‹ Weekly Natural Gas Storage Report

Notice of Additions to the Weekly Natural Gas Storage Report

Released: January 19, 2017

EIA will begin including estimated measures of sampling variability for weekly estimates of working natural gas stocks and the net change in working natural gas stocks as part of the Weekly Natural Gas Storage Report (WNGSR), on January 26, 2017. These weekly estimates provide increased transparency into the statistical properties of the working gas figures published in the WNGSR.

EIA is not changing the methodology for producing estimates of stocks and net changes published in the WNGSR. The inclusion of weekly estimated measures of sampling variability in the WNGSR does not affect the sampling or estimation methodologies for computing the weekly estimates of working gas. The data product publication procedures and schedule also will remain unchanged. This addition will not affect most WNGSR web products, including the wngsr.txt, wngsr.csv, and wngsr.json.

The publication of estimated measures of sampling variability will affect the WNGSR web products in the following ways:

  • A new table containing estimated measures of sampling variability will be added at the bottom of the WNGSR main webpage.
  • A new data product (cvhistory.xls) will contain historical data series of the estimated measures of sampling variability for weekly estimates produced from the current WNGSR sample, starting with the week ending April 10, 2015.
  • Updated methodology and revision documents will be released on January 26, 2017.

EIA has provided answers to frequently asked questions about incorporating estimated measures of sampling variability for published estimates of weekly stocks and their net changes in the WNGSR.

Initiative to increase transparency in the Weekly Natural Gas Storage Report

The publication of estimated measures of sampling variability is part of EIA’s initiative to increase transparency into the statistical properties of the working natural gas estimates reported in the WNGSR. On March 1, 2016, EIA published a report on the types of possible survey errors in the working gas estimates in the WNGSR. These potential errors are inherent in all collected data and include sampling error, coverage error, nonresponse error, adjustment error, and processing error. That report also included a discussion of the bootstrap methodology used to estimate standard errors for the WNGSR, as well as historical estimated measures of sampling variability for weekly estimates of working gas from April 10, 2015, through November 13, 2015.

In the triennial performance evaluations on the WNGSR, EIA has always included estimated measures of sampling variability for the weeks that most closely correspond to the end of the calendar months. Including estimated measures of sampling variability in the WNGSR release provides data users with more information on a weekly basis regarding the statistical properties of the estimates published in the WNGSR. As a Principal Federal Economic Indicator, WNGSR must meet the Office of Management and Budget requirement to disseminate error estimates so that information is publicly available on the statistical properties of the published estimates. Changes to the WNGSR in this announcement will allow data users to obtain the most recent week’s estimated measures of sampling variability with each release of the WNGSR.

Upcoming additions to the Weekly Natural Gas Storage Report

EIA will add a new table with estimated measures of sampling variability at the bottom of the main WNGSR web page, below the primary data table and storage graph. The changes to the WNGSR webpage will not affect the primary data table containing the weekly estimates of working gas. The new table will include estimates of the coefficient of variation (or relative standard error) for working gas stocks reported for the current and prior report weeks on a regional basis and for the Lower 48 states. The new table also will include estimates of the standard error for the weekly net changes reported, on a regional basis and for the Lower 48 states. A link to a spreadsheet containing historical estimated measures of sampling variability will appear at the top right-hand side of the new table.

Example of new table

EIA will implement these changes in the WNGSR published on January 26, 2017. EIA will not have a period of public testing before implementing the published changes because the changes will not affect the contents or location of the primary data table, or any other weekly data files that are updated in the WNGSR each week.

Revisions to estimated measures of sampling variability

The current EIA WNGSR revisions policy on working gas levels remains in effect. EIA publishes revisions to working gas levels when the respondent-reported changes are at least 4 billion cubic feet (Bcf) at either a regional level or for the Lower 48 states. We will amend the revisions policy to incorporate revisions to estimated measures of sampling variability. Consistent with the established revisions policy, estimated measures of sampling variability will be updated using the data currently available at the time of the disseminated revision. Also consistent with the current established policy, the historical database of estimated measures of sampling variability will be updated weekly and will contain the most current set of estimates available. EIA may revise these historical estimated measures of sampling variability when we revise the historical working gas levels.

For additional information contact Jose Villar at jose.villar@eia.gov or (202) 586-9613.