EIA officials disclosed in an interview with NGI on Friday that the agency is planning to make a change to its natural gas storage survey estimation parameters to improve its weekly storage estimates while avoiding frequent storage revisions.

In the near future, EIA plans to begin using monthly data from the beginning and the end of the heating season to calculate the size of its survey sample. The survey sample, which is developed using EIA’s monthly storage data, has been frozen since last May when the agency responded to criticism about the numerous revisions it was making.

“People were getting all nervous because we were revising our estimation parameters regularly,” said EIA’s Roy Kass. That’s why was it was frozen with May’s monthly number last year.

The weekly storage survey is considered one of the most important indicators of natural gas supply and demand that is available to the marketplace on a regular basis. However, the revisions were sending out confusing market signals and diminishing the credibility of the weekly storage numbers.

The original survey design forced EIA to update on a regular monthly basis the parameters used to calculate total working gas levels. As new monthly data came in, EIA would take its weekly survey sample and determine what percentage of the total monthly number the sample represented and would then use that percentage to calculate its weekly working gas totals.

Kass faulted the original survey methodology for contributing the frequency of revisions, but he also said that some respondents simply had misunderstood what EIA was asking them to do.

“It’s a very simple question — ‘How much working gas do you have in storage’ — as far as I’m concerned. But it was very complicated in terms of the responses. ‘What do you mean by saying, What do you have in storage,’ is what the issue came down to. It was a steep learning curve.”

With the confusion over what exactly EIA was requesting, Kass said, the last thing the agency wanted was its own methodology continuing to make matters even worse by triggering frequent revisions.

But by freezing its survey sample last May, EIA eliminated its ability to capture certain seasonal variations in storage operator behavior. In order to get a better representative sample, the agency now plans to begin using an average of March and October monthly storage data rather than May to determine the size of its sample compared to the total, said Kass.

“We’re changing the parameters for the sample, but it will still be a sample of the monthly operators. The sample design is going to be changed somewhat.

“We’ve run a lot of simulations, and the estimates are more stable if we take an estimator based on high and low [storage levels] rather than current. If we were to still do current, we probably would have revisions [again] so this will be more stable.”

Kass said the change will be made once EIA is convinced it is getting stable responses from storage operators, some of which are new to the survey. Another EIA official suggested the change may take place as early as next month. After that, EIA will update its survey sample parameters annually using the prior year’s March and October storage data.

©Copyright 2003 Intelligence Press Inc. All rights reserved. The preceding news report may not be republished or redistributed, in whole or in part, in any form, without prior written consent of Intelligence Press, Inc.