On December 13, 2018, the Bureau of Investigation and Enforcement (BI&E) of the Pennsylvania Public Utility Commission filed a complaint against Sunoco. The complaint alleges that Sunoco was negligent in preventing corrosion of a section of the Dragonpipe (Mariner East 1), causing a leak of ethane and propane near Morgantown, PA. I provided a brief summary of the complaint and a link to the text of the complaint in a recent post (“PUC finds that shoddy inspection procedures led to ME1 leak near Morgantown“). In this post, I will provide some of the details of what the BI&E inspectors found.

What happened at Morgantown? On April 1, 2017, a resident along the right-of-way of Mariner East 1 noticed bubbling from the ground. (The location was less than a mile from the Twin Valley Schools, which I recently wrote a blog post about. The leak could easily have been adjacent to the schools.) The pipeline was carrying ethane and propane at the time. Although liquefied under pressure, as they escaped from the pipeline they reverted to gases, resulting in the bubbling that the resident observed. The resident alerted Sunoco, who sent out a technician, confirmed the leak, and shut down the line.

Sunoco reported that 840 gallons of liquid were lost. Actually, I think that number should be taken with a grain of salt, since Sunoco did not have the ability to detect the leak and therefore could not know how long it had been going on, or how material much was really lost.

It took two days for the gas in that section of the pipeline to be purged. Only then could Sunoco representatives and a safety inspector from BI&E approach the pipeline and begin investigating the cause of the leak. Sunoco dug up the pipe, and corrosion was found along the bottom of it. An eight-foot length of pipe was cut out and sent off for laboratory inspection, which confirmed the corrosion. Sunoco subsequently replaced 83 feet of pipe in the area and put it back in service.

In the months that followed, BI&E reviewed Sunoco’s corrosion control practices. They found many problems, the majority of them having to do with Sunoco’s “cathodic protection” system. If you are not familiar with cathodic protection, you may want to read my previous post describing the basics of the technology and how it can fail. In essence, it uses an electrical system to give the pipeline a slight negative charge relative to a nearby buried “anode”. The current flowing from the anode to the pipeline prevents corrosion.

 A litany of mistakes and failures. The degree of Sunoco’s negligence and incompetence revealed in the BI&E complaint is awesome. Here are some examples:

  • Sunoco applied cathodic protection systems intended for single pipelines to the protection of each of two adjacent pipelines in the same right-of-way (the second pipeline is the 12-inch “bypass” pipeline, which runs adjacent to ME1 in this area). When cathodic protection systems from two pipelines interact, their readings get skewed and special considerations apply. Sunoco ignored this.
  • At the test station nearest the leak, Sunoco recorded a potential of -628 mV in 2016 and -739 mV in 2017. Both of these failed to meet the minimum required to prevent corrosion (-850 mV).
  • Sunoco conducted “side drain” measurements at a distance from the pipeline, intended to determine whether the required protective current (generated by the cathodic protection system) was flowing in the correct direction, whether it was adequate, and whether there were areas of unusually high current. Several of these tests indicated the current was flowing in the wrong direction (meaning that the pipeline was not being protected) but the company took no action.
  • Furthermore, federal standards say that side drain measurements must not be used in a multi-pipeline right-of-way. (Sunoco ignored that requirement.) And these standards state that localized corrosion will often be missed by the side drain technique.
  • In its “close interval potential surveys” (CIPS) to determine how well the cathodic protection was working, Sunoco only measured potentials with the power to the cathodic protection system turned on. But if there were other sources of current in the vicinity, those could affect the reading. Sunoco should have tested with the power on and off, and compared the two.
  • The locations of the CIPS readings in Sunoco’s records do not match the locations given for the test stations. So, one way or another, the location records are incorrect.
  • Sources of interference which could invalidate the readings, such as power lines and other pipelines, should be identified in the records. They aren’t.
  • Finally, not all the problems were related to cathodic protection. In 2016 and 2017, Sunoco used an “in-line inspection tool” (a “pig”) to evaluate the thickness of the pipeline. Large areas of corrosion should be detected in this way. (But see “The silent “smart pig”: why Sunoco’s inspection system won’t protect us”.) The 2016 pig run failed and didn’t produce any data at all. Sunoco didn’t redo it. The 2017 run worked, and it showed metal loss in the area near Morgantown. This should have been reported as possible corrosion in Sunoco’s inspection reports, but it wasn’t.

It is very clear that Sunoco was just going through the motions. There was no real intent to monitor the integrity of the pipeline; perhaps they just wanted to satisfy the letter of the law by doing some kind of plausible measurement (and they failed miserably even at that). Data collection was haphazard and sometimes incorrect. There was enough troubling information in the data Sunoco did capture to clearly indicate probable corrosion, and yet the company did nothing.

Sunoco frequently tells us to rely for our safety on their vaunted “integrity management plan”. The BI&E complaint shows that, however good the plan may look on paper, in this instance it proved to be worthless in actual implementation.

The BI&E complaint lists 45 separate “counts” (violations) including some beyond those I have listed above. What it says to me is: when it comes to pipeline safety, nothing Sunoco says can be trusted.

How about the rest of the state? The BI&E is clearly concerned that this is not just a local matter in the vicinity of Morgantown. They write: “While the data reviewed was largely specific to the site of the leak, [Sunoco’s] procedures and overall application of corrosion control and cathodic protection practices are relevant to all of ME1 and, thus [BI&E] alleges that there is a statewide concern with [Sunoco’s] corrosion control program and the soundness of [Sunoco’s] engineering practices with respect to cathodic protection.” Let everyone near the Dragonpipe right-of-way, across the entire state, take that as a warning!

How about the 12-inch line? The BI&E complaint does not address the 12-inch “bypass” pipeline directly, but given that it runs alongside ME1 for most of its length in Delaware and Chester County, and given that it also depends on cathodic protection, we can assume that it, too, is a candidate for corrosion problems. Was it corrosion that caused that pipeline to leak into Darby Creek in June? No one is saying, but it now seems likelier than ever.

What happens now? Sunoco has 20 days to respond to the complaint. My guess is that they will simply pay the fine of $225,000 (a trivial amount in the context of this project) and agree to the other terms proposed by BI&E, rather than go through the embarrassment of having to try to defend their corrosion control procedures to the PUC. If they go that route, they will be accepting the obligation to:

  • Conduct a “remaining life study” on ME1 and propose a retirement date
  • Increase the frequency of pig inspections to a minimum of once a year “on all bare steel and poorly coated pipelines”
  • Revise corrosion control procedures to meet federal guidelines
  • Develop procedures to “determine the adequacy of cathodic protection”, including taking into account sources of interference
  • Implement the new and revised procedures within one year. “If the results indicate [possible corrosion, then Sunoco] shall replace the impacted sections….”

For those of you that monitor PUC proceedings, the docket number for this case is C-2018-3006534. Here is a link to the docket: http://www.puc.state.pa.us/about_puc/consolidated_case_view.aspx?Docket=C-2018-3006534