Changes to the EIA-914 Sampling and Estimation ProChanges to the EIA-914 Sampling and Estimation Processes
EIA has conducted a full review of the methodology used in the EIA-914 natural gas production
program, as outlined in March 2009. In April 2010, reporting February production, we plan to
introduce the new methodology based on that review. At that time, we plan to revise January
estimates using the new methodology as well. In some cases, we expect the revisions will be
significant. For this report of January 2010 estimates, however, we will use the methodology
described in our current documentation.
January 2010 Updates, Based on Current Methods
The estimated changes in production from December 2009 to January 2010 in this report reflect
not only actual changes in reported production, but also on the normal annual updates. Updates
required under the current documentation include:
• Sample– this year’s annual sample update added 27 companies and removed 16
companies.
• Estimation – the calibration time period for the estimation model has moved ahead one
calendar year.
• Other States Ratio – the ratio used to inflate the sample reported production to the total
production for the Other States group increased slightly.
• Marketed Ratios – the annual ratios used to convert gross withdrawals to marketed
production were revised for each State based on the additional year of data.
We estimate that the updates by themselves lead to changes in estimated production of less than
1 percent for the Lower-48 United States as a whole.
Next Month: New Methodology
The new sampling and estimation processes are designed to simplify calculations and make use
of the most recent reliable information available from the States. Calculation of the EIA-914
will involve three main changes:
• Monthly Sample– moving forward, the sample will be updated every month using more
recent information.
• Estimation–estimation of production of non-sampled companies will use data that is 6 to
18 months old (rather than the previous practice that used data that was 2 to 7 years old).
• Timing –calibration will be updated monthly rather than annually.
These changes are intended to allow the EIA-914 to reflect major changes in the industry that
might affect production of non-sampled companies (such as the growth of shale gas production)
in a much more timely way.