Derailment in Washington state

[Supaned]

and yet , there is a perfectly good , working and simple solution which would have prevented Chatsworth , Amtrak 188 and this accident , and at a fraction of the cost of PTC - the UK system TPWS, which overlays with the existing signalling system to take action should a train try to pass a signal at danger, and is also installed at speed restrictions to automatically apply the brakes if a train approaches too fast.

 

I don't doubt that PTC will be great when it's all up and running , but I still maintain it's excessively complex and expensive for what it is actually needed for.

 

Naturally I extend my condolences to all involved in this latest accident.

[Titan]

and yet , there is a perfectly good , working and simple solution which would have prevented Chatsworth , Amtrak 188 and this accident , and at a fraction of the cost of PTC - the UK system TPWS, which overlays with the existing signalling system to take action should a train try to pass a signal at danger, and is also installed at speed restrictions to automatically apply the brakes if a train approaches too fast.

 

I don't doubt that PTC will be great when it's all up and running , but I still maintain it's excessively complex and expensive for what it is actually needed for.

 

Naturally I extend my condolences to all involved in this latest accident.

 

Don't make too many assumptions. American Railroads have to deal with different circumstances, which may well mean that what works well in another country would be very difficult to install. It may not be possible to consider anything other than a radio based system as laying expensive cables along many thousands of miles of track, where existing signals are few and far between, could make something like TPWS prohibitively expensive compared to the UK. There may or may not be different reasons, but the significant difference in the way that American Railroads operate has to be taken in to account when determining suitability of equipment.

[Jeff Smith]

For just top speed control, I would have thought a relatively simple autonomous GPS system with throttle and brake override could be installed on each locomotive. The route map could be continuously updated to include temporary speed restrictions. Is this too simplistic?

[AndyID]

For those unfamiliar with the area, before the sharp bend where it derailed, the train was traveling on a high speed section parallel with and clearly visible from the I5 freeway.

[AndyID]

For just top speed control, I would have thought a relatively simple autonomous GPS system with throttle and brake override could be installed on each locomotive. The route map could be continuously updated to include temporary speed restrictions. Is this too simplistic?

 

Or at least set off an alarm for the driver.

[stewartingram]

Don't make too many assumptions. American Railroads have to deal with different circumstances, which may well mean that what works well in another country would be very difficult to install. It may not be possible to consider anything other than a radio based system as laying expensive cables along many thousands of miles of track, where existing signals are few and far between, could make something like TPWS prohibitively expensive compared to the UK. There may or may not be different reasons, but the significant difference in the way that American Railroads operate has to be taken in to account when determining suitability of equipment.

As a comms engineer working with the emergency services, may I add a comment here about cables/radio?

​When working with the Fire Service, we had a certain callout system installed. Basically there are 2 types of fire stations - manned and unmanned. With the manned stations (usually in towns), it was normal to have the data from the callout system sent over a Private Wire (which was basically a dedicated permanent BT phone line). With unmanned (village) stations, where the firemen were employed in a "day job" and only called out when needed, the connection to this station equipment was also by landline; in the same way. Of course there were backups to ensure it all worked; usually the PW had a backup dialled line (just like a home phone).

Now, the point I'm getting at, is that in one county, they decided to go better than that. There was a dedicated private radio channel used for voice communications. It was decided to send the data over this radio channel, with the PW and dialled lines becoming backups. 

Now see the relevance? Perhaps TPWS might be possible over radio? I honestly don't know - but as I understand it, it is basically digital data, so in theory it might be possible? But we are not in the field of designing it for them, so.....?

 

Stewart

 

ps apologies for not mentioning names and more details, but i am bound by confidence here!

[peach james]

and yet , there is a perfectly good , working and simple solution which would have prevented Chatsworth , Amtrak 188 and this accident , and at a fraction of the cost of PTC - the UK system TPWS, which overlays with the existing signalling system to take action should a train try to pass a signal at danger, and is also installed at speed restrictions to automatically apply the brakes if a train approaches too fast.

 

I don't doubt that PTC will be great when it's all up and running , but I still maintain it's excessively complex and expensive for what it is actually needed for.

 

Naturally I extend my condolences to all involved in this latest accident.

 

It doesn't help if the line is black...as in, not signaled at all, and trains controlled by track warrant.  The Canadian certainly used to be handled that way...

 

You really have to have traveled from point A to point B in Canada or the US to get an appreciation of the actual distances involved.  It's 3 DAYS from here to Toronto (4 now...) on the Canadian.  Now, imagine placing TPWS for each curve along 7000 km of line...

 

James

[Patriot87003]

Be careful .... simple systems like TPWS, speed traps etc only work best where the braking rates of the various types of train are similar. In the UK, even where TPWS is fitted, there is no guarantee that all collision or derailment scenarios will be prevented (e.g. Heavily loaded freight train with poorer braking rate and with slippy rails). Likewise, TPWS penalises trains that have a superior braking rate in terms of performance, with extended journey times, caused by drivers driving defensively in order to avoid a TPWS activation.

 

I've read it reported that the advance warning board for the 30mph speed restriction was located two miles in advance of the actual start point of the restriction/curve. No doubt that this will be around the location where a heavily loaded freight train will require the brakes to be applied (plus a factor of safety distance allowance) to reduce safely to 30mph, but the braking distance from line speed to 30mph for a modern passenger train is likely to be much much less than two miles, probably as little as half a mile (if similar to UK passenger train braking rates). For a passenger train driver running at the full line speed, this will be around about a minute after passing the advance speed board, during which time there is plenty of time to be distracted.

[dave1905]

For just top speed control, I would have thought a relatively simple autonomous GPS system with throttle and brake override could be installed on each locomotive. The route map could be continuously updated to include temporary speed restrictions. Is this too simplistic?

 

You would have to be able to load the entire US track network into the GPS system.  Plus GPS isn't accurate enough, it can't with absolute certainty, tell which track a train is on in multiple track territory.  Since basically that's what PTC is, it probably wouldn't be any savings or any simpler.

[dave1905]

No expensive cables running for miles, just a pair of (permanent) magnets a suitable distance from the curve. On each loco a magnetic sensor (for each direction) and some spare computing capacity. 

 

But now you have to equip thousands of engines with magnetic sensors OR you have to have special pilot engines that can be used as leaders.  In the US system engines are interchanged and may operate all over the western half of the US or even the entire country.

[caradoc]

Don't make too many assumptions. American Railroads have to deal with different circumstances, which may well mean that what works well in another country would be very difficult to install. It may not be possible to consider anything other than a radio based system as laying expensive cables along many thousands of miles of track, where existing signals are few and far between, could make something like TPWS prohibitively expensive compared to the UK. There may or may not be different reasons, but the significant difference in the way that American Railroads operate has to be taken in to account when determining suitability of equipment.

 

Certainly USA circumstances are very different from the UK, but TPWS does not require thousands of miles of cables, as it is not installed at every signal, merely at those controlling junctions, plus at some speed restrictions (where the decrease in speed exceeds a certain percentage), and on the approach to buffer stops. Installing it, on and for passenger trains only, would have prevented the two overspeed derailments I mentioned previously (Edit; Post #38), plus the buffer stop collision at Hoboken last year, plus possibly (as we do not know the full facts yet) this tragic accident as well.

Edited December 19, 2017 by caradoc

[Ozexpatriate]

Just read that the NTSB has confirmed that the train was going 80, so it looks like the cause was speeding.

 

PTC is NOT universally mandated, only required on those territories that meet requirements with regard to passenger service and hazardous material volumes. 

This particular stretch of track (the Point Defiance bypass on Sound Transit tracks as opposed to the BNSF mainline) was upgraded for the express purpose of greatly improving running times through a combination of operating speed, shorter distance and fewer freight meets.

 

The S curve that crosses I5 is part of a section with a lower posted speed, (this is specifically called out in WSDOT project material online) presumably because of the specific track geometries around the highway crossing and the fact that it is quite close to the connection point with the BNSF mainline at Nisqually. 

 

Given the track upgrades taking place over the last several years this feels to me like an (almost) perfect opportunity to enable PTC. Doubtless there are some issues. Much of the rolling stock has been in place for a while now (though some new Siemens locomotives were added, it was commented here that one was on the head end) and the equipment needs to operate on the longer Vancouver - Eugene corridor rather than the Seattle to Portland stretch where train 501 was scheduled. 

[BR60103]

a small correction to my previous post. The coach next to the loco has a false side that sweeps up to match the height of the loco. It wasn't bent in the middle.

[dave1905]

As for nation-wide use of motive power, I'm aware of that and it also includes Canada as well. I'm also aware all Class I's run a considerable portion of their motive power in sets that stay coupled for a prolonged period of time and often form a semi-permanent consist with their train, including remote power units, mostly for bulk commodities like intermodals, grain/wheat, coal and ore. That means that initially only the leader can/needs to be equipped, with the remainder on a less urgent schedule for conversion. 

 

Having helped manage the locomotive fleet of a US class 1 in a former life, keeping sets together is a noble goal, but there is a LOT more churn on those "semi-permanent" consists (more emphasis on the semi than the permanent) than people think. 

 

There are a lot of things that they could have done besides PTC (like use the nearly 75 year old cab signal systems already in place on thousands of miles of US railroads), but they went with the Cadillac of PTC.  It took decades to implement air brakes, it took decades to implement knuckle couplers.  It took a decade to get rid of iron wheels, wooden underframes, etc.  It will take time to cut over PTC.  There are 10's of thousands of miles of railroad operating with PTC now (more or less successfully) and they have been working the bugs out of it for the last several years.  The line in question will get PTC, just hasn't had all the parts installed yet.

[jsp3970]

KOMO news in Seattle, WA has posted an article on line that the engineer may have been distracted approaching the curve.

 

http://komonews.com/news/local/official-amtrak-engineer-may-have-been-distracted-before-derailment

[Regularity]

That is pure speculation on the part of the Newspaper: it may turn out to be true, of course, but there is no way of verifying the source, and it would be scary in the extreme if the NTSB did not interview the driver and anyone else in the cab as part of their investigation.

 

For all we know, the trainee may have trying to tell the driver to slow down.

Edited December 20, 2017 by Regularity

[caradoc]

So, if what has been reported so far is correct: This was a new train service, hauled by a brand new locomotive, operating on a refurbished route which however included a severe speed restriction, yet there was no form of automatic speed control, whatsoever ? That sounds incredible.

[Oldddudders]

So, if what has been reported so far is correct: This was a new train service, hauled by a brand new locomotive, operating on a refurbished route which however included a severe speed restriction, yet there was no form of automatic speed control, whatsoever ? That sounds incredible.

Why? The US railroads are freight operators, and Amtrak is just a tenant. A nuisance, really. As has been said, there is still a large mileage of railway that has no signalling of any sort that we would recognise, where the working timetable is a key part of the safety system. That alone is implausible from a British perspective. Comparabilities beyond steel wheel on steel rail are limited.

[caradoc]

Why ? Because the USA has had a spate of fatal passenger train derailments on speed restricted curves (3 now in 5 years), and because it is much simpler to build in some form of speed control to a new or upgraded line, rather than overlay it onto existing systems.

[dave1905]

DId the driver know the line speed I wonder? First train on first day of service, all new.... Presume they have to sign the road as such?

 

Latest reports say the engineer was distracted by an employee in training in the cab.  Road familiarization trips carry a risk because the primary engineer has to be both training and operating.  One of the last studies I did was an analysis on when a crew had more than 2 people on it, who were the extra people on the crew.  Most of them were somebody training.  A true double edged sword.

[dave1905]

So, if what has been reported so far is correct: This was a new train service, hauled by a brand new locomotive, operating on a refurbished route which however included a severe speed restriction, yet there was no form of automatic speed control, whatsoever ? That sounds incredible.

 

30 mph isn't that "severe" of a speed restriction.  Other than PTC, there is no "automatic speed control" for track related speed restrictions anywhere in the US, why is it "incredible" there wasn't any at that location?

[Dunsignalling]

That is pure speculation on the part of the Newspaper: it may turn out to be true, of course, but there is no way of verifying the source, and it would be scary in the extreme if the NTSB did not interview the driver and anyone else in the cab as part of their investigation.

 

For all we know, the trainee may have trying to tell the driver to slow down.

Assuming the crew weren't among the dead.

[dave1905]

The head end crew were injured but survived.

[dave1905]

As has been said, there is still a large mileage of railway that has no signalling of any sort that we would recognise, where the working timetable is a key part of the safety system. 

 

Actually no.  Timtetable operation pretty much ended back about 1985.  Nobody uses that anymore.  There are unsignalled portions of track, but they are operated by some form of track warrant control, where the trains are verbally authorized to use sections of track.  On the class 1 I worked for, about half the main track mileage was CTC, 1/4 signaled TWC (track warrant) and 1/4 unsignalled TWC, yard limits or other.

[Regularity]

Assuming the crew weren't among the dead.

  

The head end crew were injured but survived.

We know they weren’t among the dead, otherwise I wouldn’t have made the comment.

 

We had a signal passed at danger here in Oakham a few years ago (early noughties), at the level crossing. Luckily nobody was injured. It was an empty DMU on a driver training run, and the driver and trainee ran past a distant and two reds, which means they must have cancelled the AWS or whatever in the cab without realising it. I only found out about it via a friend who was a signaller at another box on the line. At the time, the Nottingham-Peterborough part of the cross country run from Liverpool to East Anglia (and vice versa) used to pass through at speed, so it is possible that the driver was used to running straight through. I don’t know which train operating company was involved.

 

Anything involving the human element is capable of error, whether that be driving, building, designing a train or just how people approach grade/level crossings: we have all seen the videos...

 

I can see nothing to be gained from commenting over different automated systems applied on different continents under not only differing jurisdictions, but different railway cultures.

 

So far, we know this much:

A train derailed on a curve approaching the bridge, the resulting accident has caused fatalities and casualties.

The train was going considerably faster than it should have been.

There are consistent (with each other) reports that there were two railway personnel in the cab.

The NSTB is investigating.

Their report will endeavour to obtain the truth.

 

Most of the discussion on here is speculative at best, or from a position of ignorance at worse. A few posts have shed light on general truths, but until we get the report from the NSTB, we will not really know the full specific details.

 

I suggest we let due process take its course.