41 replies to this topic

[Trogdor]

If anything, PTC could slightly lower the average speed due to the way PTC systems enforce speed restrictions, forcing engineers to begin braking sooner than they otherwise would.

[George Harris]

If anything, PTC could slightly lower the average speed due to the way PTC systems enforce speed restrictions, forcing engineers to begin braking sooner than they otherwise would.

Very True!!!

[Amtrak George]

George Harris has some very valid points. One thing I worry about is the false sense of security PTC may give. There are a host of things that can go wrong and cause fatal accidents, even with PTC.
Example: sabotage, items on track, washouts and bridge damage that don't break track circuits (remember the Sunset bridge disaster in 1993? How about the sabotage of the Sunset in Arizona? If I recall correctly, the track circuits for the block signals were not broken in either case). Grade crossing accidents, sideswipes with frieghts, etc. etc. can also happen even with PTC.

Also, I am curious about something. Let's say a passenger train is moving slowly, holding the main, waiting for a freight to take siding. The passenger train engineer is obeying signals and the PTC system thinks things are fine. Then the passenger train engineer runs by the home signal and hits the freight head on. Will PTC in its present development prevent accidents at meeting points when trains are so close together? Sure, the system should stop the train that runs a red signal, but in such situations will it be too late?

I wonder if we need more testing and development before spending so much money on something that may not be effective.

[George Harris]

Also, I am curious about something. Let's say a passenger train is moving slowly, holding the main, waiting for a freight to take siding. The passenger train engineer is obeying signals and the PTC system thinks things are fine. Then the passenger train engineer runs by the home signal and hits the freight head on. Will PTC in its present development prevent accidents at meeting points when trains are so close together? Sure, the system should stop the train that runs a red signal, but in such situations will it be too late?

That is exactly what PTC is supposed to do. In fact, your description is essentially the same as the Chatsworth collision events. However, you always get around to what can go wrong will go wrong at some time or other. Also, the additional hardware will serve to reduce reliability as there will be more things that can go wrong, and the general "fail safe" concept in safety related hardware and rules the normal result of a hardware failure will be to bring things to a halt.

[AlanB]

Also, I am curious about something. Let's say a passenger train is moving slowly, holding the main, waiting for a freight to take siding. The passenger train engineer is obeying signals and the PTC system thinks things are fine. Then the passenger train engineer runs by the home signal and hits the freight head on. Will PTC in its present development prevent accidents at meeting points when trains are so close together? Sure, the system should stop the train that runs a red signal, but in such situations will it be too late?

That is exactly what PTC is supposed to do. In fact, your description is essentially the same as the Chatsworth collision events. However, you always get around to what can go wrong will go wrong at some time or other. Also, the additional hardware will serve to reduce reliability as there will be more things that can go wrong, and the general "fail safe" concept in safety related hardware and rules the normal result of a hardware failure will be to bring things to a halt.

Amtrak George,

Just to expand a bit further on your scenario, if the situation that you describe were happening, the PTC would still have the pax train barely crawling by the time it approached the signal protecting the switch to the siding. Remember the PTC controls speeds, not just signals. So the closer to the home signal the passenger train creeps, the lower the speed limit will be.

Assuming that the engineers did their homework, eventually the speed will be so slow that if he overruns the home signal, an emergency stop will still stop him before fowling the switch. And of course the freight will have slowed considerably too, simply to take the switch.

Yes, as George Harris noted if enough events cascade, something can still go wrong. But one is really cutting down on the odds with PTC.

[RampWidget]

Yes, as George Harris noted if enough events cascade, something can still go wrong. But one is really cutting down on the odds with PTC.

And that line of thinking is why, especially after Chatsworth (and Graniteville) Congress was receptive to the idea. But... there are other things we can do to cut down the odds on human-factor accidents that have a much better benefit/cost ratio than PTC. For example:
Rail / rail crossing separations have already been mentioned in this thread. The issue of switches left misaligned in non-signaled territory can be addressed (and is in many locations) by installing CTC "islands" at the ends of sidings, effectively making them dispatcher-controlled. The FRA has now issued and enforces civil penalties on carriers, and individuals if warranted, for unauthorized cell phone usage on board trains. (Personally, I'd prefer cell phone jammers, but I don't think FCC cares for that idea).

[jis]

Also, I am curious about something. Let's say a passenger train is moving slowly, holding the main, waiting for a freight to take siding. The passenger train engineer is obeying signals and the PTC system thinks things are fine. Then the passenger train engineer runs by the home signal and hits the freight head on. Will PTC in its present development prevent accidents at meeting points when trains are so close together? Sure, the system should stop the train that runs a red signal, but in such situations will it be too late?

That is exactly what PTC is supposed to do. In fact, your description is essentially the same as the Chatsworth collision events. However, you always get around to what can go wrong will go wrong at some time or other. Also, the additional hardware will serve to reduce reliability as there will be more things that can go wrong, and the general "fail safe" concept in safety related hardware and rules the normal result of a hardware failure will be to bring things to a halt.

A recent incident which led to a slow speed derailment is very illustrative of how failures still happen in PTC territory and how the resulting damage is more contained than it could be otherwise. Mind you I present this just as an example without having any further agenda towards either pushing for or detracting from the use of PTC where it makes sense for appropriate level of risk mitigation - afterall all safety systems are about risk mitigation, not total avoidance, which is impossible.

A few weeks back an Acela Express derailed at slow speed at CP Davis on the ACSES controlled NEC in RI. The failure involved was not of ACSES, which is fully certified PTC system, but of a human being who inappropriately allowed the train to proceed into a PTC blocked section by essentially overriding the protection without fully ensuring that it was safe to do so, and another human being proceeded to do so without following the rules for doing so to the 'T'. There is no amount of protection that could have prevented that derailment given the event chain that was produced by human ingenuity. OTOH, the residual protection prevented things from being worse by enforcing the 20mph speed limit.

The PTC regulation essentially lists a set of capabilities that must be present. It does not specify a technology. The freight railroads have chosen the GPS based technology as what they believe to be the lower cost alternative to achieve the required capabilities. One of the systems that has so far been certified however, is not GPS based but transponder and Coded Track Circuit Cab Signal based - and that is Amtrak's ACSES in the northeast. It turns out that ACSES does also have a radio link in addition to the coded track circuit channel and passive transponder for communication, and indeed things like TSRs are handled through the radio channel after attempting to do so using transponders was deemed to be too burdensome and error prone. It turns out that this same radio link can be used to provide the radio messages necessary for the GPS based PTC systems, thus opening up the possibility to add an overlay to allow interoperability for train borne segment of the GPS based PTC system on ACSES equipped sections. This will be possible for trains traveling at less than 90mph (AFAIR). That is how two very different technology based systems may be able to interoperate at least on the train segment while using very different office segment.

Purely from a techno-geek perspective I find these developments very interesting. As for what is the most appropriate thing to do under which operating conditions, that is a discussion that needs to take place with the types of information that I do not have access to, so I refrain.

Edited by jis, 29 May 2012 - 02:54 PM.

[MattW]

Jis, it never ceases to amaze me both the mass of knowledge you have on these systems, and how well you can explain it so those of us not necessarily "in" on these technologies can understand.

[jis]

I see there is quite a debate here on PTC and its use.

I'm from Michigan and I know it has been tested/used on the Wolverine route near where I live. I'll be taking the Lakeshore Limited to the east coast this summer. Is it in use in any part of that route as of now? Should I be concerned if it isn't? I know a lot of that route is owned by freight...?

Thank you!

At present it is not used in any of the LSL's route, and you should not be concerned.

[jis]

I understand there is a difference between ATC, ATS, Cab Signals, and PTC. I was trying to make the point that some sort of train control exists on most route miles in Europe and Asia and that it would be prudent to take the best of those systems in designing our version of PTC rather than starting from scratch as we seem to be doing.

The premise that "some sort of train control exists on most route miles in Europe and Asia" is incorrect. Actually only a relatively modest proportion in Europe has train control and a very very small proportion in Asia has such. Secondly, any sort of train control does not make it PTC or even promotable to PTC

I think going with a hardware system similar to TVM 430 would be best, transponders x feet/miles along the route with spacing being determined by expected maximum trains per hour. Additional transponders could be placed to enforce permanent speed restrictions such as in curves. If all of these transponders are linked via track circuits you would essentially have signals that communicate with each other, the trains and way side devices such as crossing gates, switches, detectors, RFID readers, etc. If dispatchers could control the transponders to issue temporary speed restrictions all requirements of the PTC mandate would be met.

And you have of course figured out that deploying this will be less expensive than the one being deployed presently? As George said in this thread, it is remarkably expensive, but its requirements are also to operate trains at 200mph, which to some extent justifies that expense for routes so equipped.

BTW, you may wish to read materials that are available on Amtrak's ACSES to see how such things are put together instead of hypothesizing such. Most of the devil in such systems is in the details. Putting together an abstract architecture is the easy part, and such documents contain a lot of words like "could", "should" and "would". Unfortunately there is not way to verify that such proposals work without spending a lot of resources to build prototypes, both mathematical models and real things.

Most importantly we should focus on interoperability. A train should be able to go anywhere in the national network just like you can drive your car anywhere on the highway system.

If we want to focus on interoperability, Europe should be the last place that we should be trying to emulate. There are very few trains in Europe that are able to "go anywhere".... well actually there are none. The few that can cover multiple countries do so by brute force carrying of upto six different on board systems so that they can switch to the right one for each territory they operate in, and that is an incredibly expensive way to get interoperability. Actually the Americans are doing much better in that department, and most of the proposed PTC systems that are being deployed can share a lot of the on board equipment across multiple deployments.

Edited by jis, 29 May 2012 - 10:57 AM.

[jis]

Jis, it never ceases to amaze me both the mass of knowledge you have on these systems, and how well you can explain it so those of us not necessarily "in" on these technologies can understand.

Thank you for the kind words. It is just my techno-geek hobby, and to some extent overlaps with my professional work on autonomic control systems as applied to large scale IT management.

[George Harris]

In the US, actually, the US plus Canada plus Mexico, there has been through running of equipment between railroad companies almost from the beginning and through running of power which requires compatibility of control systems at least to as far back as the beginning of diesel operation, and that is not mentioning such things as trackage rights arrangements, some of which go back at least 160 years. Since, until recently, all or almost all European operations were contained within national borders and for the most part on unified national systems, the whole issue of interoperability is a relatively new concept.

If you want to make your brain hurt and go into spasms, get your hands on the Technical Specification for Interoperability and try to understand any part of it.

Edited by George Harris, 29 May 2012 - 01:25 PM.

[jis]

Europeans are still working on a single standard coupler to replace the chain and hook coupler. So far they have failed to muster the political will to make it happen.

[George Harris]

Europeans are still working on a single standard coupler to replace the chain and hook coupler. So far they have failed to muster the political will to make it happen.

Why I used to say to some of the European guys I was working with at the time, try to join the 20th century before you lose the chance.

For those that wonder: Here it is: Act of March 2, 1893 (45 U.S.C. 1-7), An Act to Promote the Safety of Employees and Travelers upon Railroads by Compelling Common Carriers Engaged in Interstate Commerce to Equip Their Cars with Automatic Couplers and Continuous Brakes and Their Locomotives with Driving-wheel Brakes, and for Other Purposes

Section 2: need automatic couplers that can be uncoupled without man. On January 1, 1898, it will be unlawful for a common carrier used for interstate commerce to haul or permit to be hauled any car that is not equipped with couplers coupling automatically by impact and which can be uncoupled without a man going between the ends of the cars.

Other sections of the act related to brakes, grab irons, and setting of uniform coupler heights

[Eric S]

So, why is Via allowed to do 100+mph where Amtrak is only allowed to do 79mph under virtually identical conditions?

The answer is mostly political, after the Chase MD wreck the hammer came down on Amtrak, arguably the 79mph ruling was an overreaction to a very rare occurence. Via escaped the trickle down effect.

Except the 79mph speed limit long predates the Chase, MD, wreck. The 79mph limit dates to the 1940s (I believe in response to a Chicago-area wreck) while the Chase, MD, incident occurred in the 1980s.

[George Harris]

So, why is Via allowed to do 100+mph where Amtrak is only allowed to do 79mph under virtually identical conditions?

To state the obvious: It is a different country with their own set of laws.

The law in the US that gives us the 79 mph and 59 mph limits has nothing to do with the ability of the equipment to move between the countries and everything to do with the US congress making laws to regulate things they do not understand.

The Canadian's legislature saw no need to impose this sort of regulation.

Before the passage of the law giving us these restrictions in the US there were some lines that allowed as much as 90 mph with no signals at all.

I would argue that the record of performance of the railroads in Canada without these constraints is a very good arguement for their elimination, and against imposing even more stringent and expensive to obey regulations.