Not the best way to understand your underground assets…

On January 3, 2011, the National Transportation Safety Board (NTSB) issued a series of urgent pipeline safety recommendations based on preliminary findings from the September 9, 2010, gas transmission pipeline rupture in San Bruno, California, which killed eight people and caused extensive property damage. The NTSB investigation determined that pipeline records were inaccurate, and potentially resulted in an inappropriate Maximum Allowable Operating Pressure (MAOP) for the line in question. The resulting urgent safety recommendations “call on pipeline operators and regulators to ensure that the records, surveys, and documents for all pipeline systems accurately reflect the pipeline infrastructure as built throughout the United States so that maximum safe operating pressures are accurately calculated.”

The safety recommendations stem from a particular finding. At the rupture site, several short “pup” joints of pipe were found whose longitudinal seam type does not match recorded pipe properties. These pup joints have fusion welded seams; records indicate pipe in the area is seamless. This is critical, because some classes of fusion welded pipe are not as strong as seamless pipe. ASME B31.8 requires MAOP for such pipe to be reduced to 80% of the MAOP for otherwise similar seamless pipe. Thus, it’s possible MAOP for the San Bruno line was set too high. (Note that operational data indicate line pressure never exceeded the MAOP of record, and at that pressure the fusion welded pup joints should not have failed. So some unidentified ‘X’ factor may be in play.)

Eagle has no knowledge of the San Bruno system beyond published information, but this data makes for an interesting hypothetical case. Let’s examine it. Two possibilities for provenance of the fusion-welded pup joints in our hypothetical system are: 1) The fusion welded pup joints were part of original construction. 2) The pup joints were installed during later repairs. Both require consideration from the standpoint of Geographic Information System (GIS) Data Governance.

Common practice for initial population of GIS-enabled databases in older pipeline systems is to collect data from available manually drafted as-built alignment sheets; let’s assume our hypothetical GIS was populated this way. This approach facilitates GIS duplication of the existing alignment sheets, which is often the initial goal of a pipeline GIS implementation. So far so good, but as-built alignment sheets are only the tip of the pipeline data iceberg.

There are two potential pitfalls to loading a pipeline GIS solely from alignment sheets:

First, with respect to original construction, as-built alignment sheets are usually drafted at a scale that precludes depiction of features as small as pup joints. At typical alignment sheet scales, pup joints are mere dots. Furthermore, at the time of construction (1956), the potential strength limitations of aged fusion-welded pipe relative to seamless pipe were not well understood. Draftsmen would have had no compelling reason to distinguish between the predominant seamless pipe and a few small lengths of fusion welded pipe. It’s unlikely that pup joints like those at San Bruno would have been captured on our hypothetical as-built alignment sheets.

For older pipelines, information on small features like pup joints might be found in more detailed sources, like pipe tally records. However, the sophisticated computerized inventory management systems we enjoy today did not exist in the 1950’s, and the importance of tracking every joint of pipe was unrecognized. For some older systems, there simply may be no record of small items like pup joints. Regardless, digging this deep into pipeline records is not common during initial pipeline GIS data capture, so even if such data exists, it’s unlikely it would initially make its way into our hypothetical GIS.

Second, in some instances manually drafted as-built alignment sheets may not be kept up to date with ongoing repairs. In such cases the alignment sheets do not reflect pipe changes that have occurred since construction. A pipeline GIS populated from such sheets is missing the pipe repair history. Even if the GIS includes functionality to capture ongoing repairs, the prior repair history goes missing. So again, if features like our pup joints were installed during historical pipe repairs, they could be missing from our hypothetical GIS.

At this point it should be obvious that some pipeline GIS databases may be lacking data requisite to the NTSB recommendations. It’s equally obvious that greater regulatory scrutiny is on the way. However, even prior to the San Bruno event, some of Eagle’s pipeline clients recognized the value of capturing as-built information at the highest level of detail possible, and of capturing all pipe inspection and repair data. As a result, Eagle has developed a robust set of data governance tools for this type of work. We’ll explore these tools, and methods for detailed pipe data capture and data governance, in upcoming posts.

Meanwhile, Eagle stands ready and able to help your company achieve the highest standards in pipeline data governance.