct-fired" instead of blank.
Column O  -  This should be labeled as "Moisture content of the mat at dryer inlet, on a %-dry basis" not on an "oven dried basis."
Column P Target dryer Outlet... - If EPA is asking for board moisture at the end of drying, then suggest changing the instruction text from "Enter Target Dryer Outlet Moisture Content (oven dry % basis)" to read "Target moisture content of the board at the outlet of the dryer, %-dry basis."
(N) Change made.

(O) Change made.

(P) Change made.
F35
NCASI
OtherEquip Tab:
Column G Throughput units of measure  -  Drop down menu is missing ODT/yr.
Column N Estimated wood temperature at exit  -  This may not be a reasonable parameter for mandated reporting. Wood temperature at the outlet of the refiner is most likely not an operational requirement and therefore not readily available. The same applies to wood temperature coming out of a digester.
Column Q Describe...work practices... - EPA should allow mills to list sanders with common emission points together like they allow for saws (Column S) and chippers/flaker (Column V).
Column S Number of saws in emission point grouping  -  Clarify instruction text: "List saws with common emission points and controls together." EPA should clarify that "saws" is not referring to the individual saw blades here but to the whole saw-line process unit that has an individual process unit ID. EPA should also clarify whether single saw-line process units that have dedicated pneumatic dust collection systems are to be entered here or only the configuration where multiple saw line process units share a common pneumatic dust collection system.
Correct instruction text: "Do not list sawmill sanders used for lumber production".  "Sanders" should be "saws".  Is EPA indicating that saw process units at lumber mills are not be included here?
The example provided by EPA for this section of the survey has Emission Unit ID:Other-3, Emission Unit Type:Other, Emission Unit Description:other examples.  Is this the format that is to be used for sawline process units with combined exhaust point?
Column V-Y Chippers and Flakers  -  This section of the survey is mainly for information on chippers/flakers which are dry end process units. The example provided by EPA for this section lists Emission Unit ID:Flaker-1, Emission Unit Type:Chippers\flakers\hammermill, Emission Unit Description:OSB stranders which is not correct since a strander does not fit into the same category as chipper\flaker\hammermill.  EPA needs to be aware that chippers, flakers, and stranders are different types of process units.  
(G) Change made.

(N) Change made. The question asking for estimated wood temperature at the digester/refiner exit was removed as part of our effort to reduce the number of parameter-related questions. 

(Q-Y) Change made. We revised the OtherEquip tab to clarify that finishing sanders, finishing saws, and panel trim chippers are of interest, and not green-end material handling sources. We removed unnecessary columns based on this clarification. Process units are identified in the EquipDetail tab and the process Unit ID is prepolulated in the OtherEquip tab. Appendix 2 of the instructions was revised to explain that finishing saws refer to saw lines, not individual saw blades. Revised instructions clarify that facilities producing only kiln-dried lumber are not required to complete the OtherEquip tab. Thus, sanders and saws at lumber mills would not be included.
F36
NCASI
APCD Tab:
Global tab note: EPA should consider adding conditional formatting to the columns to the right of Column I that is based on the entry in Column G to facilitate data entry accuracy and efficiency.  For example, if the entry in Column G is a baghouse "BH", then all other non-baghouse related column entries should be shaded for inapplicability.
Columns B  -  G: It does not appear that rows 22 and greater are being pre-populated 
Columns K - N: EPA needs to distinguish between organic HAP APCDs and other end of pipe gas stream conditioning devices by shading the rows that do not have biofilter, RCO, or RTO in column G. 
Conditional formatting was added as requested.

(B-G) Change made.

(K-N) Change made. The HAP and THC control efficiency columns were shaded for controls that do not affect HAP or THC emissions (ESP, baghouse, cyclonic collectors, and rotary bed protectors). The HAP and THC columns were retained for incineration-based HAP controls, as well as for other controls that could potentially affect HAP or THC emissions to a lesser degree (wet ESPs, scrubbers, EFBs, and sand filters).
F37
NCASI
APCD Tab:
Columns K  -  O: 
EPA needs to distinguish between organic HAP APCDs and other end of pipe gas stream conditioning devices, such as a baghouse. Since EPA is specifically asking for HAP control device data, EPA should shade the rows that are not relevant to organic HAP APCDs, for example: organic HAP control efficiencies are not relevant to saw-line baghouses. This would significantly speed up the survey in this area.
"Actual"  -  unclear whether EPA is soliciting Inlet/Outlet test data.
Column K - Specify HAP Controlled  -  The information requested in this column is redundant since it has already been specified in Column AB of the `EquipDetail' Tab. EPA needs to pre-populate Column K with that information. If EPA is not going to pre-populate this column, then the instructions need to be clarified. Since most thermal oxidizer APCDs actually provide control of all 6 HAPS, THC, and VOC at the same time and there is no add-on control compliance option specifically for the 6 PCWP `total HAP', the first part of the instructions need to be corrected from: "If the APCD controls a specific HAP or the 6 PCWP `total HAP', list here. PCWP HAP are listed. If other HAP are measured for control, enter on blank line." to a more appropriate instruction: "Select the appropriate compliance option for this add-on control system." Then provide the same dropdown menu as provided in Column AB of the `EquipDetail' Tab.  
(K-O) We clarified the instructions for APCD columns K through O. First, as noted above, we applied conditional formatting throughout in the APCD tab to block out columns that are not required based on the specific type of control device indicated in column D. Second, we reworded and revised columns K through O to eliminate the word "actual" and to clarify our request for measured control efficiency values for formaldehyde, methanol, THC and/or PM. Respondents are only asked to populate these columns if inlet/outlet testing across the APCD has been conducted for the pollutants listed. If not, then respondents would leave the columns blank.     

As part of the edits to the control efficiency columns, the "Specify HAP Control" column (K) was removed.

These columns are not redundant with the EquipDetail tab which asks about the PCWP compliance option used, but not the measured control efficiency.
F38
NCASI
APCD Tab:
Column L and M: These columns appear to be redundant for control compliance options that rely on the 90% emission reduction strategy. If formaldehyde or methanol are selected as an add-on control compliance option, then there will be no available control efficiency data for THC. Likewise, if THC is selected, then there will be likely no information on formaldehyde or methanol control efficiencies. As a consequence, Columns L and M can be combined. If EPA is not going to combine the control efficiency data requests, then the "If known" statement in Columns L, M, and N should be stated first so that survey responders can quickly assess applicability.
Column N Actual VOC CE  -  Not sure about the applicability of Column N. Are mills going have a PCWP MACT THC compliance option requirement PLUS an additional permit requirement for VOC control efficiency?
Any permit that is going to contain a requirement for VOCs to be reported as a calculated value, such as OTM26 or similar methodology, is going to require this demonstration only at the outlet of the APCD and not the inlet. As a consequence, it is highly probable that none of the facilities will be filling this in. Note that the PCWP MACT I does NOT have a calculated VOC requirement, however, since that time PCWP facilities have been required to report VOC outlet emissions as per OTM26.
(L and M) We have retained columns L and M and revised their instructions to account for the possibility that methanol or formaldehyde or THC control efficiency have been measured for a given control device, either at the same or different times. 

We added the initial phrase "if inlet/outlet testing has been conducted" to clearly state our intent to gather measured values as opposed to more theoretical vendor-quoted values.


(N) The VOC control efficiency in column N has been removed from the APCD tab. We are not requesting calculated VOC results because compliance with the PCWP rule is based on measurement of formaldehyde, methanol, or THC (Method 25A), and does not require calculation of VOC through use of methods such as OTM-26 (as needed for permitting programs). However, we have clarified in instructions for the EmTest tab that we are asking to receive test reports where OTM-26 calculations were applied to a Method 25A THC (as propane) measurement because these THC values can be converted to THC (as carbon) for use in evaluating the outlet emissions of control devices under the RTR.   


F39
AWC
MiscCoat Tab:
EPA should ask for copies of the MSDS or the information from MSDS but not both  -  this would reduce burden.
One column asks "is the coating currently a `non-HAP' coating?" As defined in the PCWP rule, "non-HAP coating means a coating with HAP contents below 0.1 percent by mass for OSHA-defined carcinogens as specified in 29 CFR 1910.1200(d)(4), and below 1.0 percent by mass for other HAP compounds." You do not need to fill out the rest of this tab for non-HAP coatings. What is a carcinogen depends on the updated OSHA rules which were changed March 26, 2012 to align with the UN Globally Harmonized System of Classification and Labeling of Chemicals and effective May 25, 2012. It appears that 29 CFR 1910.1200(d)(4) is no longer in the OSHA rules. Respondents need a clear list of carcinogens to reference to determine if they need to fill out this tab or not. EPA should clarify its intent.
Columns F & G  -  It appears that Column F "Changes made to comply with non-HAP coating requirement..." and Column G "Is the coating currently a "non-HAP" coating?"  are not in the correct order.
With the current coating being used entered in Column D, EPA should then ask if it is a "non-HAP" coating. 
If "yes", then ask if changes were made to switch to this coating and what were those changes. EPA should CLEARLY highlight that the rest of the tab does not need to be filled out if "Yes" is selected as mentioned at the end of the instructions within cell G10 which may be over looked. 
If "No", then the rest of the tab is filled out. 
Column J (HAP 1), L (HAP 2), and N (HAP 3):  HAP list could be revised to be more reasonable and manageable; does "Coke oven emissions" apply here?
Change made.  The request for MSDS was eliminated.

Change made. The outdated OSHA reference 29 CFR 1910.1200(d)(4) refers to lists of carcinogens published by the following sources:
    (i) National Toxicology Program (NTP), Annual Report on Carcinogens (latest edition); available at  https://ntp.niehs.nih.gov/ntp/roc/content/listed_substances_508.pdf
    (ii) International Agency for Research on Cancer (IARC) Monographs (latest editions); available at
http://monographs.iarc.fr/ENG/Classification/latest_classif.php, and 
http://monographs.iarc.fr/ENG/Classification/Table4.pdf

These lists may be used for purposes of the ICR to identify non-HAP coatings.  Thus, for purposes of the ICR, a "non-HAP coating" means a coating with HAP contents below 0.1 percent by mass for carcinogens as specified by the National Toxicology Program (NTP) or International Agency for Research on Cancer (IARC), and below 1.0 percent by mass for other HAP compounds. Respondents are not required to complete the miscellaneous coating tab for non-HAP coatings. We will further review and propose amendments to the definition of "non-HAP coating" in the PCWP rule as part of the forthcoming technology review. 

(F & G) Change made. The order of the columns was adjusted, and the statement that "You do not need to fill out the rest of this tab for non-HAP coatings" was bolded. 

(J, L, N) Change made. The list was reduced to HAPs most likely to be associated with HAP-containing wood coatings: xylenes, toluene, ethyl benzene, ethylene glycol, methyl isobutyl ketone (MIBK), methanol, MDI, styrene, formaldehyde, acetaldehyde, glycol ethers, propionaldehyde, and triethylamine. Other HAPs can be written in.
F40
AWC

WW tab:
The term "wet process control" is inconsistent with the terminology in the PCWP MACT ("wet control device"). The ICR should use the PCWP MACT terminology so the definition is clear.
Column I: There should be three choices: enclosed tank, open top tank (such as a pretreatment clarifier), and open pond.
Column J: EPA should provide guidance on how to handle hard-piping that goes through lift stations that are open to the atmosphere.
Change made.  "Wet control device" is now used.

(I) Change made. A menu was added containing the choices: enclosed tank, open top tank, open pond, and NA.

(J) Change made. The question was reworded to say, "select yes or no regarding whether the process wastewater is predominantly enclosed (hardpiped) to the POTW from the facility, with the exception of lift station vents." (emphasis added)
F41
NCASI
WW tab, Column B:
FRS Site ID  -  This column is not being pre-populated as the instructions state it should be but is providing a dropdown menu. EPA should clarify.
(B) Change made. The column is now pre-populated with the ICR ID. The number of rows in the tab was also reduced to reflect that a facility-level response is requested as opposed to process-unit-level responses. 
F42
NCASI
WW tab, Column C:
Onsite vs offsite WW treatment  -  The text in cell C10 and the text in Instructions C9 do not seem to be asking the same thing.  "on-site" would apply to settling ponds but wouldn't it also apply to storm water ponds as well? "Off-site" would apply to POTWs which are not mentioned here. 
"HAP-containing process waters" needs to be qualified. At ppb levels, the response will be "yes" for all mills. If at 1 ppm as TOC or 1 ppm of methanol or formaldehyde levels, then some mills might not need to complete the remaining tab information.  
Also, wastewater operations are defined in the instructions document, but wastewater treatment is not. Some may not think of evaporation ponds or holding tanks as wastewater treatment.
Changes made. Instructions were aligned with the query to ask, "Does the mill generate HAP-containing process waters from equipment or air pollution control devices that require onsite or offsite wastewater treatment to remove HAP?" This focuses on wastewater treatment methods designed with the intent of removing HAP from the water. For purposes of the survey, the instruction row was modified to state that "HAP-containing" process waters are waters with the concentration of any HAP in excess of 1 ppm. 
F43
NCASI
WW tab:
Column E: The text "wet process control" within the phrase "...facilities use wet process control..." is confusing because this could mean controlling the process (not sure if that makes sense) or controlling process gases for HAP emissions, which would be a wet APCD.  Has EPA adequately provided a definition for a wet air pollution control device?
Column F: Instructions in F9 indicate that the remainder of the tab does not need to be completed if criteria in this cell are met.  The Instruction document, on the other hand, states that: "respondents will only need to complete the spreadsheet as far as the POTW information." These two instructions do not compare.
Column I: Is there supposed to be a drop down menu here for "Tank" and "Open pond"? Also, is this column supposed to be for wastewater collected and held before being discharged to the POTW? If not, delete.
Column W: The drop down menu is only available in the first row. Not sure if this is by design or not.
(E) Change made.  Reference to "wet process control" has been replaced with "wet control device." 

(F) Change made. Instructions were aligned to clarify that if PCWP process wastewater is sent to a POTW for treatment, respondents will only need to complete the spreadsheet as far as the POTW information in column K.  

(I) Change made. A menu was added containing the choices: enclosed tank, open top tank, open pond, and NA. The instructions were clarified to indicate that the question is asking about wastewater storage prior to discharge to the POTW.

(W) Change made. Drop down menu fixed.
F44
AWC
Primary HAP Emissions and Other HAP Emissions tabs:
EPA should limit its request to any available site-specific data; many facilities use emission factors to estimate emissions, and it is not clear how EPA will benefit from requiring sources to report conservative emissions estimates based on emission factors that may not be representative of individual source HAP emissions. If emissions estimates are needed to evaluate risk, EPA can use the throughputs provided in the ICR and its own proposed emissions tool to generate those estimates.
The Primary and Other HAP Emissions tabs were consolidated into a single HAP Emissions tab and transposed to minimize the number of columns. The tab is structured to gather emission estimates from individual process unit release points. The instructions document was updated with details on how to provide emission estimates, including how to apportion emission estimates for different process unit and emission point configurations. In addition, the EPA included columns in the HAP Emissions tab with optional provisional calculations as described in section IV of this document. The provisional calculations couple process details provided in the ICR with publicly-available emission factors to estimate emissions. Site-specific data are preferred, but the provisional calculations were added to ease burden for facilities that do not have more-representative emissions data to provide. The EPA is interested in stakeholder comments on the utility of the provisional calculations, whether they achieve the goal of reducing burden, or whether it would be more appropriate for facilities to develop their own emission estimates based on data available to them.
F45
NCASI
Primary HAP Emissions and Other HAP Emissions tabs:
Columns H, N, T, Z, AF, and AL  -  EPA should not use the term "Actual" when the results being reported have not actually been continuously measured throughout the year.  The data input would be an estimate at best based on operating times and emission factors from NCASI database, one off NCASI engineering test, or performance test.  Engineering tests and NCASI database values may not represent the emission factor value obtained from the same production rate level as used for a performance test. Performance test will reflect max production level to test the APCD, so is that reasonable to apply to the entire year?  
Change made. The term "Actual" has been replaced with "Estimated Actual." Residual risk modeling is intended to be based on actual emissions to the extent that actual (measured) emissions data are available for a given HAP.  However, when actual (measured) emissions data are not available for a given emission point, reasonable estimates of actual emissions may be used for purposes of risk modeling. We revised the terminology to reflect this intent. Similarly, we added the term "estimated" to the columns seeking allowable and maximum emissions. Facilities are asked to provide their best estimate of emissions based on information available to them.
F46
NCASI
Primary HAP Emissions and Other HAP Emissions tabs:
Columns for "Maximum Emissions"  -  Will mills not know how to estimate this?  Seems that this will be very difficult to determine and thus responses will be variable making this data useless.  
First, mills will only have lb/hr data from stack test reports which only reflect conditions during the test.  How to translate an emission factor or test report value to the highest short term emissions on a lb/hr basis is going to very difficult.  At best, one would simply use the uncontrolled dryer emissions, but uncontrolled emissions are seldom measured anymore for all 6 individual HAPs.
Second, this data should also have a max emission time frame associated with it so that the worst case emissions are not used for the entire year.
Third, EPA should waive this if the mill has been in compliance for say 90% of the year, then no short term estimate is required.
Facilities are asked to provide their best estimate of maximum short-term emissions in pound per hour based on information available to them. Maximum emissions on a short-term basis are requested for purposes of acute risk modeling. An example of maximum hourly emissions may be short term (peak) uncontrolled emissions during peak production rates or during a time when the routine control device maintenance exemption (RCDME) is used (e.g., when the 90% control efficiency is not being achieved). Within the provisional calculations, the maximum short-term emissions are calculated as the annual uncontrolled emissions divided by annual operating hours to approximate maximum lb/hr assuming there is 1 hour when the process is operating but the control device is out of service (e.g., during the RCDME). For uncontrolled process units or process units that do not use the RCDME, facilities can provide the maximum lb/hr assuming peak production or by another means according to their best engineering judgement. 


F47
NCASI
Tank tab:
Columns B through E are not being prepopulated. It appears that the VLOOKUP formula contained in these fields is returning a "#NA" from the referenced data within the `Tank Calcs' worksheet. The VLOOKUP considers this an error, therefore the formula results in a blank value. The data being referenced in the `Tank Calcs' worksheet is "#NA" because there are multiple issues with the VLOOKUP formulas being used in that worksheet. The following is an attempt to diagnose and resolve the issue within the `Tank Calcs' worksheet: 
Columns R through AC in the `Tank Calcs' worksheet are seeking incorrect references.  
First, there is a breakdown in the formula thread with the referenced `table_array' range of $A$3:$M$252.  This is incorrect because the data required to be referenced begins in Column B instead of Column A. This first issue can be corrected by replacing the `table_array' range $A$3:$M$252 with $B$3:$M$252.
The second issue becomes evident after the first correction when the Columns S through AC return incorrect data.  {See p. 6-7 of NCASI supplemental comment for VLOOKUP issues}
Column F Product Trade Name and Catalogue Number  -  This column has been changed from a prepopulated entry in the first draft to a dropdown menu option but the column is still grey.  Grey shading needs to be changed to no fill.
Column G Principal resin type  -  The instruction for this column in cell G8 are incorrect, the prepopulated entry will work when data is selected in Column F of this tab not the Resin Tab. Instructions should be corrected.
(B thru E) Change made. Prepopulated. 

(R thru AC) Change made. References fixed.

(F) Change made. Shading eliminated

(G) Change made. Instruction clarified.
F48
NCASI
EmTest tab:
The initial input for this tab should be revised. The Emission Unit ID should not be the first entry point since all test reports will be referencing outlet locations and associated results. The outlet location is the most logical point to begin data collection.  The first data to be entered, therefore, should be the Emission Release Point ID.  From there, EPA should be able to pre-populate current Columns B, E, H, I, and J.
Change made. The Emission Release Point ID column was moved to precede columns containing process unit details and pre-population was added. 
F49
NCASI
EmTest tab:
Column K Pollutant  -  The drop down menu options Total HAP, Total non-PCWP HAP, Total PM, PM10, PM2.5, and THC need to be revised for the following reasons:  
 Total HAP:  The individual concentration values for each of the six PCWP HAP cannot be added together, only mass related values can be added together. As a consequence, Columns M and N cannot be reported in this survey for total HAP. 
 Total non-PCWP HAP: Is EPA requesting that other organic HAPs be combined with HAP metals? It does not appear to be the case since the last menu selection is "Total HAP metals". EPA should revise the term "total non-PCWP HAP" to "total non-PCWP organic HAP". Note that the individual concentrations of non-PCWP organic HAP cannot be added together on a concentration basis, however. 
 Total PM: EPA should recognize that `Total PM' is an obsolete term. Since Column L provides a menu option for Method 5, EPA should use the correct terminology used in Method 5: particulate matter (PM). 
 PM10: This terminology relates to Primary PM10 or Direct PM10 that represents the combination of filterable PM10 and CPM.  Column L only allows for the entry of one method whereas reporting of this parameter requires both M201A/M202. 
 PM2.5: Likewise, this parameter represents the combined results from two methods: filterable PM2.5 and CPM. 
 THC: The term "THC" is an acronym for total hydrocarbons which is a class of organic compounds that technically only includes alkanes, alkenes, alkynes, dienes, and arenes. Method 25A is designed to determine total gaseous concentration of vapors consisting primarily of alkenes, alkenes, and/or arenes. This means that M25A will measure this subset of THCs but also will measure other organic compounds. EPA should clarify if only this subset of THCs are to be reported or if the total gaseous organic concentration (THCs and non-THC response) from M25A is to be reported. 
 Missing VOC?: The ICR Survey Instruction manual includes "VOC" on page 29, third paragraph.  But the dropdown menu does not include this option.  EPA should clarify or remove VOC from instruction manual.  
(K)(1) Change made. "Total HAP" was omitted from the menu, and replaced with the individual HAP compounds comprising Total HAP as defined in the PCWP rule. The columns M-P requesting respondents to enter numerical emissions values were deleted from the table to significantly reduce respondent burden. The purpose of the EmTest tab is for respondents to provide a detailed catalogue of emission test reports submitted by pollutant so EPA can quickly assess the breadth of test report data collected and prioritize test data extraction for compilation of an emissions data base. 

(K)(2) Change made. The terms "Total metals" and "Total non-PCWP HAP" were removed from the menu. Facilities are now requested to list the individual HAP compounds tested (e.g., toluene, xylenes). Tests involving HAP groups measured by the same test method such as Method 29 HAP metals or Method 429 POM/PAH may be listed as "HAP metals" or "POM" on one row instead of listing every specific compound. The EPA will speciate compounds when extracting the data from the test reports.

Methylene diphenyl diisocyanate (MDI) was added to the pollutant menu, and Conditional Test Method 031 was added to the test method menu for MDI (in addition to Method 320 which was already in the test method menu).  Furthermore, the "sealed caul plate" test method ASTM D7770 was added to the test method menu because this alternative method may have been used by engineered wood products producers to estimate press HAP emissions.

(K)(3) Change made. The menu was changed to refer to "PM" instead of "Total PM." The intent is to collect Method 5 filterable PM data. 

(K)(4) and (5) Change made. The combined Method 201A/202 was added to the menu to correspond to the combination of filterable PM10 (or PM2.5) and CPM. Method 201A and Method 202 were also retained in the menu for tests where the two methods were not combined.

(K)(6) Change made. The instructions were clarified to state that we are seeking the value of the Method 25A THC response, which is consistent with THC measurements conducted under the PCWP rule.

(K)(7) Change made. The reference to VOC was removed from the instructions.
F50
NCASI
EmTest tab:
Column L Test Method  -  
 Menu options provide M201A and M202, but the parameters measured by these methods, filterable PM and condensable particulate matter (CPM), are not offered as a selection in Column K.
 EPA has not provided Method 29 for metals.
 If VOC is to be included and based on a calculation technique, then is the calculation technique to be included here as a test method? Any VOC value from a calculation technique will not apply to the ppmvd only to mass emission rates.
(L)(1) Change made, as noted above.  

(L)(2) Change made. Method 29 was added to the menu.

(L) (3) Change made. References to VOC have been removed. 
F51
NCASI
EmTest tab:
Column M Concentration  -  Cannot add or average concentration values for total HAP, total non-PCWP HAP or HAP metals.  No units provided for opacity readings.
Column N Concentration Result Units  -  EPA uses "ppmdv" this should be changed to "ppmvd".
(M and N) Change made. After further consideration of the burden associated with completing the EmTest tab and the EPA's need for consistent test data entry, we elected to delete columns M through P. The EPA will tabulate the emissions test results using the test reports provided. 
F52
NCASI
EmTest tab:
Column O  -  Instructions do not address mixed set of detects or how the average total HAP is to be determined.  
For total HAP, do the individual run values get added first then the three sums get averaged and reported?  
No instructions provided for opacity
Column P  -  Mass-based production units for pneumatic dust collection systems and coating and wastewater operations will be difficult to apply or interpret. For example, finishing area process unit, such as saws and sanders, may be associated with the press production or surface area of the panels processed (with a further complexity of reporting only one or both surfaces). The problem here is that most product lines run finishing area process units independent of the press. Compiling reported results based on multiple production rate units will be challenging.
Column R Other emission units/pollutants  -  Note that the examples show PM as a pollutant but the instructions state not to include PM.
(O and P) Change made. After further consideration of the burden associated with completing the EmTest tab and the EPA's need for consistent test data entry, we elected to delete columns M through P. The EPA will tabulate the emissions test results using the test reports provided. 

(R) Change made. The examples were edited to remove reference to PM.
F53
NCASI
EmTest tab:
Column W Frequency of PCWP NESHAP Testing  -  The instructions in row 9 specify frequency only relative to PCWP NESHAP testing while the column header in row 10 refers to frequency of testing for all pollutants. EPA should resolve this discrepancy. Also, provide a "Other: (specify)" option in drop down menu.
Change made. The instructions were edited to clarify that we are interested in the testing frequency of the pollutant listed in each row of the spreadsheet, not limited to PCWP NESHAP testing. Should EPA find reason to expand the current PCWP testing requirements, it will be useful to know the testing frequency already required by permitting authorities (if applicable) as we consider incremental testing costs. The "other" option was added to the menu.

F54
NCASI
EmTest tab:
Column X Cost per test  -  Instructions on how to calculate the cost per test for the pollutant listed should be provided. For example, an itemized cost between M25A "THC" testing and NCASI Method A105 (BHA) reasonably attainable. But itemizing the cost between methanol and acrolein within the BHA Method is more difficult. Should responders divide the total cost of the BHA Method for stack testing and analytical by 6?
Change made. The instructions were clarified to indicate that for test methods that measure multiple compounds (e.g., NCASI A105.01), the total cost of the test is of interest, not the itemized lab cost for each pollutant measured by the method.  




G. Phase II: Emissions Testing
No.
Commenter
Comment Summary
EPA Response/Action Taken
G1
AWC

Boise Cascade
Concerned about the meaningfulness of any new data collected under the proposed test plan, especially at the high estimated cost of testing. It is not feasible for EPA to prescribe or enforce a numeric emission standard for many of the sources included in the proposed test plan, and many of these sources will qualify for a design, equipment, work practice or operational standard in accordance with section 112(h). In addition, data that are collected will be very limited data or of poor quality and representativeness and any limit developed will not reflect the actual level of emissions from those sources nor will it reflect the achievable level of performance across all sources and operating conditions.

The PCWP MACT remanded sources cannot be tested in a manner that assures the emissions are solely representative of the emission source or that assure emission limits established for those sources capture all the emissions attributable to that source. Even under circumstances where the tests could be reliable measures of the emissions from those units, it is not likely that the emissions tests could reliably take into account all the factors that affect the variability of emission rates at these units. It is not reasonable to attempt to test these units and establish emission limits based on the best performing units when emission performance is not based on the unit itself but rather the specific product being manufactured in that unit. It is also not reasonable to establish emission limits based on existing emissions information that does not take into consideration all these different factors. For these reasons we strongly urge EPA to forgo testing and setting emission limits for most of the remanded sources, and instead work with industry to develop reasonable work practice standards for these sources. 

While existing data may not be appropriate to establish emission limits, it may be useful emission data to support EPA's work on the residual risk at these facilities. If EPA needs additional emissions information to support the residual risk evaluation, then additional testing for that purpose may be reasonable. Such a testing program could be designed to test worst case emission rates that would be useful for a conservative risk assessment even if that data was not useful for establishing emission limits from the various remanded sources.

As discussed in Sections III and VI of this memo, Phase II of the ICR has been eliminated as a result of timing, burden, and testing feasibility considerations. The EPA now intends to rely on the emissions data gathered through Phase I to complete the RTR.


G2
CPA

Boise Cascade
Testing and Controlling Most Downstream Sources would be Inherently Unreliable.
Most of these units have emissions that are unrelated to the function of the source (e.g. sanders and trim saws), but related to the product emissions as the panels move through the production process after pressing. These sources do not have emissions generated by combustion or heating of the materials.
There is an inherent variability of emissions dependent on the type, resin content, wood species and thickness of the product.
Some of these downstream units also work off-line, i.e. they may process panels after production. Emissions could be different than when processing newly pressed panels.
Given the variability of those products and the fact that these sources typically do not generate emissions but rather experience emissions as the product flows through them, it is difficult, if not impossible to collect meaningful emissions data for the sources or establish defensible numeric emission limits for them.
Requiring estimates or devising extensive testing regimens to attempt to capture emissions data from these sources would be inherently unreliable and unnecessary for the residual risk evaluation or for establishing emission limits for the miscellaneous sources.
It is very difficult to precisely capture and measure emissions from most emission sources downstream from dryers, such as plywood presses, trim saws, and sanders. These emission units are typically fugitive emission sources, or sources that are not isolated from other sources, and they often have very low HAPs concentrations that will require extended test runs. Some of these obstacles can be at least partially overcome through construction of enclosures, construction of testing stacks, and long test runs, such requirements are often complicated for various reasons and usually rather costly.
      Without assurances that sources can be carefully controlled to represent emissions from the unit tested, it is not reasonable to establish an emission limit for the specific units based on the test data because the data may include more or less emissions than actually emitted by the source. While such data may be useful to estimate emissions for modeling, it is likely not of the quality necessary to set an emission limit for which other facilities must then test to demonstrate compliance when those facilities may also have similar or different test issues. These issues make it impracticable to measure emissions from many of the PCWP MACT remanded sources due to technical and/or economic limitations. As such, EPA should conclude that it is not feasible to prescribe or enforce emission limitations for most of the PCWP MACT remanded sources and conclude these sources are subject to work practices.
As noted previously, Phase II emissions testing has been eliminated.
G3
AWC
After the review of the Phase I data is complete, OMB and the public should have an opportunity to review and comment on EPA's decision on whether a Phase II information collection is appropriate and if so, the scope of the Phase II test plan. The public cannot fully comment on the scope of the test plan without knowing the particular facilities and sources impacted and whether there is practical utility in further data collection.
As noted previously, Phase II emissions testing has been eliminated.
G4
NCASI
Section 1.0:
Existing emissions testing data
Historical performance tests on dryers and presses would include three 1-hour runs
Highly unlikely that there would be test data on emission points associated with remanded processes that would have seven test runs as stipulated in the ICR Test Plan

Therefore, suggests that if a facility has conducted testing that is applicable to the ICR and testing was comprised of a minimum of three 1-hour runs then the facility not be required to perform additional testing for those pollutants.
As noted previously, Phase II emissions testing has been eliminated.
G5
AWC

EPA should not require testing of equipment eligible for a work practice or alternate standard approach, as not all of the listed source types are isolatable and testable and would produce measurable HAP emissions: The following sources should be removed from the test plan:
Chippers: there is nothing to suggest that there would be measurable HAP emissions from chipping wood (especially green wood).
Miscellaneous sources that currently are not subject to HAP emissions standards under the PCWP MACT
As noted previously, Phase II emissions testing has been eliminated.
G6
NCASI
Section 1.1: Emission sources:
PCWP MACT I dryers:
A technically justifiable alternative to new testing would be to use existing data for boilers in the Boiler MACT database (stokers/sloped grate units designed to burn kiln-dried biomass or suspension burners designed to burn biomass, etc.) to carry out the residual risk analysis
If new data collection is necessary, the following should be considered for softwood veneer dryers
Units have both heating and cooling zones
PCWP MACT I has emission limits for heated zones and work practice standards for cooling zones
PM (HAP Metals) data collection should be limited to the heating zones instead of all atmospheric vents
PCWP MACT II Hardwood Dryers
If testing is required only the heated zones of direct-fired (non-natural gas) dryers should be tested for HAP metals and PM.
PCWP MACT II Hardboard ovens
Section 1.1.2 indicates a requirement to collect production rate data for hardboard ovens, but hardboard ovens are not mentioned in the list of required emission sources to be tested.
As noted previously, Phase II emissions testing has been eliminated.
G7
NCASI
Emission Point Definition/Characteristics
Dry blending, forming, and finishing sources are served by pneumatic dust collection systems that are not designed to collect organic HAP emissions at 100% and are instead open systems that utilize building air to transport fugitive, sander, and saw dust.
Use of the term "emission point" underscores the fact that the emissions from these miscellaneous baghouses are expected to be non-specific and variable.
It is infeasible for facilities to certify that collocated units were operating in a normal and representative manner during emission tests as normal operations may not include all collocated units in operation.
Inability to fully enclose and isolate process units will have a significant impact on the emission data collected.
As noted previously, Phase II emissions testing has been eliminated.
G8
NCASI
Normal and Representative Operating Conditions
Please clarify and define "normal and representative" for process units associated with the "dry forming and blending emission points" and the "finishing sander, saw, and chipper emission points"
Blending  -  Composition of resinated furnish processed through face and core blending units can vary depending on the type of panel manufactured but the amount of dry wood furnish processed will likely be consistent due to the surge bins ability to receive the resinated wood furnish and can accommodate fluctuations. Blending operations should be considered as separate entity from the formers and press production areas and have stand-alone production rates. EPA should stipulate what blender throughput is considered "normal and representative" so blending operations from similar product types can be compiled and evaluated.
Forming  -  Amount of resinated furnish laid down by the former will depend on the type of product being manufactured. EPA needs to designate a normal and representative mat thickness and/or line speed for formers of the same product type. Formers and presses can be operated at different rates and should be considered separate entities and could have different throughputs increasing the complexity if they share the same emission point.
As noted previously, Phase II emissions testing has been eliminated.
G9
NCASI
Use of term "performance test":
Consider an alternate terminology that is more representative of this data collection effort such as "PCWP ICR testing." "Performance testing" could be seen as a compliance test where process unit specific emissions are evaluated against a standard.
As noted previously, Phase II emissions testing has been eliminated.
G10
NCASI
Certification that APCD was in operation during testing:
Clarify that this question refers to the "air pollution control device" that was in operation is relevant to the parameter being tested. For instance, a baghouse is an APCD for PM but not for organic emissions.
As noted previously, Phase II emissions testing has been eliminated.
G11
AWC


Enclosing/Testing Sources
   NCASI has conducted a detailed survey of various source types that documents the difficulties associated with enclosing sources for HAP emissions testing. 
   It will be challenging, if not impossible, to completely separate emissions from different types of sources that are located in close proximity to each other within a building. Building air brought in to these tests will contaminate the sample with fugitive emissions from surrounding sources.
   Many locations will not meet Method 1 criteria, especially older baghouse configurations and sources that exhaust through building vents. Facilities should not be forced to spend money to make modifications to exhaust vents to meet Method 1 criteria.
   Concerns with the required methods' ability to measure HAPs at low levels, and suspect that much of the data will be below detection limits for several sources. EPA should consider sources with high amounts of non-detects as candidates for work practices, not emission limits.
   Requirement for a temporary total enclosure would be needed not only for presses and board coolers but to all of the units for which EPA proposes to gather organic HAP data, if EPA means to gather emissions data that is associated to a particular process unit. Units that are not currently controlled under PCWP MACT are not enclosed for capture or control of organic HAP emissions. In order to set standards based on the test data, the tests must be carefully controlled to ensure that the emissions measured at the test location represent the total emissions from the specific unit and exclude any emissions not associated with the specific unit.
   Some temporary enclosures may be feasible for a few days, however permanent enclosure around the equipment may be impossible to design and install without severely hindering operations and compromising worker safety where a facility would need frequent access to the equipment for maintenance, operational adjustments, or product manipulation.
As noted previously, Phase II emissions testing has been eliminated.
G12
NCASI
Testing of alternative unit.
   The determination of what units are "more amenable to testing" has the potential to be highly subjective (and open to interpretation) especially in the case of collocated miscellaneous remanded sources. It is unclear how a facility will make this determination.
     Does this provision apply only to process unit emissions (versus those characterized as emission points)?
     Does this mean that the company must look to another facility that has more amenable individual process unit-specific emission points, for example, a former baghouse that is separate from a saw line baghouse?
     What if the more amenable emission point is serving a process unit that is actually in a worse location relative to a fugitive HAP source?
     How would the test results from the more amenable process unit emission point be related to a non-amenable emission point at another facility?
As noted previously, Phase II emissions testing has been eliminated.
G13
NCASI
Required Stack Test Methods - THC
 Provide instructions on how to adjust Method 25A response for the presence of non-THCs such as oxygenated organic compounds. Note that THC does include the contribution from methane.
 For Method 25A EPA should use terminology that will accurately describe the output they desire. Method 25A determines a "total gaseous organic concentration" (TGOC).
 Method 25A does not address the conversion of TGOC to organic mass emission rate [concentration (ppmvd) to mass rate (lb/hr)] as the ICR requires. EPA should include a standard procedure for converting concentration to a mass rate for all ICR test data
As noted previously, Phase II emissions testing has been eliminated.
G14
NCASI
"Filterable particulate matter (PM)": 
The use of the term "filterable particulate matter" in this ICR test plan needs further clarification since the term "filterable" is typically only used to describe the collection of particulate matter by EPA Method 201A. We believe EPA's intent was to not include the condensable fraction. We recommend the use of the term "particulate matter" for PM collected using Method 29 and Method 5. 
As noted previously, Phase II emissions testing has been eliminated.
G15
NCASI
Required Stack Test Methods  -  Methane for direct gas-fired dryers
  Methane can be a significant component of Method 25A response being measured at certain APCDs and other emission points that are comprised of building air. Methane measurements are therefore important to back-out the contribution of methane (non-methane TGOC) for all sources where THC measurements are required (not just direct gas-fired dryers).
  
  Mandated Concurrent Testing: If concurrent testing is mandated, methane needs to be added to the dry blending/forming and finishing emission points within Table 1.1. Board coolers should also be included in the list of sources for methane co-sampling. Not including methane in the exercise of correlating "THC" to HAP emissions will be a significant issue for most of the targeted PWCP MACT II emission sources because methane is not directly emitted by the associated process units but is being contributed by nearby fugitive methane sources. The contribution of methane, therefore, is expected to be highly variable and very much dependent on the physical location of the PCWP MACT II process unit. As the physical location of most finishing process units, for example, varies from mill to mill, the amount of methane contribution to "THC" is expected to vary. As a consequence, the correlation of "THC" to total HAP or individual HAP will not be applicable industrywide. Therefore, concurrent testing for most PCWP MACT II process units or emission points should be thoroughly reviewed and perhaps exemptions granted where deemed inappropriate. While the technical requirement of concurrent testing for "THC" and total HAP is probably not a huge issue for most testing companies, the concern with the concurrent testing issue is more to do with the cost and effort put in to collect data that will be utilized inappropriately. Furthermore, requiring seven (7) tests where "THC" and total HAP have to be tested simultaneously at emission sources with multiple emission points is going to be a technical challenge and very expensive. 
As noted previously, Phase II emissions testing has been eliminated.
G16
Earthjustice
Stack Test Methods
EPA is proposing the use of NCASI Method ISS/FP-A105.01 for quantifying the regulated organic HAP. In the NCASI method, non-traditional blanks are specified. EPA should specify the use of more traditional field blank which includes the preparation and recovery of a sample train which has been used for sampling stack gas and specify the use of duplicate sample trains which use two complete sampling trains rather than a single probe and filter box with two impinger trains.
Table 1-2 uses Method 29 to quantify emissions of metals from processes which essentially dry the wood and bind wood pieces or veneers together. It is not expected that metals would be released from fuels used to heat the presses or dryers, wood which is not combusted, or glues used as binders. However, the tests will document either the low emissions or that emissions are below the capabilities of Method 29. If EPA is attempting to determine low level pollutants, several of the processes involve heating wood to levels at which incomplete combustion may be present. As a result, polycyclic organic ("POM") compounds may be formed. Thus, we recommend that EPA use CARB 429 to quantify POM compound emissions.
Also in Table 1-2, EPA has specified sampling for filterable particulate using Method 29 or, alternatively, Method 5. Rather than quantifying only the filterable component of particulate, EPA should require use of Method 202 combined with Method 5 to quantify total particulate matter included in the sampling.
As noted previously, Phase II emissions testing has been eliminated.
G17
NCASI
Number of Test Runs for Concurrent Testing:
Requiring seven (7) tests where "THC" and total HAP have to be tested simultaneously at emission sources with multiple emission points is going to be a technical challenge and very expensive.
As noted previously, Phase II emissions testing has been eliminated.
G18
NCASI
Enclosure for Plywood Presses:
The inherent design of plywood mills is to have one layup line that feeds multiple presses within a single large building. If a temporary total enclosure (TTE) is to be constructed around one of the collocated presses, the close proximity to other press(es) would make the complete isolation of the enclosed press unachievable. While TTE might meet the associated criteria, the emissions collected from the enclosure would include emissions from the rest of the building including the layup line and other presses. This manufacturing footprint configuration is a unique characteristic of the plywood product category which distinguishes the process of manufacturing plywood from manufacturing particleboard, OSB, and MDF.
As noted previously, Phase II emissions testing has been eliminated.
G19
NCASI
Table 1.2 Summary of Test Methods: Speciated Organic HAP 
The use of the term "at least" for establishing the Sample System Minimum Measurement Level (SSMML) might be confusing to some testers.
If the "maximum" allowable SSMMLs are at the 0.2 (aldehyde) / 0.5 (alcohol) ppmvd levels provided in Test Plan, then running the BHA method according to the parameters used by NCASI for field validation (and outlined in the referenced example that is provided in the Method's Appendix A4) will not meet the ICR's test method SSMML criteria because the SSMMLs would be higher that the "at least" values stipulated in Table 1.2 of the ICR test plan, except for phenol.
To achieve the SSMML requirements in the ICR Test Plan, one or more sampling parameters need to be modified. Since the test plan is requiring 7 sample runs, the sample time of 60 minutes is likely not the parameter to modify. Increasing the sample flow rate to > 1.00 L/min will lower the methanol SSML to 0.45 ppmvd and formaldehyde SSMML to 0.2 ppmvd. Lowering the formaldehyde and methanol SSMMLs to the ICR Test Plan criteria, however, significantly challenges the quality assurance (QA) measures for the remaining compounds. Ultra-low concentrations detected at or just above the SSMML concentrations can cause the bracketed spike range between the low and high ESLs to be too narrow. Very low spiking levels are also difficult to formulate to get exact spike solutions from neat compounds and testers will have to make sure that they run the high ESL of the Bracketed Spike below 2 ppmvd for any compound detected below 0.5 ppmvd to guarantee that the proper spike recovery criteria are met.
As noted previously, Phase II emissions testing has been eliminated.
G20
NCASI
Table 1.2 - The sample line should be no longer than 2 meters
Further clarify this statement. Normally when total HAPs are tested concurrently with M25A, testers will use a heated 100 ft sample line to bring a slipstream of stack gas to a manifold that feeds both the M25A analyzer and the HAP trains. The sample gas is the conveyed from the heated manifold to the first impinger of each HAP train by a section of unheated Teflon sample line. EPA should clarify that the 2 meter sample line restriction applies only to the unheated section of sample line connected to the first impinger. Otherwise, testers will be restricted to conducting BHA testing near the test ports, which may not be feasible for some Phase II elevated emission points. Modify text to read: "The unheated section of sample line attached to the first impinger of the HAP sampling train should be no longer than 2 meters."
As noted previously, Phase II emissions testing has been eliminated.
G21
NCASI
Table 1.2 - Testing gas streams with entrained water droplets using a VOST probe 
EPA has included modifications to the NCASI BHA method for testing gas streams with entrained droplets. The BHA Method does address use of the method at saturated emission sources: "This method should not be used on sources where significant amounts of entrained water droplets are present unless it is used in an isokinetic manner." This modification in the ICR Test Plan, however does not require isokinetic sampling but does recommend that a VOST prove be used. A VOST probe is a 3-foot glass probe that is configured with a short section that has a 5/8-inch section is considered to be a "cup" where water droplets are designed to fall out and the collected moisture would presumably drain out the probe tip. Since there is a potential for aldehydes and alcohols to concentrate in water droplets within the source duct or stack, failure to capture a representative amount of water droplets can bias the results.
As noted previously, Phase II emissions testing has been eliminated.
G22
NCASI
Table 1.2 - BHA Method QA measures
All QA/QC...criteria...must be...met: The following topics relate to the BHA method's QA measures that should be taken into consideration
    The BHA Method is primarily designed to be conducted in three (3) sample runs. The field QA that is conducted within those three sample runs is one spike run or one bracketed spike (consisting of two spike runs) and one duplicate run. Since EPA is requiring seven (7) sample runs, EPA needs to state that this ICR Test Plan only requires the reporting of one field QA dataset for all 7 sample runs. That is one duplicate and one single run spike or one bracketed spike.
As noted previously, Phase II emissions testing has been eliminated.
G23
NCASI
   The BHA method has been written in such a manner as to allow the end user the flexibility to interpret and evaluate method performance. Note that there is no statement within the BHA method that links the failure of an individual QA criterion to the invalidation of the three source gas concentration determinations. There is language within the method requiring that all QA procedures are to be performed and must be within certain criteria. However, the method does not declare what to do about failed QA criteria other than to clearly describe the circumstance for transparency purposes. This designed ambiguity gives the end user the ability to evaluate the results from the numerous QA criteria which cover a wide range of measurement capabilities. Due to the unique complexities of individual organic compound measurement, the QA steps in the BHA method provide the end user with an array of performance metrics which can be used to access the validity of the source gas concentration measured by the method's normal train.
As noted previously, Phase II emissions testing has been eliminated.
G24
NCASI
   The NCASI BHA method is a self-validating method which is different than performance based methods, such as EPA Method 5. EPA Method 25A is also a self-validating method but M25A is able to utilize direct measurement system response to reference gases that meet EPA Protocol 1 criteria. The BHA Method, on the other hand, relies on neat compounds for calibration of laboratory analytical equipment and spiking for assessing method execution. Since those neat compounds are transitioned through multiple matrices and environments during method execution, the complexity of handling and recovering individual volatile organic compounds has to be recognized and appreciated. While the focus will be to meet all of the QA criteria within the BHA method, the immense scope of this ICR test effort will inevitably result in cases where one of the numerous BHA Method QA criterion might not be met. Since each of the six HAP compounds will have an individual set of QA data, there will be more than sufficient QA information to diagnose any criteria failures and distinguish whether there was a systematic method failure or an anomaly with an individual compound measurement.
As noted previously, Phase II emissions testing has been eliminated.
G25
NCASI
   EPA should also clarify whether anomalies in sample analysis can be investigated and resolved by reanalysis of samples. NCASI recommends that sample reanalysis be allowed to evaluate anomalies. If the anomaly is rectified based on the result of the reanalysis, then the rerun result is to be reported as the correct value but the original value is also to be clearly reported with a summary of the issue.
As noted previously, Phase II emissions testing has been eliminated.
G26
NCASI
   Acetaldehyde, acrolein, and propionaldehyde: Emissions of acetaldehyde, acrolein, and propionaldehyde are for the most part generated when heat is applied to wood chips during drying activities and to wood surfaces during pressing activities. Since these three compounds are not a significant component in urea-formaldehyde or phenol-formaldehyde resins, process units that handle resinated dry wood furnish or conduct finishing activities on cured panels are not expected to generate emissions of these three compounds. Based on the limited data available, atmospheric refiners, blenders, formers, saws, and sanders appear to have non-detect levels of these three compounds. The historical data, however, are based on higher detection levels than being required by this ICR test plan. To ease the burden of this testing effort, EPA should waive the run spike QA requirements for these three compounds when testing is conducted for this group of emission points. Even if one or more of these compounds were to be detected, the levels would be so low that any associated QA-spike measures would not be of any consequence. Since the run spike would still be conducted for formaldehyde, methanol, and phenol, the second QA-spike train would provide valuable duplicate data for acetaldehyde, acrolein, and propionaldehyde since these compounds would not be in the spike solution. The tester would obviously still be required to conduct the train spike for all six compounds.
As noted previously, Phase II emissions testing has been eliminated.
G27
NCASI
Table 1.2 - Total Hydrocarbons (THC): 
Change the term "THC" to better reflect what EPA is actually trying to measure. Also to avoid confusion, EPA should specify if the "THC" values includes methane or not. The stipulation that measured concentrations must be bracketed by the low and high level calibration gases is unrealistic for most low concentration "THC" sources such as finishing emission points. Most Method 25A analyzers are operated within a specified concentration range, for example 0-10 ppmv or 0-100 ppmv. Method 25A requires that the analyzer be calibrated with a zero gas and a high level gas that is 80% - 90% of the concentration range setting. The low level gas is required to be 25% to 35% of the concentration range setting. Most low concentration PCWP sources have "THC" concentrations that fall within the range of 5 to 20 ppmvd "THC" as propane and can be significantly impacted by the presence of methane. Methane is not generated by the majority of PCWP MACT 2 process units/emission points and methane can cause the "THC" levels to fluctuate between the 1st and 2nd analyzer range settings. Under this strict ICR "THC" criteria, any source within a concentration between 9 and 25 ppmvd "THC" as propane would not be testable since the source gas would be above the high level gas for 0-10 ppmv range setting and below the low gas for the 0-100 range setting. Since EPA is requiring concurrent testing of "THC" and total HAP for 7 sample runs, any source "THC" concentration drift between the 1st and 2nd analyzer range setting would invalidate the test and require the tester to recalibrate the M25A measurement system and replace the HAP train(s) for a substitute sample run. This calibration gas bracket stipulation has the potential to be time consuming, very expensive, and adds unnecessary burden to the test plan. Method 25A analyzers do provide predominantly linear responses. Therefore, it should be sufficient to allow the tester to follow the method's written QA and select the concentration range that includes the maximum expected source gas "THC" concentration. If EPA remains concerned that Method 25A analyzers will not return a reasonably linear response to measured concentrations below the low calibration gas, then the "THC" criteria could allow testers to include a 4th calibration gas that will challenge the analyzer at some lower level.
As noted previously, Phase II emissions testing has been eliminated.
G28
NCASI
Table 1.2 - Measurement of methane via Method 18:
While Method 18 may be an appropriate method for measuring methane, it is an inappropriate method for this ICR application if the "THC" value is to be adjusted for methane and/or will be used as part of the concurrent test evaluation being conducted to evaluate the correlation between "THC" and total HAP. While M18 does use a gas chromatograph/flame ionization detector (GC/FID), the issue with M18 is that it is a separate analysis method with no direct link to the M25A "THC" measurement. The M25A "THC" response will have a methane component to it and an associated methane FID response factor specifically for the field flame ionization analyzer (FIA) that has been calibrated with propane. As a consequence, basing any methane result on Method 18 will not accurately reflect the actual methane response or contribution to the "THC" measurement collected by the M25A FIA. If EPA is going to allow adjustment of the M25A concentration for Method 18 methane measurements, then clear instructions are required to show how that adjustment is to be calculated, including the impact of the FID response factor for methane. An alternative to Method 18 that would avoid this issue would be to allow the use of a M25A FIA that is equipped with two FIDs, one dedicated to measuring "THC" and the other dedicated to measuring methane. The "THC" FID would be calibrated with propane and the output would be the "THC" response on a propane basis. The methane FID would be fed through a duplicate instrument path that includes a methane cutter to allow only methane through. This FID would be calibrated with methane and the output would only reflect the direct response to methane. Since both FIDs are operating simultaneously within this FIA, this arrangement would allow for (1) the direct measurement of the "THC," (2) the direct measurement of methane, (3) the direct determination of the methane response factor on the "THC" FID, and (4) the accurate determination of non-methane "THC" concentration.
As noted previously, Phase II emissions testing has been eliminated.
G29
NCASI
Table 1.2 - Metals and PM (filterable):
EPA is requiring HAP metals and PM testing at the PCWP MACT I dryer emission sources via Method 29, with Method 5 as an option for PM. Since these PCWP MACT I dryers already have organic HAP APCDs installed, EPA needs to clarify that the testing needs to be at the APCD outlet.
EPA is requiring seven (7) sample runs to be conducted, with each run having to collect at least 105 dscf, or 3 dscm. This would imply each sampling run will have to be 2.5 to 3 hours in duration; making the testing effort labor intensive and very expensive.
As noted previously, Phase II emissions testing has been eliminated.
G30
NCASI
Table 1.2  -  Mercury: 
While Method 29 does collect total mercury, this method requires that five (5) different mercury sample fractions be collected to account for total mercury, with each sample fraction having a different detection level. Particulate bound mercury associated with the firing of sander dust is expected to be very low at the outlet of organic HAP APCDs because these APCDs normally have some method for minimizing particulate loading to the organic HAP APCD. As a consequence, particulate emissions from controlled dryers are expected to be low and EPA should allow the use of Method 30B for mercury instead of Method 29, where applicable. M30B is an approved test method in Boiler MACT. EPA should also consider conducting an initial HAP metal risk assessment based on fuel analysis to estimate maximum HAP metals emissions and then apply PM removal efficiencies to estimate applicable controlled emissions. These two estimation techniques may provide sufficient evidence for this purpose.
As noted previously, Phase II emissions testing has been eliminated.
G31
NCASI
Table 1.2 - MDI:
Target reporting units are missing in Table 1.2 under the column heading Reported Units of Measure
As noted previously, Phase II emissions testing has been eliminated.
G32
NCASI
Table 1.2  -  O2/CO2
Measurement of O2/CO2 for every run at ambient air supplied emission sources is a burden. EPA should stipulate that O2/CO2 data collection can be limited to one test result showing ambient O2/CO2 levels for ambient air supplied sources.
As noted previously, Phase II emissions testing has been eliminated.
G33
NCASI
Table 1.2  -  Moisture
Moisture determinations by wet/dry bulb should be allowed for ambient air supplied emission sources that have low to moderate levels of moisture.
As noted previously, Phase II emissions testing has been eliminated.
G34
NCASI
"You should collect emissions samples for the identified pollutants downstream of the last relevant APCD (i.e., stack or other point representing the composition of the flue gases at the exit to the atmosphere), unless otherwise indicated": EPA needs to be more specific with regard to emission sample collection points for emissions sources with non-relevant APCDs..
As noted previously, Phase II emissions testing has been eliminated.
G35
NCASI
PWCP Process Data Collection (pg 7, 2nd paragraph):
   EPA has not provided production rate criteria for "Dry blending and forming operation emission points," "Finishing sander, saw, and chipper emission points," and fiber washers. This is a significant issue because establishing a standard production rate basis for these types of process units is a major challenge. The production unit selected should be applicable to all of the process units/emission points associated with each category.
   For temporary total enclosed non-isolated process units collocated with other similar process units, such as for multiple plywood presses located within the same building and only one press is being tested, the production rate for the enclosed press should be based on the production rate from all presses.
   For collocated dissimilar process units under the same emission point or collocated process units of the same type under the same emission point, the total production rate should be based on all relevant process units.
     Hardboard ovens are not included on the list of emission sources to be tested but are required to report production rates?
     Clarify what is meant by "Emission unit operating temperature"; does EPA want the operating temperature of the process unit itself or the temperature of the gas stream associated with the process units at the emission point to the atmosphere or test location?
     EPA should stipulate that wood moisture content determinations can be reported by typical/standard mill collection methods.
     EPA needs to clarify the intent of collecting the operating parameters for APCDs, i.e., is EPA soliciting 1) the "operating parameters relevant for the APCD" or 2) "operating parameters relevant for the APCD relevant to the test parameter/pollutant"?
baghouse parameters relevant for organic HAP and "THC" testing
RTO parameters relevant for metal HAP and PM testing
     It is our recommendation that EPA provide exemptions for instances where APCDs are not applicable for the pollutant being tested
As noted previously, Phase II emissions testing has been eliminated.
G36
Boise Cascade
PCWP MACT sources are highly variable. Among other factors:
   The emissions generated at the presses are due to heating of wood and resin. Because the wood has been previously heated during drying, most, but not all, of the emissions from the wood have been driven off. However, as the wood is reheated during pressing, additional HAPs are emitted from the wood.
   Most of the HAPs emissions generated at the press come from the resin applied to the veneer. 
   In addition, the resin application rate, the thickness of the panel, and the number of veneers in the panel greatly affect the HAPs emission rate.
   Press temperature and press time can affect HAPs emissions during pressing. 
   The variability of emission rates due to the range of conditions at the press would likely carry over to all downstream emission units such as the trim saws and sanders as the hot panels pass through those emission units.  Cool panels passing through those same units would likely emit less HAPs as HAPs emissions tend to decrease as the panels cool.

While we have concerns with, and recommend EPA not pursue emissions testing of most of the remanded sources, we think the test plan omits some important data that must be collected to assure that tests data can be understood in terms of the variability of the processes. For example, EPA should request that all the product and operating parameters described above should be required to be included in the test reports.
As noted previously, Phase II emissions testing has been eliminated.
G37
NCASI
Appendix A
The guidance on detection levels does not include handling of NCASI BHA minimum measurement level (MML) or sample system minimum measurement level (SSMML). Since the ERT does not support the BHA method, does EPA plan to incorporate these flags in the "PCWP Testing Supplement" Flag column?
Since M29 has five sample fractions for Hg and each sample fraction has its own detection level, EPA should provide guidance on how to combine non-detect and detect fractions into a total Hg result. NCASI recommends the use of Method 30B for Hg testing given the potential for high bias
As noted previously, Phase II emissions testing has been eliminated.
G38
SLMA
EPA has requested that facilities provide emissions data in response to the Phase I ICR survey. Capturing and quantifying emissions from lumber kilns is difficult, particularly for continuous kilns which remain open during operation. We urge EPA to focus on the development of work practice standards for lumber kilns and to revise the ICR survey to achieve this objective.
As noted previously, Phase II emissions testing has been eliminated.
G39
Masonite
We believe that EPA will learn from the ICR Phase I Questionnaire that many of 19 sources included in the test plan have not been tested, and cannot be tested due to several reasons inherent to the wood products industry including:
the lack of isolated emission points,
::	the fugitive nature of the emissions
::	the inability to capture and direct the emissions through an emission point,
the very low concentrations of HAPs,
the high cost and of enclosing the sources and modifying the emission points to make testing possible.
The wood products industry has been successful in identifying the primary sources of HAP emissions in our industry. This can be substantiated through scientific understanding of the processes, including parameters that impact HAP emission such as temperature, moisture content, wood species, and process additives. We urge EPA to use this information to develop work practice standards rather than to employ costly and potentially non-representative testing for these remaining low-emitting sources where testing is not technically feasible.
As noted previously, Phase II emissions testing has been eliminated.
G40
Weyerhaeuser
Eliminate testing for remanded presses
EPA should carefully consider the safety, technical and cost issues with attempting to enclose and test emissions from presses that were not regulated with process unit specific emission limits under the current PCWP NESHAP for technical feasibility reasons.
EPA toured a Weyerhaeuser engineered wood products facility in 2008, observing and documenting the configuration of very large multiple, continuous, laminated veneer lumber (LVL) presses. Discussions with Weyerhaeuser staff at the facility pointed to the difficulties that would thwart enclosing such presses to isolate and capture the low level, largely fugitive emissions in order to quantitatively test and/or convey the emissions to control devices. Industry and university investigators developed the heated sealed caul plate test method, considered the most accurate method for estimating emissions from difficult-to-test presses. Emission factors developed from data based on the sealed caul method are appropriate for EPA's RTR evaluations. The history of the development of the sealed caul method is evidence of the appropriateness of an EPA decision to not require attempts at full-scale press enclosure and emissions testing.
As noted previously, Phase II emissions testing has been eliminated.
G41
Earthjustice
Suggests that EPA modify the ERT to document the NCASI ISS/FP-A105.01 Method.
As noted previously, Phase II emissions testing has been eliminated.
G42
Earthjustice
EPA should require additional testing for the two categories for which it is proposing testing in Phase II; requiring only a small number of tests from only 4-5 units, as EPA has proposed, is unlikely to produce sufficient data to inform EPA's rulemaking duties. EPA estimates the number of facilities to include:
235 sawmills with lumber dry kilns,
92 plywood and veneer facilities,
36 OSB facilities,
22 MDF facilities,
21 particleboard facilities,
4 hardboard facilities,
4 fiberboard facilities, and
11 engineered wood product facilities.
On page 28, EPA is proposing a total of 80 emissions tests to characterize 20 different processes, with only two to five tests of each individual process (as shown in Table 3 of the Supporting Statement). If the facilities to be tested will be randomly selected, we believe that this is an insufficient number of tests to characterize the average emissions of the best performing 12% of the facilities and the lowest emitting source. In addition, if EPA intends to assess if there are significant differences in the type of wood being processed or the process employed or if the wood types or processes can be combined for regulatory purposes, the number of samples would likely only identify order of magnitude differences. We suggest that EPA re-evaluate both the number of tests proposed for individual process types and the selection of the facilities for testing such that the tests represent the better performing sources.
As noted previously, Phase II emissions testing has been eliminated.
G43
Earthjustice
EPA should ensure that Phase II can cover all additional HAP data needed going beyond just the proposed categories. It is extremely likely that, in response to Phase I, EPA will receive little or no emission test data, or that some tests will be 13 years old. EPA should include in Phase II additional tests for all known HAPs emitted that are currently covered by the NESHAP, so that it gets an accurate picture of what the existing standards are controlling and the health risks under those standards.
As noted previously, Phase II emissions testing has been eliminated.
G44
Earthjustice
EPA should request testing for additional metal HAPs and others that may potentially be emitted even if not known to be emitted. As EPA knows from other rulemakings, such as the Wool Fiberglass rule, tests may show previously-unknown HAPs (like hexavalent chromium, in the case of the Wool Fiberglass rule) are being emitted, requiring EPA regulation. EPA should ensure this ICR gives the agency an accurate picture of the emissions inventory from this sector so that community members are not subject to hidden health threats without EPA doing the investigation needed to discover those threats that may well be coming from unregulated HAPs emitted. The situation near a plywood facility in Crossett, Arkansas, is an example of why broader emission testing is needed in this ICR to ensure appropriate pollution limits are in place for this and other similar facilities.

In particular, in order to assess the actual and potential air emissions from wastewater systems for enhanced biodegradation units and wastewater sources, EPA should require both testing of the effluent for specific parameters and air emission testing for all known and potential HAPs, including but not limited to for methanol, phenol, benzene, xylene, and others. This would provide information that EPA and stakeholders could evaluate to fulfill all of the stated objectives of this ICR.
As noted previously, Phase II emissions testing has been eliminated.
G45
Georgia Pacific
The Crossett plywood facility was idled in October 2011 and has since been permanently closed. It has not been the focus of concerns about health impacts or air pollution. 
As noted previously, Phase II emissions testing has been eliminated.


G46
Boise Cascade
If requiring additional testing it is very important for EPA to collect information on the cost and complexity of installing temporary enclosures and temporary test stacks on the various sources that must be tested. Photos of the enclosures would be helpful. EPA should also require documentation that the enclosures either meet the definition of temporary total enclosures or provide measures of capture efficiency for each temporary enclosure.
As noted previously, Phase II emissions testing has been eliminated.
H. Other/General
No.
Commenter
Comment Summary
EPA Response/Action Taken
H1
AWC
CPA
Many of the identified miscellaneous sources are not susceptible to control and the application of measurement methodology is not practicable due to technological and economic limitations. Work practice standards would be the better statutory standard setting approach for many of these sources at our members' facilities. An appropriate work practice standard is readily available to limit HAP emissions  -  mill production of panels with emissions consistent with those required under Phase II of the California Air Resources Board's Air Toxic Control Measure (ATCM) for Formaldehyde Emissions from Composite Wood Products. California Air Resources Board (CARB) for work practice standards Model work practice standards after CARB II production management system to document and verify the minimization of HAP emissions. Resin-based work practice standards would be appropriate for many of the remanded sources.
We acknowledge that there are technological and economic limitations associated with testing many of the miscellaneous sources. The Phase I ICR results will be used to more fully understand equipment configurations, emission levels, and potential work practices for reducing HAP emissions such as the CARB ATCM or related resin-based work practices. Phase I of the ICR will provide information on resins used in PCWP mills today following implementation on the CARB ATCM and the TSCA formaldehyde rule.
H2
CPA, WWPA, SLMA, Masonite
Supports AWC comments.
The commenters' support for AWC's comments is acknowledged.
H3
APA, Hardwood Federation, Weyerhaeuser
Supports AWC and NCASI comments.
The commenters' support for the comments from AWC and NCASI is acknowledged.
H4
Boise Cascade
Supports AWC, NCASI, CPA, and APA comments.
The commenter's support for the comments from AWC, NCASI, CPA, and APA is acknowledged.
H5
Earthjustice
Public Information: The instructions for this ICR recognize that EPA has reviewed certain information (e.g., emission tests, compliance reports, permits, and industry-association technical bulletins from the National Council for Air and Stream Improvement). See Survey Documents and Spreadsheets Appendix 10. EPA should put into the docket for this ICR all information that it reviewed to create this ICR; commenters were hampered from evaluating the ICR and being able to provide an informed comment due to a lack of supporting material from EPA. In addition, EPA should ensure that all data received from this ICR, and for which a CBI claim has not been proven and substantiated by EPA, are made publicly available as soon as possible on EPA's website or in the docket for this ICR on regulations.gov.
The types of information the commenter mentions (emission tests, compliance reports, permits, and data from NCASI technical bulletins) were docketed as part of research to support prior PCWP rulemaking actions. General industry knowledge of the PCWP industry gained from prior rulemaking research was used to construct the ICR. The purpose of the ICR will be to gather information (e.g., test reports, compliance reports, emissions estimates, and permit information) which will be tabulated and included in the docket for the proposed PCWP RTR.  Technical memoranda and reports documenting our analyses of the collected information will be available in the docket and on the web site for the proposed RTR. 
 

