TTC Fleet Procurement Strategy - 2022

[cplchanb]

According to Steve Munro's post on the new trains for Line 2, the TTC's requirements for acceleration are extremely low:


60 seconds to reach 65 km/h is pathetic. BART trains only take 17 seconds to reach 70 km/h.
Even the "high rate" is very slow, with 0-83 km/h in 60 seconds. BART trains reach only take 26 seconds to reach 90 km/h, and continue on to 110 km/h within 40 sec.

In 60 seconds, a Dutch local train will have already reached 140 km/h

Sorry but the comparison is apples to oranges. First the third rail for BART and especially the overhead for the Dutch train is much higher voltage. ttc is only 600v DC which itself is 40% lower than barts 1kv third rail not to mention the Dutch uses 1.5kw.

Secondly and more importantly both your examples' designed operating speed is much higher than ttcs..BART being the slower operator of the 2 is still designed for 110km/h speeds. Ttc only has a revenue speed of 75km/h... there's no point in accelerating that fast when you're not going fast anyways.they don't even have enough trains to support that to begin with.

 

[reaperexpress]

BARTs trains are configured to operate at much higher top speeds than the TTC's. The TTC prioritizes acceleration, not top speed. It also prioritizes maintainability, which is why high-rate is almost never used.

A TTC subway can go from 45km/h to a full stop, service the stop, and back up to 45km/h in 45 seconds. A BART train can not.

Dan

If it takes 60 seconds to reach just 65 km/h, then clearly they are not prioritizing acceleration. There are plenty of segments in Toronto where the speed limit is 70 or above.

The dwell time and braking rate do not impact a train's ability to accelerate. Comparing those factors is irrelevant to the topic of acceleration, which should clearly be a priority on Line 2 since it has very tight stop spacing.

And besides, the TTCs braking rate does not seen to be higher than other systems such as BART. Could you provide a source to support that claim?

I also doubt that maintainability is such a huge issue that they need to kneecap acceleration to less than a third of the acceleration rate of systems in yhe US, Europe and Asia, considering that basically every metro on the entire planet accelerates more quickly than the TTC subway.

 

[DSC]

Steve has just published a lengthy analysis of the TTC's plans for Line 2. Well worth reading, as usual. https://stevemunro.ca/2024/12/30/ttc-requests-proposals-for-new-line-2-trains-and-signalling/ Like most projects, the actual trains are only part of what is required!

 

[rbt]

If it takes 60 seconds to reach just 65 km/h, then clearly they are not prioritizing acceleration. There are plenty of segments in Toronto where the speed limit is 70 or above.

Toronto Rocket acceleration isn't linear which makes any discussion much more complicated. It starts at about 0.9 m/s2 which, after 60 seconds, would be around 190km/h. They get to ~30km/h as the tail of the train leaves the station but beyond that they accelerate quite slowly.

According to wiki, Bart trains accelerate at 1.34 m/s² and likely do so over wide range of speeds. Due to the large distances between stations they can take advantage of higher speeds.

It would be interesting to see what rush service on TTC might look like with higher acceleration. Station dwell time by the train in-front might make it largely useless though you'd expect it to help a train behind schedule catch up a bit faster.

 

[reaperexpress]

Toronto Rocket acceleration isn't linear which makes any discussion much more complicated. It starts at about 0.9 m/s2 which, after 60 seconds, would be around 190km/h.

Yes, no train's acceleration is linear. You can see a similar curve on every train, including the BART train and NS train shown here.

They get to ~30km/h as the tail of the train leaves the station but beyond that takes more and more time.

30 km/h in 140 m is also very slow.

Other systems only take about 12 seconds to accelerate from 0 to 50. Are you suggesting the TTC trains would reach 50 in 12 seconds then take 48 additional seconds just to reach 65?

 

[rbt]

Are you suggesting the TTC trains would reach 50 in 12 seconds then take 48 additional seconds just to reach 65?

I don't know what the 0 to 60 curve of a Toronto Rocket train actually is. The platform exit speed was from a platform screen door report, initial acceleration came from wikipedia.

Just wanted to note that while 0-60 might be 60 seconds (I have no idea if that's accurate or not), their 0-30 is not 30 seconds.

 

[reaperexpress]

I don't know what the 0 to 60 curve of a Toronto Rocket train actually is. The platform exit speed was from a platform screen door report, initial acceleration came from wikipedia.

Just wanted to note that while 0-60 might be 60 seconds (I have no idea if that's accurate or not), their 0-30 is not 30 seconds.

My point is about plausibility. The magnitude of the difference in acceleration between TTC subway trains and BART trains is so large that even though we don't know the exact acceleration curve, there is no way the proposed train specs could be anywhere near the acceleration rates we observe on basically every other metro system in the world.

Here are the acceleration rates from the video I posted, which are about the same as dozens of measurements I took of local trains in the Netherlands.

You can see that the train sustains acceleration rates far above 0.9 m/s/s for extended periods, and doesn't drop below 0.9 m/s/s until well beyond 65 km/h. There seems to be a low startup rate around 0.6 m/s/s, probably due to a combination of ramping up gradually for passenger comfort, and lag in the GPS measurements.

Even if the TTC train could maintain 0.9 m/s/s from 10 km/h to 64 km/h and spend the entire remaining time climbing to 65 km/h, it would still be significantly slower than the BART train throughout the entire range of speeds (orange line in the chart below).

More realistically, a peak acceleration rate of 0.9 m/s/s would produce a curve more similar to the red line.

 

[T3G]

The fact that the T1 refurb is happening after all is an interesting one. If they had gone ahead with this half a decade or so ago, there might have been less hysterics about the possibility of the SRT debacle repeating itself.

When is this program supposed to start, and how extensive it is supposed to be, if the plan is still to retire the cars as new trains become available? If most of the trains are supposed to be available by 2033, that's either going to be a very light refurb, in which case one has to wonder how effective it will be at actually extending the life of the cars (ALRVs part 2?) or it's going to be an insane waste of money to junk them after less than a decade.

 

[81-717]

BART being the slower operator of the 2 is still designed for 110km/h speeds.

@T3G Who knew subway systems with 100 km/h speeds actually exist?

The fact that the T1 refurb is happening after all is an interesting one. If they had gone ahead with this half a decade or so ago, there might have been less hysterics about the possibility of the SRT debacle repeating itself.

When is this program supposed to start, and how extensive it is supposed to be, if the plan is still to retire the cars as new trains become available? If most of the trains are supposed to be available by 2033, that's either going to be a very light refurb, in which case one has to wonder how effective it will be at actually extending the life of the cars (ALRVs part 2?) or it's going to be an insane waste of money to junk them after less than a decade.

Yeah, it's probably going to be a light refurb good for an extra 3–4 years, not 10. But the idea of using some of the first 55 trains for the SSE/YNSE rather than replacement is interesting, as makes a bold assumption that the SSE and YNSE will actually open before all 55 replacement trains arrive and the additional 15+ (specifically meant for said extensions) begin arriving.

Have the T1s been operating their entire lifespan to this point basically without any kind of overhaul, aside from SMS (Scheduled Maintenance System) like in NYC, and minor retrofits like LEDs? This would be somewhat reminiscent of the H2s operating for exactly 30 years without an overhaul (and H4s being overhauled at around the same age).

 

[reaperexpress]

Sorry but the comparison is apples to oranges. First the third rail for BART and especially the overhead for the Dutch train is much higher voltage. ttc is only 600v DC which itself is 40% lower than barts 1kv third rail not to mention the Dutch uses 1.5kw.

Okay, so voltage is different, so what? The voltage is something we can change if we decide that we want our trains to perform comparably to at least an average metro system.

If we ignore any lessons from systems with higher voltages, we will never consider the benefit of increasing the voltage and will forever keep suffering with unnecessarily slow transit.

Secondly and more importantly both your examples' designed operating speed is much higher than ttcs..BART being the slower operator of the 2 is still designed for 110km/h speeds. Ttc only has a revenue speed of 75km/h... there's no point in accelerating that fast when you're not going fast anyways.they don't even have enough trains to support that to begin with.

The closer stop spacing on the TTC compared to those other systems makes acceleration MORE IMPORTANT, not less.

The number of trains required on a line is a function of the round trip time. If you accelerate faster, you need FEWER trains.

 

[smallspy]

If it takes 60 seconds to reach just 65 km/h, then clearly they are not prioritizing acceleration. There are plenty of segments in Toronto where the speed limit is 70 or above.

And there are lots of places where even 60km/h isn't achieved, so what's your point?

The priority has been given to the lower speed regime, as the trains spend far, far more time down there.

The dwell time and braking rate do not impact a train's ability to accelerate. Comparing those factors is irrelevant to the topic of acceleration, which should clearly be a priority on Line 2 since it has very tight stop spacing.

They don't, but the acceleration rate - which you are complaining about - is the longest part of that sequence. And it goes to show that the acceleration curve is not linear.

And besides, the TTCs braking rate does not seen to be higher than other systems such as BART. Could you provide a source to support that claim?

Braking rates are approximately the same around the world. BART's are likely not substantially different than the TTC's.

The TTC used to publish pamphlets of the various vehicle types that they had in their fleet, including rather comprehensive ones on the subways, but stopped in the early 1990s. Bombardier then published their own versions of the T1 and TR cars - Steve Munro has thoughfully scanned the TR ones and you can find them here: https://stevemunro.ca/2011/06/04/the-toronto-rockets-debut/ Included in them are train performance specifications.

I also doubt that maintainability is such a huge issue that they need to kneecap acceleration to less than a third of the acceleration rate of systems in yhe US, Europe and Asia, considering that basically every metro on the entire planet accelerates more quickly than the TTC subway.

In the short time that the TTC did run the trains in high-rate mode in the 1980s, they found out that traction motors burned out twice as often. Now, this may have also been a function of the shitty equipment spec'd on the H6 cars, but nonetheless that experience has quite obviously coloured their view on high-rate mode and has resulted in the situation we have today.

Dan

 

[81-717]

In the short time that the TTC did run the trains in high-rate mode in the 1980s, they found out that traction motors burned out twice as often. Now, this may have also been a function of the shitty equipment spec'd on the H6 cars, but nonetheless that experience has quite obviously coloured their view on high-rate mode and has resulted in the situation we have today.

Ironic since one of the reasons the H6 cars were ordered in the first place was precisely because of the change from high to low rate:

In 1981, the TTC considered expanding its subway fleet by another 26 cars as the change from high-rate to low-rate operation on the Bloor-Danforth Subway required the use of more trains.

Has high rate ever been used at all at any point between the 1980s and present day? And wasn't it used before the 1980s when the M1 and H1 entered service?

 

[Northern Light]

The exchange above between @reaperexpress and @smallspy is an interesting one, as one might expect from two very knowledgeable people.

I do think its becoming a tad tinged by mutual exasperation though.

Perhaps I could invite a step back in which we bring up the same points, but organized a bit differently.

*****

The point of discussing acceleration, and/or maximum speed in a procurement thread should simply be read as can we do better, if so, what trade-offs are there, in performance, maintenance or upfront costs; is it worth it?

Sticking for the moment to the two issues above, I won't get mired in details better discussed by the above subject matter experts; but I will try to position the above to perhaps illicit from same some further insight.

Assumption/Question:

What benefits are obtained, theoretically, from faster acceleration, and/or greater maximum speed?

1) From the Rider perspective, its entirely about travel time. So what is the estimated, probable improvement in travel times, for argument's sake, from Kennedy Station to Yonge/Bloor, or from Vaughan Metro Ctr to St. George, just to provide a benchmark example of the gain.

2) From the system perspective, the theoretical gain is one of efficiency (fewer trains required to move 'x' people), which can either be used to generate additional capacity or as a cost savings that can be redeployed elsewhere.
What is the likely savings then, from achieving the lower travel time of higher speed/acceleration?

How many fewer trains are required, how much does that save? Do the higher performance metrics result in additional cost per train? How much do the higher metrics drive up maintenance and power costs?

***

Are there any one-time capital costs, trains aside, required to achieve the higher level performance? If so, what is the cost of same, and how long would it take to make the required improvements?

***

My object here is not to give our fine subject matter experts endless homework, but simply to take advantage of their knowledge such that the rest of us have a better sense not only of what could be achieved, but whether it makes financial or operational sense to achieve that.

********

One final note, since I'm here............. can I get a comment on why it takes the trains on Line 1 so long to open their doors after stopping, and is there anything to be done w/existing or new rolling stock to put a stop to that waste?

Many thanks in advance for the thoughtful contributions of the above noted, and any others who have insight to offer.

 

[reaperexpress]

And there are lots of places where even 60km/h isn't achieved, so what's your point?

My point has consistently been that TTC subway trains, notably the RFP for new trains that haven't even been designed yet, have extremely slow acceleration. The fact that trains are not reaching 60 km/h is yet another demonstration of how slow the acceleration is. As noted above, BART trains reach 60 km/h in just 14 seconds so if our trains had comparable acceleration they could reach that speed between almost every station on the system.

I have never complained about the 80 km/h top speed of the existing and proposed trains, so I have no idea why you and @cplchanb are talking about acceleration at high speeds. The fact that other trains can accelerate onwards to 110 km/h or 160 km/h is not relevant to this discussion since that does not enhance their ability to accelerate at lower speeds. In fact, higher top speeds tend to be detrimental to acceleration since they require longer final drive ratios.

The priority has been given to the lower speed regime, as the trains spend far, far more time down there.

The reason our trains spend so much time at low speeds, is that they have very poor acceleration. As noted above, it should only take 12 seconds to reach 50 km/h, so they should hardly ever need to be operating below that on straight segments between stations. Priority has evidently not been given to the lower speed regime, since TTC trains' acceleration at low speeds is also far slower than most other metro systems in the world.

They don't, but the acceleration rate - which you are complaining about - is the longest part of that sequence. And it goes to show that the acceleration curve is not linear.

Hence why I was talking about the acceleration rate. You were the one who started the straw man argument about braking rate and dwell time. I also never claimed that acceration was linear, so that's also a straw man argument.

Braking rates are approximately the same around the world. BART's are likely not substantially different than the TTC's.

Yeah that's what I thought. Hence why I was confused about your claim:
"a TTC subway can go from 45km/h to a full stop, service the stop, and back up to 45km/h in 45 seconds. A BART train can not."

Given that the braking rate is the same and BART's 1.2 m/s/s average acceleration from 0-45 km/h is significantly faster than the TTC's peak acceleration rate (let alone the average), then the only remaining variable is the dwell time, which is also irrelevant to the topic of acceleration. Again I'm just confused about what you were trying to argue by making this claim.

The TTC used to publish pamphlets of the various vehicle types that they had in their fleet, including rather comprehensive ones on the subways, but stopped in the early 1990s. Bombardier then published their own versions of the T1 and TR cars - Steve Munro has thoughfully scanned the TR ones and you can find them here: https://stevemunro.ca/2011/06/04/the-toronto-rockets-debut/ Included in them are train performance specifications.

Thanks for the link! The brochure confirms the peak acceleration of 0.9 m/s/s that we've been discussing, which is much slower than the 1.5 m/s/s found on many other metro systems, and roughly 1.8 m/s/s I observed on BART trains travelling at low speeds.

In the short time that the TTC did run the trains in high-rate mode in the 1980s, they found out that traction motors burned out twice as often. Now, this may have also been a function of the shitty equipment spec'd on the H6 cars, but nonetheless that experience has quite obviously coloured their view on high-rate mode and has resulted in the situation we have today.

Technology in the 1980's was very different, as you note. Nowadays we can use ATC to select a different acceleration profile depending on the train's punctuality, allowing higher performance to be targeted in real time to a handful of trains that need it. So the maintenance impacts would only be a fraction of what they were when every train is always in high rate. That also changes the cost/benefit analysis since reducing scheduled round trip times will save millions of dollars. Each 3 min reduction in round trip on a line with a 3-min peak headway is one fewer to purchase/maintain/operate. One train represents $13.78 Million dollars of capital cost and hundreds of thousands of dollars per year in staffing, storage and maintenance. You can pay for an awful lot of electricity and motor replacements with that amount of money. And that's before even talking about the value of making transit more competitive.

The reason I am so adament about the topic of acceleration is that the RFP for a new generation of trains is the time to think about how to speed up our existing rapid transit services. This is the moment they can think about specifying equipment that can reliably handle higher torque, can vary the acceleration rate based on real-time traffic control needs, and future-proof for other performance enhancements such as increasing the voltage from 600V to 750V.

 

[reaperexpress]

One final note, since I'm here............. can I get a comment on why it takes the trains on Line 1 so long to open their doors after stopping, and is there anything to be done w/existing or new rolling stock to put a stop to that waste?

This is a really good point. There are 76 stations in a round trip on Line 1, so if we can save 2 seconds per stop, that's 152 seconds round trip, which is about the same as the headway, so it's enough to reduce the fleet requirement by one train.

I'm definitely not an expert in vehicle operations, but some methods that have been used in other places include automating the door opening procedure. For example on the Elizabeth Line (Crossrail) in London, the train determines on its own when the train is stopped in the correct position and immediately opens the doors when the train stops (unless the operator has specifically told the train not to). This eliminates the time that guard needs to confirm that the train is stopped in the correct position, point at the marker then press the button to open the doors.

It might also be possible to adjust the programming such that the doors are unlocked below 2 km/h or so, instead of 0 km/h. That would allow them to unlock slightly before the train comes to a stop, reducing that delay time. By the time the doors are open wide enough for a person to fit through, the train would already be stopped anyway. The TTC would probably have a knee-jerk reaction that this is unsafe, but 2 km/h is slower than an escalator so even in the extremely unlikely event that the door has time to open before the train is fully stopped, it's not a danger. Many other systems such as the Paris Metro and Den Haag Trams start opening the doors before the train is fully stopped and it doesn't seem to be a problem.

The TTC and many other organizations usually overlook the safety consequences of implementing 'safety' procedures that slow down transit service. Any time transit becomes slower, more people are encouraged to use other modes, all of which are far more dangerous than transit - both for people inside and outside the vehicle. My intuition is that the injuries and fatalities resulting from even a minuscule mode shift from transit to cars would easily outweigh the extremely remote chance of a minor injury while exiting a train that was moving at less than 2 km/h.