Sunday, July 14, 2013

The Quebec Rail Disaster: Nine Notorious Necessities

In the science of logic, a necessary condition for a thing to occur is a circumstance that will stop the thing from occurring if the circumstance is not present.  Having fuel in my car is a necessary condition for getting it to run.  No fuel, no go. 

In the study of industrial disasters, you often find that instead of a single cause, there are multiple interlocked causes, each one of which was a necessary condition for the accident to take place.  The longer the chain of necessary conditions, the less likely it is that all of them will happen in a way that leads to a problem.  This is one reason why such situations are hard to anticipate.  The disaster that struck the town of Lac-Mégantic, Quebec on the night of July 5 and 6 is just such a tragedy.  While the full details have yet to be investigated, by going only a little beyond what is known I can assemble a chain of nine separate conditions, each of which had to prevail in order for the accident to happen. 

Lac-Mégantic is a community of about 5,000 on the tip of a lake of the same name, near the Canada-Maine border.  It was founded during the construction of the Canadian transcontinental railway in 1884, and a heavily-used rail line still runs directly through the center of town and extends westward up a rather steep grade to the next town of Nantes.  About 11 P. M. on the night of July 5, a five-locomotive eastbound train consisting of 73 tank cars filled with crude oil, besides other cargo, pulled into Nantes and stopped.  The engineer, Tom Harding, was done with his day’s run and planned to spend the night in a hotel in nearby Lac-Mégantic.  Before he left for the night, he followed what was apparently standard procedure in securing the train:  setting the handbrakes on all five locomotives and ten freight cars, and leaving one engine running to maintain pressure on the air brakes.

Railroad air brakes were an invention of George Westinghouse, who had the clever idea that a loss of brake pressure, such as would occur if part of a train broke away and lost its air connection to the engine’s compressor, should apply the brakes.  The same basic system is used today.  Each car has an air storage tank that provides the actual pressurized air that is applied to the brake cylinders on each “truck” or set of wheels.  The air from the tank is valved to the brakes when the main-line pressure falls, either gradually for controlled braking or suddenly, as when a train comes apart. 

But each tank holds only so much air, and so the other function of the air in the line is to maintain pressure in the tanks on each car. 

By setting the handbrakes, Harding was taking extra precautions.  In principle, as long as the engine was running to maintain air pressure, the air tanks would stay pressurized and the brakes would hold.  Satisfied that he had done his job, he made his way to Lac-Mégantic and to bed.

At 11:30 P. M., a concerned citizen reported to the Nantes fire department that the one running engine was on fire.  Local firemen reported to the scene and rousted out the nearest available rail employee, who turned out to be a track worker unfamiliar with the operation of locomotives.  The firemen and the railway employee managed to turn off the engine and put out the fire, but no one re-started the engine after the fire, and everyone left the scene by midnight.

For reasons that remain to be determined, about an hour later the train’s brakes ceased to hold, probably because the pressure in the air tanks dropped after the engine and its compressor were shut off.  The steep (1.2%) grade combined with the heavy load to send the train accelerating down the line six miles (10 km) along a straight track that executes a sharp bend in the center of Lac-Mégantic, near a strip of bars where late-night patrons were enjoying themselves.  Several people noticed the train rushing past at up to 66 MPH (100 km/hr).  When the leading locomotives hit the curve, they left the track and landed a few blocks away.  But the easily-punctured single-walled tank cars ruptured and caught fire, turning Lac-Mégantic into an inferno.  At the time of this writing (July 14), thirty-three bodies have been recovered, and seventeen more persons are missing and presumed dead. 

For this tragedy to occur as it did, you had to have

1.  A long heavy train full of
2.  Crude oil or other flammable material in
3.  Easily-ruptured tank cars attached to an
4.  Engine that caught fire and was turned off by
5.  A track worker who didn’t know how to start it again, and
6.  A steep grade under the train, and
7.  The engineer asleep in a hotel where nobody could find him, who
8.  Set enough handbrakes to keep the train from moving under normal circumstances, but not enough under the peculiar conditions that prevailed, and
9.  A town built around the closest turn in the tracks downhill from where the train started to roll.

If any of those nine conditions had been otherwise, the accident would not have happened, at least not to the extent that it did.  A shorter, lighter train might not have overcome the fading brakes.  Tanks of a non-flammable substance would have caused physical damage, but would not have burned.  Special double-walled tanks that are recommended for use in crude-oil service but not required, might not have ruptured as easily.  If the engine hadn’t caught fire, it would have kept the brakes going.  If the railroad employee had known how to start another engine, the brakes would have worked.  If the track had been flat instead of on a grade, the train might not have rolled away.  If the engineer had left a note saying where to call him in case of emergency, he might have restarted the engine.  If he had set more handbrakes, it might have prevented the runaway.  And if the bend in the tracks had been in a cornfield instead of in the center of town, no one might have died.

These “ifs” are cold comfort to those who have lost family and friends in the tragedy.  But they show what a long chain of unusual circumstances had to come about for the accident to happen.  Investigations may show that using more rupture-resistant tank cars would have reduced the chances of fire.  And there will certainly be questions raised about the advisability of crude-by-rail, or CBR, as a means of transporting large quantities of crude oil, instead of politically controversial pipelines, for example.  Let’s hope that what engineers, regulators, and the public learn from this accident will help in preventing the next one.

Sources:  I referred to articles on the accident in The Globe and Mail at and The Guardian at, a blog by Lloyd Alter at,
and a blog in Railway Age by Tony Kruglinski at, as well as the Wikipedia articles on Lac-Mégantic (the town) and the Lac-Mégantic derailment. 

Note added July 18:  Several comments indicate I did not adequately explain the operation of locomotive air brakes.  They are indeed essentially fail-safe in the short term, but not in the long term. 

An analogy can be made to the type of emergency lighting in stores and restaurants that is powered by storage batteries.  Under normal circumstances, the utility power charges the batteries and signals the emergency lights to remain off.  But when the utility power fails, the loss of voltage signals the emergency lights to turn on and operate from their storage batteries.  However, the storage batteries will eventually lose their charge if the power is not restored within a certain time.  When the storage batteries are depleted, the emergency lights will fail as well.

Here is the analogy.  The utility’s electric power is like the locomotive’s air compressor.  The emergency-light storage batteries are like the compressed-air tanks in each rail car.  The emergency lights coming on are like the brakes on each car becoming actuated by the air pressure from the tanks when the pressure from the locomotive fails (either deliberately or accidentally).  The tanks can supply only so much air (which escapes due to leaks, imperfect seals, etc.) and eventually the pressure falls to the point that the brakes no longer hold, just as the emergency lights will eventually go out. 

One reason the brakes are not actuated by a non-pneumatic method (e. g. springs) is that there are situations in which railway workers need cars to roll freely without being connected to a source of air pressure, as when the cars are sorted by being rolled one at a time over a “hump” and down a controlled grade through a sequence of switches.  If the brakes were always applied in the absence of air pressure, this kind of thing would not be possible. 

I hope this clarifies the question of how air brakes on trains work.


  1. I don't understand how loss of air pressure caused the brakes to fail, if I understand that loss of air pressure, by Westinghouse's design, is supposed to cause the brakes to apply.

    i.e. if the brakes need air pressure to hold, they are not designed to be fail-safe.

  2. I agree jpcarson. This is not fail safe. All air loss = no brakes. I wonder why the brakes are not applied with mechanical spring pressure and disengaged using the air pressure. Then air loss = full brakes.

  3. Good analogy. I "googled" the Westinghouse brake and a pretty good idea how they functioned. However, I was wondering why the brakes couldn't be applied constantly. Thanks for the explanation.