Pipeline Quality Issues

NOTE:  This example is the first we've published that utilized our new Ishikawa/Fishbone feature, released in Causelink Version 7.  You might want to read our blog on this feature and how we plan on using it.

On June 2, 2013 multiple leaks were discovered in the new process water treatment line during final hydrostatic testing. This delayed startup by two weeks and cost an additional $525,000 to repair the leaks and replace coupling gaskets. In addition to the startup delays and added costs, a problem like this could potentially lead to reportable spills, additional cleanup costs, and lost future customer contracts. This is the first problem of this type and magnitude, but there have been other build-quality issues in the past.

Before a line like this is placed into service, it must be hydrostatically tested to ensure it is leak-free. This is a process whereby the system is pressurized with water using an external pump. The system is brought up to a target pressure which is then maintained for a period of time. If no leaks occur, the system’s integrity is confirmed, allowing it to be handed over to operations. Leaks occurred in this system even before it achieved the target pressure. These leaks were found primarily in various mechanical joints (couplings). Upon examination, it was found that the leaks were caused by a combination of multiple causes.

Gaskets Did Not Seal:
Each coupling includes a gasket that forms a seal between the two surfaces. In multiple cases, gaskets were found to have been damaged and/or creased, resulting in failure to properly seal. The gaskets are made from a relatively soft material that makes it easier to establish an effective seal. If they are not properly aligned when the coupling is tightened down, they will fail to seal. This can occur if the coupling was not tightened properly and also if the gasket gets stuck/hung on parts of the pipe during installation.

The gaskets were lubricated using soapy water in order to make installation easier and more effective. But because it was so hot (Phoenix AZ in June), the water evaporated quickly, leaving the gaskets sticky rather than slippery. Exacerbating this is the fact that the pipe material is HDPE plastic and black in color. Using soapy water is an accepted method of lubricating gaskets in applications such as this. The coupling supplier confirmed that this method would be both effective and leave the materials undamaged. Plus, the soap is safe and convenient for the installation crew. Under more normal environmental conditions, the gaskets would have remained lubricated long enough to achieve proper installation. The installation team did not recognize the risk of proceeding with installation even though the gaskets were drying out. They were focused primarily on getting the job done.

Evidence was also found that the couplings were not tightened using the method specified by the supplier. These bolts needed to be tightened evenly using a cross-bolt pattern. This creates a uniform seal across the entire gasket. If the bolts are fully tightened in a circle pattern, it causes the gasket seal to be out of balance around the circumference of the coupling, and this creates leaks. The installation team did confirm the torque specifications for coupling bolts, but did not consult the exact method of tightening the couplings. One team member mentioned that they used the circumferential pattern on the last job and it had “worked fine.” And the installation teams were primarily relying on the training that had been conducted onsite rather than the manufacturer documentation. This training did not focus specifically on bolt pattern.

Pipe Ends/Surface Imperfections:
Significant imperfections were found on the ends of the pipe, as well as on parts of the exterior surface. These imperfections contributed to an ineffective seal. The pipes are cut in the field using a chainsaw. This is a “standard procedure” that is both fast and easy, but it leaves rough edges that can create a bad seal if not properly cleaned up. Also, there were many scratches and digs in the exterior surface of the pipe. The pipe is made from plastic that can become scratched if it drags across rocks and/or sharp objects on the ground. The terrain in this area was on a 35% slope. And the hot temperatures made the black plastic pipe softer. This exacerbated the tendency for the pipe to be dragged across the ground and subsequently damaged. As was the case with coupling installation, the teams largely relied on what they thought they remembered from class instead of using the published installation documentation as guides. These classes did not expressly emphasize the need to properly prep the pipe surfaces prior to installing couplings. And they used PVC pipe in their classroom demonstrations, which is fundamentally different than HDPE.

Coupling Leaks Not Discovered:
The leaks were not discovered prior to the final hydrostatic test because no incremental testing is performed. This was largely because the Superintendent didn’t ask for it. Project leadership assumed that this was being performed, but did not specifically check up on him because the Superintendent had come highly recommended.