2566ss01PartB_June_20_2018

Supporting Statement B for Information Collection Request

EPA ICR No. 2566.01, OMB Control No. 2050-New

1. SURVEY OBJECTIVES, KEY VARIABLES, AND OTHER PRELIMINARIES

(a) Survey Objectives

The Agency is proposing to administer a voluntary survey to states, tribes, and territories to collect this information. The information is intended to assist in estimating the universe of facilities that could potentially be subject to discharge prevention regulations for hazardous substances listed at 40 CFR part 116.4 (“CWA HS”). This information is intended to assist EPA in determining and refining appropriate regulatory approaches to prevent discharges of CWA HS to jurisdictional waters, as well as estimating the scale and potential impacts of any regulatory action taken by the Agency.

EPA intends to request information collected by states under the authority of EPCRA section 312. Facilities must report the maximum and average daily amounts of hazardous chemicals onsite during the preceding year to the respective state, Tribal or territorial authority (EPCRA does not require this information to be reported to EPA). The agency is also interested in reported impacts of discharges of CWA HS to surface waters (e.g., fish kills, drinking water) over the past 10 years, as well as existing programs in place to help prevent and mitigate the impact of CWA HS to surface waters. This information is maintained by states, territories and tribes, but is not collected by EPA for its use in regulatory programs, nor is it available in publicly accessible formats.

(b) Key Variables

Key information includes number of Tier II facilities with CWA HS, types and amounts of hazardous substances produced, stored or used at Tier II facilities, potential impacts of discharged CWA HS to drinking water, and the environment (e.g., to downstream drinking water intakes, fish kills), and aspects of existing state, territorial or tribal prevention and mitigation programs specifically directed at CWA HS.

(c) Statistical Approach

The potential respondent universe is 87 state, local, and tribal entities. The agency chose a census based approach for the collection. Respondents to this voluntary ICR are state, territorial, and tribal government agencies with Emergency Response Commission duties (e.g., State Emergency Response Commission [SERCs], Tribal Emergency Response Commissions [TERCs]), as well as sister agencies within the respective jurisdictions that may have additional information such as fish kills or drinking water impacts. The state SERC staff identified by EPA Regional liaisons will be the agency’s primary point of contact (POC). EPA will assist state POCs in identifying other state and tribal agencies that may have data that would assist in responding to this survey. Examples of other agencies that may assist in responding to this ICR include:

• Department of Natural Resources—e.g., fish kill investigations

• Department of Environmental Quality— e.g., drinking water alerts, fish kill investigations

• Department of Environmental Health— e.g., human health impacts; drinking water alerts or shutdowns

EPA contacted a random sample of nine states that have collection authority for Emergency Planning and Community Right to Know Act (EPCRA) Tier II (“Tier II”) facility data under Section 312. Seven of the nine states chose to participate in the consultation (listed below) and were given the proposed information collection request instrument.

1. California (R9) 2. Connecticut (R2) 3. Delaware (R3)

4. Massachusetts (R1) 5. Maryland (R3) 6. North Dakota (R8)

7.Texas (R6)

The states provided either per question burden estimates for survey, or actual data that was responsive to the survey questions.

The Agency expects to use contract support for data collection, tabulation and preliminary analyses of the data and information collected pursuant to this request. Abt Associates (10 Fawcett Street, Cambridge, MA 02138) has provided support on collection and analysis of information and baseline data related to CWA hazardous substances and their impacts in publicly available databases (e.g., National Response Center). This data and information was used to inform the draft rulemaking.

(d) Feasibility

During the consultation, EPA discussed potential obstacles related to responding to the information collection and impacts on estimated burden. Respondents identified several issues that may hinder the ability to respond to the request in a timely manner.

i. Available resources – the ability to respond in a timely manner was identified as a potential obstacle. EPA provided up to 45 days from the response to the consultation, a 15-day extension of the original 30-day timeline. This facilitated responses from 2 of the 5 states that needed extra time. Therefore, we anticipate providing at least 45 days to respond to the ICR.

ii. Level of effort for data reduction to provide EPA with only the data requested. Several states were concerned about the man-hours required to sort data for responding to the survey, particularly questions 1-3. They stated that they would likely send the unsorted parent database and rely on EPA to perform the data reduction to access the data necessary for the response. EPA indicated that would be acceptable, and had included that as an option on the survey. Based on that discussion respondents consulted indicated they would rely on EPA for data reduction to respond to survey questions 1-4. Question 1 had a 57% (4/7 respondents) preference for sending an unsorted database; question 2 had an 85% (6/7) respondent preference rate; question 3 had a 43% (3/7) respondent preference rate; and question 4 had a 71% (5/7) respondent preference rate. Four of the seven states provided unsorted databases that were responsive to one or more survey questions during the consultation process.

2. SURVEY DESIGN

(a) Target Population and Coverage

The target population for this ICR is composed of state, territorial and tribal agencies (SERCs and TERCs) that have collection authority for EPCRA Tier II facilities that produce, store or use listed CWA hazardous substances. In addition, state agencies that collect information on the impacts of discharged CWA hazardous substances to drinking water and the environment (e.g., to downstream drinking water intakes, fish kills) are targeted by this survey. This survey is intended to census the respondent population.

(b) Sample Design: Not applicable since survey will be distributed to 100% of the potential respondent population.

(c) Precision Requirements: Not applicable since a statistically-based sampling design was not applicable.

(d) Questionnaire Design

The objective of the survey is to collect information to be used to support a regulatory action for CWA HS spill prevention under CWA 311(j)(1)(C). EPA will use the information to refine an estimate of the number of potentially regulated facilities and to more accurately estimate the potential costs and benefits of a regulatory action. EPA also intends to identify and evaluate existing state requirements to prevent discharges of hazardous substances to surface waters, to identify facilities/equipment subject to these state requirements and to help inform any regulatory action.

The information requested in the ICR is intended to help accomplish three objectives:

1. Develop an estimate of the universe of potentially affected facilities;

2. Develop an assessment of recent CWA hazardous substance discharge history and impacts to health and the environment; and

3. Develop an assessment of existing state spill prevention programs.

1. Questions 1-2 request facility and chemical information (private and government-owned facilities) available in state Tier II reporting databases.

2. Questions 3-6 request information regarding CWA hazardous substances discharge occurrence and impacts (i.e., drinking water supply and fish kills) and will assist the agency in developing impact analyses.

3. Questions 7-8 request spill prevention and existing state program information which will assist the agency in refinement of cost/benefit and existing regulatory coverage analyses.

EPA selected open-ended questions requesting data and information because of the uncertainty and variability in the format in which this information is maintained by the various state agencies. The Agency also provides flexibility to respondents regarding the content of the responses. As part of survey question 1, EPA solicits the portion of the database containing the relevant information, or the entire database, from which EPA will extract the relevant information. This provides the respondent flexibility in the content that will be provided, and reduces respondent burden if EPA filters and extracts the relevant information.

(e) Addressing Potential Nonresponse Bias

To avoid nonresponse bias, EPA designed a short survey (less than 10 questions) that requests information that is already be collected by states, tribes and territories. EPA also ensured that the survey design included flexibility in how respondents can submit information to minimize the burden to the respondents (e.g., respondents can submit existing databases).

Further, EPA gathered information on state contacts from EPA regional personnel to ensure that the survey is sent to the correct individual in each state that can respond to questions on EPCRA Tier II information. EPA also gathered contact information on other agencies within each state that may have information on CWA HS discharges (and their impacts) to minimize the level of effort required to identify appropriate state contacts within sister agencies. By identifying state contacts in advance and coordinating with individuals with which EPA has an existing relationship, EPA expects to improve the response rate for the survey.

To address uncertainty introduced by nonresponse, EPA will compare the information from the voluntary surveys to information collected in publicly available databases (e.g., NRC database) to estimates of facility universe, number of CWA HS discharges and impacts of those discharges.

Universe of Respondents. To estimate the universe of facilities potentially affected by a regulatory proposal EPA intends to request EPCRA Tier II data to identify facilities that produce, store, or use CWA HS. One approach to address nonresponse would be to compare surrogate data from similar sized/population states to estimate the number of facilities, perhaps by using manufacturing output data for industrial sectors that produce store or use CWAHS (example – organic or inorganic chemical manufacturing.)

CWA HS Discharges/Impacts. Addressing nonresponse bias for CWA HS discharges and impacts is more challenging. The potential for discharges and impacts will vary from one state to another and it would be difficult to match other variables that might have a role in the number of CWA HS discharges and impacts of those discharges (e.g., distance to a surface water from a CWA HS facility).. Also, causality of impacts (such as fish kills) may be difficult to determine due to time lapse between the discharge and observations of the impacts, unless the discharge was significant and assets were deployed to respond in a timely matter to investigate the discharge and notify appropriate officials of the discharge.

In the event of nonresponse bias for CWA HS discharges and impacts, EPA will continue to rely on existing data sources (such as the NRC database) for this information.

3. PRETESTS AND PILOT TESTS

To pilot test the Hazardous Substances Survey, we randomly selected SERCs from 9 states. Seven of the nine states selected participated in the consultation. The consultation was conducted from July 1, 2017 - August 15, 2017. Respondents provided data and information, as well as burden estimates for responding to the survey. The findings from the consultation are detailed in Appendix 1.

In addition to the consultation, the draft survey was provided as supplemental information for the publication of the first Federal Register Notice (published September 21, 2017) announcing EPA’s intention to conduct an ICR. EPA used the comments received from the public to modify the survey questions.

4. COLLECTION METHODS AND FOLLOW-UP

(a) Collection Methods

EPA will collect responses electronically or on paper, at the respondent’s convenience. In collecting and analyzing the information associated with this ICR, EPA (or its contractors) will use personal computers with Microsoft Excel© based analysis. The EPA will review each submittal for accuracy, completeness, and relevance to the regulatory action, to the extent possible. The information collected pursuant to this ICR will be maintained electronically on secure EPA servers.

EPA will notify the public via the Federal Register of the ICR and survey. Additionally, the EPA maintains a closed listserv for monthly communication between EPA and state, territorial and tribal ERCs; the October 2017 newsletter was used to notify the affected entities of the EPA notice of intent published in the Federal Register on September 21, 2017. This listserv will be used to notify respondents prior to and during the public comment period allowed to respond to the voluntary ICR. In addition, contact information for EPA staff lead for the ICR is available on the survey instrument should respondents need assistance. The listserv text from October 2017 is provided here as a sample:

Clean Water Act (CWA) Hazardous Substance Spill Prevention:

On September 21, 2017, the Proposed Information Collection Request; Comment Request; Survey on Clean Water Act Hazardous Substances and Spill Impacts Notice by the EPA was published in the Federal Register.

• This notice informs the public of EPA's intention to do a voluntary information collection request and provides background information on the request subject, the respondent universe, and the estimated burden to respond to the request.

• The draft ICR questionnaire is provided in the docket as additional information to facilitate the comment process.

• Public comments will be considered and the ICR package will be modified, as appropriate, prior to submission to OMB for approval.

  • The public comment period for this ICR is 60 days.

  • Comment due date: no later than November 20, 2017.

  • This is the first of three related FR notices.

  • Read the Notice: https://www.federalregister.gov/documents/2017/09/21/2017-20170/proposed-information-collection-request-comment-request-survey-on-clean-water-act-hazardous

(b) Survey Response and Follow-up

The default target response rate for this survey is 100% because we are requesting responses from all potential respondents (87). The actual estimated response rate (68 respondents; 78%) is based on extrapolation of the consultation response rate to the universe of potential respondents. EPA anticipates prior notification to all respondents via the listserv, and follow-up to occur at least once over the allowed response period (anticipate 45 days) for the information collection.

Security issues related to Tier II data (e.g., in the context of potential threats to facility security) will increase the likelihood that States will provide data that is not aggregated (e.g. Tier II facility data may not have associated location (GIS) data). They may also decide not to respond, or to only provide partial data since this survey is voluntary. The former will result in an overestimate of the potentially affected facilities as it will increase uncertainty associated with the number of potentially regulated facilities based on proximity to jurisdictional waters since facility location is unknown, whereas the latter will impact the accuracy of agency’s estimate of the facility universe requiring a surrogate measure rather than a direct count from a state, territory, or tribe that does not respond.

5. ANALYZING AND REPORTING SURVEY RESULTS

(a) Data Preparation

EPA anticipates that most of the information and data requested will be in electronic database (likely MS Access or Excel) format. The contractor will manually enter any data not in an electronic format. The contractor and EPA will perform QC on a portion of the filtered database(s), as well as manually entered information to flag missing items, inconsistencies with baseline data, and potential keystroke errors. EPA will compare the information collected in the survey with the analysis conducted by the Agency to develop the baseline assessment supporting the proposed regulatory action to prevent discharges of CWA HS to jurisdictional waters. This information will assist EPA in determining and refining appropriate regulatory approaches to prevent discharges of CWA HS to jurisdictional waters, as well as estimating the scale and potential impacts of any regulatory action taken by the Agency.

(b) Analysis

Collected data will be summarized and compared with available baseline data and information, and a decision will be made on whether the collected data will augment the existing information, or whether the new information will be used to modify baseline assumptions used in the preparation of the draft rule. In both cases, the new information will be reflected (as appropriate) in the modified proposed rulemaking following a 60-day public comment period on the draft rulemaking to be published in June 2018.

(c) Reporting Results

EPA will produce a summary of the information collection as part of the supplementary materials released with the final rulemaking in 2019. Specific requests for information collected will be addressed by the Agency. EPA will not release a compiled database from the information collection; rather the Agency will release information to other government personnel and the public upon request, in a format that does not compromise individual facility chemical safety and security concerns.

3