Cascades 501 Wreck 18 December 17

Noel Weaver wrote:Come on Dutch would you expect them to use any good old common sense here????
Noel Weaver

Common sense flew out the window about a decade ago.

When Asiana Air dumped a 777 short of the runway at SFO, they didn't blame the autopilot. A thorough investigation was conducted. The blame was rightfully aimed at company culture, training, and the pilots.

In a recent auto accident that killed the owner of a Tesla operating under autopilot, the blame was square on the operator, not the autopilot.

For some reason, we continue to hang our hat on PTC. We blame lack of PTC for 188 and 501, as if some magic pixie dust BS is going to cure all of our ills. I suspect the investigation will reveal some combination of operator error, poor training, and company culture. Perhaps too many men in the cab, as that was noted as a contributing factor in the fatal overspeed derailment on Via's corridor a few years back.

Mark my words, one of these days there will be a mass casualty under PTC. We came darn close to one a few years back in Niles, Michigan. ITCS was "turned off" for repairs and a switch was lined for the MoW yard, and the Wolverine took it at track speed. If you can "turn off" the PTC and not see some sort of restrictions, it's not positive. It's just bull hockey.

Link to PTC-caused derailment.
http://www.marp.org/?p=4287" onclick="window.open(this.href);return false;

The real question though is would something as complex as PTC actually be required to stop stuff like this? This is a wild guess, though one based on years of studying the railroad industry and operations, but something like cab signals with ATC would prevent 90% of the collisions, 99% of the fatalities, at 20% of the cost of PTC. A key piece of evidence I use for this is Chatsworth. Ignoring the fact that the cabs would have been screaming at Mr. Sanchez on approach to the station (advance approach, followed by approach), once the train exceeded 40mph (operating under approach), the cabs would scream again. Even assuming he suppressed the cabs and kept operating at 40, as soon as he crossed CP-Topanga, the cabs would drop to restricting on both the Metrolink and the UP train either applying the brakes, or requiring the crew to apply the brakes. Based on the NTSB timeline, it seems likely that the earlier application of brakes would have slowed both trains a good bit more probably reducing the number of fatalities. I freely admit though without access to a detailed simulator, this is conjecture on my part, and have no problem with more experienced people poking holes in this.

Now I do recognize there is a difference between Chatsworth and Dupont. Chatsworth was a signal violation, and Dupont (and Spuyten Dyvil, and Frankford, etc.) were overspeed on curves, but the same immediate response to those, with the same non-PTC systems, could apply here, or even could have been implemented from the get-go.

Automatic Train Stop would help for absolute stop signals (this 1930s analog tech still exists on the ex-Santa Fe main, allowing a 90 mph top speed), but otherwise you need something that can control speeds. As of now, only PTC and full Automatic Train Control can take action to slow or stop a train, if a well-trained and qualified operator is not in control of the train and fully informed of the situation (i.e. proper signal aspects displaying, speed restrictions flagged in the ETT and on the right-of-way, visual conditions allowing.)

MattW wrote:The real question though is would something as complex as PTC actually be required to stop stuff like this? This is a wild guess, though one based on years of studying the railroad industry and operations, but something like cab signals with ATC would prevent 90% of the collisions, 99% of the fatalities, at 20% of the cost of PTC. A key piece of evidence I use for this is Chatsworth. Ignoring the fact that the cabs would have been screaming at Mr. Sanchez on approach to the station (advance approach, followed by approach), once the train exceeded 40mph (operating under approach), the cabs would scream again. Even assuming he suppressed the cabs and kept operating at 40, as soon as he crossed CP-Topanga, the cabs would drop to restricting on both the Metrolink and the UP train either applying the brakes, or requiring the crew to apply the brakes. Based on the NTSB timeline, it seems likely that the earlier application of brakes would have slowed both trains a good bit more probably reducing the number of fatalities. I freely admit though without access to a detailed simulator, this is conjecture on my part, and have no problem with more experienced people poking holes in this.

Now I do recognize there is a difference between Chatsworth and Dupont. Chatsworth was a signal violation, and Dupont (and Spuyten Dyvil, and Frankford, etc.) were overspeed on curves, but the same immediate response to those, with the same non-PTC systems, could apply here, or even could have been implemented from the get-go.

Matt, you make some really good points here. I look at Chatsworth as a texting-and-driving incident. Many studies have shown that such activity impairs one's ability to operate a motor vehicle to the same degree that heavy drinking does. I have little doubt that if the engineer had put his darn phone away before he got in the cab, we'd have no incident. It escapes me why, until that time, cell phones weren't a de-certifiable offense.

Tadman wrote:Matt, you make some really good points here. I look at Chatsworth as a texting-and-driving incident. Many studies have shown that such activity impairs one's ability to operate a motor vehicle to the same degree that heavy drinking does. I have little doubt that if the engineer had put his darn phone away before he got in the cab, we'd have no incident. It escapes me why, until that time, cell phones weren't a de-certifiable offense.

Railroading is a conservative business that reacts to threats conservatively (I.e. after an incident), hence why we are still talking about PTC in the abstract and then only as a GPS-based system rather than a nationwide landline system more similar to ACSES. I’m glad the regulators and railroads recognize the threat of electronics now.

Just to be clear, were electronics a factor in the Cascades wreck? I hadn’t heard such and want to make sure this is just an abstract, related discussion.

No reported use of cellphones from an early NTSB report/scan thru of the Charger's in and out view cameras.
Note: BNSF has the Seattle Sub PTC online,but may not have the Sounder/Amtrak locos setup in their computers yet!

Tadman, I must say given your background, I consider that a high compliment!

mtuandrew, I do agree that the engineer must be in full control of the train, but I believe even ATS would have provision for approaching permanent speed restrictions. All it really is is a binary indicator, for a permanent speed restriction, that indication just happens to be always on. If the train is already at the speed, then great, nothing needs to happen, if not, then the engineer is reminded to apply the brakes. So effectively this becomes the big flashing light board mentioned a number of posts back. I believe the UK rail system has a slightly more advanced version that can detect if the train is below the speed and only activate if it's still going too fast. But, the more advanced cab signaling I talk about in my first post has been used for at least 60 years, and maybe longer. Even if it couldn't necessarily enforce a positive stop, the fact that it would still reduce the speed of both trains in a head-on collision would go a long way to reducing the damages from collisions.

ATS with timed circuits would have worked but requires more signal bungalows. The Santa Fe does have a timed circuit with its timed circuit going into LAX Believe if train exceeds 65 ? then inductor stops train before descending a grade ?

ATSF used "inert" inductors on various parts of the "Passenger mainlines" to enforce speeds on curves,grades,and where needed by the ATS system.
On both the Glorieta and Raton Subs,they enforced speed restrictions in curves in the Passes.

One of the advantages of PTC was supposed to be, the railroads would gain additional capacity by being able to run freight trains faster and closer together, and do it safely. A pilot program about twenty years ago on BN in northern Minnesota worked very well. Early on Union Pacific was a big supporter of PTC, saying it would not only speed up operations (saving fuel and crew costs) but make scheduling more reliable plus permit better locomotive and car utilization. BN had plans to expand PTC use but there was a change in senior management and the plan was cancelled. As for UP, and a few of the others, PTC requires spending a lot of money and I think there was some pushback from the investment community, and the railroads backed down a bit. I think essentially that was why Congress finally got involved.

re: Trackage ownership

Can someone point me to a citation as to ownership of the trackage in question?
I know an ex-Hill staffer who helped write the Staggers Act, and at dinner yesterday, he was sure it's BNSF.

Railroading operates in two-ish dimensions. Aviation works in three, with air currents making significant differences to direction and handling, and yet to my mind has embraced automation to a more significant degree - e.g. cat 3b autoland. To my mind it is shocking that road automation has progressed further than rail when road is a significantly more chaotic environment. But posters here (and other rail interest venues of course) will reliably throw up dire warnings about trusting machines and the effect on crew awareness (which is not entirely unfounded - see "children of magenta" - but all too often throws out the baby with the bath water)

It does not help that North American rail control operates in isolation from the rest of the world (more or less) whereas standardization is forced on more diverse environments like the European Union which also does not have the same degree of private capital control of infrastructure - imagine getting something like the European Railway Packages through Congress or as a NAFTA agreement with K Street to contend with!

At the very least, can we agree that a relatively simple locomotive fitted system which would have automatically activated based on GPS if the speed approaching this curve (based on GPS mapping similar to car speed limit warning systems) would require emergency braking (and therefore trigger a review of crew actions) would be something that would strike a balance between ultimately preventing catastrophic loss and keeping crews mindful that failure to adhere to limits would have career consequences?

farecard wrote:re: Trackage ownership

Can someone point me to a citation as to ownership of the trackage in question?
I know an ex-Hill staffer who helped write the Staggers Act, and at dinner yesterday, he was sure it's BNSF.

The WA State Rail Ownership Map is a bit dated (2011) but indicates the line as being owned by Tacoma Rail (itself owned by the city of Tacoma). Wikipedia indicates the Lakewood Division was purchased from BNSF in 2004.

The old route around Point Defiance is BNSF, but the new route is Tacoma Rail with Sound Transit trackage rights.

I'm a little puzzled by the reports that the second employee in the cab was a conductor being trained.

Why would a passenger train conductor need to learn the route in that level of detail?

Or was this an instance of a young reporter mistakenly referring to a train operator as a conductor? I see this quite often from young people discussing urban rapid transit. They seem aware that driver isn't the proper American term, and so they call the person at the controls the conductor, based, I suppose, on hazy memories of watching Shining Time Station.

Conductors must be qualified on the territory just as engineers are.