The departure of one large gas storage company, Bismark,ND-based Williston Basin, from the American Gas Association’sstorage survey last week, both shocked and puzzled industryobservers. Williston, one of the largest gas storage operators inthe western region, if not the largest, with 193 Bcf of working gascapacity, wouldn’t say why it was exiting the survey. AGA said itwas told the company was cutting costs by reducing time spentreporting storage information to numerous surveyors, not just AGA.

Whatever the reason, the impact was a reported 18 Bcf of workinggas injections in the middle of winter into storage fields in theConsuming Region West. That perplexed some observers and led themto question the validity of AGA’s report. Even a spokesperson atWilliston Basin said she couldn’t understand how the associationarrived at an injection. AGA said in a footnote had the company notdropped out the survey would have shown 10 Bcf of withdrawals.Another odd result was the reported 81 Bcf of total U.S. workinggas withdrawals, a small number for the middle of winter. AGA saidthe number would have been higher, 109 Bcf, had Williston notdropped out.

The loss of Williston Basin dropped AGA’s Consuming Region Westsurvey sample to 73% of the estimated total storage in the regionfrom 96%, putting it about on par with the sample AGA collects forthe Producing region. Its Consuming Region East survey sampleremains at 95% primarily because most of the storage is held byAGA’s LDC members who are willing to contribute the information.

AGA calculates working gas levels based on whatever percentageof an estimated full level it believes its survey samplerepresents. When the companies in its Consuming Region West reportworking gas levels, AGA multiples the total number reported by thepercentage (73% as of last week). That provides its weekly workinggas level for the region.

If a new company comes in or is taken out of the survey, AGA’sstorage methodology is designed so that it shouldn’t effect thesample much, particularly if every company in the sample isoperating in a similar way, said AGA’s Chris McGill, director ofgas supply. “But if one company is acting a little differently,than when you take them out of the sample, it shifts the percentfull relationship for that week compared to the previous week ofhaving them in.

“That means that one company can have an impact. It happened ata time when storage levels are being drawn down. It happened with aparticular company that is significantly different that the rest ofthose in the pool. This company may have gone into the winterheating season with much less gas than everybody else. It’s not anissue of gas having gone anywhere. It’s how the numbers arereported and the context that you put them in,” he said.

“I have to assume everyone is operating similarly to do this theway we do it. When that doesn’t happen or there’s a set ofcircumstances or the stars cross and the moon lines up andeverything happens all at once, it gives you something that canresult in this kind of magnitude. I’m here to tell you this is anexception rather than the rule.”

McGill dismissed observer claims that AGA is swayed by theagenda of its members in reporting storage data. Some observerslast week noted the report sent an unusually bearish signal to themarket. “There’s no sway to it,” said McGill. “The numbers are whatthey are. There’s no LDC group or pipeline group or marketer groupin my ear. The numbers simply are what they work out to be.”

He said he would not have reported the change had it not been solarge. AGA typically reports that a company has left the surveysample only when it alters working gas levels by more than 1 Bcf.”If there’s more than a 1 Bcf/d change because they’re out – thatis we know there’s going to be a little blip in the curve, adiscontinuity in the data compared to the previous week – then wetell the companies. Consistency of sample is important to me, butremember what I’m really trying to capture here is the change fromweek to week.. I’m not looking at the past; I’m looking forward..What I want to have as accurate as possible is the slope of thecurve.”

McGill said for a comparison between last week’s result andhistorical data, analysts should use the 109 Bcf drawdown for theentire U.S. as reported in the footnote “because that drawdownreflects the sample of companies that have been associated with thepast. If I were going to go forward with this data and I had tohave a number for this week., I’d use the 81 number because that’sthe number that’s going to go forward in time as per the sample.”

He doesn’t think the loss of one company, despite its size, willdiminish the value of the survey. “I think it creates a one-weekproblem for me. The change that we report next week will be basedon the number that we recorded in our matrix of numbers this week,not the numbers in the foot note. And the magnitude of that delta -which will be a drawdown, we all know – will be as accurate as anyof the magnitudes of delta that we have had in the past.

“We still have 73% of the gas in the west. If companies want totake our information now with more of a grain of salt than theyused to, that’s okay,” he said.

AGA hopes to lure back Williston Basin as it had done with aproducing region company last year when it decided to drop out ofthe survey. “As they explained to me – whether it’s true or not Idon’t know – but they are stopping reporting to all sources. I wantto work with them and I want them to make us the exception ifthere’s one to be made.”

Rocco Canonica

©Copyright 1998 Intelligence Press, Inc. All rights reserved. Thepreceding news report may not be republished or redistributed inwhole or in part without prior written consent of IntelligencePress,Inc.