Toronto Ontario Line 3 | ?m | ?s | Metrolinx

SELTRAC is easily capable of that, and better, should the track configuration allow. A lot of the ATC/ATO products on the market can, too.

Hell, even the crappy old fixed-block signalling system on the Toronto subway can achieve that in sections, and for short spurts.

Dan

Yes, you can achieve 90 second headways, in short spurts, if everything goes perfectly. If everything goes perfectly…

This is the real world though, and everything will not go perfectly. I’d be impressed if there are any rapid transit lines with 90% utilization rates that reliably hit 90 second headways. That’s a very tough task when you have 50,000+ imperfect human passengers doing imperfect human things on your rapid transit system at any given time. Everything from passengers holding up doors, to poor passenger circulation in stations and trains will impact the headways.

We were discussing this same topic in this thread a few months ago. Someone posted a PDF document comparing headways and utilization rates of various rapid transit lines around the world. If my memory serves me correctly, the only systems that were able to consistently hit that target were high capacity systems with relatively low ridership. It’s easy to hit 90 second headways, even without ATO, when you have a transit system running empty trains, with few opportunities for passengers to screw up the operation. This used to happen all the time during the early days of Line 1, even without ATO. However, it’s a totally different situation on lines that will be at near 100% capacity like the Ontario Line. There is an inverse relationship between the utilization rate of a rapid transit line and reliably achievable headways.

You’d definitely know better than me though. I’d be really curious to know if there are any metro lines in the world that reliably achieve 90 second headways while also having a 90%+ utilization rate. I'm not really interested in systems that achieve this with lower utilization rates. I don't recall being finding any example, but I might very well be wrong

But even if this was achievable elsewhere, I still say it would be unwise to assume it would be replicated here in Toronto, simply due to the local culture. We all know commuters in Toronto have little reservation about holding train doors at stations. All the time at across the system, even when trains aren't all that crowded, I see train operators having to press the door close buttons multiple times due to passengers rushing the doors. The problem is so bad that the TTC has had to hire announcers to practically beg customers not to delay trains at Bloor-Yonge. A few of incidents of passengers delaying a train by 10 seconds will quickly erode away your theoretically achievable 90 second headways

Also, given how cheaply this line is being built, I don't have a ton of faith that these stations will be designed with proper circulation to eliminate the crowding that induces passengers to hold doors open. These stations will likely look a lot like the Line 1 stations, with few points of egress resulting in passengers lumping themselves in one part of the train or platform. We see this all the time on the Yonge Line, where the north end of trains will be crowded with passengers, while the south end is empty, because passengers refuse to move to the south end of the Line 1 platforms at Bloor-Yonge Station (this is because the transfer to the Line 2 platforms are at the north end of the platform). Even though the TTC's ATO system is fully capable of 90 second headways, and even if track geometries were modified to be compatible with 90 second headways, I don't believe for a second that we'd ever see rush hour services with 90 second headways on the TTC, unless we drastically reduced ridership and redesigned the Line 1 stations to allow for better passenger flow through stations. Perhaps that would be achievable off peak, when 90 second headways aren’t needed in the first place, but certainly not on peak in busy sections when those headways would actually be beneficial

So long story short, I think any claims that 90 second headways will significantly boost real-world capacity in Toronto needs to be taken with an incredible grain of salt. Just because a system is technically capable of it, doesn't mean it will ever actually happen when it's needed most (trains and platforms are full)
 
Last edited:
^ I've seen this concern before. Why do some believe 90-second headways aren't possible? What's the challenge to meeting them if the technology is available? Crowding?

It gets harder physically. You want to make sure the trains are at a sufficient stopping distance between them, which for a train is a fair distance. Also at 90 seconds you've really hit the point where any single train delay will throw the whole system off kilter, it's harder and harder to evenly space without having stop and go train traffic.
 
What was the capacity of this proposal? It's using the same cars as the SRT, which has a capacity of under 5,000 pphpd under four-car transits. The trains depicted in this image are larger, utilizing 6-car sets, but even then that is a capacity of under 7,500 pphpd. The ICTS trains shown here take up the full length of the station platforms, so it's not even like they planned the stations with room for longer trains, as we've done with Line 4. But even if we were able to double the length of these stations (probably an impossible task), that's still a capacity of under 15,000 pphpd.

If you look closely at the image of the train it seems a fair bit larger than ICTS. Three doors, four windows, vs two doors three windows. It also has a tapering toward the roof which ICTS also doesn't have. I think what the image was supposed to show is the notion of a "futuristic" looking train, along the lines of 70-80s era BART or DC Metro.
 
^ I've seen this concern before. Why do some believe 90-second headways aren't possible? What's the challenge to meeting them if the technology is available? Crowding?
All empty trains can achieve this without problem as long as it clears each station within less than 90 seconds. In theory this would work even if the deceleration, dwelling and acceleration time is 85 seconds assuming all trains are identical (i.e. brakes wear and tear, motor outputs, etc.). Seeing that in person with trains leaving and the next entering the station at the same time at full speed would be a freighting sight.

Practically the safety bubble between each train needs to be spaced well enough which is controlled by ATC/ATO system. Wear and tear affects performance hence all trains aren't the same. Second issue is dwell time. Uneven dwell time between trains would cause big gaps if the slower train can't catch up. For the TTC, it's dwell time cause people hold doors and cause them to break or be delayed. Third is the weight of the train. Lighter trains can stop and accelerate faster which helps the OL.

A Maglev system would perform a lot better than conventional rail as there is no wear and tear plus precious control over positioning. The Skytrain/SRT also uses a magnetic rail to guide the train thus they can achieve closer headways than conventional rail.
 
If you look closely at the image of the train it seems a fair bit larger than ICTS. Three doors, four windows, vs two doors three windows. It also has a tapering toward the roof which ICTS also doesn't have. I think what the image was supposed to show is the notion of a "futuristic" looking train, along the lines of 70-80s era BART or DC Metro.

Early concept ICTS MKII cars had the tapered and bulging sides (they were eventually built wider, without the angularity, and with 3 doors):

 
Early concept ICTS MKII cars had the tapered and bulging sides (they were eventually built wider, without the angularity, and with 3 doors):


Neat. Looks like something from Total Recall. Though this MkII concept came out well after the 1985 report when the Mk1 were still new (or not even rolled out yet). Guessing mid 90s.

That being said I do like what came out of Skytrain / Bbd's capacity upgrades for the Innovia line of vehicles. Trains are roomier, but with the same narrow floor width. Still pretty certain that Innovia forms the basis for Michael Schabas and Metrolinx's Ontario Line. The guy was a UTDC salesman, now he's back to close that one sale got away all those years ago.
 
The study "assumed an ICTS technology to provide a framework for geometric planning and environmental impact evaluation", and that "with a limited number of changes in geometry a subway technology could be readily adopted."

In the ICTS section of the description of technology options, it mentions either use of the vehicles we have today on the Scarborough RT or the longer and wider vehicles proposed for the then planned GO ALRT lines. Their numbers say that a line capacity of 18,000 pph using six Scarborough RT style trains at a headway of 90 seconds could be achieved, and a line capacity of 30,600 pph using six GO ALRT style trains running at a headway of 120 seconds. The forecast ridership at the time of the study predicted 18,000 pph on a "peak season basis" by 2001.

The cross-section of the proposed Pape-Lipton Avenue station suggests a platform length of 140 meters, as does the Union Station cross section.
If you look closely at the image of the train it seems a fair bit larger than ICTS. Three doors, four windows, vs two doors three windows. It also has a tapering toward the roof which ICTS also doesn't have. I think what the image was supposed to show is the notion of a "futuristic" looking train, along the lines of 70-80s era BART or DC Metro.

Good observation. In that case, the 18,000 pphpd capacity should be accurate

That said, this ICTS Relief Line proposal still seems incredibly short sighted to me. As per the report, that would’ve had the DRL overcapacity barely 10 years after it opened. The line would’ve had no capacity to absorb future growth, let alone accommodate future extensions.

I really don’t understand this desire to build the DRL with as minimal capacity as possible. With both the Ontario Line and the ICTS DRL proposal, these lines wouldn’t reach 15 years old before they’re overcapacity and no longer able to relieve the Yonge Line. I just don’t see the value proposition in this. Why not spend the billion or two today, so that we’re not having to plan another $10 Billion line mere years after the DRL/OL opens?

The wonderful thing about the Relief Line South is that it built in the capacity to support extensions beyond Sheppard for decades to come. It recognized that we’re not just building this to line for ourselves, but also for Torontonians long after we’re gone

I can’t help but feel like if the Yonge Line were designed today, it would’ve been designed with a capacity of 10,000 pphpd, killing any opportunity to ever extend the line. Surely none of us would think that would be a wise decision. Yet these ICTS DRL and OL proposals are trying to do exactly that. I don’t get it
 
Good observation. In that case, the 18,000 pphpd capacity should be accurate

That said, this ICTS Relief Line proposal still seems incredibly short sighted to me. As per the report, that would’ve had the DRL overcapacity barely 10 years after it opened. The line would’ve had no capacity to absorb future growth, let alone accommodate future extensions.

I really don’t understand this desire to build the DRL with as minimal capacity as possible. With both the Ontario Line and the ICTS DRL proposal, these lines wouldn’t reach 15 years old before they’re overcapacity and no longer able to relieve the Yonge Line. I just don’t see the value proposition in this. Why not spend the billion or two today, so that we’re not having to plan another $10 Billion line mere years after the DRL/OL opens?

The wonderful thing about the Relief Line South is that it built in the capacity to support extensions beyond Sheppard for decades to come. It recognized that we’re not just building this to line for ourselves, but also for Torontonians long after we’re gone

I can’t help but feel like if the Yonge Line were designed today, it would’ve been designed with a capacity of 10,000 pphpd, killing any opportunity to ever extend the line. Surely none of us would think that would be a wise decision. Yet these ICTS DRL and OL proposals are trying to do exactly that. I don’t get it

As chinesehorse mentioned and in the report, the ICTS option included ICTS mkI and GO-ALRT type cars in it. So it wasn't specifically a medium capacity option, more a medium and high capacity one. Also the station lengths in the blueprints were still for ~150m. So hypothetically if we did somehow use SRT type cars we probably would've ended up with trains of many cars. And I think it's safe to say the inclusion of ICTS and GO-ALRT was politically influenced.

But yes the Ontario Line station lengths of 100m are a problem for me. Obviously high capacities are achievable. However I've argued for years about the merits of building subway lines for 100m trains elsewhere in the city and beyond, and never would've claimed they can carry more than 25k pphpd. Logic would dictate if 36k for 6-car, then 24k for 4-car. Meanwhile the Prov is claiming they can do +30k pphpd? Hmm. I'm down with narrower different trains. But def longer stations to accompany it.
 
As chinesehorse mentioned and in the report, the ICTS option included ICTS mkI and GO-ALRT type cars in it. So it wasn't specifically a medium capacity option, more a medium and high capacity one. Also the station lengths in the blueprints were still for ~150m. So hypothetically if we did somehow use SRT type cars we probably would've ended up with trains of many cars. And I think it's safe to say the inclusion of ICTS and GO-ALRT was politically influenced.

But yes the Ontario Line station lengths of 100m are a problem for me. Obviously high capacities are achievable. However I've argued for years about the merits of building subway lines for 100m trains elsewhere in the city and beyond, and never would've claimed they can carry more than 25k pphpd. Logic would dictate if 36k for 6-car, then 24k for 4-car. Meanwhile the Prov is claiming they can do +30k pphpd? Hmm. I'm down with narrower different trains. But def longer stations to accompany it.
Technically ML's logic works as the math works out. If you can reduce the headway to 90 sec or less, you make up the difference with smaller, lighter and faster trains like a ~100m Mark III could carry 780-820 riders or 32k ppdph. With PSDs trains can run at full speed into the station and accelerate faster than the slower TTC subways. An 8 car Mark III train would be 140m long and would carry over 40k. More standing room would help capacity.

However ML doesn't understand how rapid transit or local service works. It's totally different than having a GO train sit at Union to load up 2k riders over 10 minutes. Can they achieve 90 seconds headway when the trains are over 75% packed? They might need to pass a law that can arrest people for purposely plying doors opening, delaying, causing damage to the train and charge them with mischief.
 
Technically ML's logic works as the math works out. If you can reduce the headway to 90 sec or less, you make up the difference with smaller, lighter and faster trains like a ~100m Mark III could carry 780-820 riders or 32k ppdph. With PSDs trains can run at full speed into the station and accelerate faster than the slower TTC subways. An 8 car Mark III train would be 140m long and would carry over 40k. More standing room would help capacity.

However ML doesn't understand how rapid transit or local service works. It's totally different than having a GO train sit at Union to load up 2k riders over 10 minutes. Can they achieve 90 seconds headway when the trains are over 75% packed? They might need to pass a law that can arrest people for purposely plying doors opening, delaying, causing damage to the train and charge them with mischief.

IMO the only way 90 second headways is really achievable is if they go with a Spanish Solution (drink) at the busy stations. Everybody gets off on one side, and on on the other in order to reduce dwell times.
 
Technically ML's logic works as the math works out. If you can reduce the headway to 90 sec or less, you make up the difference with smaller, lighter and faster trains like a ~100m Mark III could carry 780-820 riders or 32k ppdph. With PSDs trains can run at full speed into the station and accelerate faster than the slower TTC subways. An 8 car Mark III train would be 140m long and would carry over 40k. More standing room would help capacity.

However ML doesn't understand how rapid transit or local service works. It's totally different than having a GO train sit at Union to load up 2k riders over 10 minutes. Can they achieve 90 seconds headway when the trains are over 75% packed? They might need to pass a law that can arrest people for purposely plying doors opening, delaying, causing damage to the train and charge them with mischief.

On top of that, they’re trying to fit more people per square meter onto the trains than anywhere else on the network. They’ve really built absolutely no tolerance for error whatsoever into this proposal
 

Back
Top