Today's [5 SEP 2023] Trains News Wire article discusses ways roads use to mitigate "Loss of Shunt" [LOS] issues with lighter weight trains.
https://www.trains.com/trn/news-reviews/news-wire/the-quest-to-counteract-loss-of-shunt-special-report/
"These may include setting an “axle count,” a minimum number of passenger cars plus the locomotive for each train; mandating the type of equipment used and its corresponding weight; reducing maximum allowable speeds; or any combination of these, based on evidence showing what has successfully prevented the issue from surfacing. "
Has any railroad added non-track-circuit sensors/equipment into this mix of solutions?
My thought, FWIW, is along the lines of a power garage door stop sensor shining a light to a reflector. Counting & timing when enough breaks in the light occur could feed into the crossing controller, dispatcher's system, …
I know there is no one safe place, height above rail or distance from crossing. to make my idea "plug & play" with existing ROW or safety systems.
Your thoughts?
=============================
Update 09/06/23 Trains New Wire article
https://www.trains.com/trn/news-reviews/news-wire/seeking-answers-on-loss-of-shunt-special-report/
Contains: (the complete article has more)
Canadian National has been sponsoring the tests over the last several years, with financial and technical assistance from Amtrak, the Federal Railroad Administration, the Illinois and California departments of transportation,and other Class I railroads.
In a lengthy statement to Trains News Wire, CN says, “Extensive testing up to this point indicates that the random loss of shunt detection events experienced by single-level passenger equipment in the US is due to the absence of sufficient contact area between the wheels of passenger trains and the rail head.
Denver, Colorado has had fun with their PTC crossing gate controls.
https://www.cpr.org/2023/02/10/rtd-not-responsible-for-100m-plus-in-commuter-rail-costs-judge-says/
"RTD initially lauded the lines as pioneers because they were the first U.S. commuter rail lines to have PTC-compliant systems built in “from the ground up.” But the crossing-gate system was not ready by the time RTD opened the A-Line in April 2016. Gates came down too early and stayed down too long — and in one case in March 2016, there was a near miss with two vehicles."
I've encountered problems with light engine moves.
From what I've read and heard, getting crossing protection sensing is a balancing act. Moisture, salt, anything that will change the conductivity between the rails presents problems.
We start out our season dealing with rusty rail. Until we've polished things up a bit, we may or may not get an activation, even with an entire train (for us, 6-8 cars).
One would think that PTC and GPS would make life easier, but it does depend on all the communications links working properly.
Larry Resident Microferroequinologist (at least at my house) Everyone goes home; Safety begins with you My Opinion. Standard Disclaimers Apply. No Expiration Date Come ride the rails with me! There's one thing about humility - the moment you think you've got it, you've lost it...
abdkl: there are other systems out there in use, but they have their own gremlins. Pneumonic systems (motion sensor/PMD) being at the top of the list.
Not sure if there is a rail system out there that is anything but track circuit dependent. Most railroads have a mix.
rdamon: Denver RTD tried to go all-in on an untried and untested system without any trials to confirm what they had. Experienced railroaders warned what was bound to happen and it did. Designers failed to understand how GPS works and when the predictor systems failed because of poor placement of sensors and antennae - the result was chaos. Even though, for the most part, the track circuits worked but could not overcome the other clusterfluffs in the system which continually crashed the PTC system.
The toy train amateurs at RTD still have managed to create create hazardous outcomes where even PTC can't save them (or some of Denver's finer citizens who continue to put themselves at risk).
Thanks, everyone. Interesting stuff. Does anyone know IF the Track signal is actually lost, or simply (?) too weak to be seen, and processed, by the signal receptor
Subsequent thoughts…
Maybe it's time to re-invent PRRs TrainPhone, upgraded with 21st-century micro-sensitive technology. Maybe that would permit WiFi for pseudo-GPS for the Dispatcher's displays and for AMTRAK passengers through the tunnels and around mountains, and canyons. The "pseudo part generated in the cab via speed and accelerometers to function off of the unit's GPS signal lost point.
Revision: I do understand that the Pennsylvania Railroad's Trainphone operated via a low-frequency carrier, and THAT would prevent my idea of having wifi in tunnels from being practical for streaming at computer speed, but my thinking -might- still permit tracking trains within tunnels using something like the old wire-borne signals used to locate wire breaks in telegraph wires. And yes my radio knowledge barely extends beyond the on/off button.
But when I started learning data processing most businesses were only just beginning to get rid of punch cards and electro-mechanical relays. Accounting machines were programmed by plug wires. Some of you may remember seeing those in person. Computers back then, mostly, were octal, not hexadecimal, and computer memory in hundreds of characters, not terabytes. Remote computer terminals Teletype (TTY) and IBM 1050 machines. And asynchronous (telephone), not digital, communications.
I have no idea of what is inadequate in the signal systems of those carriers that are requiring 32 axles.
Amtrak has been operating the Capitol Limited for the past year or so between Chicago and Washington with 20 axles. Engine, baggage car and three passenger cars.
B&O and subsequently CSX restrict single engines to 30 MPH as at higher speeds, the equipment can move through the circuits faster than realys could react. Of course, the newer generations of signal equipment have done away with relays and replaced them with electronic circuits - which to my mind should activate faster than relays.
Never too old to have a happy childhood!
abdklDoes anyone know IF the Track signal is actually lost, or simply (?) too weak to be seen, and processed, by the signal receptor
Traditional crossing protection operates by placing a pulsed signal on the rails in the vicinity of a crossing. If there is no train (hence no shunt) there will be nothing sensed.
On regular systems, the circuit ends with a rail gap. On predictive systems, not so much. Predictive systems activate the crossing protection based on speed and distance from the crossing. Doppler comes into play - it's sort of a hardwired radar system.
When a train shunts (ie, shorts) the standard circuit, the crossing system will detect that and activate the crossing protection. It's really quite elementary in principle.
The problem arises when the sensitivity of the system has to be adjusted for other factors, like moisture, and in northern climes, road salt. Then comes the fine line between being sensitive enough to trip when there's a train in the circuit and not so sensitive that a little rainfall will activate the protection.
It's possible that the sensitivity could be turned down far enough that it would not sense a short train, or that other factors like leaves on the tracks or rust are affecting the quality of the "short".
The usual disclaimer - I have no association with the following site - it popped up with a search: https://worldwiderails.com/how-are-railroad-crossings-activated/.
Along with all of this, the block occupancy signals have to work, too. I do not envy the signal folks the task of figuring out all of the logic involved for a complicated interlocking that also includes road crossings.
There are several books that have been written on the entire topic. It can get complex.
When BART was built there was a concern about loss of occupancy, partly due to the light weight of the equipment. To address this they installed a system called SORS. Sequential Occupancy Release System. This system does not release occupancy, even if detection is losses, unless the train has been detected moving into an adjacent block. The down side of this system is that FOs have to be manually released. In over 50 years there has never been a loss of occupancy accident at BART. I am not aware that this system has been used anywhere else.
In Europe with their lighter weight cars, axle counters are more common than track circuits on any route that has been resignaled in the last 20 or so years.
The PATCO line in NJ/Philadelphia uses coded track circuits just like the "big boys" and runs light weight single car trains on ATC all the time. Has been since 1967... on all sorts of track conditions.
They are not the only transit system with lightweight equipment and functioning track circuits.SEPTA subway surface line functions with single trolley cars.
I'm sure there are others. I'll bet the solutions are expensive for host roads...
-Don (Random stuff, mostly about trains - what else? http://blerfblog.blogspot.com/)
David1005 When BART was built there was a concern about loss of occupancy, partly due to the light weight of the equipment. To address this they installed a system called SORS. Sequential Occupancy Release System. This system does not release occupancy, even if detection is losses, unless the train has been detected moving into an adjacent block. The down side of this system is that FOs have to be manually released. In over 50 years there has never been a loss of occupancy accident at BART. I am not aware that this system has been used anywhere else.
WMATA got in trouble because they didn't have such a system. A bad track circuit lost a train and they let the supervisory system run a train right into a train that stopped in the bad section.
So, the longer the track circuit, the lower the impedence shunt you need to reliably shunt it. Rain, snow, etc. will shunt the circuit if the sensitivity is set to high. Freight RRs can get away with really long blocks, so can have really long track circuits with very low sensitivity. Perhaps with PTC installations, the RRs went to longer track circuits (to save money) without caring too much how it would effect Amtrak?
JayBee In Europe with their lighter weight cars, axle counters are more common than track circuits on any route that has been resignaled in the last 20 or so years.
I think there's some confusion in this thread about exactly what track circuits we're talking about. From everything I've heard - including the Trains Newswire article referenced in the original post - the loss-of-shunt issue typically affects highway crossing signals, not block signals.
A modern ("constant warning time" or CWT) approach circuit in a crossing system is far more complicated than a coded track circuit used for block occupancy. These circuits not only measure whether there is a shunt in the circuit, but where the shunt is and how fast it's moving.
Specifically: an AC voltage is applied across the rails at one end of the circuit; a permanent shunt at the other end connects the rails. When a train enters, it shortens the circuit, which reduces the impedance of the circuit and causes more current to flow. The circuit measures the rate at which the current is changing, and uses that to estimate the train speed and arrival time. That lets the system activate the crossing signals a constant amount of time before the train arrives regardless of train speed.
But obviously there are other factors that cause the impedance to change from time to time. If you make the system too sensitive to changes, you get false activations, which are considered a safety hazard as well. It's like any other system that has to try to detect a signal against a noisy background. The only way to really make things better is to make the signal stronger - with the train providing a better, more reliable shunt.
You could develop visual, laser, or radar systems to try to measure train presence and speed instead of relying on the track circuit, but I don't know that the signal-to-noise ratio would be any better for those (especially if you want them to work when, say, you're shoving a cut of flat cars across the crossing).
The only potential way out is that, instead of having an autonomous system that independently evaluates when a train will arrive, you could have the crossing integrated into the train control system, so the train can tell the crossing when it will be there and the crossing can tell the train when the crossing signals are ready for the train to arrive.
That's how the crossings on the Illinois and Michigan high-speed lines work, and the Denver RTD lines, are SUPPOSED to work. And often they do. But at this point in time these communication-based systems are FAR too unreliable to serve as the only means of crossing protection. So they all have backup plans, and the backup plan always involves a track circuit. On the IL line the backup is a CWT circuit (but the train has to slow to 80 MPH if the communication-based activation fails). On the RTD line the backup is a constant-distance circuit instead (where the crossing activates as soon as the train enters the circuit, regardless of speed). Because most RTD trains are not operating at maximum track speed most of the time, these tend to produce really long warning times. And the communication failed often enough that the local FRA office took exception to those long warning times.
You don't get as many issues with block signals, because a.) they're doing a much simpler task ("is a train present" versus "how fast is the train moving"), and b.) the fail-safe mode (false occupancy) doesn't create as many safety concerns as the fail-safe mode on a crossing (false activation).
Dan
For what it's worth, in Canada CN does not impose any special restrictions on VIA trains. We only have basic restrictions, a single unit or self-propelled car is restricted to 30 mph when operating on bonded track (signalled territory or approaching crossings with warning devices in dark territory) and two units or one unit and one car are allowed 50 mph. If you have three pieces of equipment there are no speed restrictions as far as I know, VIA should be allowed to run a single P42 and two LRC coaches at 100 mph on the Kingston Sub. I know they have run such short trains between Toronto and Sarnia or Windsor, where the passenger track speed is 80 mph or better for long stretches.
Greetings from Alberta
-an Articulate Malcontent
dpeltier From everything I've heard - including the Trains Newswire article referenced in the original post - the loss-of-shunt issue typically affects highway crossing signals, not block signals.
From everything I've heard - including the Trains Newswire article referenced in the original post - the loss-of-shunt issue typically affects highway crossing signals, not block signals.
I post this and then two days later NTSB opens their docket on the fatal collision on the Yuma sub a year ago. A train was routed into a controlled siding and collided with 74 intermodal cars that had been stored there for months. The signal logs indicate that the track circuit on the siding had intermittently cleared a number of times in the weeks before the accident. No conclusions yet, but looks likely that loss of shunt (due to rust) on the standing cars may have played a role, probably along with other contributing factors.
See https://data.ntsb.gov/Docket?ProjectID=105885#.
dpeltier dpeltier From everything I've heard - including the Trains Newswire article referenced in the original post - the loss-of-shunt issue typically affects highway crossing signals, not block signals. I post this and then two days later NTSB opens their docket on the fatal collision on the Yuma sub a year ago. A train was routed into a controlled siding and collided with 74 intermodal cars that had been stored there for months. The signal logs indicate that the track circuit on the siding had intermittently cleared a number of times in the weeks before the accident. No conclusions yet, but looks likely that loss of shunt (due to rust) on the standing cars may have played a role, probably along with other contributing factors. See https://data.ntsb.gov/Docket?ProjectID=105885#. Dan
From a operational standpoint - under every form of rules I operated under for 51 years - a cut of cars placed on a signalled track for any reason for any period of time were to be identified by a 'track tag' as well as having the control point controlled switches Blocked for movement away from the blocked track.
Irrespective of 'loss of shunt' such tags and switch blocks are not to be removed by the person operating the control point(s) without on the ground verification that all cars have been removed and the track can be used in accorcdance with operational needs.
If a train operating under signal indication runs into cars occupying a track - there are many more failures than just 'loss of shunt'.
The false shunt problem takes a delicate tweaking to get proper shunting. Near here CSX had an older crossing activation system that was experiencing many false activations. Final resolution? Was not even CSX's fault. The commercial power feed to the signal case had a bad commecial power ground.
I would hope that CN has not had false actiivatios with these suspect crossings and tweaked the crossings to where they do not detect high speed short Amtrak trains. Also has CN examined the commercial power feeds to the signal bungaloes? Also when did IC (CN) go to local commercial power from pole line power?
blue streak 1The false shunt problem takes a delicate tweaking to get proper shunting
I just do not understand how a Railway that has owned and run the electrical circuitry since installation has no idea what is causing what amounts to an electrical issue and can't resolve it fast. I am failing to grasp that whole concept with this issue specifically. I presume they have a large staff of electricians just to keep the railroad running or at least union crafts with electrical expertise.
blue streak 1I would hope that CN has not had false actiivatios with these suspect crossings
So on this point and I am not sure you realize it or not but false activations and inactivations happen on every railroad with crossing signals. It's more common than you realize and this is why they stress that you should STOP, LOOK and LISTEN and not rely 100% on the crossing signal activation. A lot of people presume that crossing signals are 100% reliable but long ago when I was a railfan I obseved time and time again when they would fail to activate or would deactivate before the train had cleared the crossing. That was back in the 1970's when the railroads had signal maintainers everywhere. I can imagine that the reliability has improved with technology but the cut back in employees makes me wonder how much has changed with overall reliability.
CMStPnPI just do not understand how a Railway that has owned and run the electrical circuitry since installation has no idea what is causing what amounts to an electrical issue and can't resolve it fast. I am failing to grasp that whole concept with this issue specifically. I presume they have a large staff of electricians just to keep the railroad running or at least union crafts with electrical expertise.
Ever notice that when you're watching TV and a commercial comes on, that you have to turn down the volume?
An unusually hard winter may result in more road salt being used. A wet spring may wash a lot of that salt away from the tracks.
Fines in the ballast may change the balance
And the list goes on.
The railroads know how to deal with it - it just takes a little fine tuning from time to time.
Our community is FREE to join. To participate you must either login or register for an account.