Failure of a hydrogen pipeline
A steel hydrogen pipeline failed with a circumferential rupture, at a location where a leak had been already detected and repaired. The leak was discovered during a one-call response for planned third party excavation one month before. At that time, the leak was isolated, a section of pipe removed and replaced with new pipe. Following the repair, the entire pipeline was successfully hydrostatically tested, cleaned and dried. The line was returned to service three week later (November 23). Third party excavation around the pipeline commenced December 2, 2002. The excavation terminated December 3. Two hours after the line had been backfilling, the pipe failure occurred. A plausible explanation, confirmed by metallurgical examination, is that the failure initiated in the heat-affected zone of a butt weld on the original pipe material that had become hardened and susceptible to sudden stress relief failure. The material failed due to the application of heat, because it hardened and made the high strength steel brittleness. [Note of the event validator: the PHMSA report does not mention the role, which the excavation works may had in building stress in the materials and in initiating cracks. It is clear nevertheless that both this and the re-pressurisation actions may have brought the stress above the limit of the welded materials, which had been compromised by the repair and welding works one month before.

INITIATING CAUSE was the failure of a welding under sudden stresses. The ROOT CAUSE was the repair work consisting in a weld which had changed the mechanical characteristics (hardening) of the original pipe high strength steel. This may be related to a mistake on the execution of the welding and/or in shortcoming in the operative procedure guidelines.