Presentation to the International Heavy Haul Association (IHHA)
Lessons Learned from Accident Investigation of Longer, Heavier Trains
Transportation Safety Board

Delivered by:
Jonathan Seymour
Calgary, Alberta
June 20, 2011

Click here to see PowerPoint Presentation (PPT)  [2133 KB]

Check Against Delivery

Slide 1: Title Page

Good afternoon, and thank you for the opportunity to speak today on the lessons learned from TSB investigations of accidents involving longer, heavier trains.

I am accompanied today by Mr. Daoxing Chen, Senior Railway Dynamics Engineer, from our Engineering Laboratory. Mr. Chen, who is the author of the "case study" papers submitted, will be available afterward to answer questions regarding the papers.

Slide 2: Outline

Today my discussion will:

  • provide information about the TSB;
  • explain our concern with respect to inappropriate marshalling and handling of longer, heavier trains, and why is this a Watchlist Issue;
  • provide an overview of 2 illustrative case studies (Brighton, Drummondville);
  • outline the findings from other relevant TSB investigations and the lessons learned.

Then, following a review of the progress made in addressing these concerns, I will take a look ahead from a TSB perspective.

So, let me start with background information about the TSB.

Slide 3: About the TSB

The TSB is an independent government agency. We are not the regulator, and we have no regulatory power�that's Transport Canada. Our mandate is to advance transportation safety in the air, marine, rail and pipeline modes that are under federal jurisdiction by:

  • conducting independent investigations
  • identifying safety deficiencies
  • identifying causes and contributing factors
  • making recommendations
  • publishing reports

We exist for one reason�to advance transportation safety.

The TSB is comprised of approximately 230 employees in 8 regional offices nationwide. Investigators' qualifications include both operational and technical experience in each of the modes. Investigators are supported by professionals in various disciplines such as Human Factors, Macro-Analysis, Communications and other support groups. We have our own well equipped Laboratory facilities but also make use of contracted technical services as required.

We receive notice of roughly 4000 occurrences per year. In 2010, just over 1000 of these were in the rail mode. When that happens, we deploy our investigators, assess the situation and, if we feel there is a safety lesson to be learned, we launch a comprehensive investigation.

Slide 4: Watchlist

In March 2010, the TSB released a Watchlist, which highlights nine critical safety issues posing the greatest risk to Canadians based on findings from our investigations.

The Watchlist originated when TSB investigators began reporting troubling patterns in their work, arriving at the scene of an accident only to find repetitive safety issues as causal or contributory. We therefore developed it as a "blueprint for change" � a way to restate the TSB's safety messages, stimulate discussion, and generate further action on the part of industry and regulators: that is the "change agents."

On the slide the nine issues are itemized in brief. There are 2 Marine, 2 Rail, 3 Air and 2 Multi-modal issues. The one of particular interest today is highlighted in its abbreviated form as "Operation of Longer, Heavier Trains".

Slide 5: Watchlist (continued)

Underpinning these nine Watchlist issues are a series of 41 recommendations, and numerous investigation findings. Action taken to date by the relevant modal "change agents" has not fully addressed all of them, and more still needs to be done.

In the rail mode, the highlighted issue in its full wording reads, "Inappropriate handling and marshalling can compromise the operation of longer, heavier trains." In the case of this issue, progress has and is being made, but ongoing initiatives are still necessary.

Today, this is the focus of my talk.

Slide 6: Why This Is An Issue

Freight trains cross the country every day, and today's mixed-merchandise manifest freight trains are longer and heavier than ever: over 25 per cent from just 1995. These can and do stretch three kilometers and more.

As they continue to grow, so too can the in-train forces resulting from locomotive configuration, train marshalling, equipment characteristics, train handling and territorial characteristics. The relevant TSB investigations have primarily been related to such mixed manifest freight trains, as opposed to single-commodity or unit trains.

Since 2000, the TSB has investigated at least 12 such accidents�three of them in 2007 alone. As these longer and heavier trains see expanded use across Canada, including into the country's busiest traffic corridors, the consequences of any derailment can become magnified. It is therefore important that those who manage, identify, monitor and oversee the risks�especially operators�mitigate them.

Now I'd like to illustrate some of the issues by providing an overview of Daoxing's papers and then follow with a summary of other relevant TSB findings. For anyone interested, detailed investigation information can be accessed through the TSB website.

Slide 7: Brighton Case Study #1 (R09T0092)

On a March morning in 2009, a 137-car freight train was headed east through undulating terrain near Brighton, in eastern Ontario. It had a gross weight of 11845 Tons was about 8850 feet in length, and powered by three head-end locomotives.

Just over half of its 137 cars were loaded, with the majority at the rear of the train. As a consequence, there were a number of lighter empty cars marshaled between blocks of heavy loaded cars.

The terrain resulted in accordion-like slack action within the train. At one point a weak knuckle failed, causing a break-in-two and triggering an emergency brake application. A subsequent collision between the two separated train segments then occurred. Although no one was hurt and there was no loss of product, six cars derailed�including three loaded with molten naphthalene.

This multi-track subdivision accommodates inter-city passenger trains operating at up to 100 mph, as well as many freight trains daily. In addition, it passes through many communities and over or adjacent to several waterways.

Slide 8: Brighton train/track profile

This slide gives an indication of the train tonnage and the track profile.

Slide 9: Brighton: Findings

During the investigation, the derailment site was carefully surveyed, evidence was collected, and the data from the locomotive event recorder were downloaded and analyzed.

Simulations confirmed the train was proceeding through undulating territory at about 50 mph when a moderate run-out of train slack resulted in a broken knuckle on the 107th car. This caused a separation into two parts and an emergency brake application. The train separated into a head-end portion of 107 cars and a rear-end portion of 30 heavy loaded cars.

The Board's findings included:

  • the heavier rear portion�which was on a descending grade�did not brake as effectively as the head-end portion, and collided with the slower moving head-end portion
  • following the collision, in-train forces were sufficient to result in the derailment
  • bail-off of the independent brake was unable to reduce the buff force to a safe level
  • the simulation also demonstrated that with different marshalling and where an undesired emergency brake application occurred, the maximum in-train buff force could be significantly reduced.

Slide 10: Drummondville Case Study # 2 (R07D0009)

In February 2007, a freight train, while proceeding eastbound at about 30 mph between Montreal and Quebec City, derailed 8 cars at Mile 98.8 of the single-track Drummondville Subdivision. The train, which was approximately 7006 feet in length, was composed of five head-end locomotives and 105 cars, for a total train weight of 10 815 tons.

Of the 80 loads on the train, a 50-car block of heavy loaded grain cars was marshalled on the rear end. A broken knuckle on the 75th car (one of the loaded grain hoppers) caused a train separation and an emergency brake application to propagate in both directions from that point. There were also dangerous-goods "residue" cars involved, but there was no release of product, and no injuries were reported.

As is normal during a TSB investigation, the data from the locomotive event recorder were downloaded and analyzed. A train dynamics simulation was conducted to determine the in-train forces and the contributions from train marshaling, independent brake, and other factors.

This single-track subdivision accommodates inter-city and the eastern trans-continental passenger trains operating at high speeds, as well as many freight trains daily. In addition, it passes through many communities and over or adjacent to several waterways.

Slide 11: Drummondville train profile

This slide depicts profiles of the train tonnage.

Slide 12: Drummondville: Findings

The Board's causal and contributory findings included the following:

  • he train was marshalled with empty cars, residue cars, and light loads on the head end, and a block of 50 loaded grain cars on the rear of the train.
  • The front portion of the train was on an ascending grade when the emergency application signal arrived in the lead locomotive, whereas the rear portion was on a relatively flat portion of track.
  • A derailment ensued because of high in-train buff forces resulting from a combination of the heavy rear-end marshalling, and the locomotive engineer's late "bail off" of the independent brake.
  • The simulation confirmed that, if the train had been marshaled in a reverse profile of heavy head /light rear, the buff forces at most cars would have been minor.

Slide 13: Other Investigation Findings

These two case studies underline the significance of managing in-train forces. Other TSB investigations have also identified additional matters that can result in excessive in-train forces and further underscore the need for their effective management, including:

  • Inappropriate use of throttle, dynamic braking and automatic braking
  • Emergency brake application initiated from the head-end only
  • Equipment that has non-alignment controlled couplers
  • Long and short car combinations
  • Equipment with end-of-car cushioning
  • Using distributed power on manifest trains can significantly reduce in train and on track forces�particularly in curved and undulating territory.

Slide 14: Lessons Learned

I've identified a number of findings in the two case studies and from other occurrences investigated by the TSB since 2000.

It is important to make a distinction. Length and tonnage alone are not the ultimate issue. These are just two of the variables that can affect in-train forces. The most significant lesson to be learned from TSB investigations is the need to effectively manage in-train forces and how a train interacts with the track. Application of technology can greatly assist in different ways. This requires a systemic approach by operators, including risk assessments that are carried out in accordance with company safety management systems.

Given that this audience is well-versed in operations and that there will be related industry presentations, I won't go into detailed discussions regarding different factors such as marshalling, handling, distributed power, defenses in case of undesired emergency applications, and the use of relevant technologies. Rather, I would like to discuss the "state of play." In other words, having published our findings over the years, and having issued our Watchlist in 2010 to encourage continued progress � I ask these three questions:

  • Where are we now?
  • What is happening?
  • And are the lessons learned being applied?

Slide 15: Progress by Industry

With respect to the industry, this is what we have seen taking place:

Both CP and CN have taken and are taking action.

Given the significant affect of car marshalling, CP has developed TrAM, a proprietary computer program to reduce in-train forces and their effects. TrAM, which stands for Train Area Marshalling, considers factors such as the type of car, as well as its length, weight, the length of adjacent cars, track grade, and curvature. Although the company also uses destination-block marshalling, this does not take precedence over TrAM restrictions. CP also uses distributed power to reduce the forces within the train and on the track.

In addition, CP has adopted enhanced Train Information Braking System (TIBS) technology that includes an end-of-train auto-emergency feature and has equipped 94 per cent of its locomotives. All newly purchased road locomotives are to be equipped with the new technology.

All told, these measures have contributed to significantly reduce the number of CP rail accidents related to in-train forces.

CN, in addition to increasing its use of distributed power on mixed merchandise trains, has also developed a computerized marshalling system aimed at reducing in-train forces. These, too, take into account various factors such as car length, equipment characteristics, the train's overall weight distribution, and territorial characteristics. CN too has adopted end-of-train braking technology and equipped many of its road locomotives.

Each of these proprietary marshalling management systems identifies deviations from the "rules" for correction prior to departure. En-route validation for adherence to the rules is also available to crews related to pick-ups or set-offs.

Moreover, both CN and CP have also augmented training for locomotive engineers, providing them with tools such as guidance and/or train handling aids which are both general and specific to a territory. Technology to improve locomotive engineer train handling includes smoothing throttle action such as with "Trip Optimizer."

Technological defenses, such as wayside systems that protect against many other factors, also afford defenses against undesired emergency applications. These include hot-box detectors, hot- and cold-wheel detectors, wheel impact load detectors, and wheel profile detectors. These are augmented by enhanced approaches to the inspection and maintenance of track integrity.

Slide 16: Progress - Regulator

There are practically no Canadian regulatory requirements for train marshalling or tonnage distribution within a train, aside from the Transportation of Dangerous Goods Regulations. However, Transport Canada has supported TSB's views and so informed the companies they regulate.

In addition, TC has sponsored research on related topics. For example, one project involves the issue of train separation on the busy Kingston Subdivision.

A second major project is aimed at finding ways to improve the handling of longer trains.

To this end, TC has initiated a research project aimed at in-train forces and marshalling practices to enhance safety. The research will be used to develop relationships between train marshalling and track geometry that include consideration of track grade, track curvature, train length, train weight, total power requirement, power placement within the train, longitudinal in-train forces, wheel lateral forces and the potential for train derailment. The research results will be used to establish policies and guidelines relating to train marshalling, power distribution and related track geometry/superelevation factors that minimize or optimize in-train forces to ensure safe train operations.

Slide 17: Progress: a TSB Perspective

Over the past few years, and in addition to the findings from our investigations, the TSB has issued numerous safety communications dealing with the safe operation of these trains.

In 2004 for example, a TSB recommendation urged Transport Canada (TC) to "encourage railway companies to implement technologies and/or methods of train control to ensure that in-train forces generated during emergency braking are consistent with safe train operation."

In 2007, the Board expressed concern about the frequency of derailments caused or accentuated by in-train forces, particularly in transportation corridors that traverse densely populated areas.

It was just over a year ago that our Watchlist gave additional prominence to and identified inappropriate marshalling and handling of longer, heavier trains as a critical safety issue. Happily, we can now say that although the TSB would have preferred to see more uniform progress over the intervening years, we are seeing significant advances across the industry and with support from the regulator.

Slide 18: What's Next?

It is the operators who are responsible for effectively managing the safety of their operations and operating practices. Transport Canada's regulatory role is to oversee the safety of railways.

The TSB's role is to investigate occurrences, identify and report on safety deficiencies, make recommendations, and promote safety improvements.

Consequently, the TSB will continue to do the following:

  • monitor progress
  • investigate occurrences
  • publish our findings
  • make appropriate recommendations
  • advocate for any necessary changes to ensure that safety deficiencies are effectively addressed.

Slide 19: Summary

So, this afternoon I spoke about the TSB and explained our concern with respect to inappropriate marshalling and handling of longer, heavier trains, as well as why this was identified as a key "Watchlist" issue.

As the case studies of both Brighton and Drummondville clearly illustrate, the most significant lesson to be learned is the need to effectively manage in-train forces and how a train interacts with the track. Application of technology can greatly assist in different ways, and this requires a systemic approach by operators.

The good news is that there has been some progress in addressing the TSB's concerns. Both CN and CP are taking action and, as we go forward, the TSB will continue to monitor that progress, investigate future occurrences, and report publically. Where change is necessary, we will advocate for it, so that safety deficiencies are effectively addressed.

So with that, I thank you once again for the opportunity to speak with you today.

Now we'll take questions.

Slide 20: Questions?

Slide 21: End