Anhydrous Ammonia and Roadways

I ran across an interesting editorial about a recent anhydrous ammonia pipeline leak. The focus of the editorial was on the problem of aging pipelines, but it briefly addressed an issue that I have touched upon on several occasions; the problem of anhydrous ammonia leaks near major roadways.

The Incidents

The most recent incident was on October 18^th near Tekamah, NE (see news reports here, here, here and here). There was a major leak on an above ground portion of an 8” diameter pipeline carrying liquefied anhydrous ammonia. The resulting vapor cloud crossed US 75. A motorist drove through the cloud and died. The accident is being investigated by the National Transportation Safety Board (NTSB).

The other incident that I have reported on was in August of 2009 in South Carolina where there was a hose rupture during a tankwagon unloading incident. Again, the resulting vapor cloud crossed a major highway (US 321). A mother of two drove into the cloud and died.

In both cases the facility owners properly notified authorities of the incidents, but the response was not quick enough (or perhaps not well planned enough) to have stopped the victims from driving into the cloud.

PHMSA and Leak Detection

Back in 2010 while reporting on an advanced notice of proposed rulemaking (ANPRM) from the DOT’s Pipeline and Hazardous Material Safety Administration on hazardous liquid pipelines I did a special post of the comments that I submitted on that rulemaking. In those comments I noted that pipelines carrying poisonous inhalation hazard (PIH) chemicals (like anhydrous ammonia) pose a special hazard in the case of leaks. I suggested that:

“Any time that a PIH pipeline traverses an area near major thorough fare the PHMSA regulations should provide treatment similar to the HCA provisions even if the roadway is in an otherwise rural area. Once again, any above ground portions of the PIH pipeline in these areas will have an even larger potential affect.”

I also noted that:

“Once again, I would like to suggest that any place where a PIH pipeline is above ground externally based leak detection sensors are the only technology that would provide adequate warnings of the relatively small leaks of PIH materials that could affect unprotected civilians.”

In its notice of proposed rulemaking (NPRM) on the hazardous material pipeline revisions PHMSA responded to my comments (and similar comments by others) about HCA provisions for pipeline segments near roadways by saying:

“PHMSA is not proposing to designate major road and railway crossings as HCAs, but will consider whether the pipeline IM requirements should be applied to these areas when completing the study that Congress mandated under section 5 of the Pipeline Safety Act of 2011. PHMSA notes that the pipelines at such crossings would be afforded additional protections under the other proposals made in this proceeding, including the requirements for the performance of periodic internal inspections and the use of leak detection systems.”

On the external leak detection issues, PHMSA responded:

“PHMSA commissioned Kiefner and Associates, Inc., to perform a study on leak detection systems used by hazardous liquid operators. That study, titled “Leak Detection Study,” [4] was completed on December 10, 2012, and was submitted to Congress on December 27, 2012. PHMSA is considering, in a different rulemaking activity, whether to adopt additional or more stringent requirements for sensitive areas in response to this study.”

It should be noted that in the recent incident, it appears that the pipeline operator did have flow-based leak detection active on the affected pipeline section. One news report stated that: “the company’s remote sensing system detected a pressure drop on the portion of the pipeline that runs through Burt County. A pressure drop means a release may have occurred, he said.” Automated valves were then closed and authorities notified, just not soon enough to stop the one victim from driving into the ammonia cloud.

Emergency Notification

In both of these incidents the facility owner made all of the appropriate notification when the leak was discovered and there is no indication that the emergency response was not prompt. Still, in both cases an innocent third-party, with no connection to the facility, died as a consequence of the leak. In both cases they drove into a vapor cloud that looked no different than innocent bank of fog. Once in the cloud the auto’s motor stopped (for lack of oxygen) and the people died either from ammonia exposure or lack of oxygen.

While it is easy to say that better something would have prevented the leak (and the NTSB investigation will tell us that), it is even easier to say that if the person driving the car had been warned not to drive into the cloud, they would not have been killed. We have to expand our definition of emergency notification.

For local residents, notifications can be made by a reverse 911 notification system. In the most recent incident that would have saved the victim; he was a local who apparently was in search of the source of the ‘pungent odor’ of ammonia that he had smelled. If he had been notified by a reverse 911 system, he may never have left the house, and almost certainly would not have driven into the cloud if the message had been properly crafted.

For out of area personnel traversing highways near such incidents, the reverse 911 system is much more problematic. More advanced systems use phone location for notifications not sign-up addresses. But, then again, safety people are trying very hard to stop people from answering phone calls while driving. Something else is needed.

I proposed in an earlier blog post that signs could be posted on major roadways near fixed facilities and pipelines that handle PIH chemicals. These would be digital signs that would flash a warning not to proceed when a local PIH chemical detector detected a chemical cloud near the roadway. When not warning of a PIH leak, the signs could display other safety messages.

Moving Forward

The editorial that peaked my interest in this incident concludes by saying:

“Nebraska’s congressional delegation needs to work together to ensure the agency [PHMSA] is giving Nebraska proper attention, particularly in the case of anhydrous ammonia, to avoid a repeat of the tragedy caused by the leak in Burt County.”

The Nebraska congressional delegation does have some influence. Rep. Fortenberry (R,NE) is on the House Appropriations Committee (but not the Transportation Subcommittee). Sen. Fischer (R,NE) is an influential member of the Senate Commerce, Science and Transportation Committee and Chair of the Surface Transportation and Merchant Marine Infrastructure, Safety and Security Subcommittee (at least until December 31^st).

The most immediate thing that will influence any legislative or regulatory action on the pipeline safety issues involved in this incident will be the outcome of the NTSB investigation. The preliminary investigation by NTSB has not yet resulted in the incident being added to the list of current investigations. This means that there may not be a formal NTSB investigations. If, the NTSB does not take up this investigation then the effects of this accident on future legislative or regulatory actions will be very limited.