New GO Train Control+Signalling (PTC, CBTC, ETC) -- Safety & Subway-Like Frequency

[mdrejhon]

UPDATE: New November 22, 2017 News:

GO is going to deploy ETCCS -- Enhanced Train Control and Conventional Signaling:
Parsons secures Metrolinx ETCS Contract

This is a hybrid between European ETCS and traditional railroad signalling. It has a moving-block upgrade path. ETCS Level 3 is equivalent to CBTC with moving-block ability. This means Metrolinx/GO is currently deploying a signal system upgrade path that keeps the door open to moving-block capabilities found in ultra-short-headway metro systems.

UPDATE: Post updated April 2016:

Hello,

GLOSSARY:
PTC = Positive Train Control
CBTC = Communications-Based Train Control (can also do positive train control safety responsibilities too).
ATC = Automatic Train Control, often piggybacked on PTC or CBTC

CBTC (or a variant) is now part of Metrolinx requirement for GO RER

Positive train control does an automatic slowdown/stop of your train for safety if train tries to exceed speed limit for any reason (grade, error, inattention, traffic) and/or if it detects known danger ahead (such as a train ahead, a curve ahead, TSOs, PSOs). This saves everybody's asses when the train driver does not manually slow down the train on time for any reason.

Shorter Headways
These requires more advanced control systems. All major RER systems worldwide with tight headways (e.g. 3-minute, 5-minute) now use an enhanced control system such as PTC or CBTC or an Europe variant called ECTS. Paris RER, German S-Bahn, all use variants of these systems. The GO $13.5bn electrification budget includes the cost of this upgrade. Such systems are particularly important for the "SmartTrack" route of GO RER, which may have the tightest headways on the GO network.

TTC Uses It
TTC already has been using a PTC variant for a long time. They are upgrading to CBTC to allow ~2-minute headways on Yonge. The addition of the system to GO's system is one of the pieces necessary to make GO feel more like a metro system, similar to other RER/RER-like systems worldwide.

GO already is researching 5-minute headways on some RER routes (like the "SmartTrack" routing)
Several Metrolinx documents mentions 5-minute headways on some RER routes (like the SmartTrack route). While there is a lot of debate whether or not this is possible with the first $13.5bn of GO RER, there is definitely no doubt given sufficient money, a CBTC system makes this path possible as has already occured elsewhere in the world.

How much is PTC or CBTC?
The assumed cost of installing CBTC into the GO network is estimated by Metrolinx to be $800 million, embedded as part of the existing $13.5bn GO RER/electrification budget. PTC is estimated only $200 million, but Metrolinx is planning on going beyond and planning CBTC for RER.

Saves Lives / Amtrak Disaster
With the Amtrak disaster, and the fact that positive train control would have prevented deaths:
www.npr.org/sections/thetwo-way/2015/05/14/406652406/positive-train-control-the-tech-that-couldve-prevented-amtrak-derailment
www.cnn.com/2015/05/15/us/philadelphia-amtrak-train-derailment/
The train and track was equipped with this safety system, but wasn't turned on yet because they hadn't yet gotten permission to turn on the train control radios yet (radio frequency permission issue)!

Automatic Train Control?
This is often an optional feature of advanced train control systems (variants of PTC and CBTC). GO will not use this unsupervied due to unions and safety issues, especially given platform overcrowding, but the technology provides that option for some systems worldwide. In reality, automation is like an autopilot, which the train driver supervises. This is a speed optimizer, to keep a train more easily in its headway between trains, accelerating and decelerating precisely.

Union Station Bottleneck?
Recent 2015-2016 Metrolinx documents mention other planned major upgrades (Union Station Rail Corridor) plan to increase train throughput at Union station to approximately 50 trains per hour (by ~2031). This includes 45mph crossovers in the east and 30mph crossovers in the west. With this speed and signalling upgrades made available, combined with CBTC, trains will no longer need to crawl slowly towards Union for several minutes before entering the beginning of a Union platform. Combined with short-dwelling, and (potentially expensive) optimizations elsewhere, this potentially allows at least one combined route to have approximately 5-minute same-platform headways at Union, presumably on one of the wider platforms. Whether the debate of ultimate headways are 15-minute or 5-minute, it is assumed that headways will progressively decrease as upgrades are rolled out, and will likely further continue beyond the initial completion from the $13.5bn GO RER budget.

GO RER requires CBTC -- See Newer Post
The short headways in GO RER objectives require very tight train headways. During frequent train services, this can mean increased risks and safety issues (collisions, etc). CBTC will help greatly permitting trains to follow trains at braking distances. TTC, and many other frequent-service subways, already has positive train control on their subway train for a very long time now, and are upgrading to CBTC.

How urgently should positive train control / communications based train control arrive on GO's network? Let's discuss.

 

[vic]

How about VIA? It likely would have helped in the 2012 derailment in Aldershot.
http://en.wikipedia.org/wiki/2012_Burlington_VIA_derailment

 

[BurlOak]

My $100 GPS has a setting where the unit beeps if the speed limit is exceeded. A simple thing like that would alert the operator that they are exceeding the speed limit. Actually controlling the trains speed is a more complex operation, but if we have an operator, why not alert him/her of the speed limit. It would not work for all cases, but I can easily recall some where plain and simple they went too fast.

 

[mdrejhon]

My $100 GPS has a setting where the unit beeps if the speed limit is exceeded. A simple thing like that would alert the operator that they are exceeding the speed limit. Actually controlling the trains speed is a more complex operation, but if we have an operator, why not alert him/her of the speed limit. It would not work for all cases, but I can easily recall some where plain and simple they went too fast.

That works for static stuff such as slowing down for curves in railroads, but unfortunately, it is not that simple for railroads. GPS can also be used to help positive train control, but it also involves signalling, too.

The speed limits the train need to folow, are dynamically changing all the time, based on track state and conditions (e.g. trains ahead, switch state, rail/ballast condition, etc). For example, if a switch (crossover) is activated, the speed limit might suddenly fall to 15mph for one track, while an adjacent track without a switch activated, stays 75mph. And then the next moment the configuration of crossover switches change, it's vice-versa. Passing a crossover between two tracks too fast will cause you to derail as switching tracks must often be done more slowly. And there's TSO (Temporary Slow Orders) and PSO (Permanent Slow Orders), weather conditions, etc. And if the train ahead stops on the same track as you are, or a level crossing gate failed to come down, you're required to stop (e.g. defacto speed limit of 0). Maintenance staff needs to fix one track, a system can signal the train to automatically stop if they're on that out-of-service track. Defects are detected in a section and speed limit is temporarily lowered until fixed. And if a berth is taken at Union, the train can automatically stop before crashing into that train (something that has happened in the past). And fail to follow the speed limit, or you may derail or crash into something. Therefore, simple GPS and static databases will never work for train speed limit control, for many reasons including because your speed limit is changing all the time as tracks switches changes. So signalling must always occur, so the train control/protection system can know what to do.

A very good, complete and safe ATP (automatic train protection) / PTC (positive train control system), combined, can pretty much factor all the above, and know what speed you should be going at, and slow the train down if you're going too fast. There's so many types of train systems including ATP (Automatic Train Protection), PTC (Positive Train Control), ATC (Automatic Train Control), etc, that have less complete to very complete protection or even automation, in various countries -- train protection systems.

GPS may be utilized in some these systems, to dynamically set a speed limit for your train depending on the train ahead/behind of you, so you're slotted safely with enough stopping distance for all trains involved in the event of emergency, and that might even be weather-dependant, too.

More advanced versions of these kinds of systems help make possible continuous moving blocks -- that allows trains to follow each other at exact braking distances -- rather than old-fashioned stationary blocks/sections. TTC is supposedly upgrading their system to allow moving-block operation -- this is the goal of TTC's long-term resignalling project to permit sub-2-minute headways between trains.

 

[vegeta_skyline]

The speed limits the train need to folow, are dynamically changing all the time, based on track state and conditions (e.g. trains ahead, switch state, rail/ballast condition, etc). For example, if a switch (crossover) is activated, the speed limit might suddenly fall to 15mph for one track, while an adjacent track without a switch activated, stays 75mph. And then the next moment the configuration of crossover switches change, it's vice-versa. Passing a crossover between two tracks too fast will cause you to derail as switching tracks must often be done more slowly. And there's TSO (Temporary Slow Orders) and PSO (Permanent Slow Orders), weather conditions, etc. And if the train ahead stops on the same track as you are, or a level crossing gate failed to come down, you're required to stop (e.g. defacto speed limit of 0).
Maintenance staff needs to fix one track, a system can signal the train to automatically stop if they're on that out-of-service track. Defects are detected in a section and speed limit is temporarily lowered until fixed. And if a berth is taken at Union, the train can automatically stop before crashing into that train (something that has happened in the past). And fail to follow the speed limit, or you may derail or crash into something. Therefore, simple GPS and static databases will never work for train speed limit control, for many reasons including because your speed limit is changing all the time as tracks switches changes. So signalling must always occur, so the train control/protection system can know what to do.

Generally you are correct in you assessment, though I don't necessarily agree with referring to every type of speed change as being dynamic.
PSO's are exactly that - permanent. Also the maximum track speed called the "zone speed", can change every couple of miles on some subdivisions. Basically those two things are no different than driving your car down the road, if so one could also refer to the speeds that cars need to follow as dynamically changing all the time too. TSO's are different in that they can change on a daily basis but we are made aware of their location before we go out there and their speed does not change once implemented, though they can be removed.

The more important consideration is that trains brake far more slowly than cars and so they need advance warning and need to start slowing down well ahead of a restrictions.

I would agree though in referring to following a train ahead of you, crossing malfunctions and crossing over tracks
as dynamically changing speeds. And while cars may be faced with similar circumstances i.e. a traffic suddenly not working, because of their ability to slow down quickly, similar restrictions are insignificant for road vehicle. It's definitely quite true that we can't know for sure what out routing will be ahead of time. The routing can be different each time we pass the same location in the same day. For one run we might be routed on the same track all the way to union but the very next run, in the same direction, might be routed through a 15mph crossover.

However, it wouldn't be accurate to say that "suddenly" happens. Normally all the RTC does is input the starting location of a train and the destination. The CTC system then routes a train through the territory based on other trains it is expected to encounter. Normally the quickest possible route is chosen i.e. crossing over the least number of times. This is unless there are conflicts, in which case the system will reroute the train where needed along the way. Under these circumstances there will be an advanced warning of at least 3 signals before reaching the location were the train is to crossover. So in that sense it is not a "sudden" change.

However a minor delay of a minute or two to a single train can affect multiple trains down the line. This is where the RTC can step in and change the routing of a train from the original route the system had program for it. But he would only do so if it would be beneficial in some way(or at least it will be to CN). I swear sometimes... However in this instances he can only change the route if a train is still 3 blocks(signals) away from the controlled location he wants to change and would involve changing the aliment of the switch.

A controlled location is a place where a train can crossover in signaled territory, though the name actually represents the fact that the signals indications can be controlled by the RTC or the CTC system itself and not the fact that there are crossovers present in that spot. Most of the signals you can see along the tracks are not controllable in this way i.e the RTC or the system cannot change them directly but rather they change by themselves to preset indications based on what the signal at the controlled location is and/or because of the presence of other trains nearby or problems with the track circuit itself.

The RTC can change the signal in less time under two circumstances- A) if he contracts the crew, tells them exactly what he wants to do and the train crew acknowledges that they still have time to be able to react to the change or B) in case of an emergency.

It's also important to note that there are only a limited number of unchanging locations where trains can crossover tracks and the maximum absolute speed for these crossovers do not change, unless they are physically removed and replaced of course. The vast majority are rated for 45mph while there are a few that are only good for 15mph. However it is true that the signal system may indicate a lower speed then the maximum speed at that crossover. I may get an indication telling me to only do 30mph through the exact same crossover(s) that I went through at 45mph earlier in the day. This however does not mean that the crossover itself is suddenly not capable of supporting 45mph. In fact that speed indication has nothing to do with the condition of that crossover seeing as how the system is not programed in this way. But rather the slower speed signal is an advance warning of a train several signal blocks ahead of me. Obviously that's not to say I can just blow off the signal and say whatever, I'm doing 45 here because I know I physically can. Although I can tell you that's exactly what many CN crews did back when they were running the show, that was there idea of bending but not breaking a rule. I can assure you the expectation today is of a strict adherence to the signal indication in all cases.


Anyhow, on a day to day basis things usually remain unchanged. Some trains will always generally be routed straight down one track and while other trains will have to cross over multiple times along the way, but generally use the same crossovers at the same controlled location, every day. This can create a false sense of security which appears to be the case with what happened to VIA 92.

That crew would stop at Aldershot GO/VIA station and then would occasionally crossover at Burlington West but almost never at Aldershot East. These are the two controlled locations between Aldershot and Burlington GO stations.

Aldershot East is almost never used to crossover passenger trains since having to crossover using the slow speed(15mph) crossovers there would greatly impact train schedules. So passenger trains will not be crossing over there unless there is no other choice. And that's only if there going in the right direction/right track - trains going east can only crossover to tracks to the south and trains going west can only crossover to tracks to the north. A train going east on track 3 cannot crossover to any track and trains going west on track one cannot crossover to any other mainline track, but can enter into Aldershot yard though under no circumstances would a GO or VIA train be routed there. CN basically put in a controlled location here just so that freight trains can access Aldershot yard without having to block the mainline too long. There's a whole explanation for that too, but its rather long and not relevant.

Aside from the crossovers at Aldershot East, slow speed crossovers(15mph) from one mainline track to another also exist at Oakville yard(at Trafalgar rd. in Oakville), Campa(30th st. in Toronto) Fort York (Strachan Ave. in Toronto) and of course all over the USRC(Union Station). The first 3 for the same purpose as the ones at Aldershot East - to access yards. They've slowly been removing these crossovers from the system since they obviously pose a hazard. In fact the TSB(Transportation Safety Board) recommended that passenger trains should be restricted from using these crossovers at all. But Transport Canada, basically the 'old guard' that I mentioned in a previous post, rejected that recommendation. http://teamstersrail.ca/February_25_2013.htm#.VVa_JkZ5bnA Now why would they reject a seemingly simple life saving recommendation like that? Because they are catering to the freight companies. Approving that recommendation would of been detrimental to CN & CP by placing restrictions on where train meets with passenger trains could happen and thus reduce their operational flexibility. This would in tern slow down the movement of goods and ultimately cost them money. And fFor CN & CP it all comes down to the money. And that Money is power, the power to lobby the government and influence their agenda. I not suggesting some big conspiracy, rather its pretty simple. The government is more beholden to their wants and needs because they are a big driver of the Canadian economy, much more so than VIA rail who's contribution is insignificant in comparison. And metrolinx does not have significant influence on the federal level. Now I referred to it as life saving, because had those crossovers been the typical 45mph version, chance are(well over 90% probably around 98-99%) that train would not have derailed. How do I know this? Because over the last 10 years I am aware of about a dozen instances of trains going through 45mph crossovers at speeds in excess of 68mph - in some cases almost 80mph, without derailing(nevermind flipping over to its side). And obviously if they did derail, you all would of known about it too.

Point being, those slow speed crossovers are there but are almost never used by GO or VIA trains, it is considered usual for GO & VIA trains to use them. Then there was the issue of being distracted by and concerned for the work crew they spotted working ahead of them on the track they were operating on. There was no rule 42 in effect(advance notice of a foreman's limits/ basically a track work order), which is allowed because the RTC instead had signal protection in place i.e the train was routed to crossover to track 3 from track 2. Signal protection is not the preferred means of doing any kind of track work. But is utilized when there is short notice of a defect that needs to be corrected asap.


It's not unusually to see track workers working under signal protection alone. But what is usual is for these track workers to not move off to the side when the see a train is approaching. Normally they take that extra precaution. From a train crews perspective its very alarming when they don't. I certainly feel nervous when ever I see that. Sometimes I'll be coming around a bend at well over 100kph and suddenly I'll see a work crew working on an adjacent track hammering away. It's a very jarring experience since they are literally only a couple feet away from the trains path and thus instant death. Which is unfortunately what happened to Rick McColl, a CN foreman who was fatality struck by a VIA train back in 2009. The circumstances were a little bit different in that he wasn't even working under signal protection. At one point CN allowed track work to be done on a live track with only a spotter to watch for oncoming trains. Though the track Rick and his college were working on was not the track the VIA that hit him was running on. Unfortunately when they saw the VIA train coming they moved to get off the track but Rick went the wrong way and crossed over in front of the VIA trains path.

What's relevant about this situation is that they only had 6 seconds to react from the time they saw the train to the moment it passed their location due to the trains speed and the curvature of the track. This is why its always jarring as an operator to anyone, trespasser or workers by the track. With the limited amount of time to react, there is essentially nothing you can do to prevent an incident from happening. This wariness carries over into circumstances where you see a crew working on the ground well ahead of you who don't move to step off to the side or in some cases don't react in any noticeable way as you move closer and closer to them and thus the amount of time to react keeps decreasing. I've come across situations like that and it definitely can be very disconcerting and thus distracting. Because even though I can see that the signal is routing me onto another track, and I'm obviously slowing down for that crossover ahead of time, I never feel 100% sure about the safety of those track workers until I can actually see that the switch points are lining me onto the next track over. You can see a signal from miles away, but that doesn't physically effect you train. It's the swtich points that physically route the train over and they can only be seen from about a 1/2 mile away at best. Deep down I think most people don't explicitly trust a system that is out of their control 100% of the time, though that feeling might only be subconsciously perceived. So no matter how hard you try, that sense of danger is still present, even if only on a subconscious level.

So if its present why not just slow down if one is not completely confident you ask? Well for one that sense of danger is not for the safer of ones self or ones passengers, its for the safety of the crew working on the ground. And on the railroad everyone has to take responsibility for there own safety first and foremost. It's a very unforgiving industry when it comes to mistakes. And for that matter everybody knows, or at least should know, that unlike a vehicle on the road a train cannot move out of the way - you must move out of its way. That carries over into these circumstances. Anyone looking at this situation from a logical point of view would expect the workers on the ground to step off to the side of the track, which would be no more then 4 or 5 feet to the left or right and would take a second or two. Not to mention expend millions of times less energy the then energy lost from stop and starting a train again. It's simple logic. I'd also like to believe that everyone realizes, especially a track worker, that a train takes far longer to stop than a vehicle. I previously mentioned how I manged to stop a GO train traveling at 57mph/92kph in the length of a platform - that is not applying the breaks until I hit the start of a platform. This was was done with practically the best breaking GO train I've even had - brakes well worn in and a very sensitive air braking system that was responding almost instantly which indeed is very rare, as normally there is a few second delay before you feel any braking. Plus dry rail and an slight uphill grade which helps in braking. So with that 'prefect' combination, the best braking heavy rail train in Canada still took about 1,200ft/360m to stop from 57mph. Yet the average vehicle can stop from 60mph/96kph in about 180ft/60m(http://www.bbc.co.uk/schools/gcsebitesize/science/add_gateway_pre_2011/forces/motionrev3.shtml). If you include reaction time or rather "thinking distance" that increases to 240ft/73m - which is still 5 times better then pretty much the absolute best performance possible from a train.

The whole point of that is to show just how reinforced and ingrained the concept of trains having the right of way is. On top of that we are trained not to stop the train when we something in front of us - even if its a person. Obviously when it is clear that person is not moving or if we encounter an obstruction that will cause significant harm to the train, which few mobile objects can, we will start breaking. But not before then for all the reason I stated before. Like I said, this concept is still diametrically opposed to the natural human sense of concern for others.

There is of course a degree of tolerance in how strictly each individual adheres to that. But if you can't adhere to that to a certain margin, you are not considered a viable candidate to operate a train. I've seen it, people have not make the cut as engineers because they were entirely too cautiousness. Right or wrong, trainers take that as you not being confident in your ability to operate and the company will not view you as a productive employee in time sensitive industry like the railroad.

To me its clear what happened to VIA 92, they knew they had a permissive signal - that is any signal other than a stop signal. Every train crews biggest fear is going by a stop signal. Because that means at a minimum a huge suspension or termination and at worst, death due to collision. And a stop signal is instinctively very obvious - all red. For whatever reasons overspeeding crossovers don't instil that same fear and panic, although I think that's changing. So they knew, at least on some level that they had something other than a stop which is why they approached that light(at Aldershot East for the slow speed crossovers) so aggressively. But that fear for the safety of the track workers who were not stepping off to the side and possible not even acknowledging the trains approach, clearly distracting their attention from the realization of what exactly the signal was telling them to do. Compounded by the fact that they never, or practically never crossed over using the slow speed crossovers at Aldershot East. And they had a schedule to keep, in a time sensitive industry where there is a constant pressure to perform and maintain the schedule or risk disciplinary measures. And lastly there was 3 people in the cab which could also have served as an additional distraction. Imo the more people in the operating area the more potential for distraction. Even two people can serve as a distraction though that is countered by the ability to help each other maintain focus, but 3 or more has no positive benefit since its human nature to want to socialize when forced together into groups. This likely reinforced the fear for the track workers safety they had been feeling when they vocalized this concern to each another, which is very regular discussion that crews would have when operating for hours at a time in a locomotive and encountering people out on the track.

Of course all of that in no way absolves or excuses a crew from their responsibility to following instructions as provided by a signal indication. But it is the most plausible explanation and on some level it gives a sense of why something that seemed so obvious could have been missed.

Anyhow I suppose I've meandered off track a little.
The good thing about the GO system is that most of the corridors are tangent(straight). The biggest danger is a sudden decreases in speed due to curvature & thankfully there are few places like that on the system. For the lines which are not very straight(Richmond Hill, Stouffville) speed are generally low the entire way & the biggest sudden change in speed on those two lines is no more than 15mph. There are a few significant drops in speed for PSO's on the Lakeshore West but generally they're over stated, again there's been a few violations where crews have unfortunately exceeded the limitation, with no damage sustained. TSO's are more of a concern, though they are overstated even more so than PSO because it is more likely to miss something that changes on a day to day basis, the worst violation I know of being 80mph over a 30mph TSO, with no damage sustained. OF course if everyone did that eventually the track is going to buckle but thankfully its a rare occurrence somewhere in the area of 1 in 1000 trains movements. Right now the biggest safety hazard out there on the GO system are those 15mph crossovers in otherwise high speed(I mean relatively, not real HSR of course) territory. However those could be removed fairly easily. I'm shocked they didn't do it after VIA 92. Otherwise PTC biggest benefit would be enforcing stops. Not just for stop signals but for emergency situations like the washouts that have occurred at Dixie rd. in recent years. We're lucky no trains went over those or it would of been a disaster.

 

[CDL.TO]

Thanks for such an informative post, vegeta.

 

[mdrejhon]

Same!

People complaining of a malfunctioning switch delay need to give your post a read. Hope all the old switches, including the 15mph ones, are replaced eventually.

 

[dowlingm]

Fort York and presumably Oakville can't be laid at the door of the freight operators now that Metrolinx owns those tracks?

The first thing I'd say about PTC is that it's a shame there hasn't been more obvious joint development between the Europeans and North America. ERTMS is rolling out across Europe whereas US signalling seems to be still quite split up with freight roads resisting the cost of ACSES and ITCS being installed as an alternative on 110mph projects. Furthermore, PTC requires (depending on method) quite a bit of radio spectrum and the implementers are being obliged to purchase it while the U.S. Feds are providing a pittance in contribution.

If PTC is rolled out in Toronto/Canada there are going to be interesting issues around technology, scope (will CP install it on Milton and CN on Richmond Hill?) and funding (who will pay for VIA to get PTC - themselves, the Feds or will Metrolinx have to fund it?) Then, if people (other than in the U.S.) notice that VIA is operating 90-100mph services without advanced signals, what will be the reaction of CN to PTC between Quebec and Windsor?

edit: Railway Gazette article where APTA say PTC deadline won't be met south of the border
http://www.railwaygazette.com/news/...tion-by-december-not-possible-warns-apta.html

 

[vegeta_skyline]

Shameful how that deadline will likely be pushed back, there certainly was ample time to implement it.

And it just boggles my mind why Metrolinx hasn't replaced those all the 15mph crossovers at Aldershot East, Oakville Yard, Campa & Fort York yet. Especially since they own the tracks through most of them(the exception being Aldershot East), its now entirely their responsibility. I'll give them credit for replacing the ones at 9th line(between Clarkson & Oakville stations) and at Mimico East a few years back when they added additional tracks through each of those areas. But why wait to couple it on to another project when they pose a very real and dangerous safety hazard now? Especially when that safety hazard can be easily rectified. There certainly is more than enough space to extend the signal plants at Aldershot, Oakville & Campa so that's no excuse. There are space constraints at Fort Yard, but really that controlled location does not even need to be there since trains can crossover to any track within the nearby USRC. In fact its a huge operational constraint when trains do have to crossover at Fort Yard.

Metrolinx would have to pay for there replacement regardless of whether or not they owned the track or how rarely GO trains used the crossovers of concern(i.e. Aldershot East). As stated, they aren't commonly used(except for Fort York it seems) because their use results in train delays. But their continued existence will always pose a danger. CN meanwhile is just fine with keeping them they way they are because their trains have to slow down anyways before entering into the freight yards at Aldershot, Oakville & Mimico. But they wouldn't care one way or another if they were replaced and long as the project timeline was coordinated with them, which is a given.

 

[mdrejhon]

Yeah, I can imagine even unactivated 15mph switches is still a risk factor when speeding through Aldershot. If the switches are malfunctioning, frozen in ice with heater broken, or accidentally activated I imagine that it can be quite an unexpected derailment danger to any fast train. I presume there's several failsafes that will alert you to slow down if the switch ends up in a hazardous mode.

Wasn't it also planned to upgrade the switches in the USRC too as well?

Also, if there is wide implementation of train control systems (positive train control) and level crossings were completely eliminated, and on a corridor that does not have freight trains (e.g. SmartTrack route), how would you feel operating a non-FRA train such as an European-type commuter train? Assuming Transport Canada lets it happen under those safer conditions? I know you expressed concern being in tin boxes.

 

[crs1026]

As I understand it, the vanilla PTC that is being implemented on freight railroads in the US manages not only track speed limits but also authority to proceed (red vs green signals, to be simplistic) and distance to next stopping/slowing point. I don't know if it also manages scheduled platform stops for passenger trains. For GO, speed authority and stopping control are all needed. This might affect the choice of technology and a more subway-style technology might be preferred.

At the end of the day, it's a question of risk vs priority. Would we prefer that GO defer some other project eg RER and spend the available money on PTC instead? So far, GO has always had a very commendable safety record and I would argue there is no 'crisis' driving us towards PTC. However, near misses do happen infrequently (most people never hear about them) and if there ever were a major tragedy, questions would arise about whether it was prudent to have operated without a greater safety margin. In the US, the prevalence of train control on commuter railroads varies from route to route. There is no reason to believe that GO has fallen behind the pack, but clearly the US regulator is moving to insist on PTC.

I would argue that PTC does belong somewhere in GO's priority list, but incrementally and no faster than proven technology is available on the market. (One hears rumbles that the US technology still has bugs - better to wait til the product is assured.) Over the long term, it will become as commonplace and as must-have as seatbelts in cars.

- Paul

 

[smallspy]

And it just boggles my mind why Metrolinx hasn't replaced those all the 15mph crossovers at Aldershot East, Oakville Yard, Campa & Fort York yet. Especially since they own the tracks through most of them(the exception being Aldershot East), its now entirely their responsibility. I'll give them credit for replacing the ones at 9th line(between Clarkson & Oakville stations) and at Mimico East a few years back when they added additional tracks through each of those areas. But why wait to couple it on to another project when they pose a very real and dangerous safety hazard now? Especially when that safety hazard can be easily rectified. There certainly is more than enough space to extend the signal plants at Aldershot, Oakville & Campa so that's no excuse. There are space constraints at Fort Yard, but really that controlled location does not even need to be there since trains can crossover to any track within the nearby USRC. In fact its a huge operational constraint when trains do have to crossover at Fort Yard.

In some of those cases - Canpa and Oakville Yard come to mind - those crossovers were installed long ago primarily to allow freights to access yards or lower speed areas. If you look at the track arrangements, the more recent crossovers installed at either of those locations (to TMC at Canpa, and to track 4 at Oakville) were all #20s/45mph because they were for the passenger trains and thus the speeds were also expected to be higher through them.

As for replacing the crossovers - you know that it's not as easy as simply pulling out the #12s and putting in #20s. A #12 crossover is about 325 feet long, from point-to-point. A #20 is over 550 feet long. The plant at Fork York is scheduled to be upgraded when they extend the A1 track past Exhibition, but it's also going to lose a bunch of crossovers because there simply isn't the real estate there to allow for all 45mph crossovers across 5 tracks in both directions.

Metrolinx would have to pay for there replacement regardless of whether or not they owned the track or how rarely GO trains used the crossovers of concern(i.e. Aldershot East). As stated, they aren't commonly used(except for Fort York it seems) because their use results in train delays. But their continued existence will always pose a danger. CN meanwhile is just fine with keeping them they way they are because their trains have to slow down anyways before entering into the freight yards at Aldershot, Oakville & Mimico. But they wouldn't care one way or another if they were replaced and long as the project timeline was coordinated with them, which is a given.

Ultimately, anything that prevents you from continuing at full speed - and this includes anything engineering-wise - could be considered a danger. I don't think that a #12 crossover is any more or less dangerous than some of the geometry-based PSOs on the Oakville Sub, and if anything they could be regarded as safer because there are signals to help reinforce them (versus the flags, which could be forgotten, covered over or removed). At the end of the day, any additional crossover that exists, regardless of its speed rating, is one that allows for more flexibility.

The bigger issue, in my mind, is one of efficiency. A train that has to slow down from 60mph to 15, and then back up to 60 is forced to expend a huge amount of energy to do so, not to mention the time wasted as well.

One thing that I would like to see done, and which apparently is being looked at by the Ivory Tower, is the installation of even higher-speed crossovers in select locations. There exist a couple of AREMA standards for very-high-speed crossovers - #27.5 (Amtrak 60mph) and #32.7 (Amtrak 80mph). Yes maintenance will be higher, but the time savings for some trains - in particular expresses and VIAs - that are no longer forced to slow down from 80 and 90mph to 45mph can be fairly substantial.

Dan
Toronto, Ont.

 

[vegeta_skyline]

Yeah, I can imagine even unactivated 15mph switches is still a risk factor when speeding through Aldershot. If the switches are malfunctioning, frozen in ice with heater broken, or accidentally activated I imagine that it can be quite an unexpected derailment danger to any fast train. I presume there's several failsafes that will alert you to slow down if the switch ends up in a hazardous mode.

Indeed the signal system is designed to revert to the most restrictive indication in the event of a failure in the track circuit and it won't give a 'permissive signal' unless the switch is locked in. While I've never heard of a switch failing in this way, I wouldn't go so far as to say its completely impossible. I have seen and heard of signals showing the wrong indication at times, though so far thankfully they've all been more restrictive indications rather than less restrictive.

Wasn't it also planned to upgrade the switches in the USRC too as well?

The switches & signal system in the USRC are to be upgraded to allowed for 25mph verse the current 15mph limit as part of the Signal Modernization contract. Unfortunately that work was mishandled by the previous contract and will not be in fully implemented until 2019. Dan can tell you all about the mismanagement that went on there. What won't change is the 10mph speed limit within the shed and there might be additional restrictions added in the near future as of a result of Daniel Panacci unfortunate death.

Also, if there is wide implementation of train control systems (positive train control) and level crossings were completely eliminated, and on a corridor that does not have freight trains (e.g. SmartTrack route), how would you feel operating a non-FRA train such as an European-type commuter train? Assuming Transport Canada lets it happen under those safer conditions? I know you expressed concern being in tin boxes.

haha tin boxes indeed, its one of the two main reasons why I prefer operating the locomotive, the other being that the brake system just seems to react quicker and is easier to modulate from there. If the level crossings were removed and PTC was implemented I would feel quite comfortable operating non-FRA trains.

 

[vegeta_skyline]

In some of those cases - Canpa and Oakville Yard come to mind - those crossovers were installed long ago primarily to allow freights to access yards or lower speed areas. If you look at the track arrangements, the more recent crossovers installed at either of those locations (to TMC at Canpa, and to track 4 at Oakville) were all #20s/45mph because they were for the passenger trains and thus the speeds were also expected to be higher through them.

I was going to note that they had changed a couple of the crossovers at those locations and I did mention that they normally do not route us through the #12 crossovers but my post was already becoming exceedingly long and over detailed. The main point was simply that slow speed #12 crossovers still exist there.

As for replacing the crossovers - you know that it's not as easy as simply pulling out the #12s and putting in #20s. A #12 crossover is about 325 feet long, from point-to-point. A #20 is over 550 feet long. The plant at Fork York is scheduled to be upgraded when they extend the A1 track past Exhibition, but it's also going to lose a bunch of crossovers because there simply isn't the real estate there to allow for all 45mph crossovers across 5 tracks in both directions.

To be fair I didn't suggest that it would be that simple, which is why I said that there is nothing to prevent them from extending the signal plants themselves(save for Fort York) and not simply that there is space to extend the crossovers within the signal plants, which there is not. Extending the track circuits and relocated the signal masts does greatly increase the costs but its certainly more than worth it to prevent casualties, not to mention the benefit of increase efficiency as you noted.

I don't think that a #12 crossover is any more or less dangerous than some of the geometry-based PSOs on the Oakville Sub, and if anything they could be regarded as safer because there are signals to help reinforce them (versus the flags, which could be forgotten, covered over or removed). At the end of the day, any additional crossover that exists, regardless of its speed rating, is one that allows for more flexibility.

I have to respectfully disagree. There have been numerous incidents over the years which would lead me to concluded otherwise. Trains have operated at or exceed 80mph through #20 crossovers, the second slowest geometry based PSO on the Oakville sub(the 45mph PSO on track 1 at Port Credit) and through a 30mph TSO, all without incident. However anything above 50mph through a #12 crossover will almost assuredly result in a derailment.

The bigger issue, in my mind, is one of efficiency. A train that has to slow down from 60mph to 15, and then back up to 60 is forced to expend a huge amount of energy to do so, not to mention the time wasted as well.

I agree completely, though I feel the safety aspect is just as important.

One thing that I would like to see done, and which apparently is being looked at by the Ivory Tower, is the installation of even higher-speed crossovers in select locations. There exist a couple of AREMA standards for very-high-speed crossovers - #27.5 (Amtrak 60mph) and #32.7 (Amtrak 80mph). Yes maintenance will be higher, but the time savings for some trains - in particular expresses and VIAs - that are no longer forced to slow down from 80 and 90mph to 45mph can be fairly substantial.

I had heard that they were considering higher speed crossovers at one time but that they had put the idea on the back burner. Good to hear they are still looking into it.

 

[k10ery]

What's a ballpark cost estimate for PTC on the full GO network? $500 million? $5 billion?