Discussion Page   Home     Help   Index     Login
Armchair Analysis - Train Wreck at Spiral Tunnels #5322
Back To Discussion List Written: 2019.02.06 by: Robin Tivy

You may have seen various reports on the disaster at the Spiral Tunnels. It was a runaway train. The brakes failed to hold the train at Partridge siding, which is above the upper spiral tunnel. The train gathered speed, went thru the tunnel, and derailed at high speed. The three people in the lead locomotive were killed. The pictures in the paper show the grain cars piled up in a heap of wreckage just below the highway.

Many of us have explored the spiral tunnels while driving to Alberta from BC on the trans canada highway. So there is a certain amount of interest in figuring out what went wrong. Of course, months or years from now an official report will emerge, but in the meantime, here's some basic ideas to chew on.

The official story starts at the Partridge siding, where the train was parked for crew change. Partridge siding is a section of double track just above the upper spiral tunnel. The train was westbound, so it was going downhill from Kicking Horse pass to Field BC. It had been parked for 2 hours at Partridge siding waiting for a crew change. The new crew had just gotten into the train when it started moving on it's own. Supposedly even with the emergency brakes on, whatever that means. I'm not sure of the procedure regarding brake checks when changing crews. There is usually a comprehensive brake test when the train is assembled at it's starting point. In this case there was a brake check in Calgary. In my experience as brakeman I don't remember any formal brake checks when we changed crews on the fly. I just remember getting up into the engine, and then we were underway. But that was long ago, one would think that at the spiral tunnels there would be stricter procedures.

Once it started rolling, the train gathered speed. I would assume the crew did whatever they could to try and slow the train down, perhaps attempting to recharge the reservoir cylinders in each car to increase braking power. But the braking power was insufficient. The went through the upper spiral tunnel, gathering speed as it went, such that by the time it emerged, it was going well above the 20 mile per hour limit. It derailed between the two tunnels, on the section of track that goes under the highway. That's where most of the pictures show the wreckage. It derailed just before the upper bridge over the Kicking Horse river. I think the lead engines may have gone down the bank into the river, and about 80 cars piled up behind them. Two of the bodies were outside the locomotive and one inside. Presumably they may have attempted to jump at the last minute.

Now some technical facts. I used to be a brakeman for CNR so I have some experience with the basic system. Or read all about this in Wikipedia. The confusing thing to many people is the fact that the main trainline pressure holds the brakes off, and when you release air in the trainline, the brakes go on. The braking force is provided by a air reservoir in each car. Unlike trucks and buses, the brakes are not applied by a spring, they are applied by the air pressure in the reservoir. So if the reservoir leaks, you have no brakes.

In normal operation on the level, the trainline is at full pressure (90 lbs) and the reservoirs are charged from that. If the engineer wants to slow the train, he lets some air out of the main trainline, for example, reducing from 90 to 80 PSI. This triggers the reservoir to put the brakes on. But each brake application reduces the reservoir pressure as well. They only get recharged fully when the train line is back at 90 lbs, which unfortunately means the brakes are off while this happens. Normally that is ok because the train is starting out again. Except on a long steep hill, where you still want some braking even as you start. It seems from the 1996 runaway incident report that normally it all works out - you can recharge the reservoirs fast enough while you are moving. That's what the engineer tried to do in 1996, and he had probably done it successfully numerous times before. But not that time.

If the brakes were repeatly applied over a short timespan, the reservoir doesn't get a chance to recharge fully. So there is less pressure available to put the brakes on. If there is no air in the system in either the trainline or the reservoir, you have no brakes.

There is also a second reservoir which is the emergency reservoir, triggered with a sudden loss of trainline pressure. Both reservoirs are charged up via the train line. If there is a sudden drop in train line pressure, the emergency reservoir applies the brakes, and hard. A sudden drop in pressure can be caused by a break in the trainline. Or by the brakeman/conductor turning the angle cock too fast when picking up additional cars. I remember connecting a few cars one night and there was a big "wooosh" from the whole train, and then the brakes were on emergency, with trainline pressure zero. It took about 20 minutes for the engineer to pump the line back up to sufficient pressure to move the train, during which the engineer chewed me out. So after that I was always real careful to have one hand on each side of the angle cock, and tap it slowly to the open position.

Now here's a key fact: the reservoirs normally stay charged up from the trainline. On their own, they will leak out after a few hours, depending on temperature. So since there had previously been an emergency brake application to stop at Partridge, the reservoirs were not fully charged.

Hand brakes are very rarely used on operating trains, they are activated by big metal wheels that pull a chain. The wheels are near the top of the cars so it's a major job to climb up to several of them. The disaster at Lake Megantic in Quebec happened when the engine was turned off due to a fire, and then the train sat overnight, and enough air leaked out of the reservoirs over several hours such that the train started moving. The crew were supposed to apply handbrakes before leaving the engine, but normally this would be unnecessary if the engine was running.

But at Partridge siding, presumably the engines were running. But how do the cylinders get recharged? Was the trainline at zero pressure, as if the brakes were applied in emergency? If so, what keeps the reservoirs charged?

From the description of how it works, here's what I think: I think the reservoirs were partially discharged on numerous cars, as the train sat at Partridge for two hours. The only way they get recharged is via the trainline. If the trainline was blocked, they would not get recharged. Some reports say the brakes were on emergency at Partridge. My understanding of that is that the trainline pressure would be zero, so I don't know how the reservoirs would get recharged. It seems to me that if the engineer is pumping the line back up, there would be no braking power while the reservoirs recharged.

In the old days, the brakeman in the caboose kept an eye on the brake pressure as well. These days, there is no caboose, so it would all be by remote sensors and radio to the head engine. Furthermore, there was another robot locomotive in the center, and another at the end.

There is something called a retainer valve on each car, such that braking power can be retained, even though the cyllinders are being recharged. But those need to be set manually. I never did anything with those when I was a brakeman. Despite the title "brakeman" or "conductor" we did whatever the engineer told us and were not formally trained on anything. I don't have experience on the Spiral tunnels, where they presumably have all sorts of procedures. But there is a report online which describes a runaway train incident on this exact hill in 2006 from which you can get an understanding. In that case, they just lined up all the switches through Field, and the runaway train roared through to the west end of the yard before stopping.

  • 1996 Railway Investigation Report R96C0086

    The report also talks about the diesel fumes in the spiral tunnels. That reminds me that in the 1970's, Tom Tiedje and I came up with the stupid idea to walk through the lower spiral tunnel. But we hadn't gone too far into the tunnel when the fumes caused us to abandon the project.

    Anyway, back to the disaster. What I'm sure the official investigators will be asking is all about the retainer valves. And also whether or not one of the stopcocks could have gotten turned in the train line. And also what data was available as to the trainline pressure at the robot locomotives. What mechanism or procedures exist to make sure the reservoirs are properly charged up. What would happen if the trainline got iced up? Apart from the air brakes, all of the locomotives are equipped with massive regenerative braking, with the energy being burned off by giant resistors on the roof of the locomotives. It is the regenerative braking that you hear when the trains are going downhill there - just a massive whining.

    The key question is: how can you recharge the airbrake reservoirs after an emergency stop without releasing the brakes? When you start pressurizing the line, at some point that causes the brakes to exhaust. (which is what you want normally, but not when you want to retain braking while charging. For that you need the retainers to be flipped manually)

     You can see the track layout in Bivouac GMap by looking up "Mount Stephen" and clicking GMap. The best basemap to use is T5 - Topo Canada. The location on that map matches what you can see in the satelite view. It shows the double track sections as well. And to read various newspaper reports, here are the milepost locations, starting from the east at Kicking Horse Pass (Stephen (Mile 123.0). For each one, I give a source because many of the newspaper reports are garbled, and it took me a long time to figure it out.

    1. Mile 123.0 Stephen (Kicking Horse Pass
        51.4512,-116.2890 1643m

    2. Mile 127.3? Partridge siding
       "Partridge siding is labelled in 1996 report "Schematic of Field Hill" and is clearly east of the entrance to the upper spiral tunnel. It is also labelled in a fabulous book written by Engineer/Mountaineer J F Garden's overlay of the 1:50K map, page 182. The label is at 51.426033,-116.405735, just southeast of where the label CANADIAN PACIFIC appears on the standard 1:50K map. This is ABOVE the upper spiral tunnel.

    3. Mile 128.8 Entrance Upper Spiral Tunnel
       I guess this is the mileage when coming from the east. (a westbound train). My source is the 1996 Investigation says under "Other Information" "In the event of a train stopping in the upper and lower spiral tunnels (miles 128.8 and 131.0)...

    4. Mile 130.6 Derailment
       This is shown in newspaper reports as just below the underpass of the trans-canada highway, between the two spiral tunnels. It is just before the upper bridge over the Kicking Horse river.

    5. Mile 131.0 Lower Spiral Tunnel
       (From "1996 report, see mile 128.8). I guess this is the westbound entrance.

    6. Mile 131.9 East Switch at Cathedral
       Tis is in the Railway Investigation Report R96..

    7. Mile 132.4 Cathedral siding
       J F Garden, Page 222
       -page 226 says it is mileage 132.4
       1:50K shows it below the lower tunnel. There is a campground.

    8. Mile ... Field BC, end of hill

  • Comments

    #6436 - 2019.02.07 Robin Tivy - CP changes procedure, says 25 handbrakes must be applied
    Today CP published a new procedure, requiring hand brakes to be applied in such a situation after an emergency stop. The new procedure is to apply 25 hand brakes in any similar situation. I think this is a stopgap solution. Climbing up 25 ladders and wrestling with 25 of those manual brake wheels is going to be a huge job, especially in -20 weather conditions. It won't happen reliably. And it doesn't address the question of why the emergency brakes had to be applied before Partridge. The old Westinghouse brake system has worked for 150 years most of the time, but it's not fail safe.

    But in the long run, to be really fail safe on that particular hill under Canadian winter conditions, there needs to be a backup system independent of the air brakes. One possible solution is to use the locomotives dynamic braking. And make sure they have control of all locomotives, even in the tunnels. In that case, every train going down the Field Hill must have enough locomotives equipped with dynamic braking to control the train even with air brake failure. How many locomotives? The same number as would be required to pull the train uphill. Dynamic brake resistor grids have the ability to absorb the same horsepower as the locomotive under full throttle. They do that in testing. Therefore if we can generate that much power, the retarding force would be as great as if the engines were pulling the train uphill at full throttle. So physics tells us that the retarding force would slow the train,just as if the engines were in reverse. However I read in a Trains magazine that an SD40 locomotive generates about 600 horsepower while regenerating.

    It seems to me that whatever number of locomotives it takes to pull a train up a hill, that same number of locomotives could prevent it from gathering much speed going downhill. However, it is possible they don't need to have enough locomotives to pull a loaded train up the hill. The trains are often empty going eastbound. Of course this wouldn't be sufficient to stop a fast moving train, but a train slowly crawling down the hill would be different. I'm assuming that in this case, whatever engines were on the train did NOT have the ability to control the train once the air brakes were compromised. How many had dynamic braking? I'm looking into this.

    The Westinghouse air break system is not a simple fail safe system. Its a quirky system that has many variables. Knowledgeable crews can work around the flaws, but it's too easy to make an error. Just read the 1996 runaway report on this exact hill, or many other railway airbrake disaster reports. In sudden unusual circumstances, it's too hard to figure out exactly what procedure might be the best.

    Of course there are other alternatives. The Canadian National mainline through Yellowhead pass does not have the steep grades of the Canadian Pacific. So winter grain trains could be sent via that route. Of course if we can land a man on the moon, you could redesign the whole brake system from scratch so that you didn't have to rely entirely on the traditional system. But the existing system is standard all over North America. The cars can come from anywhere on the system, owned by all sorts of different companies. So convincing everyone to change would be a daunting job. For most of the continent, the traditional system is good enough. For that reason, I think the answer lies in something the individual railway has control over. The locomotives.