2170ss09

SUPPORTING STATEMENT

Air Emissions Reporting Requirements (AERR)

EPA ICR # 2170.09 OMB Control No. 2060-0580

Short Characterization/Abstract

The Environmental Protection Agency (EPA) most recently promulgated revisions to the Air Emissions Reporting Requirements (AERR) on February 19, 2015 (FR Vol 80, No. 33, 8787). Since then, the Agency has twice renewed its Office of Management and Budget (OMB) clearance -- under OMB control number 2060-0580 -- to enforce reporting and recordkeeping requirements associated with the AERR regulations. The current term of approval for this collection was approved via EPA’s information collection request (ICR) # 2170.08 on January 31, 2022.

This supporting statement and ICR are being submitted in conjunction with proposed revisions to the AERR which, if finalized, would be effective for three years starting with the 2023 inventory year. Air emissions-related information would be due to be reported within 15 months after the inventory year, depending on the data category. Thus, this ICR covers the years 2024, 2025, and 2026. The EPA has provided the detailed calculations used to estimate collection costs in the attachment “AERR Revision Burden Estimates v9 For Docket.xlsx.”

Under the current AERR ICR, 54 state and territorial air quality agencies, including the District of Columbia (DC), and an estimated annual average of 26¹ local and tribal air quality agencies, must annually submit emissions data of oxides of nitrogen (NO_x), carbon monoxide (CO), sulfur dioxide (SO₂), volatile organic compounds (VOC), particulate matter less than or equal to 10 micrometers in diameter (PM₁₀), particulate matter less than or equal to 2.5 micrometers in diameter (PM_2.5), and ammonia (NH₃). The current AERR, which can be found at 40 CFR Part 51 Subpart A, defines which emissions sources that state, local, and tribal (SLT) agencies must submit individually as “point sources.” The point source definitions are different depending on the year, and SLTs report more point sources every third year for the “triennial inventory.” For triennial inventory years, states² must also submit emissions for stationary nonpoint and some nonroad mobile sources, and they must submit model input data for onroad mobile and nonroad mobile equipment. For mobile sources, California has different requirements because it uses different models, and California must submit emissions (rather than model inputs) for these data categories of the same pollutants listed above.

The proposed revisions would continue the existing collection from states of these criteria air pollutants and precursors (CAPs) described above. In addition, the proposed revisions would clarify expectations for reporting data for airports, rail yards, commercial marine vessels, and locomotives; change requirements for nonpoint sources when EPA has published emissions methods; add a requirement for completing a nonpoint survey; change nonpoint source deadlines; change reporting requirements for nonpoint data when an Indian tribe reports; and make a variety of clarifications and administrative changes.

Some of the proposed AERR revisions would not take effect until 2027 but would require certain planning and preparation to occur during the period of this ICR. The primary proposed revision would define a new approach for mandatory collection of hazardous air pollutants (HAP) from facilities and provide air pollution control agencies the option to implement requirements consistent with the proposed AERR and report emissions on behalf of facilities. This proposed revision would require mandatory emissions reporting by owners/operators outside the states’ implementation planning authority (e.g., within certain Indian country) in 2026. Further, EPA would expect SLTs to prepare during the period of this ICR for HAP reporting by those SLTs that would choose to report to fulfill the mandatory requirement for their owners/operators in 2027. Also during this period, owners/operators would need to gather locations (i.e., latitudes and longitudes) for each release point, in anticipation of reporting these starting in 2027.

Similarly, the proposed AERR revisions would also add requirements for reporting activity data for prescribed fires starting in 2027 (for the 2026 inventory year). To be prepared for this requirement, during 2024-2026, SLTs would need to develop a collection approach for prescribed fires activity data. Other revisions proposed to take effect in 2027 would require all point sources to be reported every year; phase in earlier deadlines for point source reporting; and add requirements for reporting fuel use data for certain sources of electrical generation associated with peak electricity demand.

For owners/operators of facilities that meet criteria described by this proposal, the proposed revisions would require emissions reporting of HAP directly to EPA starting in 2027 for most sources, except when a state is approved to report on their behalf; would require owners/operators of facilities that are located outside of states’ implementation planning authority to report emissions to EPA starting in 2026; and would require reporting of performance test and performance evaluation data to EPA electronically for certain tests conducted after the promulgation date of this proposed action that are otherwise not reported electronically to EPA.

The proposed AERR revisions would also clarify existing connections between the AERR and reporting requirements for State Implementation Plans (SIPs) for meeting the ozone and PM_2.5 national ambient air quality standards (NAAQS). While these connections between regulations and statutory requirements already exist, this proposed ICR includes (in Appendix C:) new estimates of burden associated with preparing emissions inventories for SIPs. This represents an effort to move toward a clearer identification of the inventory preparation burden as a discrete part of the SIP preparation burden, as that component had not been expressly identified in prior ICRs. It is provided as an appendix in this as part of the AERR process as a matter of convenience, to ensure public review of these additional SIP-related costs not previously quantified as part of the ICRs developed for the SIP requirements regulations.

The annual emissions data collected through the AERR are, used by the EPA Office of Air Quality Planning and Standards (OAQPS) to support development of the National Emissions Inventory (NEI). The EPA uses the NEI in developing ambient air quality emission standards, performing regional and national modeling, providing air quality management support (e.g., state implementation plan development) to state agencies and multi-jurisdictional organizations (MJOs), and preparing national trends assessments and other special analyses and reports. Currently, the same reporting mechanisms used for the CAPs listed above are also used for voluntary reporting of HAP and greenhouse gas (GHG) emissions. These data are collected by the air agencies for their own purposes, and EPA encourages providing such data to EPA when it is available. The proposed collection of performance test and performance evaluation data (hereafter referred to as “source test data”) would be used by the Agency to develop emission factors, which in turn support estimates of emissions from stationary sources. The source test data would also be used to facilitate annual emissions reporting for those owners/operators who have reported it.

1. Explain the circumstances that make the collection of information necessary. Identify any legal or administrative requirements that necessitate the collection. Attach a copy of the appropriate section of each statute and regulation mandating or authorizing the collection of information.

In past years, the current AERR has served to coordinate the various state emission inventory reporting requirements and has streamlined the activities involved in submitting emissions data to EPA. The proposed AERR revisions would (1) continue this coordination to enable EPA to achieve uniformity and completeness in a national inventory to support national, regional, and local air quality planning and attainment of NAAQS and planning needed for meeting regional haze requirements, (2) greatly improve HAP data collections that are voluntary under the current AERR, (3) fill other gaps in emissions inventories for sources outside of states’ implementation planning authority (e.g., sources located within certain Indian country) and for prescribed fires nationally, and (4) greatly improve the availability of data necessary for creating emission factors. Each of these data needs are addressed separately with respect to the legal and administrative requirements that necessitate the collection.

NAAQS and Regional Haze Analysis Support. As with past AERR collections, CAP emissions collections support states and EPA efforts to address national, regional, and local air quality issues with cost-effective practices. For NAAQS implementation, an ongoing need exists for states to develop consistent inventories and to share their emissions inventory data across the air agencies. For example, for air quality modeling done for NAAQS implementation, a state needs emissions data from other states surrounding it. The proposed collection includes several additional elements that will further improve NAAQS support, including daily fuel use or heat input from some small generating units, specific requirements for special cases of emissions collection for airports, locomotives, and commercial marine vessels, and requirements for providing activity data and documentation for nonpoint (i.e., county-wide total), onroad, and nonroad sources.

The proposed AERR revision includes a proposed collection of daily fuel use or heat input data from small generating units that operate, for example, to meet or offset peak electricity demand. As described in the preamble to the proposed AERR revision, these units have been implicated in certain NAAQS nonattainment situations, and the proposed collection requirement would help fill a longstanding gap in EPA’s and states’ ability to understand and potentially resolve any impact of such sources on nonattainment.

The CAA provides EPA ample authority for acquiring such data. Emissions data are of vital importance to EPA for fulfilling a host of monitoring, standard-setting, rulemaking, reviewing, and reporting duties. For example, section 110 requires each state to prepare a plan that provides for implementation, maintenance, and enforcement of the primary standard for each pollutant for which air quality criteria have been issued. This plan must include provisions for periodic reports identifying sources and listing amounts of emissions. Further, section 301(a) authorizes the Administrator to promulgate regulations necessary to carry out the CAA. In addition, section 114(a)(1) of the CAA authorizes the Administrator to, among other things, require certain persons (explained below) on a one-time, periodic, or continuous basis to keep records, make reports, undertake monitoring, sample emissions, or provide such other information as the Administrator may reasonably require

More specifically, statutory support for collecting and reporting emissions data is demonstrated in three sections of the CAA. Section 110(a)(2)(F) requires that each state provide for periodic reports on the nature and amounts of emissions of criteria pollutants from stationary sources. Sections 182(a)(3)(A) and 187(a)(5) of the CAA specify periodic inventory requirements for ozone and CO nonattainment areas, respectively. Section 182(a)(3)(A) requires states with ozone nonattainment areas to submit a current inventory of actual emissions of VOC, and NO_x every 3 years. Section 187(a)(5) requires a similar inventory of actual CO emissions for CO nonattainment areas. Periodic inventories include emission estimates for all point, nonpoint, onroad mobile, nonroad mobile sources, and fires (also called “events” in the NEI). Section 172(c)(3) provides the Administrator with discretionary authority to require other emissions data as deemed necessary for SIP development in nonattainment areas to meet the NAAQS. Section 169A also directs EPA to provide regulations necessary for the protection of visibility, and the regulations that EPA has promulgated include some requirements for states that rely on emissions inventories.

NAAQS State Implementation Plans. To assist with fulfilling the CAA-mandated need for comprehensive data on emissions to support NAAQS implementation, EPA has promulgated several regulations that set requirements for states. These regulations are the Ozone SIP Requirements Rules (40 CFR part 51, Subparts X, AA, and CC) and the PM_2.5 SIP Requirements Rule (40 CFR part 51, Subpart Z). These rules refer to the AERR for some of the emissions reporting requirements. While the ICRs for those rules include preparation and submission of SIP documents, they did not expressly identify or separately account for the additional burden associated with collecting and reporting additional emissions information to meet those requirements. This additional burden is now expressly included as a separate burden estimate for public review in Appendix C: to this ICR for all NAAQS pollutants for which EPA anticipates SIP inventory preparation will occur over the three years covered by the ICR.

HAP Emissions. Under the current AERR, HAP emissions reporting from states is voluntary. The EPA has evaluated the effectiveness of this voluntary collection as described in the preamble to the proposed rule. HAP emissions data are used extensively throughout EPA’s regulatory and informational programs to protect public health and inform communities of potential risks from these pollutants. As discussed in the preamble to the proposed AERR revision, EPA has significant evidence that the current voluntary reporting program from states is insufficient to meet these needs. In addition to the statutory support described above for collecting and reporting emissions data, the CAA includes various requirements for EPA to implement. These include Residual Risk analysis under CAA 112(f)(2), technology reviews under CAA 112(d)(6), and review of source category listings under CAA 112(c)(5).

Further, Executive Order (EO) 12898 (59 FR 7629, February 16, 1994) directs federal agencies, to the greatest extent practicable and permitted by law, to make environmental justice part of their mission by identifying and addressing, as appropriate, the disproportionately high and adverse human health or environmental effects of their programs, policies, and activities on minority populations (people of color) and low-income populations. Part of the impact on communities resulting from EPA’s regulatory actions is to improve air quality by reducing emissions of HAP and other pollutants with local impacts. Under the current voluntary program, some states submit extensive HAP data, while other states submit little or no HAP data. The proposed AERR revision would help to close the gap in understanding impacts of HAP and other pollutants on communities and will, therefore, assist EPA with fulfilling the goals of EO 12898.

In addition, to be able to assess risks, EPA develops information about pollutant toxicity and reviews pollutant impacts under the Integrated Risk Information System (IRIS) program. Given the huge number of chemicals released into the air, it is necessary to prioritize which pollutants are investigated by the IRIS program. The EPA uses modeled risks and exposures to communities to help set those priorities, which requires detailed HAP data and release parameters that are not sufficiently available under the current voluntary program.

To support IRIS and other programs, EPA has developed nationwide risk information for all pollutants with the AirToxScreen program³ and its predecessor, the National Air Toxics Assessment (NATA) program. One use of these risk analyses is for prioritization of compliance and enforcement resources. EPA Regional offices and states use risk data to determine communities and facilities for review. For facilities emitting HAP, the current voluntary HAP data collection approach does provide some benefit; however, a more comprehensive HAP emissions collection program would further enhance the prioritization by supporting more complete and more detailed risk and emissions data than are currently available.

Additionally, HAP emissions reporting would further support NAAQS implementation, especially for modeled attainment demonstration requirements included in CAA 182(b) and 189. The CAA specifically identifies VOC as a precursor to ozone, and VOC is additionally a precursor to PM_2.5. Thus, emissions and anticipated reductions of VOC are inputs used for certain air quality modeling. Having detailed information about VOC HAP compounds reduces the need for EPA to make assumptions about the composition of VOC, which is the approach otherwise taken to support air quality modeling. As with VOCs, PM HAP provides more detail about PM_2.5 components that similarly allow for better data to support air quality modeling.

Direct collection of HAP emissions data from owners/operators is allowed by CAA 114(a)(1), which authorizes the Administrator to, among other things, require certain persons on a one-time, periodic, or continuous basis to keep records, make reports, undertake monitoring, sample emissions, or provide such other information as the Administrator may reasonably require. Given the broad scope of CAA section 114, it is well within EPA’s authority to gather the emissions data required by this proposed action. Further, the collection of these data is authorized and appropriate because such information is relevant to EPA’s ability to carry out a wide variety of CAA provisions, as illustrated by the previous description of the uses of such emissions data by EPA.

Other Gaps in Emission Inventories. The preamble to the proposed AERR revision explains EPA’s analysis of gaps that exist in EPA’s ability to create emissions inventories to support the NAAQS and HAP-related needs described above. First, the current AERR does not include collection of stationary source emissions from facilities operating outside states’ implementation planning authority (e.g., on certain Indian country). For these sources, EPA relies largely on voluntary reporting by a few tribes, which leaves gaps for the other tribes with potential emissions sources of interest for air planning purposes. As described in the preamble to the proposed AERR revision, EPA now has the technical capacity to collect emissions data directly from owners/operators via the Combined Air Emissions Reporting System (CAERS). The longstanding absence of data from some sources outside states’ implementation planning authority and the new ability to easily collect that directly from owners/operators has led to these proposed additions to this collection.

Second, EPA’s current voluntary program for collecting activity data about prescribed fires has not been successful for estimating emissions of many prescribed fires in many states, particularly on private, state, and military lands. As described in the preamble to the proposed AERR revisions, EPA’s experience over the past decade has determined that while the satellite data used by EPA is very informative, without more ground-based data, it is not possible to accurately differentiate prescribed burning from other types of fires in most states. The satellite approaches are also limited in their ability to estimate the size of the fires and have no way to characterize pile burns properly. Filling these gaps would support analyses of increasing importance within the context of recent increases in the frequency of damaging forest fire events. With several of the worst wildfire seasons occurring within the past 5 years, EPA anticipates that more prescribed burning may be needed for improved future fire management schemes. While these prescribed burns are controlled and limit emissions as compared to wildfires, they still produce significant emissions of CAPs such as PM_2.5 and VOC, HAP, and carbon dioxide, all of which are important contributors to environmental health risks and climate change.

Emission Factors. Emissions factors are a key tool used in the creation of emissions inventories. An emissions factor can be used to estimate air pollutant emissions from a normally operating point-source process or activity (e.g., fuel combustion, chemical production). In 2006, a review of EPA’s emissions factors program by the National Academy of Sciences and EPA’s Office of Inspector General resulted in a report⁴ that urged EPA to improve both emission factor quality and quantity. The EPA’s most recent approach to develop emissions factors has been prepared in response to that report, and it relies on source tests of emissions releases from stationary-source processes and activities (i.e., a sample of the process emissions is collected and analyzed).

With this proposed AERR revision, EPA is seeking to improve emissions factors to support improved emissions inventories via the proposed collection of additional source test data. The EPA has recently completed updates to the WebFIRE system that automates most of the emission factor development processes and makes that data available to the public. As a result of these efforts, EPA issued its first set of revised emissions factors for public review in November 2021.⁵ Now that the development procedure infrastructure is largely completed, EPA seeks to increase the amount of source test data by obtaining information from the thousands of emissions processes and hundreds of pollutants included for stationary sources in the NEI, as a logical progression in emission factor improvement. As is the case for collection of annual emissions data described above, direct collection of source test data from owners/operators is allowed by CAA 114(a)(1).

2. Indicate how, by whom, and for what purpose the information is to be used. Except for a new collection, indicate the actual use the agency has made of the information received from the current collection.

The data included in this collection fall into two categories: annual emissions data and related information and source test data. Each of these categories has different uses and so these are described separately below.

Annual emissions data and related information. Annual emissions data about stationary point and nonpoint sources, as well as onroad mobile and nonroad mobile sources, are routinely used by OAQPS, EPA Regional offices, and states in carrying out a variety of activities needed for implementing provisions of the CAA. The data are compiled into the NEI, which in turn supports development of an emissions modeling platform. Both the NEI and the emissions modeling platform support regulatory functions as well as functions that are more programmatic in nature, such as trends analyses. Such functions include:

• Basic information to support the boundaries selected for designation of nonattainment areas;

• Evaluation of control strategies included in SIPs;

• Evaluation of control strategies for states and larger areas, including applications of regional and national scale models;

• Preparation and publication of national summaries of emissions including trend analyses and international reporting obligations;

• As a database to assist in the identification of important source categories for possible future regulation;

• Development of national control strategies and preparation of Regulatory Impact Analyses through application of air quality modeling;

• As a key part of air quality modeling platforms, which EPA uses for research and planning and provides publicly for use for state and MJO air quality modeling efforts;

• As a building block to assist states with compliance with their requirements for preparing SIPs;

• Evaluation of locations for ambient monitoring when regulations require monitoring near sources with emissions above certain levels (e.g., for lead);

• Risk assessments and technology reviews; and

• Preparation of the stationary source portion of a report to Congress on SO₂ emissions. This report is required by Section 406 of the CAA and is due on a 5‑year cycle that began on January 1, 1995. The report must contain an inventory of national annual SO₂ emissions from industrial sources (as defined in Title IV of the CAA).

In addition to supporting the efforts listed above, emissions data are used to respond to numerous requests for reports on emission sources. Typically, the data are provided freely through EPA’s website. In some cases, specific requests for data not available on EPA’s website are also made by email and rarely, under the Freedom of Information Act. Requests come from the general public, teachers, contractors and consultants; Congress; the press; domestic and international universities; and others involved in research of many types.

Source Test Data. As previously described, source test data are used by EPA to calculate emission factors. The EPA additionally provides source test data to states and owners/operators via WebFIRE to allow the data to be considered for use in emissions inventory calculations.

3. Describe whether, and to what extent, the collection of information involves the use of automated, electronic, mechanical, or other technological collection techniques or other forms of information technology, e.g., permitting electronic submission of responses, and the basis for the decision for adopting this means of collection. Also describe any consideration of using information technology to reduce burden.

The EPA has established a central repository of emissions inventory data for all states called the Emissions Inventory System (EIS). Emissions inventory data reported electronically are stored in the EIS, and the database is used by EPA and by states for obtaining emissions data reports and input files for air quality modeling.

The EPA maintains the EIS as a central repository of inventory data for all states, and much of the data must be supplied by the state and local agencies in electronic form. The EPA requires that submitting agencies use a defined eXtensible Markup Language (XML) schema for electronic data reporting using EPA’s Central Data Exchange (CDX). Instructions for reporting data to the EIS via CDX are provided on EPA’s website via the link “How Do I Submit to EIS Using the Web Client?”⁶ To assist SLTs in preparing their submission for EIS, EPA has developed and provides a Bridge Tool, which allows for compiling data in Microsoft® Access® to convert to the XML format. In addition, the EIS provides on-screen data editing via a web browser in a module called the EIS Gateway. These tools have been in use for over a decade and have greatly streamlined the emissions collection and distribution processes for states.

To collect source test data, EPA has established and maintains the Compliance and Emissions Data Reporting Interface (CEDRI). In addition to access and review of submitted data by EPA, CEDRI provides states a mechanism to access the data collected electronically. Because this proposed AERR revision would require most Federal- and state-required source tests to be reported to CEDRI electronically, the CEDRI tool will allow for EPA and all states to comprehensively have electronic source test data available for their air quality management needs. CEDRI also supports states’ review of source tests prior to their use by EPA. Because CEDRI automatically passes data to WebFIRE, the emission rates from the source tests are also available to the public.

The EPA has also established and maintains CAERS to support collection of emissions data from owners/operators and sharing of that data with states and the Toxics Release Inventory (TRI). Ongoing CAERS development is expected to include sharing data, including characterization of the facilities, with the Greenhouse Gas Reporting Program (GHGRP) and CEDRI.

CAERS supports burden reduction in several ways. CAERS allows owners/operators to enter data in a single location, and then CAERS distributes any data elements that are also needed by other emissions collection systems so that owners/operators would not need to enter the same information in two places. For states and owners/operators, CAERS would allow entering emissions data using online forms and/or Microsoft Excel,® and CAERS converts that data to the XML format needed for the EIS. As described in the preamble to the proposed AERR revision, CAERS would also connect to source test data in CEDRI and emission factors in WebFIRE to support owners/operators using the best available data to calculate emissions (using the best available data is a proposed requirement of the AERR revisions). CAERS will provide owners/operators a mechanism to identify the location within their facility for which a source test is performed, thus associating those measurements to the part of the facility (e.g., the emissions unit and release point) and supporting a reduction in burden for subsequent emissions inventory calculations. Additionally, CAERS can reduce burden for a state that chooses to use CAERS to collect point source emissions data and report it to EPA rather than develop and maintain their own emissions reporting data system.

Additionally, to support burden reduction for nonpoint sources, EPA develops emissions calculation tools and provides nonpoint activity and emissions data that states can review, comment on, and/or accept to meet their nonpoint reporting requirements. Under the proposed AERR revision, states would be able to meet their reporting obligations by (a) reviewing and providing comments on EPA’s estimates for EPA to incorporate, (b) revising the estimates by reporting activity data and optionally reporting emissions and associated documentation, or (c) accepting the estimates after review. For onroad mobile sources and nonroad equipment, EPA develops and runs the latest mobile source emissions model using state-supplied model inputs and provides the resulting emissions for onroad and nonroad vehicles for review and comment by the states. For aircraft emissions, EPA compiles landing and takeoff data, provides to states for review and comment, and then computes emissions using the latest Federal Aviation Administration emissions model. For commercial marine vessels, EPA compiles activity data using the Automatic Identification System (AIS) database and computes emissions for review and comment by states. Finally, for locomotive emissions, EPA works with industry representatives (on a voluntary basis) to compile activity data and draft emissions for review and comment by states.

For the voluntary reporting of wildfires and proposed mandatory reporting of prescribed burning activity data, EPA compiles available satellite and ground-based fire observation data from numerous sources around the country, and provides those data for voluntary state review, edit, and comment. The EPA also estimates the wildfire and prescribed burning emissions based on these revised data for inclusion in the NEI. In addition to the data mentioned above, EPA loads into EIS for use in the NEI several national datasets, including TRI, the GHGRP data, and data on emissions from drilling wells from the Bureau of Ocean Energy Management.

Pictures and instructions of the various data entry tools are available as follows:

• The CDX web client instructions and user interface pictures are available on the “EIS User’s Manual and ‘How to’ website”.⁷ The links for “How Do I Submit to EIS Using the Web Client?” and “How do I Log on To the Network Exchange Services Center?” provide the necessary information.

• The EIS Gateway instructions and user interface pictures are available in the EIS Users’ Guide.⁸ This includes a section with instructions and pictures for the EIS Bridge Tool.

• The CAERS instructions and user interface pictures are available in the CAERS Version 3.0 Users’ Guide.⁹

• The CEDRI instructions and user interface pictures are provided via the CEDRI Job Aids website.¹⁰

4. Describe efforts to identify duplication. Show specifically why any similar information already available cannot be used or modified for use for the purposes described in Item 2 above.

Previous reporting requirements have occasionally caused state agencies to have inefficient collecting and reporting activities. The AERR was promulgated specifically to simplify previously existing emission inventory reporting by states to EPA, offer options for data collection and exchange, and unify reporting dates for various categories of inventories to avoid duplication of effort. For example, under the NO_x SIP Call rulemaking, EPA required states to submit annual inventories for all NO_x sources for which states had adopted control measures to meet their NO_x budget. In addition, statewide NO_x inventories of all controlled and uncontrolled sources were required every 3 years. The Consolidated Emissions Reporting Rule (CERR), which was the precursor to the AERR, also required annual and triennial emission inventory reporting of many of the same data elements. The publication of the original AERR aligned the reporting dates and combined data from these two previous collection activities to avoid duplication of information collected from sources, minimize the burden on industry, and reduce the effort for state and local government agencies to compile the data. Additionally, other regulations including the Ozone SIP Requirements Rules (40 CFR part 51, Subparts X, AA, and CC) and the PM_2.5 SIP Requirements Rule (40 CFR part 51, Subpart Z) refer to the AERR for some of the emissions reporting requirements specified in those rules.

More recently, EPA has worked in partnership with states as part of the E-Enterprise for the Environment¹¹ program to develop the CAERS as described above. States can voluntarily participate in CAERS using a variety of approaches that wholly or partly offset the need for states to develop and maintain their own electronic point source emissions data collection system. CAERS also allows states to avoid purchasing such a system or from using less efficient paper forms. Any electronic data collection system has a user “front-end” for data entry and a “back-end” database to store and manage the data, and CAERS provides states a way to replace the front-end and/or back-end of the states’ systems. As a result, when states participate in CAERS, they can wholly or partly offset the cost of maintaining their own electronic data system.

Additionally, EPA recognizes potential duplication across certain federal emission data collection programs. States collect data from facilities, including HAP emissions. The EPA collects data from states for the NEI (including voluntarily submitted HAP emissions), and EPA also collects HAP emissions data from owners/operators as part of the collections required for TRI (40 CFR part 372). For this reason, CAERS makes any HAP data reported by owner/operators to states using CAERS available to the TRI-MEweb software. TRI-MEweb allows owner/operators to import the data they submitted to CAERS (for the NEI) to avoid having to re-enter it into TRI-MEweb, which reduces burden. With this approach, owners/operators can use CAERS to report any state-required HAP emissions (that also flows to the NEI), then CAERS maps and aggregates that data to the resolution needed for the TRI. This approach promotes burden reduction for owner/operators and data consistency across agencies. While this feature is not appropriate to be used in all cases because of different facility definitions, there are many cases in which the state definition of facility and the TRI definition align and, therefore, promote burden reduction for owner/operators.

The EPA’s proposed AERR revision includes provisions that avoid duplication for three newly proposed collections. First, for the proposed collection of daily fuel use or heat input from small generating units, the proposed AERR revision would exclude any small generating units for which such data are already reported to EPA. Second, for any owners/operators that would report to EPA directly under this proposal, those owners/operators would only report emissions to EPA when those emissions are not reported to EPA by a state, local, or tribal air agency with implementation planning authority (see preamble section III). In addition, this proposal avoids duplication with EPA Region 10 tribal emissions reporting requirements (40 CFR 49.138)¹² in Idaho, Oregon, and Washington by exempting owners/operators from having to report to EPA under this proposed approach. The EPA is further considering how reporting data via CAERS could be used to meet reporting requirements under 40 CFR 49.138. Third, the proposed AERR revision would help ensure owners/operators do not need to report their HAP emissions data to both states and EPA separately by allowing states to be approved to report on behalf of the owners/operators. The EPA has further avoided duplication by designing CAERS to be able to support a single flow of HAP data from owners/operators to NEI, TRI, and states.

5. If the collection of information impacts small businesses or other small entities, describe any methods used to minimize burden.

State agencies are not considered to be small entities. According to EPA’s ICR Handbook, OMB’s definition for a small entity includes small governmental jurisdictions with populations of less than 50,000. According to 2010 population data from the U.S. Census Bureau, no state or territory has a population below this threshold. In addition, none of the local air agencies that report under the AERR for their jurisdictions represent an area with population of less than 50,000.

EPA convened a Small Business Advocacy Review (SBAR) Panel for the proposed revisions and received input from several small entity representatives (SERs). The EPA used much of that input to improve the AERR proposal to reduce impact on small businesses. For the panel, EPA estimated that about 43,000 small businesses (based on the SBA definition) could be affected by these requirements, thus steps that EPA can take will have a significant impact for many facilities. This number of small businesses included several sources of reductions during the panel process. Most notably, based on feedback from SERs prior to the final panel, EPA eliminated about 15,000 small businesses from reporting for the automobile collision repair industry. The small business accommodations incorporated into this proposal would apply to an estimated 34,800 small businesses (i.e., firms not establishments), based on the CAA definition of small businesses as described in the preamble to the proposed rule.

Prior to the SBAR Panel, EPA’s draft approach already included providing the CAERS system to owners/operators to calculate and submit their emissions. In addition, CAERS stores each facility’s details from year to year, so that owners/operators do not have to re-enter the same information each time. Further, owners/operators can copy and edit their report from previous reporting years. Finally, CAERS includes built-in quality assurance that helps to avoid multiple re-submissions of data because of quality concerns. All these existing CAERS features would help reduce reporting burden.

In addition, EPA’s approach described to the SBAR Panel and small entities included limiting the pollutants that non-major sources would need to report. Since most small businesses would also be non-major sources, this approach would reduce burden. However, small entities provided many concerns that further informed the Panel recommendations and this proposal.

Based on the final SBAR panel report, which is included in the docket for the proposal (see EPA docket number EPA-HQ-OAR-2004-0489-0090, EPA has included the panel recommendations in this proposal, as follows (see proposal preamble for further explanation of these included recommendations):

1. The EPA has excluded small businesses in the Automotive Body, Paint, and Interior Repair and Maintenance industry (NAICS¹³ 811121).

2. The EPA has considered both the SBA and CAA definitions of small businesses and while proposing the CAA definition, has requested comment on using the SBA definition, which would further reduce burden by allowing about 10,000 more small businesses (firms) to use AERR provisions to reduce burden.

3. The preamble of this proposed rule includes a description of an emissions estimation tool that EPA would develop and make available to owners/operators of small businesses to estimate and submit emissions. The preamble also describes a process, recommended by the SBAR Panel, for developing and updating the emissions estimation tool in collaboration with small businesses and states.

4. The preamble of this proposed rule explains that small businesses would not be expected to report emissions for pollutants when EPA does not provide a way to estimate emissions and there is no other readily available data for those emissions.

5. The preamble of this proposed rule describes an option that EPA is considering adding to the AERR to collect cross-program facility identifiers that would facilitate information sharing across EPA programs and help avoid duplicative data collection.

6. The proposed regulatory text reduces the scope of reporting for certain small businesses by providing an option to limit the facility inventory data to unit level and emissions data to facility emissions totals. EPA would treat such emissions as fugitive, ground-level emissions because no stack information would be provided. Rather than report emissions for each unit, process, and release point, small businesses could provide emissions as a total facility value, as separate totals for stack and fugitive, or some other less detailed approach. A small business would only be required to report emissions data with full unit/process/release point detail in a subsequent year of reporting if EPA’s risk modeling from a prior year estimates cancer risk of 20/million or higher.

7. The proposed regulatory text includes a process that seeks to avoid duplicative reporting to both states and EPA by providing options to states to align emissions reporting practices with this proposal. The EPA describes processes for a SLT to adopt AERR requirements and CAERS in lieu of its own requirements and systems and an alternative process SLTs can use to ensure their requirements align with the AERR when they report HAP.

6. Describe the consequence to Federal program or policy activities if the collection is not conducted or is conducted less frequently, as well as any technical or legal obstacles to reducing burden.

The data included in this proposed collection have two categories: annual emissions data and related information and source test data. Each of these categories has different uses, so these are described separately below.

Annual emissions data and related information. If annual emissions data and related information were not collected, EPA would not have data about CAPs from stationary sources of these emissions. These data form the foundation of implementation of many CAA requirements on EPA and states to define NAAQS and reduce emissions levels to meet those standards. Similarly, these data support EPA and state implementation of requirements to improve visibility in Class I areas associated with the regional haze program. The EPA uses this collection to create the NEI, which is a foundational source of data used for various key aspects of the NAAQS program including ambient monitoring locations, air quality modeling for NAAQS and regional haze, analyses to support transport SIPs, and other critical needs to implement requirements of the Act. If EPA did not collect these data, it would be extremely difficult, if not impossible, for the Agency to carry out many of its statutory obligations under the Clean Air Act related to the NAAQS and visibility.

Additionally, the proposed collection of annual emissions data for HAP has been designed to allow EPA to meet the requirements of multiple provisions in the CAA, for example, including the provisions in CAA 112(f)(2) for Residual Risk analysis and to promulgate numerous regulatory actions; in CAA 112(d)(6) for completing technology reviews every 8 years; and in CAA 112(c)(5) to review the list of Section 112 source categories and list new source categories and subcategories according to statutory criteria. Further, this collection supports EPA in complying with EO 12898. This EO directs federal agencies, to the greatest extent practicable and permitted by law, to make environmental justice part of their mission by identifying and addressing, as appropriate, the disproportionately high and adverse human health or environmental effects of their programs, policies, and activities on minority populations (people of color) and low-income populations. Many other additional needs for these data are described in the preamble to the proposed rule.

Other additions included in the proposed AERR revision fill data gaps that EPA has identified as limiting both EPA’s and states’ ability to understand emissions sources in specific ways. As described in the preamble to the proposed AERR revision, these additional proposed revisions would allow the Agency and states to better fulfill their responsibilities under the Act to improve air quality for the protection of human health and the environment.

This proposed action includes several elements that support burden reduction for all sources, including those owned by small businesses. The EPA considered burden reduction opportunities associated with (1) the form and the frequency of reporting, (2) the timing of reporting, (3) the facilities that would be required to report emissions data, and (4) the data elements to be reported. Each of these aspects has been evaluated by EPA for burden reduction opportunities. The EPA addressed these aspects in its formulation of the proposal and each is supported by the available descriptions in the preamble to the proposed rule.

In addition to the burden reductions included for small businesses described in Section 5, aspects of this rule that reduce burden for states and owners/operators that would report to EPA under the proposed action are listed here.

1. CAERS supports burden reduction for owners/operators across four EPA air emissions programs and state emissions collections. CAERS includes numerous collection features that reduce burden for both owners/operators and states, including avoiding owners/operators needing to the same data to EPA through multiple reports.

2. This action proposes to formalize EPA’s role to provide emissions calculation methods for states for nonpoint sources, which requires states to report activity data rather than calculate emissions. States can comply with this requirement simply by reviewing and accepting EPA-provided data, or by commenting on that data. Doing so is lower burden than the current AERR requirement for states to generate such data themselves.

3. This action would maintain the three-year reporting requirement for nonpoint and mobile sources and would keep the reporting of wildfire data voluntary. This action proposes to additionally make emissions reporting for agricultural fires voluntary rather than mandatory.

4. This action specifies EPA’s expected minimum reporting requirements for mobile sources model inputs, which reduces the data that states may believe that EPA expects to be reported under the current AERR that lacks such specificity.

Source test data. Source test data are necessary to calculate emission factors, which industry, states and EPA use for calculating emissions needed to support the various activities based on the Act as described above. An emissions factor can be used to estimate air pollutant emissions from a normally operating, point-source process or activity (e.g., fuel combustion, chemical production). An emissions factor relates the quantity of pollutants released to the atmosphere from a process to a specific activity associated with generating those emissions. Without emissions factors, the burden for owners/operators and states to report emissions inventories could be higher and/or the quality of the emissions data would be lower. This proposed action strikes a reasonable balance for burden by requiring electronic reporting of source test data that is already being created based on existing Federal and state requirements to perform those tests. This action does not propose to add source testing requirements other than the use of an electronic form for reporting the results and provision of test conditions needed to make the information useful for emission factors. Additionally, the proposed action furthers EPA’s implementation of corrective action in response to a review of EPA’s emissions factors program by the National Academy of Sciences and EPA’s Office of Inspector General.¹⁴

7. Explain any special circumstances that would cause an information collection to be conducted in a manner:

• requiring respondents to report information to the agency more often than quarterly;

• requiring respondents to prepare a written response to a collection of information in fewer than 30 days after receipt of it;

• requiring respondents to submit more than an original and two copies of any document;

• requiring respondents to retain records, other than health, medical, government contract, grant-in-aid, or tax records, for more than three years;

• in connection with a statistical survey, that is not designed to produce valid and reliable results that can be generalized to the universe of study;

• requiring the use of a statistical data classification that has not been reviewed and approved by OMB;

• that includes a pledge of confidentiality that is not supported by authority established in statute or regulation, that is not supported by disclosure and data security policies that are consistent with the pledge, or which unnecessarily impedes sharing of data with other agencies for compatible confidential use; or

• requiring respondents to submit proprietary trade secrets, or other confidential information unless the agency can demonstrate that it has instituted procedures to protect the information's confidentiality to the extent permitted by law.

This ICR does not violate any of OMB's guidelines for information collections.

8. If applicable, provide a copy and identify the date and page number of publication in the Federal Register of the agency's notice, required by 5 CFR 1320.8(d), soliciting comments on the information collection prior to submission to OMB. Summarize public comments received in response to that notice and describe actions taken by the agency in response to these comments. Specifically address comments received on cost and hour burden.

Describe efforts to consult with persons outside the agency to obtain their views on the availability of data, frequency of collection, the clarity of instructions and recordkeeping, disclosure, or reporting format (if any), and on the data elements to be recorded, disclosed, or reported.

Consultation with representatives of those from whom information is to be obtained or those who must compile records should occur at least once every 3 years - even if the collection of information activity is the same as in prior periods. There may be circumstances that may preclude consultation in a specific situation. These circumstances should be explained.

As this action is a proposed collection, EPA’s efforts to consult with persons outside the agency to obtain their views on the various aspects of the rule are part of the notice-and-comment proposal process. Additionally, EPA held a SBAR panel as described above and incorporated that information into its cost estimates and proposed collection approach.

9. Explain any decision to provide any payment or gift to respondents, other than remuneration of contractors or grantees.

This information collection does not provide payment or gifts to respondents.

10. Describe any assurance of confidentiality provided to respondents and the basis for the assurance in statute, regulation, or agency policy. If the collection requires a systems of records notice (SORN) or privacy impact assessment (PIA), those should be cited and described here.

Collection of data through EIS, CAERS, and CEDRI includes Personal Identifiable Information (PII) for users to register with the CDX. This information is name, role (certifier, preparer, SLT reviewer), work address, work email, and work phone number. The EIS has a SORN and PIA. CAERS is built within the CDX and shares its SORN with CDX. CAERS additionally has a PIA and shares the email addresses collected from owners/operators with SLTs using CAERS to collect data from those owners/operators.

The remainder of the data collected under the proposed action meet the definition of emission data as defined at 40 CFR 2.301(a)(2)(i) and are, therefore, not subject to confidential treatment.

11. Provide additional justification for any questions of a sensitive nature, such as sexual behavior and attitudes, religious beliefs, and other matters that are commonly considered private. This justification should include the reasons why the agency considers the questions necessary, the specific uses to be made of the information, the explanation to be given to persons from whom the information is requested, and any steps to be taken to obtain their consent.

This information collection does not ask any questions concerning sexual behavior or attitudes, religious beliefs, or other matters usually considered private.

12. Provide estimates of the hour burden of the collection of information. The statement should:

• Indicate the number of respondents, frequency of response, annual hour burden, and an explanation of how the burden was estimated. Unless directed to do so, agencies should not conduct special surveys to obtain information on which to base hour burden estimates. Consultation with a sample (fewer than 10) of potential respondents is desirable. If the hour burden on respondents is expected to vary widely because of differences in activity, size, or complexity, show the range of estimated hour burden, and explain the reasons for the variance. Generally, estimates should not include burden hours for customary and usual business practices.

• If this request for approval covers more than one form, provide separate hour burden estimates for each form and aggregate the hour burdens.

• Provide estimates of annualized cost to respondents for the hour burdens for collections of information, identifying and using appropriate wage rate categories. The cost of contracting out or paying outside parties for information collection activities should not be included here. Instead, this cost should be included under ‘Annual Cost to Federal Government’.
  
  

The emissions data proposed to be required by the proposed AERR revision are submitted by state, territorial, local air pollution control agencies and certain tribes.¹⁵ The affected NAICS code is 924110 – Administration of Air and Water Resources and Solid Waste Management Programs, which includes governmental environmental protection and control agencies, and pollution control agencies. To meet requirements of the current AERR, states require owners/operators to submit data to the state, which the state then submits to EPA. This indirect burden on owners/operators is included in the subsection below to describe burden for owners/operators.

With this action, EPA is additionally proposing to collect certain data directly from owners/operators for sources that meet the point source emissions thresholds of the proposed action. Further, EPA has estimated costs not only for the first 3-year collection period (from 2024 through 2026), but also for 2027 because some of the key proposed collections would not start until 2027. Appendix A provides the assumptions and estimated costs for the period after the 3-year period covered by this ICR.

1. Labor cost assumptions

For this ICR, labor rates were developed using the U.S. Department of Labor, Bureau of Labor Statistics Web site as of May 2021 as accessed in March 2022. Hence, the labor costs assessed in this report are in 2021 dollars. Table 1 below provides the rates for state government as well as the rates for industries. An overhead rate of 110 percent was applied to all rates to derive the loaded rates (i.e., including fringe benefits) to be used in the cost estimates. This is consistent with ICRs prepared for other EPA rulemakings.

Table 1: Labor rates

Employee Type	Employer	Mean Hourly Wage	Loaded Hourly Rate	Source
Environmental Engineer	State Government	$43.25	$90.83	https://www.bls.gov/oes/current/naics4_ 999200.htm#17-0000
Architectural and Engineering Managers	State Government	$56.64	$118.94	https://www.bls.gov/oes/current/naics4_ 999200.htm#17-0000
Network and Computer Systems Administrator	State Government	$38.58	$81.02	https://www.bls.gov/oes/current/naics4_ 999200.htm#17-0000
Environmental Engineer	Any	$48.18	$101.18	https://www.bls.gov/oes/current/ oes172081.htm
Architectural and Engineering Managers	Any	$76.43	$160.50	https://www.bls.gov/oes/current/ oes119041.htm

2. Number of state, local, and tribal (SLT) respondents

Under the proposed AERR, 54 states (including the District of Columbia and 3 territories) and, depending on the reporting year being annual or triennial, additionally between 23 and 31 local and tribal air agencies would be subject to the national reporting requirements. For the 2024-2026 period covered by this ICR, these state, local, and tribal (SLT) air pollution control agencies would be required to compile and report emissions information for large stationary point sources on an annual basis, and for smaller point sources, stationary nonpoint and onroad and nonroad mobile sources on a 3-year basis. As described in Appendix A to this Supporting Statement, the AERR proposal includes collection of point sources and prescribed burning reports every year starting with the 2026 inventory year, reported in 2027 and therefore after the period of this ICR. For certain reporting activities, a fewer number of state, local, or tribal agencies are required to report, or voluntarily do so. These lower numbers are reflected in the relevant tables of this section and in the summary table provided in Section 13.1.

Additionally, based on expressed interest in CAERS from state and local agencies to date, EPA estimates that 12 state/local agencies will use CAERS for reporting in 2024. Based on the proposed AERR revision requiring some owners/operators to use CAERS for reporting HAP, EPA projects that 30 state/local agencies will use CAERS for reporting in 2025 and 54 in 2026. Furthermore, reporting that occurs in 2025 and 2026 is for the “smaller” set of sources due to higher proposed thresholds for those years in Table 1A to Appendix of the proposed revisions.¹⁶ Reporting in 2024 includes more point sources because 2023 is a triennial reporting year. As a result, EPA has assumed an average 32 state/local agencies will use CAERS for reporting across the 3 years (i.e., (12+30+54)/3) for the burden calculations associated with sources reported in 2025 and 2026.

3. Burden for SLT respondents

The SLT respondent burden for complying with the proposed AERR revision includes burden to meet both the annual and the 3-year (triennial) cycle reporting. Within the annual and triennial reporting requirements associated with the proposed AERR revision, EPA has estimated the burden separately for one-time activities, annual reporting, and triennial reporting. The subsections below describe each of these individual elements.

The SLTs reporting to EPA under the AERR maintain their own air quality management programs, which include permitting programs and annual emissions fee programs for their point sources. These fees help offset costs associated with running these emissions programs. Nevertheless, this ICR includes as part of the burden estimates, those SLTs’ efforts to collect and manage emissions inventory data for these purposes, much of which occur irrespective of the AERR. However, this ICR does not include certain efforts of SLTs unrelated to requirements of the AERR or the associated burden on their owners/operators. Table 2 below provides a summary of the included and excluded elements of this ICR. In this table, the last row represents costs associated with SLTs reporting data to EPA that are voluntarily collected and reported along with their required data. These types of voluntary reports include additional facilities that do not meet the AERR point source thresholds and emissions of HAP.

Table 2: Summary of efforts included in this ICR

		Owners/operators report		SLT collects from	SLT reports
Point Source…	to EPA		to SLT	owners/operators	to EPA
Data collected because of proposed AERR requirement	Included		Included	Included	Included
Optional data fields associated with pollutants required by proposed AERR	Included		Included	Included	Included
Data collected because of SLT requirement	N/A				Included

The proposed AERR revisions would lead to SLTs needing to make two key decisions that would impact how they implement any final requirements. While there is no proposed requirement to participate in CAERS, an SLT’s choice of whether to participate or not could significantly impact the costs of compliance and the mechanism of compliance with point source reporting requirements. As shown in Figure 1 below, EPA recommends that states first determine whether they intend to participate in CAERS and in what way, and then determine whether they intend to report HAP on behalf of owners/operators. As illustrated in the figure, SLTs should decide whether to retain the user interface (“front end”) of their current emissions data collection system and whether to retain the database (“back end”) of their system. The front end is the user interface (often web-based) that owners/operators use to submit the data. Occasionally the SLT system interacts with an SLT electronic permitting system. The back end is the master storage location for the data collected by the SLT, and often interacts with other SLT data systems. CAERS is being constructed to support different ways that SLTs can use CAERS features from the front end, back-end, both, or neither. Even SLTs that choose not to participate in any of the CAERS cases shown can choose to reduce burden on facilities via collaboration with the CAERS features, such as expected quality assurance services, shared code tables, and other necessary aspects of electronic data collection and compilation.

Figure 1: Decision tree representing SLT decisions about implementing proposed requirements

Once an SLT has determined their plans for interacting with CAERS, the state should then determine whether it will report HAP on behalf of owners/operators. The combination of this decision and their decision on CAERS would then need to be part of the process of updating SLT regulations to collect emissions. The EPA expects most states would need to update their emissions collection regulations to comply with aspects of these proposed revisions, even if the SLT chooses not to participate in CAERS and not to report HAP on behalf of owners/operators. For example, this action proposes new requirements to collect information regarding latitude/longitude of release points, Title V permit identifiers, and regulation applicability. The EPA does not believe that SLTs will meet those requirements without collecting at least some new information. The choice of CAERS case impacts the overall burden on states as described in Section 12.3.1.

1. SLT burden for one-time activities under proposed AERR

To prepare for proposed changes to the AERR that would take effect for the 2026 reporting year, SLTs would have both required and voluntary one-time efforts that would occur during the period covered by the ICR. One-time activities would be to prepare for proposed changes (that would take effect in 2027) in point source reporting and in prescribed fire activity data requirements. For point sources, these activities depend on three SLT choices:

• Whether the SLT will adopt CAERS to support point source emissions collection or keep the SLT point source emissions collection system,

• Whether the state will report HAP on behalf of owners/operators, and

• Whether the state will maintain or develop their own HAP collection program.
  
  

These activities A through F and associated choices are:

States could choose activity A-1 (higher burden) or A-2 (lower burden):
A-1. Update SLT point source emissions collection system to accommodate new AERR requirements consistent with SLT regulation update.	| OR | | |	A-2. Adopt CAERS as SLT point source emissions collection system (case 1 or 2).
States could choose activity B-1 through B-3 (higher burden) or B-4 through B-7/B-8 (lower burden)
B-1. Revise SLT emissions collection regulation to include HAP reporting consistent with AERR requirements.	| | | | OR | | | | | |	B-4. Rely on EPA HAP collection via CAERS (cases 1-4). and, for states with HAP collection program that they want to create or maintain: B-5. Create and deliver training to owners/operators. B-6. Curate list of facilities to remove duplicates. B-7. Other coordination activities including ensuring any CAERS customizations meet SLT requirements. B-8. Optionally, instrument SLT emissions collection system to receive data from CAERS.
B-2. Update SLT point source emissions collection system to accommodate new HAP requirements.
B-3. Apply to EPA for permission to report HAP on behalf of owners/operators.
C. Revise SLT emissions collection regulations to meet new AERR requirements for point sources: new data fields, newly mandatory data fields, and reporting of daily activity data for small generation units (this could exclude updates for HAP reporting depending on SLT choice to rely on EPA collection for HAP or not).
D. Develop SLT regulations to collect prescribed burning data consistent with proposed AERR.
E. Develop SLT data collection system for prescribed burning data to conform with EPA collection and reporting requirements.
F. Develop quality assurance and other techniques for prescribed burning data.

The EPA and SLTs have envisioned four cases for how an SLT could interact with CAERS, which are relevant to activities A and B above. Under CAERS cases 1 and 2, the SLT would choose to retain its data system but rely on some aspects of the CAERS system for data sharing with other emissions programs. SLTs could also choose CAERS case 3, in which the SLT uses the CAERS user interface and retains its back-end database or CAERS case 4, where the SLT uses CAERS for both the collection and the storage of the point source emissions inventory data.

For activity B above, if an SLT chooses the path represented by activity B-4 through B-8 (the CAERS path), the SLT would have various additional choices depending on their circumstances. In this case, the SLT would be electing to use CAERS in some form. For example, SLTs that do not currently have a HAP collection program or wish to eliminate their HAP collection program and rely on EPA’s collection, could choose the lower burden option B-4 alone. For states under the CAERS path that wish to maintain their HAP collection program, an SLT could choose to:

1. Adopt CAERS as the SLT data system, which would require one-time activities listed as activities B-5, B-6, and B-7 (needed for CAERS case 4);

2. Connect the SLT data system to receive data from CAERS, which would include activity B-8 (needed for CAERS case 3); or

3. Not engage with CAERS using any of the four cases.

SLTs with HAP collection programs that choose not to participate in CAERS would potentially cause at least some owners/operators regulated under the SLT HAP reporting requirements to have to report both to CAERS and separately to the SLT system.

The tables below provide estimated hours burden for one-time activities per state respondent. Table 3a provides one-time activities for point sources, and Table 3b provides the estimated burden in hours for states to do additional one-time activities to adopt CAERS case 3 or case 4. Table 4 provides the estimated hours burden for one-time activities for developing a prescribed burning collection approach. Finally, Table 5 provides the annualized burden per state across all of the one-time activities, including costs.

Table 3a: State respondent burden hours for one-time point source activities

	Hours Per Respondent
Activity	Engineering Managerial Hours	Engineering Technical Hours			IT Admin Hours		Total
Point sources - required activities
1. Revise SLT regulations to accommodate new required data fields and reporting of daily activity data for small generation units.	52	520					572
2. Update SLT data system to accommodate new point source data fields and daily activity data for small generation units.	124	200			1,040		1,364
Subtotal	176	720			1,040		1,936
Point sources - optional activities when including HAP reporting
1. Revise SLT regulations to adjust HAP reporting based on EPA requirements	104	1,040					1,144
2. Update SLT data system to accommodate new point source HAP reporting	144	400			1,040		1,584
3. Complete and submit application to EPA for permission to report HAP on behalf of facilities.	12	120					132
Subtotal	260	1,560			1,040		2,860

Table 3b: State respondent burden additional voluntary burden for one-time point source activities when using CAERS

	Hours Per Respondent
Activity	Engineering Managerial Hours	Engineering Technical Hours	IT Admin Hours					Total
CAERS Case 3 and 4 (State uses only CAERS)
1. Update and deliver training to owners/operators	24	240						264
2. Curate list of facilities to remove duplicates	16	160						176
3. Other coordination activities including ensuring any CAERS customizations meet SLT requirements.	48	480						528
CAERS Case 3 (CAERS front end and SLT database)
4. Modify SLT system to receive data from CAERS user interface.	104		1,040					1,144
Subtotal - Case 3	192	880	1,040					2,112
Subtotal - Case 4	88	880						968

Table 4: State respondent burden hours for one-time activities to develop prescribed burning data collection

	Hours Per Respondent
Activity	Engineering Managerial Hours	Engineering Technical Hours	IT Admin Hours	Total
1. Revise SLT regulations to collect prescribed burning data.	312	1,040		1,352
2. Develop data collection system for prescribed burning data to conform with EPA collection and reporting requirements.	416	1,040	3,120	4,576
3. Develop quality assurance and other techniques.	72	480	240	792
Total	800	2,560	3,360	6,720

Table 5: Annualized one-time burden per state respondent

	Manager Hrs/Yr @	Technical Hrs/Yr @			IT Hrs/Yr @			Hours/ Year			Labor Cost/ Year
Activity	$118.94	$90.83			$81.02
Prescribed Burning Required Activities
Develop prescribed burning data collection	267	853			1,120			2,240			$199,963
Point Sources Required Activities
Reporting with EIS or CAERS case 1, 2 or 3: Update regulations and data storage system	59	240			347			645			$56,862
Reporting with CAERS case 4: Update regulations	17	173			0			191			$17,805
Point Sources Voluntary Activities
Revise regulations, update SLT data system for HAP, and complete/submit application to EPA to report on behalf of owners/operators.	87	520			347			953			$85,624
Transition Tasks for CAERS Case 3	123	533			693			1,349			$119,203
Transition Tasks for CAERS Case 4	29	293			0			323			$30,131

2. SLT annual activities under proposed AERR

Annual SLT activities would be done to submit emissions data for annually reported point sources with potential to emit 2,500 tons per year (tpy) of NO_x, CO, or SO₂; or 250 tpy of VOC, PM₁₀, PM_2.5 or NH₃. The key steps to perform the work to meet the AERR requirements are:

• Maintain the state’s data system to collect data from facilities;

• Collect emissions data and other associated information;

• Train staff in coding and data submission techniques;

• Quality-assure and quality-control emissions data and resolve errors and anomalies prior to submitting to the EIS electronic quality-assurance;

• Maintain records associated with data submitted by sources;

• Extract the necessary data from the state electronic data system;

• Convert any facility inventory data (i.e., attributes of the facility including details about its units, processes, release points and controls) for new facilities into the XML submittal format;

• Convert the point emissions data into the XML submittal format;

• Run the automated quality-assurance checks provided in the EPA data system and resolve any critical errors;

• Submit the final file to EPA; and

• Respond to any follow-up inquiries and point source data reviews from EPA.

In addition, SLTs may optionally include in their submissions additional data, including emissions for facilities that are not required to be reported annually as well as HAP and GHG emissions. To accomplish this optional work, the same activities would be done as are listed above to meet AERR requirements, but that work would take incrementally more effort.

For 2 of the 3 years in each triennial cycle, the agencies submit only the largest sources as described above. To help estimate the time needed to report emissions for only the annually reported sources, we estimate the number of such sources by considering the emissions reporting thresholds. The AERR reporting thresholds are potential-to-emit (PTE) thresholds; however, EPA does not collect PTE data under the AERR. The EPA only collects actual emissions, and actual emissions are lower than PTE values. Since EPA does not collect data on PTE, it is difficult to know with certainty the number of annually reported sources. Furthermore, many states voluntarily submit many more facilities than those required. For these reasons, EPA must estimate the number of required sources for annual reporting.

Based on an analysis of the 2017 NEI, 1,055 facilities had actual emissions greater than the 2024 and 2025 inventory year PTE thresholds of 2,500 tpy of NO_x, CO, or SO₂, or 250 tpy of VOC, PM₁₀, PM_2.5, or NH₃. To adjust for the undercounting due to actual emissions, we retained the number of estimated facilities from the previous AERR ICR, which is about 2.3x the facility count based on actual emissions. As a result, we assume 2,510 of the 2024- and 2025-year sources are reported for the purposes of this analysis across 54 state/territorial and 23 local and tribal air agencies. This equates to an average of 33 annually reported sources that would be required on average per agency for 2 of the 3 years. The number of required sources can be much larger for heavily industrialized states and smaller (all the way down to zero) for some smaller states and local agencies.

To account for states that we estimate will use the CAERS for the required annually reported sources in the 2023 through 2025 emission inventory years, we have considered the reduction in effort associated with the steps for reporting to EIS necessary only when a state maintains their own data system and thus needs to convert that data for submission to the EIS. As shown in Table 1, the rightmost column shows which steps are necessary for agencies that use CAERS.

3. SLT triennial activities under proposed AERR

Triennial Point Source Activities and Assumptions

Although the AERR proposal includes removing triennial point source reporting starting with emissions for 2026, for purposes of this document, we refer to the point source reporting for 2023 emissions as triennial reporting. For triennial reporting of 2023 emissions in 2024, SLTs would have the same point source activities as described for the annual reporting above, but those activities would take longer because more sources would be involved. Rather than the PTE thresholds listed above for annual reporting, SLTs would submit additional emissions data for point sources that have lower emissions than the annually reported sources and have a potential to emit 100 tpy of NO_x, SO₂, VOC, PM₁₀, PM_2.5, or NH₃; or 1,000 tpy of CO; or that have actual emissions of at least 0.5 tons of lead (Pb). Further, the emissions reporting thresholds for facilities within nonattainment areas are even lower for triennially reported point sources, in accordance with Table 1A of Appendix A to Subpart A of 40 CFR part 51 (for which no changes are proposed).

Like annually reported sources, the triennial reporting thresholds are based on PTE values, but EPA does not collect PTE data. Fortunately, the triennial source reporting criteria are nearly the same as the major source definition for criteria pollutants and precursors, and a list of such major sources is available from the EPA Enforcement and Compliance History Online (ECHO) web application. ECHO is fed by reporting of compliance data from the state agencies. These data are sent from states to the ECHO system many times per year, which helps ensure that we are using updated information.

For the previous version of this ICR using the ECHO database, EPA determined that there are 13,408 Major Title V facilities nationwide. To adjust this facility-count for the triennial definition, we also needed to consider the triennial threshold for Pb, which is 0.5 tons of actual emissions per year (and more stringent than the major source definition). Since the Pb threshold is based on actual emissions, we used the 2014 NEI to determine that just 12 additional facilities have 0.5 tons of Pb emissions or more and are not otherwise identified as major sources.¹⁷ The resulting triennial source facility total used for this work is 13,420. Because the number of major source facilities has decreased¹⁸ since the previous ICR, these numbers provide a conservative (or overstated) estimate of the number of facilities.

Since there are 85 reporting agencies in triennial years, we estimate an average of 158 facilities to be reported per agency for the triennial inventories. The number of sources can be much larger for the large, heavily industrialized states, and smaller for some SLT agencies. Because much of the effort needed to report the point source emissions data from the state data systems to EPA involves automated data manipulations, there are economies of scale for the states with many sources. The idea of economies of scale is further supported by the actual number of facilities reported. States reported about 72 thousand facilities (which excludes most airports and railyards that are reported differently) in the most recent 2020 triennial reporting years, which is far greater than the 13,420 facilities that we estimate are required.

Other Triennial Activities for SLTs

In addition to the triennial point source collection and reporting, the AERR would include additional triennial activities for SLTs:

• For nonpoint sources, complete a nonpoint survey to indicate plans for reporting each nonpoint category;

• For nonpoint sources with EPA-provided emissions calculation tools (excluding commercial marine vessels and locomotives), either submit nonpoint tool input data or review, comment on, and accept EPA-provided nonpoint tool inputs. This includes compiling and reporting total point source activity data for those data categories for which EPA provides templates for use in reconciliation between point and nonpoint sources to avoid double counting (e.g., industrial, commercial, and institutional boilers);

• For nonpoint sources without EPA-provided methods and tools, estimate emissions, run quality assurance checks, format data into XML format, and submit emissions data and documentation.

• For nonpoint sources in states that overlap with tribes that submit data, adjust nonpoint submissions for tribal boundaries.

• Either submit airport activity data (i.e., landings and takeoffs) or review EPA-provided data, submit comments on that data, and/or accept that data.

• Either submit rail yard activity data and associated documentation or review EPA-provided data, submit comments on that data, and/or accept that data.

• For commercial marine vessels and locomotives, either report annual actual emissions and associated documentation, provide comment on EPA-provided emissions, or accept EPA-provided emissions.

• For all states except California, develop inputs to the MOtor Vehicle Emissions Simulator (MOVES) for onroad mobile and nonroad mobile sources. Review and revise draft data from EPA and/or collect such data, review and edit that data, format data into required XML format, run quality assurance checks, and submit the data to EPA.

• For California, develop and report statewide inventory emission estimates for onroad and nonroad mobile sources for all criteria pollutants. Develop model inputs for California’s mobile source model(s), run the California mobile source model(s), run quality assurance checks, format data into XML format, and submit emissions data and documentation.

Additionally for triennial years for this ICR period, SLTs could perform several additional voluntary activities under the AERR collection:

• For nonpoint sources with EPA-provided methods and tools, estimate emissions, run quality assurance checks, format data into XML format, and submit emissions data and documentation.

• For aircraft, ground support equipment (GSE), and/or rail yards, voluntarily estimate and submit emissions and documentation of the associated calculations.

• For prescribed fire, agricultural fire, and wildfires, review, comment on, and/or accept EPA-provided activity data and emissions data or submit emissions.

4. SLT burden for annual and triennial years

The SLT burden for annual and triennial years is presented in this section as an average per year. The burden hours are provided separately for data system activities, point source reporting, and nonpoint reporting. Furthermore, required activities are separated from voluntary activities.

Use of these averages should provide an overly conservative (larger) estimate of total burden hours because the burden values for the smaller agencies are being overestimated since they will have fewer sources than average, and the average burden values do not include the economies of scale experienced by the larger agencies. The idea of economies of scale is further supported by the actual number of facilities and pollutants reported in these years as compared to those that we believe are required.

To account for states that we estimate will use the CAERS for the required annually reported sources in the 2023 through 2025 emission inventory years, we have considered the reduction in effort associated with the steps for reporting to EIS necessary only when a state maintains their own data system and thus needs to convert that data for submission to the EIS. As shown in Table 6a and Table 6b, the rightmost column indicates which steps are necessary for agencies that use CAERS.

Maintaining SLT point source collection system

Table 6a summarizes the average hour burden estimates for operation and maintenance (O&M) of the SLT data system for collecting point source data from owners/operators in the state. The table includes 50% of a full-time employee (FTE) for information technology (IT) administration and additional hours for an engineer to provide guidance to IT administration, making minor annual updates to the data system, and user support. Major data system updates have been covered previously in Section 12.3.1 as a one-time activity during the period of this ICR and not included in Table 6a. The hours for the engineering activities are estimated to be about 20% of an FTE’s time. Engineering managerial hours have been estimated as 10% of the engineering and IT administrative hours associated with each activity. The table includes estimates of O&M adjusted for estimated reductions in labor associated with CAERS cases 3 and 4.

Table 6a: SLT data system operation and maintenance hours for NEI Collection from owners/operators

	Hours Per Respondent
Activity	Engineering Managerial Hours	Engineering Technical Hours	IT Admin Hours	Total	Applies to CAERS Cases?
1. Collection system operation & maintenance (O&M)	112	80	1,040	1,232	case 3 @ 80%
2. Update collection system with new codes, emission factors, and other new information for reporting year	12	40	80	132	case 3 @ 80%
3. User support for point source emissions data reporting	36	320	40	396	case 3 @ 50%, case 4 @ 50%
Subtotal for System O&M EIS and Case 1 & 2	160	440	1,160	1,760	Hours Reduction
Subtotal for System O&M with CAERS Case 3	117	256	916	1,289	27%
Subtotal for System with CAERS Case 4	18	160	20	198	89%

For states that choose CAERS case 3, EPA estimates that the burden of activities 1 and 2 are reduced by about 20% because the state would no longer need to maintain the public-facing user interface for their collection system. Activities 1 and 2 are eliminated for CAERS case 4. In both CAERS cases 3 and 4, EPA also assumes that user support is reduced by 50% based on the streamlined processes put in place. The user support reduction would be averaged over the course of the 3-year period and would not be realized until the second and third years of CAERS implementation. Further, EPA has attempted to include only those hours associated with the sources and pollutants that EPA requires to be collected for reporting under the AERR (see also explanation provided in Section 12.3, Table 2). In other words, if SLTs incur additional burden (e.g., more help desk requests) associated with collecting emissions data from facilities that the SLT chooses to collect, this ICR does not cover that burden. Based on this information, EPA estimates that the overall estimated O&M burden reduction for SLTs, on average, would be a 27% reduction for CAERS case 3 and an 89% reduction for case 4.

EPA recognizes that many SLTs subcontract their point source emissions collection systems to a third party; however, the cost approach taken in later sections uses the hours estimates assumed in Table Table 6a as the basis for data system costs. This will be further addressed in Table 8b later in this section.

Annual and Triennial Point Source Reporting

In addition to the point source data system activities, Table 6b provides the average hour burden estimates for an SLT to perform point source reporting for the proposed AERR during the period of the ICR. The activities listed in this table match with those point source activities described in Sections 12.3.2 and 12.3.3 above.

In the hour estimates included in Table 6b, EPA has not distinguished between collection and reporting of those sources and pollutants required to be reported versus those sources that SLTs report voluntarily. Unlike the incremental burden for SLTs to collect emissions from facilities not required by the AERR, the incremental burden to report these additional facilities is small. This is because states who report many additional sources and pollutants voluntarily do so using automated processes to export, convert, and send the data to EPA. Over many years of collecting data from SLTs, EPA has heard numerous times from such agencies that it’s harder for these SLTs to exclude facilities and pollutants than simply to report both required and voluntarily provided facilities in every submission. Because of these considerations, EPA has not tried to separate out the hours by required and voluntarily reported facilities and pollutants, but rather (in this table) has attempted to estimate hours to reflect both required and voluntarily reported sources and pollutants.

Table 6b: SLT point source reporting burden hours by activity

	Hours Per Respondent
Activity	Engineering Managerial Hours	Engineering Technical Hours			Total				Applies to CAERS Cases?
Point sources - Annual (required and voluntary)
4. Quality assurance of submitted data and revision support	2	24			26				3 @ 100%, 4 @ 50%
5. Extract data from the state data system		4			4				3
6. Convert data into the XML format – facility attributes information		8			8				3
7. Convert data into the XML format – annual emissions information		4			4				3
8. Run EIS quality-assurance checks and resolve critical errors	2	24			26
9. Submit final file to the EPA		2			2				3, 4
10. Respond to follow-up inquiries from the EPA	2	4			6				3, 4
Subtotal Annual Point Source Reporting via EIS	6	70			76				Hours Reduction
Subtotal Annual Point Source Reporting via CAERS case 3	4	46			50				34%
Subtotal Annual Point Source Reporting via CAERS, case 4	3	16			19				75%
Point sources - Triennial (required and voluntary), additional hours
4. Quality assurance of submitted data and revision support	12	120			132				3 @ 100%, 4 @ 50%
5. Extract data from the state data system	0	4			4				3
6. Convert data into the XML format – facility attributes information	0	16			16				3
7. Convert data into the XML format – annual emissions information	0	8			8				3
8. Run EIS quality-assurance checks and resolve critical errors	12	120			132
9. Submit final file to the EPA	1	2			3				3, 4
10. Respond to follow-up inquiries from the EPA	10	20			30				3, 4
Subtotal Triennial Point Source Reporting via EIS – all point sources via EIS	35	290			325				Hours Reduction
Subtotal Triennial Point Sources Reporting via CAERS case 3	23	170			193				41%
Subtotal Triennial Point Source Reporting via CAERS case 4	16	80			96				70%

To create the hours estimates in Table 6b, EPA conservatively estimated that the additional hours needed for activities 4, 8, and 10 in triennial years will increase by a factor of 5 compared to the annual facility reporting. This factor is derived by dividing the average triennially reported facility count per agency (158) by the average annually reported facility count per agency (33). Activities 6 and 7 are conservatively estimated to require just twice the effort needed for the annually reported sources, because the activity is largely the same regardless of the number of sources. The EPA estimates that activities 5 and 9 would require the same amount of effort in both triennial and non-triennial years.

To account for the states forecast to use the CAERS for triennial reporting for the 2023 inventory year (reported in 2024), EPA has considered the reduction in effort associated with certain activities. The rightmost column of Table 6b indicates EPA’s assumptions about whether the activity is relevant for CAERS cases. For CAERS case 3, the state would use the CAERS user interface to collect the data and send it to the state for further processing and submission back to EPA. This approach would have the effect of running the quality assurance checks while the owners/operators were reporting in CAERS. Thus, activity 8 is essentially eliminated because all data collected via CAERS will already be able to pass EIS QA checks. For CAERS case 4, an SLT is using only CAERS, which eliminates activities 5 through 8 and 50% of activity 4.

Based on these numbers, EPA estimates SLTs reporting point sources without CAERS would spend 76 hours for annually reported sources and 325 hours for triennial reported sources. SLTs using CAERS case 3 would have a burden reduction of 41% in triennial years and 34% in other years. Finally, SLTs using CAERS case 4 would have a burden reduction of 70% in triennial years and 75% in other years.

While Table 6b includes both hours for reporting both required and voluntary pollutants as a total, EPA has made assumptions about the proportion of activity occuring for CAP and HAP, depending on each of the reporting cases available. During the 2024-2026 period, the HAP reporting is voluntary, and thus the information in Table 6c is used below in Table 8a when providing cost information broken out by required and voluntary costs. To create Table 6c, EPA assumed that the total reporting burdens from the summary rows of Table 6b were divided as follows. For reporting without CAERS, 30% of the effort is associated with HAP reporting. For reporting with CAERS case 3 or 4, 20% of the effort is associated with HAP reporting. This lower fraction of burden for HAP reporting via CAERS is based on the integrated nature with which CAERS provides for HAP reporting. This table allows for voluntary versus required costs splits provided later in Table 8a.

Table 6c: Split of burden for CAP and HAP reporting.

	Hours Per Respondent		Hours Per Respondent
	CAPs		HAPs
Activity	Engineering Managerial Hours/yr	Engineering Technical Hours/Yr	Engineering Managerial Hours/Yr	Engineering Technical Hours/Yr
2024 and 2025 emissions reporting without CAERS	4.20	49	1.80	21
2024 and 2025 emissions reporting with CAERS, case 3	3.20	36.80	0.80	9.20
2024 and 2025 emissions reporting with CAERS, case 4	2.40	12.80	0.60	3.20
2026 emissions reporting, 3-year average triennial increment without CAERS	8.17	67.67	3.50	29
2026 emissions reporting, 3-year average, triennial increment with CAERS, case 3	6.13	45.33	1.53	11.33
2026 emissions reporting, 3-year average, triennial increment with CAERS, case 4	4.27	21.33	1.07	5.33

Additional Triennial Reporting for Nonpoint, Mobile, and Event Sources

In addition to the triennial point source reporting, additional activities are required for other source categories as listed in Section 12.3.3 above. Table 7a provides the average hour burden estimates for states (not local agencies or tribes) to perform the steps that would be required by the AERR or that could be done voluntarily by states on triennial years for nonpoint sources, airports, railyards, locomotives, commercial marine vessels, and onroad and nonroad mobile sources. This table also provides the assumed number of states and territories for which each activity would apply. Where these values do not equal the total number of states or territories, it is because with the many ways to comply with the AERR requirements, states and territories choose different approaches. Similarly, Table 7c provides the average hour burden estimates and affected entities for local and tribal agencies.

Table 7a: State nonpoint, mobile, and other reporting burden hours by activity

		Hours Per Respondent
Activity	State count	Engineering Managerial Hours	Engineering Technical Hours	Total
Required Activities
Complete Nonpoint Survey	54	2	40	42
Report inputs for EPA nonpoint tools or review, comment and/or accept EPA data.	54	62	1,231	1,293
Report emissions and documentation for sectors not included in nonpoint tools	18	12	240	252
Adjust nonpoint submissions for boundaries of Indian country	4	4	64	68
Submit or review, comment, and/or accept EPA airport activity data	54	2	40	42
Submit or review, comment, and/or accept EPA rail yard activity data	43	1	16	17
Submit CMV and locomotive emissions data and documentation or review, comment, and/or accept EPA emissions estimates.	42	4	80	84
For all states but California, report MOVES inputs	53	6	120	126
For California, report onroad and nonroad emissions and documentation	1	9	180	189
Average hours per state, required activities	54	80	1,592	1,672
Voluntary Activities
Report emissions for sectors included in nonpoint tools, including documentation	13	44	880	924
Report emissions for aircraft, ground support equipment, and/or rail yards, including documentation	5	12	240	252
Comment on prescribed fire and wildfire activity data, submit activity data, or submit emissions	20	8	160	168
Average hours per state, voluntary activities	20	40	792	832

Regarding the number of states impacted by each of the activities in Table 7a, EPA made several assumptions based on past collections of NEI data from states. For all estimates in this table, EPA assumes the managerial hours to be about 5% of the engineering technical hours, rounded up to the nearest hour.

For activities other than 2 and 3 in Table 7a, EPA used expert judgement based on EPA’s implementation of the AERR for 15 years to specify the engineering technical hours. The number of states affected by these tasks are based on the following. Under the proposed revision, all states (including the District of Columbia and 3 territories) would be required to complete the nonpoint survey (activity 1). Four states overlap Indian country for tribes that reported to the 2017 NEI: Arizona, Colorado, Idaho, and Montana (activity 4). All states have airports and, under the proposed revision, they would all be required to act on these sources (activity 5). Forty-three states have rail yards (activity 6), and 42 states have waterways with commercial marine vessels (activity 7). California is excluded from MOVES inputs reporting leaving just 53 states/territories (activity 8) and that is the only state required to report emissions (activity 9).

For activity 2, EPA considered more detailed tasks associated with these activities to build the hours estimate provided. First, EPA estimates an average per state of 1,231 engineering technical hours for activity 2 based on calculations included in Table 7b below. Actual state hours burden depends on implementation choices that the state would have to comply with the AERR revisions. These calculations include state activities for three types of tools: the Wagon Wheel,¹⁹ which is the primary tool covering the bulk of the nonpoint sectors, (2) the oil and gas tool, and (3) four other stand-alone spreadsheet tools for agricultural fertilizer, livestock, fuel containers, and stage II gasoline.

Table 7b: Assumptions and calculations for state nonpoint tool submissions (activity 2)

Sub-Task	No. States	Basis	Average Submitted Templates or Sectors per State	Average Hours/ Template or sector	Total for All States Performing Sub-task	Average Hours Across All States
Prepare/submit Wagon Wheel input templates	36	Per template	12	20	8,640	1,168
Review/accept Wagon Wheel input templates	54	Per template	84	12	54,432
Prepare/submit oil and gas tool inputs	14	Per sector	1	40	560	15
Review/accept oil and gas tool inputs	22	Per sector	1	12	264
Review/comment/accept other tool inputs	54	Per sector	4	12	2,592	48
Total						1,231

As shown in Table 7b, EPA identified that 36 states submitted Wagon Wheel input templates to EPA for the 2020 NEI and for these submissions, just 6 of the 92 possible templates were submitted by each state on average. Since these counts were made before the 2020 NEI process had been completed, EPA conservatively estimates that a total of 12 templates would be submitted by each state.²⁰ EPA expects that all 54 states (including District of Columbia and 3 territories) will accept at least some of the 92 EPA-provided templates. To calculate the average number of templates states would review/accept rather than submit (84), EPA averaged the 80 templates for review/accept by the 36 states with the 92 templates for review/accept by the remaining 16 states. The EPA estimates that a state would spend an average of 20 hours to prepare and submit a Wagon Wheel template and 12 hours to review each template. Based on these estimates, states would spend an average of 1,168 hours per state on Wagon Wheel template activities.

In addition, for activity 2, Table 7b provides more information for the oil and gas tool activity. The EPA used the 2017 NEI submissions for the oil and gas tool to determine that 6 states submitted tool inputs, 8 states submitted emissions, and 22 states reviewed and accepted EPA oil and gas tool inputs and emissions. Because the proposed AERR revision would require all states to submit tool inputs, EPA summed together the counts of states submitting tool inputs and submitting emissions to assume that 14 states would submit oil and gas tool inputs. The EPA estimates 40 hours to prepare and submit oil and gas tool inputs and 12 hours to review and accept such inputs. Finally, since the total hours as used in Table 7a will be multiplied by the total number of states submitting nonpoint sources, EPA divided by this total number to determine that, on average across all states (including those that do not have these sources), oil and gas tool activities account for 16 hours.

To complete the hours estimates for activity 2, EPA also used estimates of burden for the four other nonpoint tools. The EPA expects that based on the proposed AERR revisions, all states would participate in review/comment/acceptance of those data, and this would take each state on average 12 hours per tool. Based on these assumptions, states would spend an additional 48 hours. The sum of the 1,168 hours from the Wagon Wheel, the 15 hours from the oil and gas tool, and the 48 hours for other tools provides the final average hour count for activity 2 of 1,231 hours.

For activity 3 in Table 7a, states would report emissions for sectors not included in EPA’s nonpoint tools. The EPA estimates that about one-third of the states (18) will, on average, report emissions for 2 sectors for which EPA does not have nonpoint emissions tools. Each sector is estimated to take 120 hours to estimate and submit, which is greater than the burden for other sectors because the state cannot benefit from an EPA-provided tool. Based on these assumptions, each state staff person would spend 240 hours to estimate and submit these emissions, and with manager hours included, a total of 252 hours.

In total, Table 7a shows that states would spend at minimum 1,503 hours (activities 1, 2, 5, and 8) and at maximum 1,987 (if California were to do activities 1-7 and 9) performing required activities for nonpoint and mobile source submissions in triennial years. We have also computed the average hours per state for required activities separately for each labor category. To do this, we multiplied the total number of states expected to perform each activity by the number of hours for that activity. Then, we summed the total number of hours across all activity-state combinations and divided by the total number of states. The number of hours expected for each state would depend on the choices they make for meeting the AERR requirements. The total average number of hours for required activities, including manager hours, is 1,672 hours.

In addition, several voluntary activities could be performed by states. For voluntary activity 10, EPA estimates that, under the proposed approach, just 25% of states will still submit emissions for nonpoint sources with EPA tools. Not many states are expected to take this voluntary step because it is an additional burden beyond the proposed new AERR requirements. For states that do take this step, EPA estimates 880 hours for states to report an average of 11 sectors taking 80 hours per sector. The average of 11 sectors per state was derived from EPA observations during the 2020 submission period based on 26 states submitting 290 state-sector combinations. Because the AERR required emissions submissions for the 2020 cycle (rather than only tool inputs), this estimated number of sectors per state may be an overestimate for this ICR since states would not be required to report emissions under the AERR revision.

For voluntary activity 11, states have rarely submitted airport emissions data except as part of their point source submissions for the largest airports. Even so, to capture the burden associated with this voluntary activity, EPA assumes that up to 5 states may choose to do so. The EPA estimates that, including manager hours, this labor-intensive step would take 252 hours per state.

Finally, for the last voluntary activity in Table 7a (activity 12), EPA determined from the 2020 NEI process that 20 states voluntary reported prescribed fire and/or wildfire data to EPA in the 2017 NEI cycle. The EPA assumes that these efforts take about 160 hours staff time per state. Based on these estimates and including manager hours, states could spend on average an additional 832 hours on voluntary activities associated with nonpoint, mobile, and fire emissions data for the triennial NEI.

In addition to the burden for states/territories, Table 7c separately provides estimates for local and tribal agencies that also report to the NEI. The assumptions made in compiling Table 7c are generally the same as the assumptions described previously for Table 7a, with several notable exceptions.

Table 7c: Local and tribal nonpoint, mobile, and other sources burden hours by activity

		Hours Per Respondent
Activity	Local/ Tribe count	Engineering Managerial Hours	Engineering Technical Hours	Total
Local and Tribal Reporters
Complete Nonpoint Survey	30	1	20	21
Report inputs for EPA nonpoint tools or review, comment and/or accept EPA data.	30	19	372	391
Report emissions and documentation for sectors not included in nonpoint tools	10	12	240	252
Adjust nonpoint submissions for tribal boundaries	7	2	37	39
Submit or review, comment, and/or accept EPA airport activity data	23	2	40	42
Submit or review, comment, and/or accept EPA rail yard activity data	20	1	16	17
Report MOVES inputs	23	2	40	42
For local agencies, coordinate with state agencies to complete stationary nonpoint, nonroad mobile, and onroad mobile sources for all pollutants	23	4	80	84
Average hours per entity, required activities	30	31	612	643
Voluntary Activities
Report emissions and documentation for sectors included in nonpoint tools	9	22	440	462
Report emissions for aircraft, ground support equipment, and/or rail yards, including documentation	1	12	240	252
Average hours per entity, voluntary activities	9	23	467	490

In Table 7c, the local/tribal counts are provided rather than state counts. These values reflect the 2017 NEI process which included 23 local agencies and 7 tribes reporting emissions. For the proposed AERR revisions, EPA assumes that all these agencies would complete the Nonpoint Survey (activity 1) and report inputs for EPA nonpoint tools or review, comment and/or accept EPA data (activity 2). Because local agencies and tribes may have fewer sectors to report, the technical engineering hours for completing the nonpoint survey are assumed to be half of the burden as for states. Similarly, because local agencies and tribes may have fewer sectors and do have fewer areas (i.e., counties), EPA has assumed that the technical engineering hours for activity 2 is 30% lower than the hours for states. The EPA additionally assumes that only local agencies would need to act on airport activity data (activity 5) and submit MOVES inputs (activity 7) because past tribal submissions did not include this information. Further EPA has found just 20 local agencies and no tribes have rail yards (activity 6).

As a result of these assumptions, EPA estimates that local agencies would need to spend between 597 hours (activities 1, 2, and 5-8) and 849 hours (including activity 3) on required activities. Tribal agencies that are affected by the AERR would need to spend between 451 hours (activities 1, 2 and 4) and 703 hours (including activity 3). The average number of hours for required activities, computed in the same way as for states, is 643 hours.

Table 7c also includes voluntary activities for local and tribal agencies. The EPA estimates that 2 local agencies and all 7 tribal agencies that have previously reported nonpoint data would continue to report nonpoint emissions voluntarily (activity 9). This assumption for Indian tribes accounts for the possibility that rather than do activities 1 and 2 and report nonpoint tool inputs, tribes will report emissions using techniques they have used in the past. Since those tribes are also accounted for in burden estimates for activities 1 and 2, but those tribes may not be required to do those activities, these estimates of voluntary burden may represent some double counting of burden with an overestimate on the required burden for activities 1 and 2. Even so, the impact on the overall burden estimates are small. The estimate of 440 engineering technical hours for activity 9 is created by halving the estimate for states. Finally, EPA assumes just 1 local agency may report aircraft emissions and that it would take the same number of hours as for a state to do so. Based on these assumptions, EPA estimates that local agencies and tribes could voluntarily spend between 462 and 714 additional hours (including manager hours) providing emissions data under the proposed AERR. Based on EPA’s calculations, EPA also expects that the average hours for voluntary activities by local and tribal agencies would be 490 hours.

Costs of Annual and Triennial Emissions Reporting and Associated Voluntary Activities

In addition to the hours per task as described by the tables above, EPA has computed the annualized average costs for SLTs to submit annual and triennial emissions data to EPA. Table 8a provides respondent annualized hours and costs for SLTs that use EIS rather than CAERS to collect point sources (cost reductions from CAERS are provided separately). This table includes operation and maintenance (O&M) costs for the point source data system as introduced previously in Table 6a. For point sources annual and triennial labor costs, Table 8a uses the EIS and CAERS cases 1 and 2 subtotals from Table 6b. For other data categories (i.e., nonpoint and mobile), the table relies on the average hours per entity as provided by Table 7a and Table 7b. To estimate annualized hours and costs for triennial activities, we divide the burden estimate by three to estimate the annualized burden spread over a 3-year period.

Table 8a also includes the number of entities for each activity. The EPA has used these values to compute average costs per SLT, which are provided in the table. In Table 8a, EPA has assumed that 56 out of a total of 84 agencies report point sources using CAERS cases 1 and 2, while 54 state and 30 local and tribal agencies report nonpoint and mobile sources. These assumptions are consistent with previous tables. The additional SLTs reporting via CAERS cases 3 and 4 are reflected in subsequent tables.

As shown in Table 8a, EPA estimates that the largest cost associated with this collection is the data system operations and maintenance (about $153K). This cost had not been included in previous ICRs for the AERR but has been occuring under the current AERR and is therefore not attributable to the proposed revisions. The EPA estimates additional annualized labor costs for required activities of about $63K for states ($5K + $7K + $51K) and about $38K ($5K + $7K + $20K) for local agencies and tribes.

For the resulting operations and maintenance costs, EPA attempted to verify the costs of SLT data collection systems and posed the question to a CAERS workgroup. Prior to the work done to make these cost estimates, EPA heard back from just a single state that their collection system costs ranged from $10K/year to $80K/ year, with an average of $55K per year. Based on this feedback, the estimated costs shown in Table 8aError! Reference source not found. of about $153K should be conservative. The data system cost estimates can be further revised in the final ICR based on any additional input provided by SLT agencies.

Table 8a additionally provides annualized costs for voluntary activities. The annual submission of HAP is reflected using information from Table 6b and 6c. The costs range from $15K for local agencies/tribes to $26K for states. Based on the expected number of states and local agencies to participate in voluntary activities in triennial years, EPA estimates an average annualized cost for voluntary activities of $22K for the 29 SLT agencies expected to submit data voluntarily.

Table 8a: Annualized Burden of NEI submission per Respondent for EIS Approach and CAERS Cases 1 and 2

Information Collection Activity	State, local, or tribal count	Manager Hrs/yr @ $118.94/Hr		Engineer Hrs/yr @ $90.83/Hr		IT Hrs/yr @ $81.02/Hr		Total Hours/ Year	Cost/ Year
Annual Required Activities
Point source data collection system operations and maintenance (see Table 6a)	56	160		440		1,160		1,760	$152,975
Submit annually reported point source CAPs with EIS or CAERS cases 1 or 2 (see Table 6b)	56	4		49				53	$4,950
Point Source Triennial Required Activities
Submit additional triennial point source CAPs with EIS or CAERS cases 1 or 2 (see Table 6b)	56	8.17		67.67				76	$7,117
Average Burden per Entity, Required Point Source Activities	56	172		557		1,160		1,889	$165,042
Other Triennial Required Activities
States: submit triennial nonpoint, onroad mobile, and nonroad mobile sources (Table 7a)	54	26.67		530.67				557	$51,370
Local agencies/tribes: nonpoint, onroad mobile, and nonroad mobile sources (Table 7c)	30	10.33		204.00				214	$19,757
Average Burden per Entity, Required Other Triennial Activities	84	21		414				435	$40,099
Triennial Voluntary Activities (hours from other tables divided by 3 to annualize)
State annual and triennial voluntary point source HAP reporting with EIS or CAERS cases 1 or 2 (see Table 6b and 7c)	56	5		50				55	$5,172
State voluntary triennial data reporting activities (see Table 7a)	20	13		264				277	$25,564
Local and tribal voluntary triennial data activities (See Table 7c)	9	8		156				163	$15,050
Average Burden per Entity, Triennial Voluntary Activities	56	11		169				181	$16,720

Table 8b provides the hours and cost burden reductions associated with SLTs using CAERS to submit point sources. These cost reductions are consistent with the difference between the EIS hours and CAERS hours provided in Table 6b. As shown in Table 8b, EPA estimates that SLTs implementing CAERS case 3 would save about $48K per year while SLTs choosing CAERS case 4 implementation would save about $147K per year. This significant difference between cases 3 and 4 results from the additional cost savings SLTs would realize under case 4 for eliminating the need to operate and maintain a point source emissions collection data system.

Table 8b: Annualized Burden Changes per Respondent of NEI Submission for CAERS Cases 3 and 4 Approach

Information Collection Activity	Manager Hrs/yr @ $118.94/Hr	Engineer Hrs/yr @ $90.83/Hr	IT Hrs/yr @ $81.02/Hr	Total Hours Change/ Year	Cost Change/ Year
CAERS Case 3 Burden Changes
Point source data collection system operations and maintenance (see Table 6a)	-43	-184	-244	-471	-$41,571
Annual point source CAP reporting (see Table 6b)	-1	-12.2		-13.2	-$1,227
Triennial point source CAP reporting (see Table 6b)	-2.0	-22.3		-24.4	-$2,270
State annual and triennial voluntary point source HAP reporting with CAERS case 3	-3.0	-29.5		-32.4	-$3,029
Subtotal Case 3	-49	-248	-244	-541	-$48,097
CAERS Case 4 Burden Changes
Point source data collection system operations and maintenance (see Table 6a)	-142	-280	-1,140	-1,562	-$134,682
Annual point source CAP reporting (see Table 6b)	-1.8	-36.2		-38.0	-$3,502
Triennial point source CAP reporting (see Table 6b)	-3.9	-46.3		-50.2	-$4,672
State annual and triennial voluntary point source HAP reporting with CAERS case 4	-3.6	-41.5		-45.1	-$4,198
Subtotal Case 4	-151	-404	-1,140	-1,695	-$147,054

EPA recognizes that many SLTs subcontract their point source emissions collection systems to a third party, while EPA’s cost estimation approach assumes the system is operated and maintained using in-house resources. However, EPA assumes that the cost of in-house systems are higher than outsourcing costs because SLTs are unlikely to outsource such a system unless costs would be reduced. Since EPA’s estimates for data system operations and maintenance in Table 6a, Table 8a, and Table 8b assume in-house systems only, we believe that we have not only included outsourcing costs, but may have overestimated such costs in this ICR. This approach would also potentially overestimate burden reduction associated with CAERS case 4.

4. Number of owners/operators responding

Various provisions of this proposed rule impact certain owners/operators, and to estimate the burden that the proposed requirements could have, EPA has estimated the number of facilities (not owners/operators) associated with activities that would be necessary if the proposed requirements were finalized. Table 2 has previously provided the data flows that are covered by this ICR and the associated relationships between states and owners/operators. It is necessary to use facilities to estimate this burden because much better information about facility counts is available than counts of the owners/operators of those facilities. Therefore, to estimate burden on owners/operators during the 3-year period of this ICR, it is necessary to estimate the following:

• The number of facilities that would be required to report annual total CAPs to SLTs under these proposed requirements; and

• Reporting emissions data from facilities to EPA:

  • The number of facilities outside states’ implementation planning authority that would be required to report CAP and HAP emissions under these proposed requirements (in 2026);

  • The number of facilities that would participate in a one-time collection from owners/operators for data related to High Electricity Demand Day (HEDD) events under these proposed requirements;

  • The number of rail companies from which EPA would continue to collect data about rail yards on a voluntary basis; and

  • The number of source test data reports that owners/operators would submit to EPA under these proposed requirements.

In addition to these estimates, Appendix A includes additional estimated numbers of facilities associated with proposed AERR provisions that would impact burden in 2027 and beyond.

1. Estimated number of facilities reporting emissions data to SLTs

To determine the number of facilities required to report to SLTs for the NEI, EPA has used the existing reporting information from SLTs to EPA and the estimated number of Major Title V sources from the previous AERR ICR. The design of the AERR point source reporting requirements is that the facilities that are required to report are these Title V Major sources plus any additional non-major sources that meet the 0.5 tpy actual emissions threshold for Pb emissions.

The total number of major sources required to report to states under this proposed action has been adjusted from that used in the previous ICR: 13,420, which includes all major sources available from the EPA Enforcement and Compliance History Online (ECHO) web application, plus an additional 12 sources that have 0.5 tpy of Pb or more and are not otherwise identified as Major Title V sources. Since this number was developed several years ago, and the total number of major sources tends to decrease over time, we believe this number is conservative. Since the ECHO database does not indicate whether the facility is a major source due to its CAP, HAP (or both), EPA has further refined this count to split out the CAP major (including CAP/HAP major) facilities from those that are only HAP major sources, which allows for better quantification of burden for the mandatory requirements versus burden for reporting that SLTs do voluntarily. Any facility that is not a CAP major source but is reported by the state is considered a voluntarily reported source.

To calculate the number of CAP major facilities, EPA performed additional analysis using the 2017 NEI,²¹ Integrated Compliance Information System for Air (ICIS-AIR),²² and a compilation of Residual Risk and Technology Review (RTR) data. Both designations of major sources as well as actual emissions in these databases were used. This approach further identified each NEI facility as best as possible regarding whether it is a CAP major, CAP/HAP major, or HAP major source. More information on this analysis is available in Section 2.1 of the Technical Support Document (TSD) for this proposal.²³ This approach identified 10,831 major sources, with 9,991 of these either CAP major or CAP/HAP major. To estimate the total number of CAP major for purposes of this ICR, EPA multiplied the 13,420 total major sources by the ratio of the 9,991 CAP major to the total 10,831 major. This approach resulted in an estimated 12,379 CAP major sources, which is the number used for this analysis for facilities that would be required to report CAPs to SLTs under the proposed rule.

The proposed AERR would continue to require fewer facilities to report for the 2024 and 2025 inventory years, using higher PTE emissions reporting thresholds and excluding Pb from the thresholds that require states to report point sources in those years. The EPA has chosen to use the same number of interim year facilities as was used in the previous ICR, which is 2,510 facilities. This origin of this number is described in Section 12.3.2. Across the 3 years covered by this ICR, the average number of facilities per year is (2 x 2,510 + 12,379) / 3 = 5,800.

2. Estimated number of facilities reporting emissions data to EPA

EPA estimated 4 values to quantify the possible reporting directly to EPA for the 3-year period covered by this ICR: (1) the number of facilities within Indian country potentially subject to a revised AERR, (2) the number of facilities that would participate in a one-time collection related to HEDD events, (3) the number of rail companies, and (4) the number of source test reports that EPA would expect to receive under these proposed requirements. The paragraphs below explain each of these separately.

To estimate the number of facilities within Indian country potentially subject to a revised AERR, EPA identified 85 major-source permits for facilities within Indian country from EPA databases, which reflects the complete list of such major sources. To estimate the number of non-major sources that could be potentially subject, EPA multiplied the count of 85 by the estimated number of non-major facilities expected nationally starting in 2027 (115,835) and divided by the total number of major facilities expected nationally (13,420). The calculation 85 x 115,835/ 13,420 yields an estimated 733 facilities, with a resulting total of 819 facilities.

As described in the preamble for the proposed AERR revisions, EPA proposes a “One-time Collection Option” that would require Curtailment Service Providers (CSPs) and other operators or aggregators of small generating units to report certain data to EPA. CSPs are entities that administer electricity demand response programs by working with companies that use and generate electricity to decrease electricity demand by deploying capacity from smaller units like backup generators that can reduce demand from the electricity grid. Reducing demand from the grid can involve deploying temporary electricity generation units that cause emissions and can impact air quality.

To estimate burden for the One-time Collection Option, EPA estimated the number of CSPs and similar entities. To do this, EPA first contacted the Federal Energy Regulatory Commission (FERC) to determine what data are available about the number of CSPs in the U.S. A FERC representative indicated that there is no national database of such entities. The Energy Information Administration does not require CSPs to file with FERC and the last voluntary survey available was in 2012 and, therefore, very outdated. FERC staff indicated that the best available data could be found from online lists for each of the regional transmission organizations (RTOs) and Independent System Operators (ISO). Figure 2 below provides a map of the RTO/ISOs.

Figure 2: RTO/ISOs and associated states.

Source: FERC, 2022 https://www.ferc.gov/power-sales-and-markets/rtos-and-isos.

While online lists may be incomplete according to FERC, EPA determined that no better data were readily available. In addition, for states that are not a part of an RTO or ISO, EPA reviewed an available list of demand response programs and assessed which of the programs listed could cause small unit generation that would need to be reported under the One-Time-Collection Option. If the same electric company ran a similar program in multiple states, that company was counted for each program rather than as a single company to help make the estimated number more conservative. Table 9 provides the list of RTO/ISOs and associated entity counts compiled from the sources shown, which results in an estimated 235 entities.

Table 9: List of RTO/ISOs and estimated number of respondents for the One-Time-Collection Option.

RTO/ISO/ State	Entity Count	Source
PJM	97	https://www.pjm.com/markets-and-operations/demand-response/csps
CAISO	31	http://www.caiso.com/documents/listofdemandresponseparticipants.pdf
ERCOT	18	“Demand_Response_Providers.xlsx” linked from https://www.ercot.com/services/programs/load
MISO	31	https://www.potomaceconomics.com/wp-content/uploads/2021/05/2020-MISO-SOM_Report_Body_Compiled_Final_rev-6-1-21.pdf (page 8)
SPP	0
NYISO	22	https://www.nyiso.com/documents/20142/1398619/Demand-Response-Providers-List.pdf/a9943929-edf6-4b5a-c16f-2c42bdebd18d
ISO-NE	0	https://www.iso-ne.com/markets-operations/markets/demand-resources/about
AL	2	https://www.energy.gov/eere/femp/demand-response-and-time-variable-pricing-programs-southeastern-and-midwestern-states
AR	4
FL	4
GA	2
IN	3
IA	2
KY	1
LA	1
MI	3
MS	1
MO	3
NC	1
OH	1
OK	1
SC	2
TN	2
WI	3
Total	235

Another voluntary aspect of the proposed AERR is participation by rail companies to provide data to EPA regarding rail yards. The EPA has worked with rail companies in past years and is aware of 7 rail companies that could participate. Thus, the number of rail companies used for the purposes of voluntary cost estimates for this ICR is 7.

3. Estimated number of facilities collecting release point latitude/longitude

On a one-time basis, certain facilities reporting under the proposed AERR would need to collect the latitude/longitude locations for each release point. Collecting such data would allow facilities outside states’ implementation planning authority to report such information in 2026 (for the 2025 inventory year) and in 2027 (for the 2026 inventory year). Even though much of the reporting of these data would occur starting after the three-year period of this ICR, the EPA assumes that the facilities would collect the latitude/longitude data for release points during the period covered by this ICR.

The EPA estimated the number of facilities per year starting with the total number of facilities expected to report, which is included in Appendix A: of this ICR based on an estimation approach described in the TSD for this proposal referenced above. The number estimated to need to report starting in 2027 is 129,255 facilities. The EPA adjusted this number downward by 8,309 facilities to account for the number of facilities for which states are already reporting release point latitude/longitudes to EPA via the states. The EPA derived this number by analyzing the 2020 NEI data to identify all facilities for which the reported latitude or longitude was 0.0005 degrees or more from the latitude or longitude (respectively) that represented the whole facility.

Since facilities must only collect this information once because release points generally do not move, EPA divided the resulting number of facilities (120,946) by 3 to calculate the number of facilities per year that would need to collect this information. This calculation gives an estimated 40,315 facilities per year that would need to collect release point latitude/longitude during the 2024-2026 period.

5. Burden on owners/operators

   1. Estimating burden of source testing

Finally, EPA has developed an approach to estimate the burden for reporting source test data. To calculate the number of hours for such reporting, EPA has used the formula:

Hours burden = N ´ T ´ H

Where, N is the number of facilities, T is the average number of tests per facility per year, and H is the average number of hours to prepare the electronic form to submit each test. Because major sources are those sources that would typically be required to perform tests, EPA used the same estimated number of major sources for required emissions reporting, or N = 13,420.

To estimate the number of tests per facility, EPA relied on information from selected states about their current source test collection, since source test data for state and federal purposes are collected and managed by states. The EPA contacted 9 states for input on how many source tests have been historically collected by states. Then, EPA compared the number of total source test reported by states to the number of major sources within those states. Since major sources often have testing requirements, it is reasonable to expect that the number of major sources might be a useful predictor of the number of source tests. Table 10 shows the raw data collected from the 6 states who replied with the number of major sources and source test counts.

Table 10: Number of source tests versus number of major sources provided by selected states.

State	Number source tests in 2020	Number of major sources
Illinois	450	557
Connecticut	131	56
Massachusetts	53	116
North Carolina	250	327
Washington (Island, Skagit and Whatcom Counties)	120	21
Maine	100	55
Texas	6938	870

Using these data, EPA evaluated the linear regression and determined that the coefficient of determination (R²) is 0.71 with a ratio of 5.35 tests per major source per year. The Texas test number seemed to be an outlier because it was much higher than all the other states compared to the number of major sources in Texas. Dropping the Texas data results in a linear regression, forced through the origin, with an R² of 0.93 and a ratio of 0.81 tests per major source per year, but this result had a significant underprediction bias at the low end of the data. Since neither linear regression was ideal, EPA took the midpoint between the 5.35 result and the 0.81 result, which gave 3.08 tests per facility per year. Based on this result, EPA used an estimated 3 tests per facility per year, or T= 3.

Finally, EPA polled several source testing experts within EPA, who have previous source testing experience for industrial contractors, regarding the number of hours it takes to complete a source test report and submit to CEDRI. The range of estimates received was from 2 to 6 hours. The EPA selected the midpoint of this range of 4 hours, or H = 4. The product of the number of tests per facility per year (3) and the number of hours per test (4) provides the estimate of 12 hours per facility that is included for activity 5 of Table 11 below.

2. Burden for owners/operators for emissions reporting

The burden for owners/operators to comply with the proposed AERR revision is driven by both mandatory and voluntary collections. For mandatory collections, burden is for owners/operators to report to SLTs so that SLTs can comply with the proposed AERR requirements for annual and triennial reporting requirements. During the 2024-2026 period of this burden estimate, the EPA expects that owners/operators would also need to collect latitude/longitude data for their release points, in anticipation of reporting these data in 2026 (for facilities outside of states’ implementation planning authority) and in 2027 for facilities within states. Additional burden from mandatory activities would include reporting of certain source test data, reporting emissions data to EPA for certain facilities located outside states’ implementation planning authority starting in 2026, and reporting needed if EPA finalizes the AERR proposal’s One-Time-Collection option for HEDD related data. For voluntary collections, burden would include activities by rail companies to provide rail yard data.

While different burdens exist for owners/operators reporting to a given state collection approach versus the approach from another state, EPA is unable to reflect those distinctions in this ICR because data are not available about burden from each of those systems. Similarly, we do not try to quantify the difference in facility burden for those states or local agencies who have adopted CAERS as their collection approach.

EPA has estimated burden for owners/operators to reply to report annual and seasonal emissions inventories in compliance with the proposed AERR, which includes both workflows to states as well as directly to EPA. Table 11 provides the estimated number of facilities and number of hours for each facility to respond to the data collection by a state. Although some owners/operators who operate multiple facilities may report those data centrally and have efficiencies that reduce the burden, these estimates assume that all facilities report individually.

For items 1 through 3 in Table 11, these hours cover reporting CAPs to states. Any time taken for HAP reporting for the 2023-2025 inventory years (covered by this ICR) result from state requirements and are not driven by AERR requirements. The number of hours included is for reporting emissions data only and includes the time that staff at facilities may need to spend to answer follow-up questions from the state. The time taken by facilities to collect necessary data (e.g., throughput, source testing) to comply with the reporting requirements is assumed to be a part of state permitting, compliance, and other requirements, which go beyond the scope of this ICR. Since most facilities reporting during the period covered by this ICR have been reporting emissions data for many years, this ICR does not include the additional hours associated with collecting facility attributes, except for release point latitude/longitudes that have not been previously required.

In Table 11, the hours shown are the estimated hours needed to accomplish the task within a single year (not the hours averaged over 3 years). To estimate an annual burden per facility even though different activities would occur within each year, Table 11 uses the average facility count over 3 years. Within each 3-year period, activity 1 occurs just once and activity 2 occurs twice. Thus, the average facility count shown reflects those frequencies (i.e., [12,379 + 2,510 + 2,510] / 3 = 5,800 facilities).

Since each row of Table 11 includes entities in separate categories, the hour estimates listed here are not cumulative in some cases. The respondents for activities 1 and 2 overlap, meaning some respondents do one or both of these activities. Thus, the total number of respondents for activities 1 and 2 are 12,379, with 2,510 of them expected to also perform activity 2. Thus, the range of hours for such facilities reporting to states is between 25 and 50 hours.

Table 11: Annual burden per facility for owners/operator reporting

		Annual Ave.		Hours per Facility
Activity	Facility Count in 1 Year	Facility Count Over 3 Years	Manager Hrs/yr @ $160.50/Hr			Engineer Hrs/yr @ $101.18/Hr			Total									Facility Cost/ Year
Required activities
1. Report annual CAPs by facility to states for use in triennial (2023) AERR report	12,379	5,800	1			24			25									$2,589
2. Report annual CAPs by facility to states for use in 2024 and 2025 AERR report	2,510		1			24			25									$2,589
3. Report annual CAPs and HAP to EPA by facility in 2026 (for facilities within Indian country that meet NAICS and reporting thresholds)	819	271	2			40			42									$4,368
4. Report source test data to EPA	13,420	13,420	0			12			12									$1,214
Required One-Time Activities
5. Gather release point latitude/longitude for future reporting	40,315	40,315	2.1			10.5			12.6									$1,399
Sub-total weighted average per year:	40,315	40,315	2.26			18.20			20.48									$2,206
Required activities for One-Time-Collection option for HEDD
6. Report facility attributes and daily fuel use or heat input for small generating units	235		10			120			130									$13,746
Voluntary activities for triennial inventory years
7. Provide rail yard data to the EPA for 2023 (in 2024)	7		2			10			12									$1,333

The entities performing activity 3 are expected to be different from those performing activities 1 and 2, since the activity 3 facilities are within Indian country and outside the scope of states’ implementation planning authority and are, therefore, not reporting data to states. The major sources reporting in activities 1 through 3 overlap with the same respondents performing activity 4. Thus, the 12 hours per respondent for activity 4 would be added to the total hours per respondents reporting to states (activities 1 and 2) and the respondent reporting to EPA (activity 3).

Activity 5 overlaps with the facilities performing activities 1 through 3 and includes additional facilities beyond those reporting activities. In most cases, the facilities included in activity 4 will also overlap with those included in activity 5. However, these activities are only performed one time and are shown in the table as occuring for one-third of the facilities each year, as described in Section 12.4.3.

Activities 6 and 7 are distinct from the other activities, so while some entities such as EGUs may need to perform activity 6 in addition to activities 1 through 3, those CSPs that are not electricity generators are additional entities not otherwise reporting under the proposed rule revisions and would only have the requirement for activity 6. Finally, since the rail companies are distinct from other types of entities reporting emissions, activity 7 is not expected to be cumulative with other activities.

13. Provide an estimate for the total annual cost burden to respondents or record keepers resulting from the collection of information. (Do not include the cost of any hour burden already reflected on the burden worksheet).

• The cost estimate should be split into two components: (a) a total capital and start-up cost component (annualized over its expected useful life) and (b) a total operation and maintenance and purchase of services component. The estimates should take into account costs associated with generating, maintaining, and disclosing or providing the information. Include descriptions of methods used to estimate major cost factors including system and technology acquisition, expected useful life of capital equipment, the discount rate(s), and the time period over which costs will be incurred. Capital and start-up costs include, among other items, preparations for collecting information such as purchasing computers and software; monitoring, sampling, drilling and testing equipment; and record storage facilities.

• If cost estimates are expected to vary widely, agencies should present ranges of cost burdens and explain the reasons for the variance. The cost of purchasing or contracting out information collections services should be a part of this cost burden estimate. In developing cost burden estimates, agencies may consult with a sample of respondents (fewer than 10), utilize the 60-day pre-OMB submission public comment process and use existing economic or regulatory impact analysis associated with the rulemaking containing the information collection, as appropriate.

• Generally, estimates should not include purchases of equipment or services, or portions thereof, made: (1) prior to October 1, 1995, (2) to achieve regulatory compliance with requirements not associated with the information collection, (3) for reasons other than to provide information or keep records for the government, or (4) as part of customary and usual business or private practices.

The total burden estimates for this proposed action are separated into two categories of respondents: SLTs and owners/operators. In each case, optional activities covered by this ICR are listed separately from mandatory activities.

1. State/local/tribal total annual cost burden

As described in previous sections, SLT burden includes burden for both required and voluntary activities associated with one-time tasks (Section 12.3.1)and annual and triennial tasks (Section 12.3.4). This section brings together both these burden estimates and includes capital and associated maintenance costs, which will provide annualized hours and costs for SLTs.

As previously described in Section 12.2, EPA forecasts that 29 SLTs, on average, will use CAERS during the period of this ICR, and the remaining 56 would use their existing reporting approaches. For the purposes of this ICR, EPA also has forecasted which CAERS cases SLTs may elect to adopt. To date, CAERS cases 3 and 4 have been of most interest to SLTs, presumably because these cases have the lowest burden estimates overall. Although case 4 has greater burden reductions than case 3, many SLTs seem to prefer the autonomy that case 3 provides (the states retain their back-end point source database). Based on these considerations, EPA assumes 10% of SLTs will select cases 1 or 2, 30% case 3, and 60% case 4. Starting with the average of 29 SLTs using CAERS, these percentages map to 3 SLTs using cases 1 or 2, 10 using case 3, and 19 using case 4.

The EPA has estimated annualized capital costs associated with workstations needed for SLTs to submit data required or voluntarily submitted based on the proposed requirements. The EPA assumes that each agency would require 5 workstations to comply with the reporting provisions of the AERR (1 for point sources, 1 for nonpoint sources, 1 for onroad and nonroad mobile, 1 for wildfires and prescribed fires, and 1 for managerial/coordination activities). The number of workstations has been assumed to be unaffected when states participate in CAERS because although data system maintenance is reduced or eliminated, agency staff still need a workstation to access CAERS to perform their data oversight and submission functions.

The cost for replacing a workstation including new basic software and peripherals, when replacement becomes necessary, is assumed to be approximately $1,500 per agency. For this ICR, it is assumed that 20 percent of the workstations will be replaced each year. Thus, the costs of replacement per agency would be:

5 workstations/agency x 20% replacement/year ´ $1,500/workstation = $1,500/agency/year

Cost of workstation replacement for all agencies equals: $1,500/replacement costs/year ´ 85 agencies/year = $127,500/year

Workstation maintenance costs are attributed to the normal maintenance of the workstations used to submit the required annual and triennial reports to EPA. This includes annual software costs, service costs, and warranty costs. It is assumed that the total cost of ownership over 5 years is four times the original purchase price, or $6,000. Thus, the annual maintenance costs are $6,000 minus the $1,500 capital cost divided by 5, or $4,500/5, which is $900/year per workstation. We conservatively assume (that is, more likely to overstate than understate) that one-third of the workstation annual maintenance cost can be attributed to the AERR. The resulting estimated costs associated with AERR are estimated to be approximately $300 per workstation per year, which is $1,500 per agency per year. Total maintenance costs for the respondents are estimated to be:

$1,500/agency/year x 85 agencies = $127,500/year.

As a result, the total capital and maintenance costs per year are $3000/agency/year.

Table 12 provides a summary of all costs that would be incurred by SLTs during the 2024-2026 based on the proposed action. The source of the data for each row is provided here:

• For the one-time required activities for prescribed burning, EPA assumes that 50 states and 2 territories would create a data system to collect that information. Local agencies within the state would use their state’s system. The per-state hours and costs for developing such a system are from Table 5.

• For the one-time required activities for point sources, the three summary rows for point source from Table 5 are all used: 59 SLTs are associated with hours and costs from the “EIS and CAERS cases 1 and 2” summary row, 16 SLTs are associated with the “case 3” summary row, and 10 SLTs are associated with the “case 4” summary row.

• For the annual required activities for point sources, the number of SLTs are the same as for the previous row, and the hours and costs are taken from the annual reporting estimates from Table 6a. For the EIS and CAERS cases 1 and 2 columns, the values are used as-is from the “submit annually reported point sources” row of Table 8a, whereas for the CAERS case 3 and case 4 columns, the appropriate burden reduction is subtracted from that using the values in Table 8b.

• For the triennial required activities for point sources, the calculations are made in the same way as for the annual required activities for point sources, but the triennial hours and costs are included from Table 8a.

• For the triennial required activities for other sources, the calculations are the same for EIS and all CAERS cases. The number of SLTs is the same as the previous row, but the average hours and costs are taken from the other triennial activities summary row of Table 8a.

• For the point data collection system O&M, the number of SLTs are the same as for the annual and triennial required activities for point sources, and the hours estimates are taken from the “point source data collection system operations and maintenance” row of Table 8a.

• For the capital and maintenance costs, the number of SLTs are the same as for the previous row, and the costs are $3,000 per entity as described earlier in this section.

• For the one-time voluntary activities, SLTs would update their reporting rule and apply to EPA only if they will be reporting HAP on behalf of owners/operators. For SLTs using EIS or CAERS cases 1 or 2, EPA assumes that the same fraction of states that currently reports HAP would take these voluntary steps. The EPA estimated this fraction as 88% using the current number of SLTs reporting HAP (75) divided by the total number of SLTs (85). Then, EPA multiplied this fraction by the 59 SLTs expected to use EIS or CAERS cases 1 or 2 to give 52 SLTs that EPA expects would continue to report directly to EIS (including CAERS cases 1 and 2). The EPA assumes that 4 (25%) of 16 SLTs using CAERS case 3 would report HAP on behalf of owners/operators and that 6 (19%) of the 32 SLTs using CAERS case 4 would do so. The hours and costs for these values are taken from the optional activities row of Table 5.

• Finally, for the triennial voluntary activities, the number of SLTs, hours, and costs are from the summary row of Table 8a for triennial voluntary activities. These values are not split out for CAERS cases 3 and 4.

Table 12: Annual Total SLT Burden and Cost by Activity

	EIS, CAERS Cases 1 and 2			CAERS Case 3					CAERS Case 4					Total
Information Collection Activity	No. of SLTs	Total Hours/ Year	Total Cost/ Year	No. of SLTs		Total Hours/ Year	Total Cost/ Year		No. of SLTs		Total Hours/ Year	Total Cost/ Year		Hours/ Year	Costs/ Year
One-Time Required, Prescribed Burning1	52	116,480	$10,398,053											116,480	$10,398,053
One-Time Required, Point Sources	37	23,877	$2,103,905	16		10,325	$909,797		32		6,101	$569,750		40,304	$3,583,452
Annual Required, Point Sources2	56	2,979	$277,199	10		400	$37,230		19		288.8	$27,512		3,668	$341,942
Triennial Required, Point Sources	56	4,247	$398,563	10		515	$48,469		19		486	$46,457		5,248	$493,489
Triennial Required, Other Sources	84	36,540	$3,368,347											36,540	$3,368,347
Labor Subtotal (Required)		184,123	$16,546,068			11,240	$995,496				6,877	$643,720		202,240	$18,185,283
Point Data Collection System O&M	56	98,560	$8,566,596	10		12,892	$1,114,039		19		3,762	$347,574		115,214	$10,028,208
Capital and Maintenance	56		$168,000	10			$30,000		19			$57,000			$255,000
Total (Required)		282,683	$25,280,663			24,132	$2,139,535				10,639	$1,048,293		317,454	$28,468,492
One-Time Voluntary: HAPs	49	47,106	$4,230,819	4		3,813	$342,495		6		5,720	$513,742		56,639	$5,087,056
One-Time Voluntary: CAERS	0	0	$0	16		21,589	$1,907,247		32		10,325	$964,193		31,915	$2,871,440
Annual and Triennial Voluntary1	56	10,136	$936,320	10		229	$21,425		19		194	$18,492		10,558	$976,237
Total Voluntary		57,242	$5,167,139			25,631	$2,271,167				16,239	$1,496,427		99,112	$8,934,733

¹ Costs associated with this activity are not broken out by CAERS cases. All costs are included with the group for EIS, CAERS cases 1 and 2.

² Excluding point source collection system O&M, included later in this table.

2. Owners/operators total annual cost burden

As described in Section 12.5, owners/operators burden includes burden for both required and voluntary activities. The required activities are reporting annual CAPs by facility to states and gathering release point locations and the costs additionally include reporting annual CAPs and HAP to EPA by facility for facilities within Indian country that are not reported by states, meet NAICS and reporting thresholds, and report source test data to EPA. Finally, the costs include the activities for a possible one-time collection of data for small electric generators, which is an option in the proposed AERR revisions.

Table 13 provides a summary of all costs that would be incurred by owners/operators during the period of this ICR based on the proposed action. The source of the numbers of facilities was previously provided in Section 12.4. While the number of facilities for any given year would vary, the average number of facilities from Table 11 has been used to properly calculate the annual total burden and costs.

Table 13: Annual Total Owner/Operator Burden and Cost by Activity

Information Collection Activity	Number of Facilities		Total Hours/Year	Total Cost/ Year
Required activities
1. Report annual CAPs by facility to states for use in triennial (2023) AERR report	5,800		144,993	$15,014,213
2. Report annual CAPs by facility to states for use in 2024 and 2025 AERR report
3. Report annual CAPs and HAP to EPA by facility in 2026 (for facilities within Indian country that meet NAICS and reporting thresholds).	273		11,466	$1,192,498
4. Report source test data to EPA	13,420		161,040	$16,293,705
Required one-time activities
5. Gather release point latitude-longitude for future reporting	40,315		507,973	$56,418297
Required Activities Sum	40,315		825,473	$88,918,714
Required activities for One-Time-Collection option for HEDD
6. Report facility attributes and daily fuel use or heat input for small generating units.	235		30,550	$3,230,402
Voluntary activities for triennial inventory years
7. Provide rail yard data to the EPA for 2023 (in 2024)	7		84	$9,330

14. Provide estimates of annualized costs to the Federal government. Also, provide a description of the method used to estimate cost, which should include quantification of hours, operational expenses (such as equipment, overhead, printing, and support staff), and any other expense that would not have been incurred without this collection of information. Agencies may also aggregate cost estimates from Items 12, 13, and 14 in a single table.

The EPA activities associated with the AERR include:

• Maintaining a database of emissions factors (e.g., WebFIRE) for use by states and the point sources regulated by states;

• Developing guidance and training materials for states for each emissions inventory reporting cycle and maintaining communication through EPA’s website and other methods, including providing in-person, webinar-based, and self-guided online training;

• Evaluating the adequacy of existing emissions estimation methods and models, developing method and model revisions, and publishing updated methods and models as appropriate;

• Preparing nonpoint emissions data for review and possible use by states;

• Preparing onroad and nonroad mobile model inputs for review and possible use by states;

• Preparing data for review of participating agencies, including landing and takeoff data at airports and fire activity data and emissions;

• Receiving, reviewing, and storing emission inventory data submitted by each state;

• Processing and updating data submitted by states, including performing quality assurance of data and coordinating efforts to resolve errors and anomalies;

• Fulfilling technical assistance and information requests;

• Developing technical documentation of the resulting emissions inventories created from compiling the collected data;

• Maintaining the EIS and associated electronic reporting approaches;

• Developing, operating, and maintaining the CAERS;

• Developing, operating, and maintaining the emissions estimation tool for small businesses; and

• Maintaining reporting codes to use in emissions inventory databases to identify various aspects of emissions inventories such as emissions unit types, release point types, source category classifications, and geopolitical entities.

The EPA’s costs that relate to this collection can be grouped into 7 areas:

1. Maintaining a database of emissions factors for use by states and the point sources regulated by states;

2. EIS annual operation and maintenance costs;

3. CAERS annual development, operation, and maintenance costs;

4. Preparing and providing guidance, plans, and training to states;

5. Revising emissions estimation methods and models to reflect the best available science, including mobile model updates related solely to support of AERR implementation;

6. Preparing draft nonpoint emissions and mobile-source model inputs;

7. Review, documentation, and publication of data; and

8. Information requests.

As of fiscal year 2022, the annual operation and maintenance costs for EPA’s efforts to maintain emissions factors in support of the NEI program is 2 FTE positions. No data system costs for the emissions factor program are included in this ICR because these costs are covered in the ICR for the CEDRI system.

As of fiscal year 2022, the EIS annual operation and maintenance costs are estimated to be: 2.7 FTE positions, $300,000 in Working Capital Funds and $625,000 for an information technology contractor.

As of fiscal 2022, the CAERS annual development, operation, and maintenance costs are estimated to be: 3 FTE positions and $1,200,000 for information technology contracting support. The EPA assumes an additional $300,000 in Working Capital Funds for capital costs associated with CAERS.

The projected estimated annual development cost for the emissions estimation tool, identified as part of reducing burden for small businesses, is estimated to be 0.5 FTE and $400,000 for data analysis (of emission factors for facility-wide emissions estimations) and information technology contracting support. No additional capital costs associated with this tool are included because EPA expects to build this tool as a module of CAERS.

The labor costs of preparing and providing guidance, plans and training to states is 1 FTE annually.

The labor costs of reviewing and revising emissions estimation methods and models to reflect the best available science for nonpoint emissions methods is 2 FTE annually.

The labor costs of preparing draft nonpoint emissions and mobile-source model inputs include the costs associated with developing updated emissions methods, overseeing contractor resources, quality assuring contractor results, developing documentation, and distributing data and draft documentation to states. The costs of reviewing data submitted by states include costs relating to data review, coordination of efforts to resolve any errors or anomalies, and updating of the data after the quality assurance and reconciliation assurance efforts have been completed. The costs associated with technical documentation include: compiling summaries of emissions, reviewing methods documents and notes, word processing, and section 508 compliance steps. For these activities, EPA requires approximately 1 FTE for point sources, 1.3 FTE for mobile sources, and 3 FTE for nonpoint sources to prepare draft data and review data submitted by states. In addition, the OAQPS requires 1 FTE for information requests. The EPA also incurs a $800,000 annual cost to have environmental engineering contractors assist with developing emissions methods, building data tools, and keeping input data current.

In addition to the primary roles within OAQPS, EPA Regional Offices annually use about 1 FTE in total across the 10 Regions to coordinate state efforts in making their submissions, quality reviews, and outreach and communication on behalf of the data collection program.

Thus, the total number of EPA FTEs is 18.5 (6.2 for the data systems and 12.3 for outreach, data methods, handling, and publication). Since most of the FTEs for this estimate work in Research Triangle Park, North Carolina, we used the pay rates from the General Services Administration (GSA) with locality adjustment for the Raleigh-Durham-Chapel Hill area. We conservatively estimated that the average EPA worker for these purposes is a GS-13, step 7 with a salary rate of $117,866 per year. In addition, a 26 percent increase in this amount was included to adjust for benefits paid by the government. The resulting annual FTE cost assumed is rounded to the nearest thousand dollars to $149,000. Thus, the total resulting EPA annual impact for 18.5 FTE is 37,440 hours and $2,682,000.

Table 14 summarizes the government costs along with the respondent costs from the previous sections. For SLT costs, the assumptions about SLT for CAERS usage are included, but voluntary activities including preparations for adopting HAP reporting requirements and reporting HAP voluntarily are not included. For owners/operators, the costs of both the required activities are included, but not the other optional costs or the costs of the voluntary activities for rail companies to provide data. The annual capital costs for EPA sum together the $300,000 each for EIS and CAERS.

Table 14: Total Estimated Respondent and EPA Burden and Cost Summary

Burden Element/Cost	SLTs	Owners/ Operators	EPA	Total
Number of Respondents	85	40,315		40,400
Total Hours Per year	202,240	825,473	38,480	1,066,192
Annual Capital Cost	$127,500	$0	$600,000	$727,500
Annual O&M Cost	$10,155,708	$0	$3,025,000	$13,180,708
Total Annual Capital and O&M Costs	$10,283,208	$0	$3,625,000	$13,908,208
Labor Cost Per Year	$18,185,283	$88,918,714	$2,756,500	$109,860,497
Total Cost Per Year	$28,468,492	$89,918,714	$6,381,500	$123,768,706

15. Explain the reasons for any program changes or adjustments reported on the burden worksheet (in hour or cost burden.)

As compared to the previous ICR for the AERR, this ICR covers substantially more activities, some of which are unrelated to the revisions being proposed to the AERR. In addition to the AERR revisions, this ICR has been modified to include the costs of emissions data systems that the states would likely run irrespective of this rule (to collect permit feeds) and the burden on owners/operators for reporting data to the SLTs so that SLTs can report to EPA as would be required under this proposed action. These additions make the ICR more complete but are not adding real-world burden to SLTs or owners/operators when compared to activities they are already doing. These additions are simply covering gaps in the previously approved AERR ICR (and the SIP implementation costs in Appendix C). Put another way, while the total estimated costs in Table 16 appear to be large, these figures do not simply reflect the costs that will be incurred due to the proposed revisions to the AERR. These figures also include costs that SLTs, Owners/Operators, and EPA are already incurring, and would continue to incur regardless of the proposed changes, by way of complying with existing laws and regulations (costs associated with complying with the existing AERR without the proposed changes).

In addition to the additional burden coverage described above, the proposed updates to the AERR would affect SLTs in ways that both add burden as well as providing opportunities to reduce burden. For owners/operators, the proposed changes add burden, but that burden can be offset to some degree by the choices that SLTs make regarding their use of CAERS. Additionally, some of the burden impacts would occur starting during the 2024-2026 period covered by this ICR while others would occur in the subsequent ICR period. The figure below illustrates the key elements of the revised AERR that impact burden and how SLT choices could impact burden for both SLTs and the owners/operators within each state, local, or tribal boundary.

SLTs must make the following critical choices under the proposed AERR provisions:

1. Whether to report HAP on behalf of owners/operators;

2. Whether and how to incorporate CAERS into SLT data flows for point sources.

For the choice of whether to report HAP on behalf of owners/operators, Table 15 provides the various scenarios for impact on the burden included in this ICR. If a SLT chooses to report HAP on behalf of owners/operators then, during the 2024-2026 period of this ICR, the SLT would have additional burden to implement the HAP reporting requirements previously described. During the subsequent ICR period (2027-2029), the SLT would have additional burden to collect and report HAP. If an SLT chooses to not to report HAP on behalf of owners/operators, then there would be no impact on the SLT during this ICR period but in the subsequent period, there could be an impact when an SLT chooses to receive HAP data from EPA. There is no significant burden impact of this choice on facilities, unless for SLTs that continue to require HAP to be reported to the state without integrating with CAERS or accepting the responsibility of reporting HAP on behalf of owners/operators. Table 15 reflects this last point in the footnote.

Table 15: Impacts on burden depending on SLT choice of whether to report HAP

Impacts to ICR	SLT Chooses HAP Reporting	SLT Chooses No HAP Reporting
2024-2026
For SLT Estimated Burden	Table 3a, Voluntary	None
For Owners/Operators Estimated Burden	None	None
2027-2029
For SLT Estimated Burden	Collect and report HAP data to EPA	Optionally receive HAP data from EPA via CAERS
For Owners/Operators Estimated Burden	Collect and report HAP data to SLT	Report HAP to EPA1; One-time increase to learn to use CAERS

¹ In this scenario, if an SLT were to retain their own HAP reporting requirements for reporting to the SLT, then an owner/operator would have duplicative reporting requirements to both SLT and EPA.

For the choice of whether and how to incorporate CAERS into SLT data flows for point sources, Table 16 provides the various scenarios for impact on the burden included in this ICR. Section 12.3 has previously described the various CAERS cases that SLTs can consider. During the 2024 to 2026 period, SLTs retaining their point source collection system or using CAERS cases 1 or 2 do not have impacts reflected in this ICR. If choosing cases 3 or 4, SLTs have a one-time burden increase associated with implementation (Table 3b) and once implemented a reduction in burden (Table 8b). The owners/operators burden is not different due to the SLT choice.

For the 2027-2029 period, Table 16 shows that SLTs choosing CAERS cases 3 or 4 continue to experience burden reductions during this period and would have additional burden reduction associated with lower implementation for supporting the HAP collection requirements that start in 2027. In all cases in this period, owners/operators would be reporting HAP, but the impact varies depending on the SLT choices for whether and how to incorporate CAERS. For SLTs retaining their existing system or implementing CAERS cases 1 or 2, it will be a lower burden for owners/operators if they also choose to report HAP on behalf of owners/operators (previous table). Not doing so could create a duplicative requirement for owners/operators when SLTs have their own HAP reporting requirements. Finally, the burden for owners/operators for SLTs that choose CAERS cases 3 or 4 would have a one-time increase to learn to use CAERS but then owners/operators would benefit from the consolidated reporting opportunities CAERS will provide.

Table 16: Impacts on burden depending on SLT choice of whether and how to incorporate CAERS

Impacts to ICR	SLT System, As-Is or CAERS Cases 1 or 2	CAERS Case 3	CAERS Case 4
2024-2026
For SLT Estimated Burden	None	Table 3b, case 3 Subtotal; Table 8b, case 3 reductions	Table 3b, case 4 Subtotal; Table 8b, case 4 reductions
For Owners/Operators Estimated Burden	None	None1	None1
2027-2029
For SLT Estimated Burden	None	Table 8b, case 3 reductions; Additional HAP burden reductions	Table 8b, case 3 reductions; Additional HAP burden reductions
For Owners/Operators Estimated Burden	Depends on SLT Choice for HAP Reporting Approach2	One-time increase to learn to use CAERS; Burden reduction for consolidated reporting

¹ Owner/Operators would need to learn how to report to CAERS, but that part of the burden is not included in the current ICR period.

² If states do not report HAP on behalf of owners/operators but continue to require HAP reporting to the state separately from CAERS, this would cause owners/operators duplicative reporting as also noted in Table 15.

In Sections 12 and 13, EPA has previously described the choices and assumptions made to forecast the choices of SLTs. Making different assumptions would significantly impact the overall burden comparison. For example, if more states choose case 4, then there would be more burden reduction associated with the proposed rule. If more states choose case 3, then there is more one-time burden for connecting the SLT data system, but also more burden reduction over time based on case 3.

In addition to the voluntary choice for mandatory HAP reporting, and the potential one-time burden increases and long-term burden reductions via CAERS, the following proposed AERR revisions for point sources would increase burden on states during the period of this ICR:

• Preparation for collecting additional data fields for point sources (for states not using CAERS case 4);

• Clarification on the definition of “actual emissions” (because some states may not be including startup and shutdown in their emissions reports);

• Requirement to separately report upset/malfunction emissions when they occur;

• Approach for reporting aircraft data as point sources, which codifies what many SLTs are already doing voluntarily;

• Approach for reporting rail yards, which codifies what many SLTs are already doing voluntarily;

• New approach for collecting and reporting data on portable sources (one of several options);

• Inclusion of portable offshore drilling barges in state waters; and

• Clarification that offshore oil rigs in state waters should be included in point source reports.

For sources other than point sources, the following proposed AERR revisions would increase burden on states during the period of this ICR:

• Preparation for the mandatory collection and reporting of prescribed fire activity data;

• Requirement to provide documentation of emissions for nonpoint sectors that are not covered by EPA tools;

• For states overlapping tribal regions for tribes that report to EPA, the proposed requirement that states exclude activity from those tribal regions when reporting county totals;

• For states who choose to report nonpoint source emissions for sectors with EPA tools, the additional effort to report emissions and documentation in addition to the newly required nonpoint tool inputs;

• For states who choose to report agricultural fire emissions, the additional effort required to report those as events rather than as county totals; and

• For California, the requirement to provide documentation of mobile source emissions calculations using California tools.

In addition to the opportunity to use CAERS case 4, some AERR proposed and retained revisions would provide opportunities to decrease SLT burden during the period of this ICR:

• Provision to collect HAP emissions data direction from owners/operators;

• EPA providing nonpoint emissions calculation tools for SLT use rather than requiring each SLT to develop and submit emissions with their own tools;

• The proposed provision to SLTs to review and accept nonpoint emission tool data provided by EPA;

• EPA providing mobile source model inputs for all state/local agencies except California, and the proposed provision to allow state/local agencies to review and accept mobile source model inputs provided by EPA for onroad and nonroad sources; and

• EPA providing activity data for and the proposed provision to allow state/local agencies to review and accept aircraft, rail yard, commercial marine vessel, wildfire, and agricultural fire activity data and emissions.

Table 17 provides a comparison from the summary information of Table 14 with the previous ICR for the AERR. Because this ICR includes reporting from owners/operators to both SLTs (for CAPs) and to EPA (source test data and owners/operators on tribal lands in 2026), the table provides the results separately for owners/operators and for SLTs.

Table 17: Burden Change

	Currently Approved ICR	Change	Total Requested
SLTs (includes SIPs)
Annual Responses	80	+5	85
Annual Respondent Hour Burden	48,702	+153,538	202,240
Annual Respondent Cost Burden	$4,960,908	$23,507,584	$28,468,492
Owners/Operators
Annual Responses	0	+40,315	40,315
Annual Respondent Hour Burden	0	+825,473	825,473
Annual Respondent Cost Burden	$0	$88,918,714	$88,918,714
EPA
All EPA Costs	$5,589,000	$792,500	$6,381,500

These changes show an average annual increase in the number of responses from 80 to 85 for SLTs and an associated hour increase of about 154,000 and cost increase of about $23.5 million. The reasons for the large increase in hours and costs have been described previously. The increase in the number of SLT respondents reflects the requirement that applies to the 50 states, the District of Columbia, and 3 territories (Puerto Rico, Virgin Islands, and Guam).

Additionally, increased labor rates are included in this ICR as compared with the existing approved ICR. Labor rates have been updated to the May 2021 labor rates (that are the most recent) from the U.S. Department of Labor, Bureau of Labor Statistics for managers and technical staff (downloaded on 3/21/2022).

As previously described, the costs associated with the proposed AERR include, for the first time for the AERR, costs to owners/operators for reporting to states, the cost of state data system operations, and includes source test data reporting to EPA. As such, certain apparent “increases” are due to the addition of those workflows as attributable to the AERR and to the additional cost associated with owners/operators collecting release point latitude/longitude in preparation for reporting that information starting in 2027.

EPA costs included in this ICR reflect an update to assumed salary of EPA FTEs to reflect the latest General Services Administration pay table. The additional cost of developing the emissions estimation tool to reduce burden on small businesses has also been added to EPA costs. An additional $300,000 has been included for CAERS capital costs in these estimates, to reflect the planned system migration as part of overall data system streamlining by EPA, which will incur a higher cost during this three-year ICR period for an eventual cost savings. Other costs have been recently updated in the AERR ICR approved in 2022, and those have been used in this ICR.

16. For collections of information whose results will be published, outline plans for tabulation and publication. Address any complex analytical techniques that will be used. Provide the time schedule for the entire project, including beginning and ending dates of the collection of information, completion of report, publication dates, and other actions.

This ICR collects data for the NEI, which is published on an annual basis. After states submit the data, EPA quality assures the point source data, resolves quality issues with the data submitters, and publishes the point sources in the EIS within 6-9 months. The remainder of the NEI data are published in the EIS and on EPA’s website within 15 months. The NEI is used in numerous EPA activities that are described in the latest NEI Technical Support Document available on EPA’s NEI website.²⁴

This ICR also collects data for CEDRI, which transfers the data collected into the WebFIRE system for publication. The data collected undergo a review period by SLTs that lasts 30 days after receipt for Periodic and Notification reports and 60 days after receipt for Performance Test / Evaluation reports. At that time, the data is transferred to the WebFIRE database for public distribution on the WebFIRE website.²⁵ More information is available on this process through the Central Data Exchange Guide for Reviewing Reports in CEDRI, Version 1.0 (April, 2020).²⁶

17. If seeking approval to not display the expiration date for OMB approval of the information collection, explain the reasons that display would be inappropriate.

Data under this action will be collected through electronic systems. Given that the expiration date of an ICR can change based on monthly extensions near the end of an ICR approval, EPA requests that the expiration date not be required to be displayed on the EIS, CAERS, and CEDRI data collection systems. The expiration date would be included in the user guides for both CAERS and use of CDX to submit to EIS and CEDRI.

18. Explain each exception to the topics of the certification statement identified in “Certification for Paperwork Reduction Act Submissions.”

There are no exceptions.

1. Expected costs of AERR proposed revisions starting in 2027

Some of the provisions of the proposed AERR revision, if finalized, would take effect starting in 2027 (for the 2026 emissions inventory year), and some one-time provisions in the 2024-2026 period would not apply. This Appendix provides information to help explain the changes to burden that would start in 2027 and continue indefinitely.

Table A-1 provides the proposed AERR provisions that would take effect for SLTs starting in 2027 and their associated annual changes in hour and cost burden. The EPA includes the following assumptions in these estimates:

• CAERS further expands for the 2026 reporting year that occurs in 2027 with 13 SLTs using their own data system or CAERS cases 1 or 2, 24 using case 3, and 48 using case 4.

• For activities 1a through 1c, manage and staff hours increase by 10% over the hours for the 2023-2025 period because of the additional HAP pollutants.

• For activity 1a, engineer and IT staff hours increase an additional 20% because of necessary updates to the SLT point collection system.

• For activity 1b, engineer and IT staff hours increase an additional 10% because of necessary updates to the SLT point collection system. This reduction would be associated with lower system burden because CAERS would serve as the user interface.

• For activity 2, this minor update would impact the staff engineer for supporting facilities to report additional data under the new requirement

• For activity 3, this update would include 5% managerial hours and the following breakdown of activities for staff

  • Prescribed fires collection system O&M: 40 hours engineering and 520 hours IT

  • User support for prescribed fire activity reporting: 120 hours engineering and 20 hours IT

  • QA of submitted data and revision support: 80 hours engineering

  • Converting data into required format: 8 hours engineering and 2 hours IT

  • Submitting final data to EPA via CDX: 4 hours engineering

  • Responding to follow-up questions from EPA: 20 hours engineering

• For activities 4a through 4c, the work to implement the HEDD collection would have occurred during the 2023-2026 period as part of the updates to state regulations and updating the SLT data system (except for CAERS case 4). Thus, these hours represent the minor additional burden associated with collecting data from a small number of additional units in each SLT.

• Activity 4b is expected to take more hours because of the additional time needed to register more facilities in the SLT data system. This is a one-time impact that would occur during 2027. Since ICRs cover 3-year periods, the additional hours over Activity 4a are divided by three. The EPA estimates that the additional hours (above activity 1a) that would occur in 2027 would be 12 management hours and 80 engineering hours.

• Activity 5 includes additional engineering hours (7) and IT hours (16) to update the SLT data system with the additional per- and polyfluoroalkyl substances (PFAS) pollutants that would be required. This potential impact would not affect SLTs choosing CAERS case 4.

Table A-1: Annual Burden and Additions per SLT Starting in 2027 for Proposed AERR Changes

	No. of SLTs	Manager Hrs/yr @ $118.94/Hr	Engineer Hrs/yr @ $90.83/Hr		IT Hrs/yr @ $81.02/Hr		Hours/ year per SLT		Cost/ year per SLT
Activities from with changes to burden calculation
1a. Point source annual emissions system O&M, collection, and reporting for required CAP and HAP by SLTs, with SLT system or CAERS cases 1 or 2	13	189	724		1,508		2,421		$210,446
1b. Point source annual emissions system O&M, collection, and reporting for required CAP and HAP by SLTs, with CAERS case 3	24	136	371		1,099		1,606		$138,852
1c. Point source annual emissions collection and reporting for required CAP and HAP by SLTs, with CAERS case 4	48	23	168		22		213		$19,772
Additional proposed activities
2. Provision to require inclusion of certain facility-dedicated mobile sources as part of facility emissions	85	0	40		0		40		$3,633
3. Provision for states to report activity data every year for certain prescribed burns	52	41	272		542		855		$73,493
4a. HEDD Preferred approach: States to report fuel data or heat input for for small generating units	24	1	8		0		9		$846
4b. HEDD Alternative D2: expand preferred approach to include all units deployed by CSPs	24	5	35		0		40		$3,743
4c. HEDD Alternative D3: restrict preferred approach to ozone SIP states	17	1	8		0		9		$846
5. Option: Include per- and polyfluoroalkyl substances (PFAS) in required pollutants	85	0	8		16		24		$2,023
Maximum Burden Changes per SLT (activities 1a, 2, 3, 4b, and 5)	13	235	1,079		2,066		3,380		$293,338
Maximum Burden Changes for CAERS case 3 (activities 1b, 2, 3, 4b, and 5)	24	182	725		1,657		2,564		$221,744
Maximum Burden Changes for CAERS case 4 (activities 1c, 2, 3, 4b, and 5)	48	69	523		580		1,172		$102,664

In Table A-1, the hours and cost estimates for activities 1a through 1c replace the analogous estimates in Table 12, while the remainder of the rows are estimated additional burden to the burden estimates form 2024-2026. The maximum burden changes totals in Table A-1 include activities 1a, 2, 3, 4b, and 5. The SLTs using CAERS case 3 would have a lower burden that includes activity 1b rather than 1a. The burden is further reduced for SLTs using CAERS case 4 by including activity 1c rather than 1a. The reasons for these lower burdens are both the reductions as described in the main body above as well as the lower increase in burden associated with mandatory HAP reporting.

Table A-2 provides the additional proposed AERR provisions that would take effect for owners/operators starting in 2027 and their associated annual additions of hour and cost burden. These burden increases are in addition to the reporting burden included in the main body of this document. The activity numbers in the table match the numbers used in Table A-2 and, therefore, are not sequential. The EPA includes the following assumptions in these estimates:

• The total number of facilities reporting is the number of major facilities plus the number of non-major facilities estimated to report as described in a separate document prepared for the Small Business Advocacy (SBAR) Panel. This approach is described in an attachment to the SBAR Panel convening materials “Attachment 4 – Draft AERR small business estimation method.docx”. After the panel as part of continued development of the AERR proposal, that method was applied to the final list of NAICS and final emissions thresholds proposed by this rule to derive the final estimated number of non-major facilities and small businesses using the CAA small business definition.

• The estimated number of states that will allow facilities to report HAP directly to EPA is 27. Of the remaining SLTs, 26 would continue to use the SLT system or adopt CAERS cases 1 or 2. The remaining 32 states would report HAP to EPA using CAERS cases 3 or 4.

• The number of facilities reporting under activity 1a is the total number of facilities times the number of SLTs collecting using the SLT data system or CAERS cases 1 or 2 (26) divided by the total number of SLTs reporting (85).

• The number of facilities reporting under activity 1b/c is the total number of facilities times the number of SLTs with EPA collecting or that are collecting using CAERS cases 3 or 4 divided by the total number of SLTs reporting (85).

• The number of facilities reporting PFAS is based on an Environmental Working Group report “PFAS Nation: Toxic Discharges Suspected From Almost 500 Industrial Facilities across U.S.” from June 11, 2019, and revised July 2021. The report indicates that more than 41,000 facilities may use or emit PFAS. The EPA chose to include a cost estimate based on about 10% or 4,000 facilities being subject to reporting PFAS. Given the lack of information about air emissions of PFAS, this number is highly uncertain, but since the additional hours to report PFAS are low, the uncertainty does not have a large impact on the overall burden estimates.

The maximum burden totals in Table A-2 include activities 1a, 4b, and 5. The owners/operators using CAERS case 3 or case 4 would have a lower burden through using the CAERS user interface. The burden reductions included here for owners/operators using CAERS result from the inclusion of quality controls during emissions reporting, which avoid submission errors and repeated report submittals. The EPA expects additional burden reduction from CAERS because it streamlines reporting across several other data systems, but those additional burden reductions are not quantified here.

Table A-2: Annual Burden per Facility Starting in 2027 for Proposed AERR Changes

	No. Facilities	Manager Hrs/yr @ $160.50/Hr	Engineer Hrs/yr @ $101.18/Hr	Hours/ year per Facility	Cost/year per Facility
1a. Collection by SLT for facilities required to report for HAP for SLT reporting on behalf of facilities because of new AERR	16,859	1	24	25	$2,589
1b/c. Collection of required annual HAP by EPA or SLT with CAERS cases 3 or 4 from owners/operators due to new AERR	93,372	1	16	17	$1,779
4a. HEDD Preferred approach: States to report fuel data or heat input for small generating units	235	5	60	65	$6,873
4b. HEDD Alternative D2: expand preferred approach to include all units deployed by CSPs	235	10	120	130	$13,746
4c. HEDD Alternative D3: restrict preferred approach to ozone SIP states	226	5	60	65	$6,873
5. Include PFAS in required pollutants	4,000	0	2	2	$202
6. CAERS training (one-time costs hrs / 3)	93,372	0	3	3	$304
7. Contractor support for small businesses with < 20 employees or <$3M receipts	19,024	-	-	-	$8,094
Maximum Burden per Facility (activities 1a + 5)	16,859	1	26	27	$2,791
Maximum Burden per Facility CAERS cases 3 and 4 (activities 1b/c + 5 + 6)	93,372	1	21	22	$2,285
Maximum Burden per Facility for CSPs (activity 4b)	235	10	120	130	$13,746
Weighted average across facilities (across 3 maxima and activity 7)	129,490	1	19	20	$3,225

¹ This number assumes that states would collect the data from CSPs rather than from individual facilities.

To better understand the expected impact of these changes in comparison to this ICR for the 2024-2026 period, Table A-3 provides the burden associated with labor hours for SLTs, owners/operators and the total for both 2024-2026 and for 2027. The table reflects the changes list in Tables A-1 and A-2, as well as the removal of the one-time costs occuring during 2024-2026. We calculated the SLT burden (hours and costs) in 2027 by:

• Starting with the total burden from Table 14;

• Subtracting the burden for “One-Time Required, Prescribed Burning,” “One-Time Required, Point Sources,” “Annual Required, Point Sources,” and “Triennial Required, Point Sources;” and

• Adding the burden from the three Maximum Burden Changes by SLT from the Bottom of Table A-1, using the appropriate number of SLTs for each of the rows to multiply the per-SLT burden for each.

The primary reasons for the changes to the SLT costs are the increases in burden shown in Table A-1 and decreases in burden for one-time activities for all states to update their regulations and to implement a prescribed burning data collection system. An additional decrease in burden is reflected because of the additional states expected to use CAERS cases 3 and 4 in 2027 (24 and 48, respectively as compared to 10 and 19 during the 2024-2026 period). Further estimates of CAERS adoption beyond 2027 have not been included. The costs of voluntary activities to update HAP regulations are not included in the final summary tables in Section 13, and thus the removal of those costs is not captured here as part of the changes in burden for 2027.

The primary reasons for the changes to the owner/operator costs are that there are many more facilities required to report starting in 2027 (about 129,500) plus a possible additional 235 CSPs. These numbers are in contrast to the CAP major sources (and few additional Pb emitters) required in 2024-2026 (12,379). Since the AERR for the 2024-2026 period does not require HAP, the collection of HAP by SLTs from facilities is not part of the ICR for that period. The costs during the 2024-2026 period associated with gathering release point latitude-longitude for reporting starting in 2027 would be dropped from the costs associated with these changes. With the requirement of collecting HAP emissions starting in 2027, all burden of owners/operators reporting to both states and directly to EPA becomes a part of the burden estimate. Although the increase in burden appears large, states are voluntarily collecting CAP and, in some cases, HAP from nearly 59,000 facilities (based on 2017 NEI data) in addition to the 12,379 required facilities. Thus, the practical impact on owners/operators is lower than what is captured here because the voluntary SLT collections include far more facilities than the required minimum. In other words, while this proposed action would require reporting for about 117,000 additional facilities beyond the current AERR, 59,000 of those facilities are already reporting to states.

Table A-3: Total Estimated Respondent and EPA Burden and Cost Summary Differences for 2027 Compared to the 2024-2026 Average.*

	SLTs			Owners/ Operators			Total
Burden Element	2024-2026	2027		2024-2026	2027		2024-2026	2027
Number of Respondents	85	85		40,315	129,490		40,400	129,575
Total Hours Per year	298,710	294,732		825,473	2,857,426		1,124,183	3,152,158
Cost Per Year	$28.5M	$27.7M		$88.9M	$450.1M		$117.4M	$477.9M

* 2021 dollars.

2. Collection instruments

The collection instruments for the proposed AERR revisions reflect the 2-party approach for collection. To collect data from SLTs, the current AERR relies on the EIS, submissions through the EPA Exchange Network, and the EIS Bridge Tool. The proposed AERR revisions would continue to rely on those from states. In addition, the proposed AERR revisions would rely on CAERS for collection from owners/operators that would in some cases report directly to EPA. CAERS is also a tool for states to use but would transmit data from CAERS to EIS with a simple mouse click. All of the illustrations of the collection instruments mentioned below are available as attachments to the docket entry for this ICR Supporting Statement.

SLTs often start their submission process to EIS using the EIS Bridge Tool, which is a Microsoft® Access ® tool developed by EPA. The attachment to this ICR Supporting Statement “Bridge Tool with screen shots.pdf” provides the user guide and the screen shots of the user interfaces.

When reporting to EPA, SLTs also use the Exchange Network Services Center. The website for this services center allows SLTs to upload data in predefined XML formats that would flow into EIS. The directions and user interface for submitting data in this way are available in the attachment “Exchange network services center.pdf.”

Also available to SLTs is a web client approach for submitting data to EIS, which also works via the EPA Exchange Network. The directions and user interfaces for this approach are available in the attachment “EIS Webclient.pdf.”

Finally, the currently available CAERS application can be used by facilities reporting to SLTs and for SLTs to report to EPA. As described in the preamble to the AERR proposed rule, the EPA would expand CAERS to allow for facilities to report directly to EPA. The user interfaces for doing so would largely reflect the user interfaces available in the latest version of CAERS, with amendments based on the final AERR. The available documentation on CAERS including images of the user interfaces is available in the attachment “CAER System User Guide v4 01062023 Final.pdf.”

3. Costs of emissions data activities for State Implementation Plans

In addition to the burden associated with the proposed AERR revisions, this appendix quantifies the burden of activities that states/locals must do to create emissions inventory data needed to comply with certain Clean Air Act requirements for SIPs. While the activities by SLTs to submit such information to EPA are covered by the ICRs associated with the various SIP requirements rules, the activities to develop the emissions data for SIPs have not been covered in the ICRs associated with those rules. This analysis quantifies emissions data preparation costs for those expected SIP emissions inventory preparation efforts over the three-year period covered by this ICR.

1. How, by whom, and for what purpose the information is to be used

The emissions data included in SIP submissions are required as part of various CAA provisions as described in Section 1 of the ICR. These data are necessary for states and EPA to characterize the sources of emissions that may be controlled to reduce pollutants and support a state in its efforts to attain a NAAQS for any nonattainment areas and to meet regional haze requirements.

2. Consequence to Federal program or policy activities if the collection is not conducted or is conducted less frequently, as well as any technical or legal obstacles to reducing burden.

The CAA requires states to submit SIPs, requires including emissions inventories in some SIPs, and requires air quality modeling for which emissions inventories are essential. These requirements are spread throughout CAA Part D. Without the collections included in this proposed action, states would be unable to comply with those requirements, or in the very least it would make compliance extremely difficult as each state would need to develop their own method for compliance. The burden included in this ICR associated with emissions data collected for SIPs primarily reflects existing burden imposed by the requirements of the CAA, and in-fact is very likely less of a burden on states when compared to a reality without the AERR.

3. Estimates of the hours and annual cost burden of the collection of information

The analysis provided in this section is divided into four subsections. The Section C.3.1 provides the approach for estimating the number of SLT respondents, while the second Section C.3.2 provides cost estimates for these. Section C.3.3 provides estimated costs of additional reporting by owners/operators to states to provide additional emissions inventory data needed for SIPs that is in addition to the annual data collected by states for reporting under the AERR. Finally, Section C.3.4 includes the total burden associated with these SIP-related emissions inventory activities for both SLTs and owners/operators.

1. Number of SLT respondents

The number of respondents overall includes all 50 states, but depending on the activity, some states may have more than one SIP action. The EPA estimates that during the 2024-2026 period, states will prepare 58 periodic ozone season emissions inventories, 28 projected attainment year inventories, 74 base year inventories for the nonattainment area (total for both ozone or PM_2.5 SIPs), 42 emissions inventories to support modeled attainment demonstrations, and 50 emissions inventories for regional haze modeling for the third planning period.

Table C-1 provides EPA’s estimates of the SIP actions expected occur during the 2024-2026 period. The left most column of the table shows the program name followed in the subsequent columns by the various emissions-related activities (and any relevant assumptions) and the number of areas affected. The column labeled “Count by” indicates the basis for estimating the “Number affected” on that row. These “Count by” labels are for the nonattainment area (“Area”), the combinations of nonattainment area and SLTs (“Area-SLT”), or the SLT. The type of “Count by” used affects how EPA computed the hours for the burden estimates. The “Emissions activities 2024-2026” column indicates whether during the 2024-2026 period covered by this analysis, EPA expects emissions activities to be performed by SLTs.

Table C-1: Evaluation of Potential SIP Activities

Program	Activity	Number affected	Count by	Emissions activities 2024-2026
Ozone NAAQS (all active)	Periodic Emissions Inventory	58	Area	Yes
	Projected 10-year inventory for maintenance plan	14	Area-SLT	Yes
	Base year inventory for the NAA (for inventory SIP or maintenance plan)	52	Area-SLT	Yes
	Emissions for modeled attainment demonstration (base year and projected attainment year)	38	Area-SLT	Yes
Ozone NAAQS (Transport)	No specific requirements, though upwind linked states may need to use emissions data.	23	SLT	Yes
PM2.5 (all active)	Projected 10-year inventory for maintenance plan or projected attainment year inventory	14	Area-SLT	Yes
	Base year inventory for the NAA (for inventory SIP or maintenance plan)	22	Area-SLT	Yes
	Emissions for modeled attainment demonstration (base year and projected attainment year)	4	Area-SLT	Yes
	Designations, 5-factor analysis including emissions evaluations	30	Area- SLT	No
	Infrastructure SIPs/Transport	51	SLT	No
NOx SIP call	Any activity	21	SLT	No
Regional Haze	Statewide emissions inventory needed for Regional Haze progress reports due in 2025.	51	SLT	No
	Statewide emissions inventory for the 3rd SIP planning period (due in 2028). Assume modeling inventory and emissions modeling activities needed.	51	SLT	Yes
CO	Base year inventory for the NAA (for maintenance plan)	82	Area	No
NO2	Base year inventory for the NAA (for maintenance plan)	1	Area	No
SO2	Emissions for AERMOD modeling	35	Area	No
Pb	Emissions for AERMOD modeling	1	Area	No

For the activities related to the ozone standards shown in this table, EPA reviewed possible actions that could occur during the 2024-2026 period based on the classification of nonattainment areas (Marginal, Moderate, Serious, Severe, or Extreme) and the latest information about ambient conditions in those areas. For the periodic emissions inventory, EPA assumed all 58 existing nonattainment areas would need to do a periodic inventory. These inventories are not identical to those submitted to EPA for the AERR’s triennial reporting requirement previously described, because the periodic inventories have several differences. Thus, EPA counts here only the additional effort (beyond what would be done for the AERR) needed for states with existing nonattainment areas to create the nonattainment area inventories.

For the ozone NAAQS projected 10-year inventory for maintenance plans, EPA has assumed the following numbers of state-area (or “area-SLT”) combinations to arrive at the count of 14 shown in the table: 2008 ozone standard (3), 2015 ozone standard (11). A state-area combination reflects that a single nonattainment or maintenance area may overlap with 3 states, for example, the New York-New Jersey-Long Island nonattainment area overlaps with New York, New Jersey, and Connecticut. This is consistent with EPA’s assumption that all 58 existing ozone NAAQS nonattainment areas would need to do a periodic inventory. Since each state would separately need to do emissions projections, EPA has counted these separately for purposes of estimating burden.

In addition, for ozone, EPA arrived at a count of 52 new or updated base year inventories for the nonattainment areas, as follows. Each of the following occurrences would trigger a state to need to prepare such an inventory: 2008 ozone standard areas reclassified from Serious to Severe (7), 2015 standard areas reclassified from Marginal to Moderate (5) and from Marginal or Moderate to Serious (31), and areas needing maintenance plans (9). Lastly for the ozone standard, a reclassification from Moderate to Severe would cause an area/state to create a new modeled attainment demonstration, so EPA has counted a possible 38 such demonstrations requiring modeling inventories during the period of this ICR.

For ozone transport SIPs, there are not necessarily any specific submissions from states that EPA believes would occur during the 2024-2026 period. However, states may need to develop emissions inventories (or revise EPA inventories) during the period so that they could meet requirements for submissions after 2026. Thus, EPA assumes 23 upwind states based on its latest proposed rulemaking would have emissions inventory activities during the period of the RIA.

For the PM_2.5 standards, EPA reviewed possible actions that could occur during the period for this ICR based on the classification of nonattainment areas (i.e., Moderate or Serious). In the case of PM_2.5 SIPs, projected inventories could be needed either for a 10-year inventory for a maintenance plan or for a projected attainment year inventory for a Moderate or Serious SIP. The EPA anticipates that the following could occur (during the 2024-2026 period) resulting in 14 future-year projected inventories: maintenance plans for the 1997 PM_2.5 standard (3) and for the 2006 PM_2.5 standard (7), 2012 PM_2.5 standard areas bumped from Moderate to Serious (2), and 2012 PM_2.5 standard Serious area attainment plans due (2). For new base year inventories for the nonattainment areas, EPA anticipates a total of 22 of these as follows: for all the cases just described for PM_2.5 SIP projected attainment inventories above (14) and for second maintenance plans (8).

In addition to the ozone and PM_2.5 standards where most of the activities would occur during the 2024-2026 period, EPA considered whether any state activities beyond the AERR reporting activities would be necessary for the NOX SIP call (40 CFR 51, Subpart G) and determined that there are not. For Regional Haze SIPs, the inventory work needed for the 2025 progress reports would generally have been completed prior to 2024, but the emissions inventory work for the third planning period will likely occur in part during 2025 and 2026. The EPA also considered whether any emissions-related SIP work would be necessary for CO or NO₂ maintenance areas and concluded none would occur during the 2024-2026 analysis period. Finally, for SO₂ and Lead nonattainment areas, while some activities related to emissions may occur, EPA assumes that no additional emissions activities beyond what SLTs do for annual and triennial emissions reporting and permitting efforts already covered by ICRs would be needed. This is in part because emissions levels used in modeling for these SIPs are usually PTE levels, the values for which are available in permits and collected via other ICRs as previously described.

The EPA additionally has assumed that no further emissions inventory preparation burden exists for NO₂, SO₂, or Lead SIP development. This assumption has been made because SIPs for these pollutants use source-specific analyses that require detailed PTE emissions and facility emissions release parameters. The PTE information is included in facility permits and its collection is, therefore, covered by the 40 CFR Part 70 State Operating Permit Program (EPA ICR Number 1587.15, OMB Control Number 2060-0243) and the 40 CFR Part 71 Federal Operating Program (EPA ICR Number 1713.13, OMB Control Number 2060-0336). The emissions release parameters for those facilities needed for completing these SIPs are covered by the annual emissions inventory collection, for which the burden is included in this analysis. Finally, EPA has identified that no CO maintenance plans or SIPs are due during the 2024-2026 period of this analysis.

2. SLT burden for emissions-related SIP requirements

To estimate costs associated with SIP emissions inventory data preparation, this methodology uses three steps. First, as listed in Table C-1, EPA estimated the number of states or local agencies that will need to create emissions data for each emissions-related SIP activity occurring during the 2024-2026 analysis period (e.g., for the ozone program this would be the number of agencies with nonattainment planning obligations; for the regional haze program, this would be each state air agency). Second, EPA estimated the number of hours associated with creating the emissions data that are beyond hours associated with emissions work for annual and triennial submissions under the AERR. Third, EPA multiplied the number of entities and the number of hours to get the total hours, including hourly cost information in those calculations to also get annualized costs.

Based on the standards and activities listed in Table C-2, EPA grouped activities across the air quality standards and aggregated the areas affected. The "Number affected" column in Table C-2 provides the totals of affected nonattainment areas and/or states. The other columns in that table provide EPA’s estimated number of additional annual hours beyond annual/triennial reporting to perform the emissions-related steps for SIPs. A party performing these activities would, at most, perform each activity listed only once during the 3-year analysis period. So, to provide the annualized burden per entity, EPA divided the total number of hours for each activity by 3. The EPA assumed that managerial hours for these tasks were 5 percent of the technical hours. For preparation of periodic ozone season emissions, EPA calculated that there are on average 1.3 states associated with each ozone nonattainment area. Since the row included periodic inventories counts by area, the hours were set to be 30 percent higher in this row than for similar activities in the subsequent two rows.

Table C-2: Annualized Burden per SLT respondent for SIP emissions activities

Activity	Number affected	Count By	Manager Hrs/yr @ $118.94/Hr	Engineer Hrs/yr @ $90.83/Hr	Total Hours/ Year	Labor Cost/ Year
1. SLT prepare periodic ozone season emissions for point, nonpoint, mobile, and events	58	Area	10.67	208	219	$20,160
2. SLT prepare any projected year NAA emissions for point, nonpoint, and mobile	28	Area and State	8.67	173.33	182	$16,774
3. SLT prepare any base year inventory for the NAA	74	Area and State	17.33	346.67	364	$33,548
4. SLT prepare emissions for any modeled attainment demonstration (includes future emissions and processing)	42	Area and State	8.67	173.33	182	$16,774
5. SLT prepare emissions for regional haze modeling for 3rd planning period	51	State	17.33	346.67	364	$33,548
Average Burden per State	51	State	67	1,324	1,391	$128,175

To calculate an average value across entities, EPA chose to make this calculation by state (including the District of Columbia), even though some of the burden is also shared by local agencies and tribes. This averaging approach allows for comparison between this analysis and subsequent analyses that would be needed for future SIP-related ICRs, for which the number of estimated entities per activity may change. The EPA recognizes that all emissions-related activities do not apply to all states and that some such activities are performed by local agencies and tribes. The total burden associated with the rule will not be affected by this calculation, but rather, we provide this state average only for comparison purposes with other estimates in this analysis and with subsequent ICRs.

3. Burden on owners/operators related to emissions inventories for SIPs

To estimate burden on owners/operators that is related to emissions inventory submissions in support of SIPs, EPA has estimated the number of affected facilities and the number of hours to report seasonal emissions needed for SIPs that go beyond the effort for annual reporting associated with the AERR (and covered in the primary RIA). These efforts are to report the seasonal data, such as ozone-season-day emissions.

To estimate the number of facilities affected, EPA has estimated the number of facilities that SLTs collect to comply with requirements to create inventories for nonattainment areas for Ozone SIPs and PM_2.5 SIPs. Only those standards are included because reporting the season-day emissions (that can be used for emissions inventories for those standards) would impose additional burden on facilities and states to estimate those emissions. To estimate the number of facilities, EPA analyzed NEI facilities within nonattainment areas. Facilities were counted as being directly or indirectly required to report for SIP reasons if any of the following were true:

1. The facility is listed as a “CAP Major” or “HAP/CAP Major”.

2. The facility has emissions of 5 tons or more of NOx or VOC.

3. The facility had 10x or more the minimum value of NOx or VOC of any Major facilities within the area. This step is done because NOX and VOC emissions reporting thresholds are based on PTE rather than actual emissions. The minimum values for NOX and VOC for all Major facilities within any given area are usually much less than 1 tpy, so this approach is conservative.

Based on this approach, EPA estimated 8,947 facilities were required to report based on nonattainment area emission inventory requirements. This value was rounded to 8,950 for the burden calculations made for this proposed rule.

To estimate the number of hours, EPA considered that the bulk of emissions reporting burden is associated with reporting annual emissions to states. The incremental burden would be low, and EPA assumed this would be 15 percent of the annual reporting burden (from activity 1, Table 11), or 4 hours per facility for an engineer and 0 hours for a manager once within each three-year period.

4. Total estimated costs per year of emissions inventories for SIPs

Based on the information in the previous sections, Table C-3 provides the resulting annual costs associated with creating the emissions inventories for SIPs. The estimated total costs to SLTs are about $6.5 million per year while costs to owners/operators are about $1.2 million per year. As with the other annual costs in this RIA, these costs are in 2021 dollars.

Table C-3: Total Annual Cost of Emissions Inventory Preparation Activities for SIPs

Activity	Average Number Entities Per Year	Total Hours Per Year	Total Costs Per Year
SLTs collect required SIP emissions data	51	70,941	$6,536,925
Owners/operators report season-day CAPs by facilities within nonattainment areas to states	2,983	11,933	$1,207,391

1 Thirty-one local and tribal agencies reported in 2017 and 23 in 2018. 26 is a weighted average with triennial years assuming 31 and interim years assume 23 (i.e., [31 + 23 + 23]/3 rounds up to 26).

2 Except where noted, future reference to “state(s)” in this supporting statement includes 50 states, 3 territories, 31 local and tribal agencies, and the District of Columbia (DC). As prescribed by the Tribal Authority Rule (63 FR 7253, February 12, 1998), codified at 40 CFR part 49, subpart A, tribes may elect to seek Treatment as State (TAS) status under the Clean Air Act and obtain approval to implement rules such as the AERR through a Tribal Implementation Plan.

3 https://www.epa.gov/AirToxScreen

4 See https://www.epa.gov/office-inspector-general/report-epa-can-improve-emissions-factors-development-and-management.

5 See https://www.epa.gov/air-emissions-factors-and-quantification/documentation-supporting-draft-and-final-emissions-factors.

6 https://www.epa.gov/air-emissions-inventories/emissions-inventory-system-eis-users-manual-and-how-tos

7 https://www.epa.gov/air-emissions-inventories/emissions-inventory-system-eis-users-manual-and-how-tos

8 https://eis.epa.gov/userguide/index.html

9 https://www.epa.gov/combined-air-emissions-reporting/combined-air-emissions-reporting-system-caers

10 https://www.epa.gov/electronic-reporting-air-emissions/cedri#guide

11 E-Enterprise for the Environment is a collaborative partnership through which EPA, states, and tribes work together to improve the way we protect the environment and human health. See also https://www.epa.gov/e-enterprise.

12 This regulation is part of a set of regulations that have been incorporated into federal implementation plans for 39 Indian reservations for those three states. The set of regulations is known as the Federal Air Rules for Reservations (FARR) in Idaho, Oregon, and Washington. The EPA has proposed revisions to the FARR on October 12, 2022 (87 FR 61870).

13 North American Industry Classification System code.

14 See https://www.epa.gov/office-inspector-general/report-epa-can-improve-emissions-factors-development-and-management.

15 As prescribed by the Tribal Authority Rule (63 FR 7253, February 12, 1998), codified at 40 CFR part 49, Subpart A, tribes may elect to seek treatment in the same manner as a state (TAS) status and obtain approval to implement rules such as the AERR through a Tribal Implementation Plan (TIP), but tribes are under no obligation to do so. However, those tribes that have obtained TAS status for this purpose are subject to the Subpart A requirements to the extent allowed in their TIP. Accordingly, to the extent a tribal government has applied for and received TAS status for air quality control purposes and is subject to the Subpart A requirements under its TIP, the use of the term state(s) in Subpart A shall include that tribe.

16 This smaller set of sources has been referenced as “Type A" sources in previous versions of the AERR, as well as the current version. However, EPA is proposing to eliminate the Type A and Type B terminology because the proposed revisions would require point sources to report every year starting with the 2026 inventory year.

17 This analysis was repeated with 2017 NEI data and only 6 such facilities were identified, but the difference is so small we have retained the facility count based on 2014. 2020 NEI data were not available when this ICR was prepared.

18 As described in Section 12.4.1, the revised estimate for CAP major facilities based on 2017 NEI and additional data sources is 12,379.

19 The 2020 NEI Wagon Wheel tool and templates are available on EPA’s website at https://www.epa.gov/air-emissions-inventories/2020-nei-supporting-data-and-summaries.

20 This assumption can be revised for the final ICR because more information will be available based on final template submissions for the 2020 NEI.

21 2017 National Emissions Inventory (NEI) Data, U.S. EPA, https://www.epa.gov/air-emissions-inventories/2017-national-emissions-inventory-nei-data.

22 Integrated Compliance Information System for Air (ICIS-AIR), U.S., EPA, https://www.epa.gov/enviro/icis-air-search.

23 U.S. EPA, Technical Support Document: Revisions to the Air Emissions Reporting Requirements (Proposal), June 2023, Docket number EPA-HQ-OAR-2004-0489-0091.

24 https://www.epa.gov/air-emissions-inventories/national-emissions-inventory-nei.

25 https://cfpub.epa.gov/webfire/.

26 https://dev.epacdx.net/FAQ/ViewDocument?documentNumber=Phx8CgcKgTs%3D.

17