Category Archives: Transportation History

Twelve Years After Colgan 3407, FAA Still Hasn’t Implemented Pilot Records Database

By Chairman Robert Sumwalt

I grew up in the South, and people sometimes say we do things slowly in that part of the country. Whether there’s any validity to that claim, I can’t say with certainty. What I can say with great certainty, however, is that speed isn’t an attribute commonly associated with the Federal Aviation Administration (FAA), an agency within the US Department of Transportation. Below is a sad, but true, example of the glacial pace of the FAA’s rulemaking processes—even in the wake of a congressional mandate to get something done. Perhaps the new secretary of transportation can give a needed boost to this untenable situation.   

On this date 12 years ago—February 12, 2009—while on approach to the Buffalo‑Niagara International Airport in New York, Colgan Air flight 3407, a Bombardier Q-400 turboprop, plunged from the sky. Fifty lives were lost, including that of a man who died when the turboprop crashed into his home.

The NTSB’s year-long investigation revealed that, as the airplane slowed on approach, the captain became startled by the activation of the aircraft’s stall warning system. In response to something that should have been easily dealt with, the captain inappropriately manipulated the elevator controls, forcing the aircraft into its fateful dive. Our investigation found that the captain had a history of piloting performance deficiencies, including having failed several flight tests. Possibly more troubling, he concealed these performance deficiencies from Colgan when he applied for employment.

The Colgan crash was the deadliest US airline disaster in the past 19 years.

In response to this tragedy, the NTSB issued safety recommendations to the FAA to strengthen the way airlines ascertain a pilot applicant’s background, including requiring previous employers to disclose training records and records of any previous failures.

Congress took note of these recommendations and included them in a bill signed into law in August 2010. This law required the FAA to establish a pilot records database (PRD), and stipulated that “before allowing an individual to begin service as a pilot, an air carrier shall access and evaluate . . .  information pertaining to the individual from the pilot records database.” Items required to be entered into the PRD, and considered by hiring airlines, included “training, qualifications, proficiency, or professional competence of the individual, including comments and evaluations made by a check airman . . . any disciplinary action taken with respect to the individual that was not subsequently overturned; and any release from employment or resignation, termination, or disqualification with respect to employment.” Congress appropriated $6 million per year for the next 4 years to help facilitate creation of the PRD—a total of $24 million.

The FAA’s response reminds me of my college’s football team—they get off to a good start, but after scoring on the opening drive, they have difficulty executing for the rest of the game.

In early 2011, the FAA established an aviation rulemaking committee (ARC) to develop recommendations on the best way to implement the PRD. Despite the ARC completing its work and issuing a report to the FAA in July 2011—just 6 months after being tasked with developing recommendations—it wasn’t until September 2015 that the FAA began a phased approach to implementing the PRD.

By July 2016, Congress had become impatient with the FAA’s lack of progress. After all, it had been 6 years since the FAA was required to create the PRD, and there was still no appreciable progress. Congress gave the FAA a new deadline: it mandated the PRD be in place by April 30, 2017.

Unfortunately, April 30, 2017, came and went. Still no PRD. Meanwhile, 40 days after that deadline, a young pilot applied for employment at Atlas Air and was hired shortly thereafter. As with the Colgan Air captain, this pilot concealed his history of performance deficiencies, which deprived Atlas Air the opportunity to fully evaluate his aptitude and competency as a pilot. He struggled with training at Atlas, but after failing his check ride, he was retrained and passed. Tragically, on February 23, 2019, on what should have been a routine cargo flight from Miami to Houston, this pilot, like the Colgan Air captain, encountered something that startled him. He overreacted and put the Boeing 767 into a fatal dive. The commonalities between the Colgan Air crash and the Atlas Air crash are striking: Both pilots had a record of poor performance prior to their employment, both pilots concealed that information when applying for airline employment, and both pilots misapplied the flight controls following events they weren’t expecting. Events that should have been easily corrected. Events that, tragically, led to their aircraft plunging to the ground.

Neither of these sad events was an isolated case. Including these two crashes, the NTSB has investigated 11 air carrier accidents over 3 decades in which pilots with a history of unsatisfactory performance were hired by an airline and then were later involved in an accident attributed to their poor piloting performance.

After years of foot dragging, last March, the FAA provided its first visible indication of moving forward with the PRD, publishing a notice of proposed rulemaking (NPRM) to give the public a glimpse of what the proposed rule may look like—10 years after Congress initially mandated it, and 3 years after the April 2017 deadline that Congress eventually imposed.

The NPRM indicated that the PRD should be implemented sometime this year; however, the NPRM also proposes allowing a 2-year phase-in period. This puts complete implementation somewhere around a 2023 timeframe, assuming this proposed timeline holds. If that’s the case, we will finally have the PRD 14 years after the Colgan Air disaster, 13 years after Congress mandated it, 5 years after the deadline imposed by Congress, and 4 years after the Atlas Air crash.

A crash is a tragedy. It’s even more tragic to see a similar crash happen again and again and not have the regulatory agency responsible for safeguarding the skies take corrective action in a reasonable timeframe. We’re past the point of reasonable, and the traveling public deserves better.

We Can Do Big Things. Just Look at Positive Train Control

By Member Jennifer Homendy

After 50 years of investigation, advocacy, and persistence by the NTSB, positive train control (PTC) is now a reality across the country!

This video highlights the NTSB’s more than 50 year effort in investigating PTC-preventable accidents and advocacy for this life-saving technology.

PTC systems use GPS and other technology to prevent certain train collisions and derailments. It could have been lifesaving in the 154 rail accidents that have killed more than 300 people, and injured more than 6,800 passengers, crewmembers, and track workers in major accidents stretching across the nation, from Darien, Connecticut, in 1969, to Chatsworth, California, in 2008, to Philadelphia, Pennsylvania, in 2015, and DuPont, Washington, in 2017.

But let’s step back and marvel at this real achievement—and the effort it took. Safety improvements are never easy or quick. It took more than 50 years of advocacy by the NTSB and historic action by Congress to make PTC a reality. For many of these years, the NTSB was a lonely voice for safety, pushing for PTC despite opposition from railroads over the price tag and technological hurdles.

I know how tough the battle was because I was there. As staff director for the House subcommittee charged with overseeing rail safety, I played a role in ensuring that any effort to move legislation forward to improve rail safety included the NTSB’s recommendation to implement PTC. When I got to the NTSB, one of my priorities was to ensure that mandate was implemented.

It truly is remarkable in Washington to keep such clear focus on PTC across so many administrations, through so many changes in Congress and at the NTSB.

Earlier this month, I had the honor of moderating a panel of current and former NTSB leaders and staff who recalled the long, bumpy road to PTC implementation. NTSB Chairman Robert Sumwalt and former agency heads Chris Hart, Debbie Hersman, and Jim Hall recalled their own contributions and noted how remarkable the agency’s sheer persistence was in a time of short attention spans and quickly changing priorities.

It was so uplifting to hear their personal reflections of their time on the Board fighting for PTC, and their continued commitment to the agency and its critical safety mission. But it was the staff panel that really defined persistence. Generations of rail investigators and other staff worked every one of the 154 PTC-preventable accidents over the decades, launching to horrific crash scenes only to discover similarities pointing to the same solution: PTC. They spent holidays working. Missed birthdays and anniversaries. Completed their important jobs regardless of on-scene obstacles and personal priorities.

Recording of the January 14, 2021, NTSB live‑streamed discussion about Positive Train Control implementation.

The public doesn’t often see what goes on behind the scenes at accident investigations, after investigations are completed when recommendations need to be implemented, and the tremendous work required to keep those recommendations at the forefront of discussions to improve safety. As stated in the first panel, board members come and go, but it’s the staff that keep these critical safety issues alive. It was truly remarkable and heartwarming to hear their reflections of the agency’s work and how that work has impacted public safety, as well as how it affected them personally. I hope it gave the public a sense of what it takes to stay focused on an issue for five full decades.

Was it worth it? You bet. PTC will save lives.

Other safety improvements have also taken many years to implement. Midair collisions were dramatically reduced by the Traffic Alert and Collision Avoidance System (TCAS). That took decades to put in place. Airliner fuel-tank inerting systems, which addressed fuel tank explosions like the one that brought down TWA Flight 800 in 1996, also took years. And let’s not forget about the long fight for airbags and seat belts in passenger vehicles. All these transportation safety improvements were strongly and relentlessly advocated for by the NTSB.

We can do big things in America. We can save more lives on our rails, in the sky, in communities where pipelines are located, on the water, and on the highway. But major safety improvements like PTC take time, money and, perhaps most of all, incredible perseverance.

San Bruno Victims and Their Families Deserve Long-Overdue Action

By Member Jennifer Homendy

Today marks 10 years since the devastating natural gas pipeline rupture that shattered a residential neighborhood in San Bruno, California. The September 9, 2010, explosion destroyed 38 homes and damaged 70 others. Even worse, 8 people were killed, 10 people sustained serious injuries, and many others suffered minor injuries.

The Accident

When I think of San Bruno, I struggle with the ‘right’ words to describe the horrific events that unfolded shortly after 6:00 p.m.—a time when many families across our nation are just sitting down for dinner.

In the moments after the rupture, calls flooded into 911, with reports of what many thought was a plane crash, a gas station explosion, or some combination of the two. One caller said it felt like an earthquake, and a fire captain who was on scene said, “It looked like Armageddon.” In fact, the rupture was so explosive that it produced a crater about 72 feet long by 26 feet wide and launched a 28-foot section of failed pipe about 100 feet south of the crater. The released gas almost immediately ignited. Emergency responders arrived within minutes to battle the ensuing inferno, yet it took Pacific Gas & Electric (PG&E) an astonishing 95 minutes to shut off the flow of gas that was intensifying the destruction. Firefighting efforts continued for 2 days, with 600 firefighters and 325 law enforcement personnel on scene.  

San Bruno, CA, accident scene with the crater in the foreground and the ruptured pipe section in the background
San Bruno, CA, accident scene with the crater in the foreground and the ruptured pipe section in the background

NTSB Warnings

I’m not going to get into the numerous failures at PG&E that led to the rupture. I want to focus on those 95 minutes. In December 1970, the NTSB released a Special Study of Effects of Delay in Shutting Down Failed Pipeline Systems and Methods of Providing Rapid Shutdown. You read that right—1970. We found that delays in shutting down pipelines increase the magnitude of catastrophe, and that, when the flow of gas or hazardous liquid is stopped soon after an initial rupture, the effects of many accidents would have been minimized or eliminated. In other words, numerous lives could’ve been saved, and injuries prevented.

Our report highlighted the 1968 rupture of a medium-pressure gas line in front of a daycare in Hapeville, Georgia. Construction crews on scene were unable to locate the buried valve to shut off the gas flow. A few minutes later, an explosion occurred inside the daycare. The ensuing fire engulfed the building and nine people were killed, including seven children. Three other children were seriously injured.

Nine other incidents—all involving failures to shut down pipelines—were cited in the report, and many more have occurred since it was published. The common theme? What we said in 1970 held true in San Bruno and holds true today: “For every one of the accidents cited, there are devices or equipment currently available which probably could have prevented the accident or greatly minimized its effect.”

We’ve been urging federal regulators to require those devices for 50 years! In fact, they’re still on our Most Wanted List of transportation safety improvements.

The San Bruno Investigation

Getting back to San Bruno. In those crucial 95 minutes during which the gas continued to flow, PG&E control center staff knew there had been a rupture along the pipeline, but never once called 911. The three PG&E employees who first arrived on scene, two of whom were supervisors, had no idea how to operate mainline valves. They had to call people who were qualified to operate them, and by the time those mechanics located the valves and got to the first one, it was 7:20 p.m., over an hour after the rupture occurred. Meanwhile, the fire, described by NTSB investigators as a massive blowtorch, was still raging.

Because gas was being supplied to the break from both the north and the south, the shutoff valves closest to the break had to be closed to shut down and isolate the rupture. The shutoff valves were located about 1.5 miles apart, on either end of the break, and they had to be shut manually. Had PG&E installed readily available technology—valves with remote closure capability or ones that would automatically shut off the gas flow in response to pressure changes in the line—the amount of time the fire burned, and thus, the severity of the accident, could’ve been significantly reduced. In fact, this technology could’ve stopped the flow of gas the moment the rupture was detected.

In our final report on the accident, we recommended that federal regulators—the Pipeline and Hazardous Materials Safety Administration (PHMSA)—require  pipeline companies to install automatic shutoff valves or remote shutoff valves in High Consequence Areas (defined as populated areas, drinking water sources, and unusually sensitive ecological areas).

PHMSA’s Response

On February 6, 2020, PHMSA published a notice of proposed rulemaking (NPRM), “Pipeline Safety: Valve Installation and Minimum Rupture Detection Standards,” claiming the NPRM responds to recommendations from the NTSB. It doesn’t. It requires automatic shutoff valves, remote-control valves, or equivalent technology to be installed only on newly constructed or entirely replaced onshore natural gas transmission and hazardous liquid pipelines that are larger than 6 inches in diameter.

Remember the daycare accident I mentioned? The pipeline that ruptured in that tragedy was only 1 inch in diameter. Existing gas transmission lines (like the PG&E line that ruptured in San Bruno), newly constructed or entirely replaced lines that are less than 6 inches in diameter, gas distribution systems, and offshore transmission lines are completely excluded from the NPRM’s requirements.

In other words, PHMSA’s solution won’t prevent another San Bruno disaster. Given that there are 2.6 million miles of gas pipelines in the United States, most of which date back to the 1950s and the NPRM doesn’t address any of them. With those numbers, another tragic accident is destined to occur, and if I’m the member on scene—or even if I’m not—I’ll remind PHMSA and industry, yet again, of all the ruptures we’ve investigated and all the opportunities they had to save lives.

To all those who lost loved ones in San Bruno or in another pipeline tragedy, you remain in our hearts. We are still fighting for you.

Arriving Soon: Fully Implemented Positive Train Control

By Member Jennifer Homendy

December 31, 2020—not only will it be the last day of an incredibly challenging year that I think we’ll all be happy to put behind us, it’s also a significant day for railroad safety. It’s the final deadline for all 41 railroads to fully implement Positive Train Control (PTC). It’s been a long journey to get to this point and I’m thrilled to see the great progress that’s been made over the years. There were times no one believed we’d get to where we are today, so how did we get here?

PTC is a communications-based system designed to automatically stop a train before certain accidents occur. It won’t prevent all train accidents, like vehicle-train accidents at grade crossings or those caused by track and equipment failures, but it is designed to prevent train-to-train collisions, overspeed derailments, incursions into established work zones, and train movement through switches left in the wrong position.

The concept of PTC isn’t new. In fact, the NTSB has been urging railroads to implement PTC in some form—and federal regulators to mandate it—for over 50 years. Our first recommendation related to PTC (Safety Recommendation R‑70‑20) was issued following a deadly train collision in Darien, Connecticut, in August 1969, when two Penn Central commuter trains collided head on, killing 3 crew members and 1 passenger, and injuring 43 others. Twenty years later, the NTSB included PTC on its first Most Wanted List of transportation safety improvements (MWL), and, with the exception of 4 years following enactment of the Rail Safety Improvement Act of 2008 (RSIA; Public Law 110-432, Division A), it’s remained on the list to this day.


Before the passage of the RSIA, we had been recommending this lifesaving technology for decades, yet little action had been taken to implement its use. Even the Federal Railroad Administration (FRA), which is the federal agency charged by Congress with ensuring “the assignment and maintenance of safety as the highest priority,” had rebuffed repeated NTSB calls for implementing PTC, asserting that the technology was too expensive and that it would provide little safety benefit—a claim that was seemingly inconsistent with the August 1999 Railroad Safety Advisory Committee report, Implementation of Positive Train Control Systems, which found that, out of a select group of 6,400 rail accidents that occurred from 1988 to 1997, 2,659 could have been prevented if some form of PTC had been implemented.

Most people who follow the history of PTC will say it was the 2008 Metrolink crash in Chatsworth, California, that really brought PTC to the attention of Congress, but that’s not entirely true. In fact, it was a string of PTC‑preventable accidents that occurred in the early to mid-2000s that finally caused the issue to make headway. Six of these accidents were accidents that we investigated in 2004 and 2005. They occurred in Macdona, Texas; Graniteville, South Carolina; Anding, Mississippi; Shepherd, Texas; Chicago, Illinois; and Texarkana, Arkansas.

The ones I remember most were Macdona and Graniteville; they were, in part, the reason the PTC mandate applies, not just to main lines over which passengers are transported, but also to main lines over which poisonous or toxic-by-inhalation hazardous materials are transported.

On June 28, 2004, a Union Pacific (UP) freight train struck the midpoint of a BNSF freight train traveling on the same main line track as the BNSF train was entering a siding. Chlorine escaping from a punctured tank car immediately vaporized into a cloud of chlorine gas that engulfed the area. The conductor of the UP train and two Macdona residents died as a result of chlorine gas inhalation. About 30 others were treated for respiratory distress or other injuries related to the collision and derailment.

Just 6 months later, on January 6, 2005, a Norfolk Southern train transporting chlorine encountered a misaligned switch that diverted the train from the main line onto an industry track, where it struck an unoccupied, parked Norfolk Southern train, killing the 28-year-old train engineer, Chris Seeling, and eight others as a result of chlorine gas inhalation. About 554 people suffering from respiratory difficulties were taken to local hospitals; 5,400 others within a 1-mile radius of the derailment site were evacuated for several days.

Both accidents were preventable with PTC.

I wasn’t at the NTSB at the time. I was serving as staff director of the US House of Representatives Subcommittee on Railroads, Pipelines, and Hazardous Materials. Congressman James Oberstar (D-MN) had just been named chairman of the committee, and one of his first acts as chair was to launch a series of hearings focused on rail safety. (Coincidentally, the subcommittee’s first hearing on the topic was the first time Chairman Sumwalt, then an NTSB board member, testified before Congress.)

One of those was a field hearing held in March 2007 in San Antonio, Texas, where Ralph Velasquez, a resident of Macdona, described his family’s escape from “the cloud of chlorine” and the tremendous physical, mental, and emotional toll the accident had taken on his family and the entire community. Mr. Velasquez’s words were similar to those of Chris Seeling’s parents, who had visited me and Chairman Oberstar months earlier. Both families wanted action, including implementation of NTSB recommendations.

Two months later, Chairman Oberstar delivered on his promises. The House approved legislation that mandated longstanding NTSB recommendations and established a deadline for PTC implementation. The Senate passed its bill in 2008, and we were in the process of finalizing a bipartisan, bicameral bill to include PTC when a Metrolink commuter train collided head-on with a Union Pacific freight train, killing 25 people in Chatsworth, California.

The tragedy in Chatsworth—which the NTSB later determined was preventable with PTC—gave the legislation its final push, and in October, the RSIA was signed into law, mandating PTC implementation by December 31, 2015. This deadline was later extended by Congress to 2018, and then again by the FRA on a case-by-case basis to 2020.

Since the RSIA was signed into law in 2008, the NTSB has investigated 25 accidents that would’ve been prevented had PTC been implemented, including the overspeed derailments of Amtrak passenger train 188 in Philadelphia, Pennsylvania, which killed eight people onboard, and Amtrak passenger train 501 near DuPont, Washington, which took three lives and injured more than 50 others. In all, since that first accident investigation in 1969, over 300 people have been killed and almost 7,000 others have been injured in 154 accidents that the NTSB determined could have been prevented if PTC had been operational. When people think about the 2008 mandate and how long it’s taken the railroads to implement that mandate, they get frustrated with Congress for extending the deadline, but what they fail to remember is, if it weren’t for congressional action, we wouldn’t be where we are today. The railroads wouldn’t have implemented PTC voluntarily.

Today, PTC data submitted to the FRA is looking far more positive than in the past. Since I joined the Board in 2018, I’ve visited or spoken to a majority of the 41 railroads regarding their PTC status, and they’re mostly reporting good progress. Most railroads are expected to meet the end-of-year deadline, and I’m cautiously optimistic that all 41 will succeed.

Because of the NTSB’s tireless advocacy, beginning long before I joined the Board, and the hard work of our investigators, the finish line is at last in sight. Our investigators work diligently to prevent tragedies like Macdona and Graniteville from recurring. If Chairman Oberstar was alive today, he would call them heroes for their dedication and for all they’ve done to save lives.

Today marks 154 days until the latest deadline for PTC implementation. That’s also the number of PTC-related accidents we’ve investigated over the years. For the rest of this year, stay tuned to our social media channels, where we’ll share information daily about each of the accidents (look for #PTCdeadline). These accidents are a reminder of how much we’ve lost while waiting for the implementation of PTC.




Remembering Marshall, Michigan

By Member Jennifer Homendy

This Saturday marks the 10th anniversary of one of the largest and most expensive inland oil spills in our nation’s history.

At 5:58 p.m. on July 25, 2010, a 30-inch diameter pipeline owned and operated by Enbridge ruptured, releasing nearly a million gallons of heavy crude oil into Talmadge Creek, which feeds into the Kalamazoo River – a tributary of Lake Michigan – and flowed about 35 miles downstream before it was contained.

Hazardous Liquid Pipeline Rupture and Release, Marshall, Michigan, July 25, 2010
The ruptured segment of Line 6B in the trench following the July 25, 2010, rupture. The fracture face measured about 6 feet 8.25 inches long and was 5.32 inches wide at the widest opening. The fracture ran just below the seam weld that was oriented just below the 3 o’clock position. A red circle shows a location where the coating was wrinkled and had separated from the pipe surface.

Although numerous alarms were triggered in Enbridge’s control center, located in Edmonton, Alberta, Canada, control center staff failed to recognize a rupture occurred for well over 17 hours, until an outside caller contacted the control center. Enbridge attributed the alarms to an earlier planned shutdown and column separation (a vapor-liquid void), and instead re-started the line twice for a period of 1.5 hours, pumping massive amounts of oil (81 percent of the total release, or over 600,000 additional gallons) into the pipeline. Once Enbridge realized there was a release, it was too little too late. Enbridge had only four maintenance personnel on scene; the closest trained responders – their oil spill response contractors – were 10 hours away.

At the time, I served as the Democratic Staff Director of the Subcommittee on Railroads, Pipelines, and Hazardous Materials for the U.S. House of Representatives, which had jurisdiction over the safety of oil and gas pipelines in the United States. About 24 hours after the rupture, the Chairman of the Committee, Congressman Jim Oberstar, and Congressman Mark Schauer who represented Marshall, Michigan, asked that I travel to Marshall and lead the Committee’s oversight investigation of the spill, which is different than an NTSB investigation (for example, we looked at claims and HIPAA violations).

I wish I had the right words to describe what it was like when we arrived in Marshall. Utter devastation doesn’t seem to do it justice. Oil blanketed the creek and river, the river’s banks, and flood plains, severely impacting the environment. Rescue and rehabilitation efforts for oiled birds and wildlife continued for months; river restoration went on for years. Clean-up costs totaled $1.2 billion, and Enbridge received the largest civil penalty for a Clean Water Act violation in U.S. history, and the second-largest penalty overall, after Deepwater Horizon.

July 25, 2010, Marshall, Michigan pipeline rupture
Cleanup efforts in an oil-soaked wetland near the rupture site. Saturated soil complicated the cleanup and excavation efforts. An excavator with a vacuum attachment is shown situated on wooden matting near the rupture site.

While, thankfully, no lives were lost, people lost homes and businesses, as well as income, and about 320 residents suffered symptoms consistent with exposure to crude oil.

Perhaps the most memorable moments for me were with the residents in Baker Estates in Battle Creek, Michigan, a community of about 70 mobile homes right along the river. I walked the oil-saturated river banks with them and was invited into their homes to hear about financial and medical impacts of the spill. See, no one evacuated the mobile home park. In fact, no one evacuated anyone along the river. County health officials issued a voluntary notice for homeowners to self-evacuate, which was noted in the NTSB accident report.

Meanwhile, NTSB’s investigation focused on the cause of the rupture and the oil spill response. NTSB’s former chairman Debbie Hersman was on scene. I ran into her and Peter Knudson, who is still a crucial part of NTSB’s media relations team, eating dinner one night in Marshall. Who knew we’d one day work together at the agency?

Through the investigation, the NTSB identified numerous gaps in Enbridge’s integrity management program, control room operations, training, and leak detection. To address the multitude of deficiencies, NTSB recommended that the pipeline industry develop an industry standard for a comprehensive safety management system (SMS) specific to pipelines.

Years later, I’m pleased to say that the industry didn’t just meet the intent of our recommendation; they exceeded it with the development of API Recommended Practice 1173, which also focused on safety culture and other safety-related issues. Since then, many pipeline operators have adopted and implemented the standard. The NTSB is working to encourage others, from the largest pipeline operators to the smallest municipalities, to implement SMS.

NTSB’s oil spill response investigation identified issues with the advance preparation and execution of the response that could be traced to the Pipeline and Hazardous Materials Safety Administration’s (PHMSA) regulations implementing the Oil Spill Prevention Act of 1990.

The PHMSA-approved Enbridge facility response plan did not provide for sufficient resources to deal with an oil spill of this magnitude. Furthermore, the NTSB investigation found that the severity of the oil spill could have been minimized had Enbridge focused more on source control and used oil containment methods that were appropriate for the environmental conditions. In response to NTSB recommendations, PHMSA undertook an effort to update the regulations and harmonize them with U.S. Coast Guard regulations for oil spills in navigable waterways.

Overall, there were a lot of safety gaps identified because of the Marshall spill. The NTSB’s work and the Committee’s oversight investigation led to a series of hearings that culminated in passage of sweeping legislation in 2011, which is still being implemented, albeit slowly, a decade later.

But looking back, a decade later, well after our investigators have left the scene, our final report has been issued, and recommendations are being acted upon, I think about the residents of Baker Estates and the other communities and business owners that suffered tremendous losses. For them, the work is just beginning, and Marshall will never be forgotten.

And it shouldn’t be forgotten because when the industry fails to learn from previous accident investigations and fails to make necessary changes, those accidents and the underlying issues that caused them are destined to repeat themselves.

In fact, as I sat down to write this blog, I recalled our 2005 safety study on Supervisory Control and Data Acquisition (SCADA) in pipelines. SCADA systems are essentially a computer system that allows control center staff to monitor and control the pipeline from a remote location. The study was prompted by 12 hazardous liquid accidents investigated by the NTSB in which leaks went undetected after indications of a leak were provided on the SCADA system: Brenham, Texas (1992), Gramercy, Louisiana (1996), Fork Shoals, South Carolina (1996), Murfreesboro, Tennessee (1996), Knoxville, Tennessee (1999), Bellingham, Washington (1999), Winchester, Kentucky (2000), Greenville, Texas (2000), Chalk Point, Maryland (2000), and Kingman, Kansas (2004).

Fork Shoals was eerily similar to Marshall. The pipeline owned and operated by Colonial Pipeline ruptured, releasing nearly one million gallons of fuel oil into the Reedy River and surrounding areas at Fork Shoals. Like Marshall, the SCADA alarms and alarm messages had activated, and the controller acknowledged them, but he failed to recognize that a rupture had occurred and continued pumping more and more fuel oil into the line after several shutdowns and re-starts. And like Enbridge, Colonial knew of the corrosion in the line in the months leading up to the rupture.

If you’re on the fence on SMS, I hope this prompts you to take heed and not wait for a rupture to occur to act.


NTSB Office of Rail, Pipeline and Hazardous Materials Investigations Director, Rob Hall, contributed to the writing of this blog.