‹ Weekly Natural Gas Storage Report

Differences Between Monthly and Weekly Working Gas In Storage

Latest update: November 3, 2016

Note: The weekly storage estimates are based on a survey sample that does not include all companies that operate underground storage facilities. The sample was selected from the list of storage operators to achieve a target standard error of the estimate of working gas in storage which was no greater than 5 percent for each region. Based on a comparison of weekly estimates and monthly data from January 2010 through August 2016, estimated total working gas stocks have exhibited an average absolute error of 12 billion cubic feet, or 0.4 percent.

Background

The Energy Information Administration (EIA) provides weekly estimates of working gas volumes held in underground storage facilities at the national and regional levels. These are estimated from volume data provided by a sample of operators that report on Form EIA-912, "Weekly Underground Natural Gas Storage Report." The EIA first released its estimates of underground storage for the week ending May 3, 2002, on May 9, 2002, in the Weekly Natural Gas Storage Report (WNGSR). EIA also publishes data for the end-of-month working gas volumes that it collects from the form, EIA-191, "Monthly Underground Gas Storage Report." While these surveys both result in estimates of working gas in storage, some differences between the two series occur, which are useful to understand when comparing current weekly data estimates with values in prior years or to the corresponding data from the monthly series.

Table 1 presents the difference between estimates of working gas in underground storage as reported at the end of each month on Form EIA-191 (a monthly census of all underground storage operators) and end of month estimates interpolated from the weekly estimates (based on a sample of storage operators reporting on Form EIA-912).

Table 1. Monthly-Weekly Differences in Underground Storage Estimates (Billion cubic feet) Download History (2010 to Present)
  Natural Gas Monthly Difference
Month End East Region Midwest Region Mountain Region Pacific Region South Central Region Total Lower 48 East Region Midwest Region Mountain Region Pacific Region South Central Region Total Lower 48
Jan-15 536 618 131 276 831 2,392 10 5 2 -3 0 13
Feb-15 342 345 116 273 576 1,652 5 8 1 -4 19 29
Mar-15 242 252 113 275 575 1,458 6 6 3 -4 2 12
Apr-15 312 309 116 293 750 1,780 -4 0 3 -7 3 -6
May-15 455 438 135 322 920 2,271 -6 2 3 -4 0 -6
Jun-15 570 567 157 331 1,004 2,629 1 -2 -3 5 -2 0
Jul-15 658 683 174 341 1,054 2,910 3 1 -3 3 -4 0
Aug-15 759 829 189 347 1,097 3,221 7 2 -2 1 -2 6
Sep-15 856 975 205 358 1,213 3,606 3 -1 -2 1 -7 -4
Oct-15 919 1,094 215 378 1,321 3,927 -1 6 -1 0 6 12
Nov-15 914 1,091 207 369 1,332 3,913 1 7 1 0 2 10
Dec-15 856 987 186 320 1,304 3,653 4 2 0 -3 3 6
Jan-16 632 764 156 276 1,098 2,926 3 -5 1 -9 -2 -12
Feb-16 483 607 144 263 1,024 2,520 -1 -1 1 -7 -2 -8
Mar-16 436 542 145 266 1,079 2,467 -1 7 4 -1 2 11
Apr-16 463 565 152 287 1,158 2,626 -7 3 6 -2 9 7
May-16 557 671 176 306 1,237 2,947 -7 -2 5 0 2 -3
Jun-16 655 763 197 316 1,236 3,166 -4 -2 4 -3 11 7
Jul-16 735 834 209 316 1,202 3,297 -1 -3 5 -2 2 -1
Aug-16 805 921 221 315 1,158 3,419 2 0 2 -3 5 8

Sources: Monthly from weekly values-linearly interpolated values based on volumes available with the Weekly Natural Gas Storage Report System; Natural Gas Monthly.

Note: Totals may not equal sum of components because of independent rounding.

Sources of Differences Between the Monthly and the Weekly Series

EIA determined a number of factors that can contribute to the implied differences:

  • Inventory reports by respondents to both the monthly and weekly systems can differ somewhat, sometimes simply because activities did not occur in the schedule implied by the linear interpolation of the weekly data. However, these differences historically have tended to be less than 1 Bcf in absolute value per respondent on average.
  • Reclassification of base gas to working gas contributed to differences between the weekly and monthly series, owing to the 2-month lag between the two series.
  • The estimation methodology may produce weekly volume estimates for the nonsample operators that differ from the reported monthly values.

EIA has used a number of estimation methodologies, all of which have been based on using attributes of the reported information from the operators in the sample to estimate working gas volumes for the non-sample companies. There have been periodic changes to the methodology or sample since 2002 that have impacted the resulting estimates. The initial methodology relied on fixed ratios and a fixed sample that may not have captured a change in the relative significance of sample operators and nonsample operators in total inventory in a given period. In October 2003, EIA implemented several adjustments to the weekly series that affected the observed differences beginning in July 2003. These included: a revised estimation methodology, an expanded sample, and revisions for the 17-week period from July 4, 2003 through October 24, 2003. In August 2005, EIA implemented a change to the basic methodology and revised the sample (see Methodology). In conjunction with those changes, revised estimates for the period from mid June 2005 to August were released.

To some extent, differences between the two data series likely will continue. EIA will provide periodic updates of the information in the tables above as an ongoing resource for reference or for those who may want to modify their own comparative assessment of weekly or monthly storage trends.