Citation

Cirimello PG, Otegui JL, Buisel LM. Eng. Failure Anal. 2019; 106: e104142.

Copyright

(Copyright © 2019, European Structural Integrity Society, Publisher Elsevier Publishing)

DOI

10.1016/j.engfailanal.2019.08.008

PMID

unavailable

Abstract

The failure of a 24"natural gas pipeline buried in the 1960s led to a blowout followed by fire. The damage caused the breakage and expulsion of 20 meters of tube; flames reached 30 meters high. The break occurred at the edge of a high transit route, in the opposite site of a thermoelectric plant, and under its high-voltage lines. The break was a consequence of a fracture propagated from a pre-existing defect in the reinforcement weld toe of a 3" connection to a nearby village, made 20 years before. The preexisting defect propagated from small weld anomalies, due to external loads related to soil instabilities: removal of top soil for pipe recoating and flooding due to extraordinary rains paired with the recent construction of a highway that distorted normal water flows. Lessons were also learned in terms of the role of witness recounts. Even after root causes were technically established, responsibility was erroneously attributed to high voltage cables (in the ignition of the gas cloud) and dynamic loads from traffic (as cause for leak), due in large part to the credibility given to initial descriptions. Witnesses identified three events: (1) a noise associated with the blow-out, (2) noise and flashes associated with electrical arcing and (3) the final explosion of the gas cloud, although they did not agree on the sequence of the first two events. The immediacy of the three events and the witnesses´ different places and distances from the blowout site justify the differences of time between their perceptions of light and sound.

Keywords: Pipeline transportation

Language: en

Keywords

Blow out; Failure analysis; Gas pipeline; Weld defects; Witness bias