In 1997, a research project was initiated by Southern California Gas Company, Pacific Gas and Electric Company, with support from Tokyo Gas, Osaka Gas, and Toho Gas, to investigate the cause of natural gas pipeline damage during the 1994 Northridge earthquake. As part of this research activity, extensive field and laboratory investigations were performed on a 1925 gas pipeline that suffered several girth weld failures in Potrero Canyon, a remote and unpopulated area just north of the Santa Susana Mountains. The pipeline is operated by the Southern California Gas Company, one of the principle sponsors of the gas utility research project.
The investigations into the performance of the pipeline were largely prompted by questions regarding the cause of pipeline damage. Although ground cracking and sand boils were observed in Potrero Canyon following the Northridge earthquake, there were no clear signs of permanent ground deformation near the locations of pipeline damage. Pipeline damage, consisting predominantly of girth weld tensile failure and two instances of buckling of the pipe wall, indicated that significant relative pipe-soil deformation might have occurred. Field investigations were unable to identify surface evidence of permanent ground deformation near the locations of pipeline damage and attention focused on the possibility that the damage could have been caused by wave propagation. This focus was based on the assertions of past researchers that pipelines with poor-quality oxyacetylene girth welds are susceptible to damage from wave propagation. The detailed investigation of The pipeline has concluded that wave propagation was not a significant factor in the pipeline damage and raises questions regarding wave propagation effects as a causative mechanism for pipeline damage in past earthquakes.
A simple analytical model of the transient ground deformation that may have occurred in the vicinity of the pipeline damage was found to provide insight into the cause of the ground cracking observed at the margins of Potrero Canyon, approximate magnitudes of differential ground displacements that may have occurred during the earthquake, and the reasons for the spatial distribution of pipeline damage. This model is proposed as the basis for identifying locations where similar earthquake effects can be identified in future hazard assessment studies.