Skip Ribbon Commands
Skip to main content
Bookmark and Share this page


Keynote speech at the American Gas Association’s (AGA) Executive Leadership Summit
Christopher A. Hart
Washington, DC

Thank you, Dave [McCurdy], for that kind introduction, and thanks to AGA for inviting me to speak today on behalf of the NTSB at your Executive Leadership Safety Summit.  I am particularly pleased to be here because it gives me an opportunity to give kudos to AGA regarding some progressive actions it has taken to improve safety.

But before I do that, let me give you a brief introduction about the NTSB and what we do, for those of you who may not already know.  The NTSB investigates accidents in all modes of transportation, including transportation by pipeline. We determine the probable cause of the accidents and make recommendations in hopes of preventing recurrences.

Because we are not regulators, we can’t require anybody to follow our recommendations.  Nonetheless, it is a testament to our amazing staff of investigators and our amazing staff of analysts that more than 4 out of 5 times, the recipients of our recommendations consider them to be very sound ideas for improving safety, and they act on them favorably.

Now to AGA’s progressive actions to improve safety.  Let’s start with having this meeting of industry leaders, which demonstrates your awareness that successful safety improvement programs must have support from the top.  Then there’s inviting us to speak, which demonstrates your desire to inform your safety improvement process with lessons learned from accident investigations.

So let’s talk about what we’re learning from our investigations.  One lesson we are learning is that as the industries we investigate are becoming safer, which they generally are, they face two challenges.

The first challenge is that as safety improves, complacency often rears its ugly head.  That’s not a criticism, but simply an observation of human nature.  \\

If we haven’t had an accident recently, we can easily fall into the trap of thinking that we’ve got this safety stuff figured out, so we often become less vigilant about noticing, and reacting to, potentially troubling signals that can be warning signs and harbingers of a future accident.  That is particularly easy to do if the signals are “weak” rather than glaring.  Safety, however, is not a destination, not something that we will ever reach and be done with, but a continuing journey.

The second challenge is that as our safety improves, and as we’ve taken care of the “low hanging fruit,” it becomes ever more challenging to continue improving safety.  Continuing improvement of safety often necessitates thinking out of the box and creating new improvement processes, rather than tweaking or strengthening previous processes.  The fact that a process has worked well to improve safety in the past does not necessarily mean that it will be continue to be effective in the future.  Even processes that have worked very well in that past may, with time and in changing environments and circumstances, become obsolete and ineffective.

Today I would like to focus on this second challenge of continuing to improve safety – the need for “out of the box” thinking.  In that vein, I would like to give AGA kudos for two ways in which you have progressively thought “out of the box:,” namely, commencing with safety management systems, and implementing a peer review program.

Let’s start with safety management systems.  We have recommended safety management systems in all transportation industries.  With an effective SMS in place, an organization not only takes reactive actions as a result of incidents and accidents that have already happened. The organization also takes proactive actions, actively seeking to identify hazards. Ideally, an effective SMS can result in analysis of system processes and the environment to identify potential and future problems and thus create a predictive aspect.

The SMS recommendation that I’ll talk about today resulted from the oil pipeline release from an Enbridge pipeline in Marshall, Michigan in 2010. 

In Marshall, an oil pipeline ruptured during the last stages of a planned shutdown. The rupture was not discovered for more than 17 hours, during which there were two attempts to restart the pumps. More than 840,000 gallons of crude oil was released, and more than 80 percent of that was pumped into the line after the rupture occurred.

The oil saturated the surrounding wetlands and flowed into the Talmadge Creek and the Kalamazoo River, and local residents self-evacuated.  As of last year, cleanup costs had reached $1.2 billion, making it the most costly cleanup involving an on-shore oil spill in US history.

By any measure, this was a disastrous rupture.  We found safety deficiencies in Enbridge’s integrity management program, control center staff training, leak detection processes, and emergency response and leak mitigation procedures.

We made several recommendations to Enbridge, and I am pleased to note not only that they cooperated with us very well in the investigation itself, which helped us identify what caused the release, but they have also put a great deal of hard work into implementing our recommendations.

What I would like to talk about today is not our  recommendations to Enbridge, but another very important recommendation that came out of the investigation – a recommendation that applies to all pipeline operators. This recommendation was directed to the American Petroleum Institute, and it reads as follows:

Facilitate the development of a safety management system standard specific to the pipeline industry that is similar in scope to your Recommended Practice [RP] 750, Management of Process Hazards. The development should follow the established American National Standards Institute requirements for standards development.

As many of you know, the result of this recommendation was API’s creation of ANSI/API Recommended Practice (or RP) 1173, Pipeline Safety Management Systems.

We generally endeavor in our recommendations to set the bar pretty high and encourage “stretch” thinking, so very rarely do recommendation recipients exceed what we recommend. Thus, I am pleased to report that when API issued RP 1173, we closed our recommendation as “exceeds recommended action.”

The API not only met our recommendation but exceeded it, by setting forth in RP 1173 a framework for continuous improvement in pipeline transportation safety. So what does this have to do with you?  What it has to do with you is that AGA deserves a large share of the credit for this result.  Kudos to Kate Miller, who represented the AGA in the development process. Many AGA member companies were also on the committee that drafted – and re-drafted – this standard.  Not to be satisfied with merely creating the standard, AGA is also promoting and facilitating the adoption of RP-1173 throughout the gas pipeline industry. Many AGA member companies are now engaging in a pilot implementation of the standard. They are committed to communicating with each other throughout the pilot period and sharing their stories with other AGA member companies at the conclusion of the pilot.

In drafting RP-1173, API collaborated broadly across the pipeline industry. In addition to operators of both gas and oil pipelines, and associations such as AGA and INGAA, API also involved the regulator – PHMSA – and invited the NTSB to attend meetings in an observer role. Representatives of the public with expertise in pipelines were at the table, as were representatives of two state corporation commissions.

The idea was for everybody who had a stake in the problem to have a voice in the solution.

We have seen the enormous power of collaboration in other industries, most notably in commercial aviation.  Their fatal accident rate, after declining marvelously for several decades, began to “flatten out” on a “plateau” in the early 1990’s.  Meanwhile, the FAA was projecting that commercial aviation volume would double in the next 15-20 years.  It didn’t take very advanced math to realize that doubling the volume, but maintaining the same fatal accident rate, meant that the public would soon see twice as many fatal accidents.

With the prospect of twice as many accidents looming, the industry did something it had never done before – work collaboratively to identify potential safety issues, prioritize them, develop remedies, and evaluate whether the remedies were working.  The collaboration, which operates under the name CAST, or Commercial Aviation Safety Team, included the operators, the manufacturers, pilots, air traffic controllers, and yes, even the regulator.  Bringing together all of these factions of the industry, with their differing and sometimes competing interests, was a major challenge for a variety of reasons.

The challenge was definitely worth it, however, and the CAST process has been a huge success. It resulted not only in a reduction of the stuck, flat fatal accident rate by more than 80 percent in its first ten years, but productivity was improved as well.  This flew in the face of conventional wisdom, that improving safety usually undermines productivity, and vice versa.  If that weren’t enough, it was also a success regarding another issue that commonly attends efforts to make changes in complex systems, namely, unintended consequences.  With everyone at the table, unintended consequences have been minimized.

This airline industry collaboration was enabled, in part, by a cultural characteristic of that industry that I believe your industry shares as well, namely, that anyone’s accident is everyone’s accident.  The flying public is not assuaged by the knowledge that an accident occurred on Airline X, but they are about to travel on Airline Y.  To the contrary, the public is concerned even by a German airplane crashing in the French Alps.  Similarly, the public does not focus on which pipeline operator suffered a release, or even whether the release involved oil or gas.  To the contrary, the public forms a collective, negative image of pipeline energy transportation that intensifies with every accident.  I suspect that your ability to collaborate in creating RP 1173 was helped somewhat by this cultural characteristic.

Bottom line:  kudos to you for your role in creating RP 1173, and I look forward to yet another example of the enormous power of collaboration.

Now I’d like to move to another example of progressive safety improvement thinking, and this one is more specific to AGA:  The peer review process. 

As I have noted, RP 1173 resulted from one of our recommendations.  We have not, however, made any recommendations about the peer review process.  Nonetheless, I would like to think that we played a role in it because after the gas pipeline explosion in San Bruno, we had a very frank meeting with senior AGA representatives.  One of the questions that we raised at that meeting was whether AGA looked to other industries to learn about best practices in safety.  Our question was essentially met with silence.

After that meeting, AGA began actively reaching out to other industries.  One that they reached out to was the nuclear power industry.  Nuclear power is certainly another industry in which anyone’s accident is everyone’s accident, and because of the accident at Three-Mile Island in 1979, they created a collaborative organization called the Institute of Nuclear Power Operations, or INPO. INPO demonstrated the power of collaboration long before the airline industry started CAST.  One of the many practices that INPO instituted was peer review, which at that time was largely unprecedented. 

INPO’s peer review process enables a review of one nuclear plant by experts from several other nuclear plants.  The process is voluntary, but it is widespread because the industry’s insurers pay close attention to the results of the peer review.  The process has been hugely successful because each plant has its own culture, and not only does the visited plant benefit from the look-see by experts from other cultures, but the visiting experts also benefit from learning more about the cultures of the visited plant and the other visiting experts. 

So AGA looked into this peer review process and decided to try it.  Bottom line here:  kudos to Christina Sames, of AGA, for pursuing the process of learning from other industries and for implementing the peer review process.  I fully expect that this will become yet another demonstration of the power of collaboration, both in terms of AGA’s collaboration with another industry, and the collaboration of the participants in the peer review process.

To sum up, safety is not a state of perfection. It is a continuing effort to catch imperfections before they cause any unwanted outcomes. To do so requires concerted efforts of management, employees, manufacturers, regulators, and the public, i.e., everybody with a stake in the safe transportation of energy products.

So to the AGA member companies who are implementing RP-1173, and to those who elect to be peer reviewed, I applaud you for pursuing these new directions, and I wish you the best of success.

Thanks again for inviting the NTSB – and I hope that this proves to be a productive Executive Leadership Safety Summit!