1	SURVEY OBJECTIVES, KEY VARIABLES, AND OTHER PRELIMINARIES
1(a)	Survey Objectives
For CWNS 2012, states will have the option to use the "Traditional" CWNS Method or a new "Gap Approach" option.  The primary objective of the statistical methods applied in Section B of this information collection is for EPA to identify and select a sample of wastewater systems that is representative of systems within each state that chooses to participate in the GAP Approach to estimate the needs for the 2012 CWNS.  For the selected sample of wastewater systems, states will collect information about asset inventory that will be used by EPA to estimate asset replacement cost. The representativeness of this sample of systems at the state level is critical to the CWNS because the estimated costs will be used to generate the CWNS Report to Congress, which presents costs data at the state level. 
1(b)	Key Variables
The key variables associated with selecting a representative sample of wastewater systems at the state level are: system size (by population served), CWNS 2008 needs, and system type (i.e., combined sewer system or separate sewer system). 
1(c)	Statistical Approach
In the previous data collection efforts for the CWNS a census approach (Traditional Approach) was used. For the CWNS 2012 EPA is providing states the option of using a more compressive infrastructure GAP Approach to estimate the needs for wastewater systems. This new approach will require participating states to submit significantly more information for each system, so selection of a sample of systems for participation allows for significant burden and costs savings as compared to a census approach. 
1(d)	Feasibility
EPA anticipates that the survey (the statistical sample) objectives are achievable given the existing time and resource constraints.  

   * State participation in the GAP Approach for CWNS is voluntary.  
   * There will be an electronic version of the form (with picklists and calculations coded in). 
   * The statistical approach to this data collection requires only a fraction of systems to complete the GAP form, resulting in equivalent or smaller cost and burden at the state level than would be incurred if all systems participate in CWNS 
   * The survey results will be completed in time to inform the corresponding cycle of CWNS.  
2	SURVEY DESIGN
2(a)	Target Population and Coverage
Existing public wastewater systems are the target population for CWNS GAP Approach methodology. 
2(b)	Sample Design
2(b)(i) Sampling Frame
For each participating state, EPA will develop the sample frame for the statistical selection of wastewater systems. Initial data will be pulled from the 2008 CWNS. Only those systems in operation in 2008, expected to be in operation in the future, and reporting documented needs were used to estimate the sample characteristics, but the sample size calculations was applied to all wastewater systems, regardless of their reporting of needs. The initial universe will be adjusted to account for known anomalies inventory reporting (For example, states will be able to inform EPA about existing systems that were not entered in the 2008 CWNS or that were entered but not submitted for review).
2(b)(ii) Sample Size
Final sample sizes for participating states will be determine once the state makes a decision to be part of the GAP Approach methodology. Preliminary sample sizes were estimated by EPA and are presented in Appendix A.  The sample design limits strata with less than 10,000 people to having a maximum error of 80% and strata of more than 10,000 people to having a maximum error of 40%. All facilities in the state's top 3% based on population will be included in the sample frame.

Typical sample rates are expected to be:
   * 10-25% for facilities serving less than 10,000 people
   * 30-60% for facilities serving between 10,000 and 100,000 people
   * 100% (census survey) for the largest 3% of facilities in each state
2(b)(iii) Stratification Variables
In developing the representative sample, EPA considers factors such as population served and geographic location.  For each state the sample systems will be stratified by population served (system size).
2(b)(iv) Sampling Method
The proposed sampling design is stratified random sampling. For each state the number of strata and stratification limits will be determined based on the population characteristics of the state and a representative sample of systems from each of the strata will account for different system sizes.  EPA will include at least three systems from each stratum for each state. 
2(b)(v) Multi-Stage Sampling
EPA does not anticipate the need for multi-stage sampling. Wastewater systems statuses often change little over short periods of time, so significant changes from the 2008 are not expected. 
2(c)	Precision Requirements
2(c)(i) Precision Targets
For each state the sample design is selected based on two conditions:
   * The total needs would be estimated within close to +- 10 percent of the amount of the true need within a 95% confidence interval, and
   * Each stratum consisting of less than 10,000 people would have an error of +- 80 percent, and each stratum consisting of more than 10,000 people would have an error of +- 40 percent. 

For most states one of the two conditions above drives the sample design. In cases where the individual stratum error drives the results, the total error will be less than 10 percent. Alternatively, where the total error drives the results, states will have stratum errors that are less than the 80 percent or 40 percent bounds.
2(c)(ii) Non-sampling error
Systems that do not complete the questionnaire will be considered as systems without justification for needs, and will be included as such in the analysis.  This is consistent with how these facilities are handled in the CWNS Traditional Method.  Given that many less facilities are involved under the Gap Approach, states utilizing the gap approach will have a much greater ability to follow up with facilities on completing the form.  Since non-completion of questionnaire will be assessed as a system not having justification for needs, and since CWNS 2012 results may be used to determine states' Clean Water State Revolving Fund allocations, states will have a high incentive to obtain very close to 100% form completion.
2(d)	Questionnaire Design
The GAP Data Request form was design to minimize data entry by pre-populating it with data currently in the CWNS DEP. See Appendix B for the draft form.

Other features of the form are:

   * The form will have optional extra pages for entering additional asset inventory (section 5) information.
   * The form will have a companion (~2 page) instructions document.
   * There will be an electronic, editable version of the form (with picklists and calculations coded in).  Like the hard copy form, this electronic Gap Approach form will be pre-populated with the most current CWNS data. 
   * Asset replacement costs are not requested in any versions of the form.  These costs will be calculated in the CWNS DEP based on other data entered in the form
   * All data in the form is required, except where indicated by the italic text (several rows in page 4 tables, and far right 3 table columns on pages 5-7)

The form is divided into 6 sections, each with data item request that meet a clear and specific objective:
      Sections 1 to 4: Basic system data. These four sections provide the information necessary to identify and characterize the system. Most of the information will be pre-populated but the respondents will have the ability to change any of the data items.
      Section 5: Revenues and Expenses. Data items in this section are necessary to analyze the financial situation of the system. Most data item are voluntary with only the totals denoted as mandatory. 
      Section 6: Asset Inventory. This is the most extensive section of the questionnaire and will allow EPA to estimate the type of asset, anticipated failure and replacement or rehabilitation costs. The data items required should be readily available to most operators or managers.
      Section 7: Certification. Signature requirements guarantee the validity of the data. EPA is implementing a tiered certification with smaller systems having less stringent requirements.

3	PRETESTS AND PILOT TESTS
The "Gap Approach" option is modeled after an asset management needs assessment approach used in Pennsylvania's 2008 Governor's Report on Sustainable Infrastructure. For the 2008 CWNS, the state of Pennsylvania, in collaboration with EPA, conducted a sample of their wastewater systems and estimated their infrastructure needs using a GAP methodology approach. This study in Pennsylvania served as the pilot test for the nation-wide CWNs GAP initiative to be included in the 2012 CWNS. The original Form used in Pennsylvania has been modified based on lessons learned and input from the other states that are interested in using the GAP methodology for 2012.  

4	COLLECTION METHODS AND FOLLOW-UP
4(a)	Collection Methods
CWNS coordinators from states participating in the GAP approach methodology will have flexibility to choose the collection method that best fits the characteristics of their state. EPA is providing multiple options and flexibility to facilitate the data collection effort by states:

   * Forms will being pre-populated with data currently in the 2008 CWNS.
   * Forms will have a barcode CWNS ID #, so that forms faxed in can be connected for facilities within the CWNS DEP.
   * There will be either an electronic version of the form (with picklists and calculations coded in)
4(b)	Survey Response and Follow-up
High compliance levels (97%) during Pennsylvania's effort in 2008 have given EPA confidence that equivalent levels can be achieved during 2012 in other states.  EPA plans to continue outreach and compliance assistance efforts, as needed.  

5	ANALYZING AND REPORTING SURVEY RESULTS
5(a)	Data Preparation
States are ultimately responsible for collecting and entering the information into the CWNS DEP and for the quality of the data entered. The DEP is equipped with automatic error checks to identify systems that fall outside certain normal parameter and during the data entry period of the CWNS 2012 EPA and its contractors will perform reviews of the information and documentation submitted. 
5(b)	Analysis
Data collected with the Gap Approach will be analyzed alongside other CWNS data to develop a much more complete picture of wastewater facility sustainability, with state level estimates on capital and O&M needs, revenues, and gaps projected over 20 years.
5(c)	Reporting Results
Capital needs from the Gap Approach will be reported alongside Traditional Method capital needs in the CWNS 2012 Report to Congress. Additionally, a new section will be added to the Report that summarizes data from states participating in the Gap Approach to draw sustainable infrastructure conclusions (e.g. the gap between projected capital and O&M needs and expected revenues).








Appendix A  -  Draft Sample Sizes by State


