Williams to replace Parachute pipeline

Williams is planning to replace a natural gas liquids pipeline involved in a leak that contaminated Parachute Creek several years ago, after a U.S. Department of Transportation decision that it has regulatory authority over the line.

The company is planning to install a 2,000-foot-long, 6-inch-diameter line that will travel through a 400-foot, horizontally drilled bore about 17 or 18 feet beneath Parachute Creek. The line, about 4 miles northwest of Parachute, will connect Williams’ nearby gas processing plant with a tank farm, after which the liquids are shipped by another pipeline to another processing plant in Rio Blanco County.

Another pipeline now crosses under the creek, part of it 4 inches in diameter, and part of it 6 inches. In early 2013, the company discovered a natural gas liquids leak associated with the pipeline, which it pinpointed not to the pipeline itself but to an attached, above-ground pressure gauge that had broken.

Williams estimated that about 50,000 gallons had leaked, with about 40,000 gallons vaporizing and about 10,000 gallons reaching the ground. High levels of benzene, a carcinogen, were found in groundwater. Benzene levels that briefly exceeded the designated safe drinking water
standard were found in the creek, but the benzene didn’t travel far downstream and the creek soon was declared benzene-free.

A soil and groundwater cleanup program that is ongoing has resulted in thousands of gallons of natural gas liquids being recovered.

At the time, the Department of Transportation’s Pipeline and Hazardous Materials Safety Administration didn’t regulate the pipeline beneath the creek. Instead, it assumed jurisdiction beginning with the pipeline shipping the natural gas liquids from the tank farm.

Diana Anderson, Williams’ project manager for the new pipeline, said that following communications between the safety administration and Williams after the leak, the agency determined that its authority extends to the pipeline running from the plant and beneath the creek.

Williams could have gone through a process for converting the pipeline for service as a federally regulated line, but that involves a verification process, with steps including running what Anderson called a smart tool through the line to check its condition. She said that’s particularly difficult for a dual-diameter line, and the verification work also could have required some digging in the area of the pipeline, which the company wanted to avoid while reclamation work following the leak is still going on there. So it ultimately decided that replacing the pipe made more sense, and might end up being cheaper.

The company hopes to begin the project this year. It has obtained a state stormwater permit and also needs a permit from Garfield County. On Thursday, the Parachute Town Council approved a watershed permit for the project under a process designed to protect the town’s water supply. But council members first asked several questions of Williams representatives in an attempt to obtain some reassurance about the project, with Mayor Roy McClung explaining to them that the leak “caused quite a ruckus.”

He said the incident resulted in internet blogs that were “very anti-industry.”

“In fact they were so bad they were even anti-Parachute for a while regarding the way that was handled. So we’ve got a heightened sensitivity,” he said.

He said he recognized that Williams is trying to follow best practices.

“We also need to make sure that we’re covering everything from our end and our constituents’ end as well,” he said.

Anderson said the new pipeline, like the current one, will have a leak detection system. But as a DOT-regulated line, that system will be linked to Williams pipeline control personnel in its Tulsa headquarters, providing for even more stringent real-time monitoring, she said.

She said that while there are no indications of problems with the current line, the new one will represent an upgrade.


Commenting is not available in this channel entry.
Page 1 of 1

An interesting story but, some of us would ask the following question.  How is it that Williams did not detect 50,000 of natural gas loss?  True, technology has changed over the years, but the technology to detect that type of loss has been in existence for decades.  Who would not notice if a certain volume were placed in a pipe, and at the destination, there was that type of shortage?  Even if one took just the beginning point and the end point would not notice that there was such a loss?  Somebody, notably someone at Williams was “asleep at the switch”, or perhaps they just didn’t care or used one of those flimsy excuses “Oh, it doesn’t matter” or “That loss is acceptable and we will recoup it by raising the price on the gas that does get through”.

Robert asked, “How is it that Williams did not detect 50,000 of natural gas loss?”

I, too asked that question way back when clean-up and investigation was on-going (this article say it is still on-going) AND it wasn’t natural gas. It was concentrated hydrocarbon liquids that had been distilled from NG. I sent fax and mail to the following people involved on 6/2/14:

Alina Vazquez, EPA, CID
Troy Arnold, Colorado Department of Law
Matt Lepore, Executive Director COGCC
Larry Wolk, Executive Director, CDPHE
Marc Morton, COGCC
Stan Spencer, COGCC
Peggy Tibbets, From the Styx Publisher
Dennis Webb, Grand Junction Sentinel
Bruce Finley, Denver Post  
Susan Alvillar, WPX
Kirby Winn, Garfield County O&G Liaison

In that paper I sent (• in quotes), I will highlight specific points and questions that have not been answered.

• “When I heard Dave Keylor say they “found” a burst pressure gauge, the bell went off for me. That is why I made my “game over” comment. These guys have been around for too long to write-off a burst gauge as a 25-gallon spill.  [Garfield EAB meeting, 4/4/13]
• “I did the calculations on a minimum 1/8″ pipe feeding a pressure gauge and the flow it can do in gallons/minute.” [Ans. 19.9 gal/min. – 25 gal/min.max. @ 200psi – sent to parties 4/5/13]
• “Still riding on their perceived “luck” of having groundwater flowing from the stream [influent] they are still hoping that none of this will get into the stream. But, considering the stream drops and bank features that exist downstream, their luck may run out.” [Luck ran out and spill entered the stream downstream where stream curved away from road embankment.]
• “Consider the volumes involved after the March 2103 [2013] Parachute Creek spill in western Colorado, where cleanup still isn’t done. The Williams and Bargath companies have recovered 11,800 gallons of liquid hydrocarbons, state officials said. They’ve excavated 2,275 tons of contaminated soil, hauling it for burial at the ECDC facility in Utah.” [Denver Post 5/4/14]

• “This 11,800 gal is 14.3% greater than the 10,122 gal. estimated and it does not account for the evaporation for January to March nor the remainder in the soil not collected.”

• “By this article, in their best year O&G recovered 71% and left 29% of spilled hydrocarbons. Using this best year of recovery factors, in the Parachute Creek spill, Williams say they recovered 11,800 gallons to date, which if it is 71% AND represents all they can get, then this indicates an original spill into the soil of 16,620 gallons. This 16,620 was supposed to be only 20% of the total spill with 80% evaporating. That figures out to 16620/.20 =  83,100 gallons total spill which was reported as ~50,000 gallons or only 60% of actual spill was reported. If the original spill was 50,000 by record then either the 80/20 estimate was off or they picked up more than they spilled into the ground. “ .... continued…

... the rest ....
• “I believe this discrepancy indicates Williams should produce the records by which they determined the original total spill, how it was obtained, if by real time recording – why the leak wasn’t detected earlier as it was happening, and clear up this important discrepancy surrounding this spill. The presence, lack of, or failure of any leak detecting equipment should be determined in their SCADA system. The regulatory agencies, both COGCC and CDPHE have not shown they required any accounting/investigation of these areas and whether any discrepancies are intentional, accidental or from poor practices/inadequate equipment.” [ The question here is did SCADA equipment detct leaks, record it, but alarms ignored or not answered until 1/3/13?]

Williams was not fined because spill was caused by “equipment failure”, but the question remains of any authority ever checking this aspect. This would be human failure not responding to equipment warnings until a massive spill occurred. 

As this article says, Williams is going to send leak detection data to the plant and Tulsa office on the new pipeline.

Moreover, the CDPHE never did any follow-up on the effects that might have happened as a result of over 50,000 (or over 83,000?) gallons evaporating and blowing downwind and if there was any relation to the “abnormal number of birth defects” reported by a Glenwood Birth Clinic by the end of 2013.

Page 1 of 1

Search More Jobs

734 S. Seventh St.
Grand Junction, CO 81501
970-242-5050; M-F 8:00 - 5:00
Subscribe to print edition
eTear Sheets/ePayments

© 2017 Grand Junction Media, Inc.
By using this site you agree to the Visitor Agreement and the Privacy Policy