Cover of The Natural Gas Pipeline: Steel, Pressure, Flow, a book by Ketineni RamaraoFrom the book The Natural Gas Pipeline: Steel, Pressure, Flow
Chapter 21. Read the full book free as a PDF.

Natchitoches, 1965: a failure nobody knew existed

On 4 March 1965 a 24-inch natural gas line owned by Tennessee Gas Pipeline ruptured near Natchitoches, Louisiana, and killed 17 people. Louisiana's pipeline safety history says the accident is credited with helping bring about the 1968 Natural Gas Pipeline Safety Act. Technically it mattered as much. Oil & Gas Journal reports that the first known case of stress corrosion cracking, SCC, on a buried pipeline was identified after this rupture, and that SCC on buried pipe had been unknown before. The failure began at a colony of cracks on the outside of the steel.

Carlsbad, 2000: corrosion inside the pipe

At 5:26 a.m. on 19 August 2000 a 30-inch El Paso Natural Gas line ruptured beside the Pecos River near Carlsbad, New Mexico. The gas burned for 55 minutes. Twelve people camping under the bridge that carried the pipe were killed. The NTSB found that the probable cause was a significant reduction in pipe wall thickness due to severe internal corrosion, which the company's corrosion control programme had failed to prevent, detect or control. It added that ineffective federal inspections before the accident failed to identify the programme's deficiencies.

Timeline of four gas pipeline failures: Natchitoches 1965, Carlsbad 2000, Ghislenghien 2004 and San Bruno 2010, with deaths and causes.
Four failures and their causes. Original IOI diagram; Ghislenghien figures are the book's.

Ghislenghien, 2004: a digger

The book records that on 30 July 2004 a gas pipeline at Ghislenghien in Belgium, buried 1.1 metres deep, exploded after being damaged by construction work, killing 24 people and injuring 132. These figures are the book's, from sources it cites. The book's point is that third-party damage, someone digging where they should not, is among the most ordinary causes of pipeline failure.

San Bruno, 2010: a weld from 1956

At about 6:11 p.m. on 9 September 2010 a 30-inch section of Pacific Gas and Electric's Line 132 ruptured in San Bruno, California. The NTSB says the section that failed was about 28 feet long and weighed about 3,000 pounds, and was found 100 feet from a crater about 72 feet by 26. PG&E estimated 47.6 million standard cubic feet of gas was released. The fire destroyed 38 homes, damaged 70 and killed eight people.

A hillside in San Bruno, California, after the fire: scorched trees and burnt, grey ground in the foreground, with cleared lots and standing houses beyond.
The burnt area in San Bruno, California, after the pipeline fire of 9 September 2010. Photo: Brocken Inaglory, CC BY-SA 3.0, via Wikimedia Commons.

The NTSB's probable cause has two parts. The first is inadequate quality control in a 1956 relocation project, which allowed a substandard and poorly welded pipe section with a visible seam weld flaw to be installed. The flaw grew to a critical size over decades, and the pipe ruptured when pressure rose during poorly planned electrical work at a terminal. The second is an integrity management programme that failed to find and replace the section. The board also named two contributing rule gaps: exemptions of older pipelines from the requirement for pressure testing, which would likely have detected the defect, and the regulator's failure to detect the weakness of the operator's programme.

What the cases show

Read together, they make four points. A failure mode can be unknown until it happens, as with SCC. A defect installed on the day of construction can wait half a century, as at San Bruno. Gas that is wetter or dirtier than the design allows corrodes a line from inside, which is why inspection and gas quality limits matter. And the people nearest a rupture are rarely those who made the decision that caused it. Regulatory change followed harm each time, in the 1968 law and later revisions of integrity rules; the book makes that argument, and the NTSB findings above support it for San Bruno and Carlsbad.

For how operators look inside the pipe, see what a pipeline pig does. For the first gas safety law, see why gas smells.

Quick answers

What caused the San Bruno pipeline explosion?

The NTSB found the probable cause was a substandard, poorly welded pipe section installed in 1956 and an integrity management programme that failed to find it. The pipe ruptured on 9 September 2010; eight people died and 38 homes were destroyed.

What caused the Carlsbad pipeline explosion in 2000?

Severe internal corrosion had thinned the pipe wall. Twelve people camping beneath the pipeline bridge were killed, according to the NTSB.

What was the first stress corrosion cracking failure?

The 1965 rupture near Natchitoches, Louisiana, according to Oil & Gas Journal. It killed 17 people and helped lead to the 1968 Natural Gas Pipeline Safety Act.

Sources

Natchitoches: the 4 March 1965 date, 24-inch line, 17 deaths and link to the 1968 Act are from the Louisiana DENR; first known SCC case from Oil & Gas Journal. Carlsbad and San Bruno facts and probable causes are from the NTSB investigation pages. The Ghislenghien figures (30 July 2004, 24 killed, 132 injured) are from the Wallonia history timeline and VRT News. The New York Times report that nine of the Natchitoches dead were children is the book's and is not used.

More from this book

Inspection · From the book The Natural Gas Pipeline: Steel, Pressure, Flow

What is a pipeline pig? How operators inspect a gas pipe from inside

You cannot dig up a thousand kilometres of pipe to look at it.

Safety · From the book The Natural Gas Pipeline: Steel, Pressure, Flow

Why natural gas smells: the New London school explosion of 1937

Natural gas is nearly odourless.

Engineering · From the book The Natural Gas Pipeline: Steel, Pressure, Flow

Why gas pipelines need compressor stations: pressure, friction and the sawtooth

A gas pipeline does not push gas the way a pump pushes water.

Trade · From the book The Natural Gas Pipeline: Steel, Pressure, Flow

Pipeline or LNG ship? How natural gas gets to market

A pipeline is fixed and a tanker is not.