Montréal Transit Developments

[Streety McCarface]

@Urban Sky
An extremely interesting and well thought out analysis, I must say.

I can add to the average trip length question for the Deux-Montagnes line. The 2016 ridership for the following stations were:
Canora — 431,400 passengers
Mont-Royal — 600,000
Montpellier — 634,700
Du Ruisseau — 716,300
Bois-Franc — 827,500
Sunnybrooke — 791,500
Roxboro-Pierrefonds — 1,332,300
Île-Bigras — 161,600
Sainte-Dorothée — 963,400
Grand-Moulin — 506,100
Deux-Montagnes — 1,608,300

Central is difficult to figure out, since it combines all three lines into one ridership number — 10,018,800. The sum of the 3 EXO lines that serve it have a daily ridership of 11,500,100 passengers per day, meaning that approximately 14% of riders do not travel into Central (probably more on the DMT line since the line is electrified), meaning that any estimation is going to be a little higher than assumed, but for the sake of argument, we'll say that all of the DMT line's passengers travel into Central station.

Under this argument and assuming some reduction in ridership numbers, the passenger*km distance travelled by the line users would be about 140,300,000 km/year, and that the average trip length would be 18.8 km, close to double what you assumed. Even when we consider a margin of error of 5 km, it's still almost 5 km higher than the 9.4 km assumed.

 

[BurlOak]

I would find it odd if this were not the case given that it means REM has no built-in expansion capability? That being said at a minimum the DM line stations could handle longer trains (which could be limited to that section or could simply leave their doors closed on the rest of the network). I also imagine the tunneled stations at the airport would be built with longer than required boxes at a minimum (we could also likely expand service by other means given this branch has only 2 stations.)

I have yet to hear Toronto consider this at all.
On Yong and Bloor, once ATC is in effect - can a full length 7th car (or even eighth) be added to the train. I am sure when pocket track and yard space was designed back in the day, there must have been some tolerance that more accurate train controls now can take advantage of.
For new transit lines, this same concept can be used as well - but the one or two major stations could either have all door boarding (or Spanish style), or they could have longer platforms to accommodate the full ultimate length of train, or both.

 

[p_xavier]

SkyscraperPage Forum - View Single Post - The Great Canadian Transit Thread

SkyscraperPage.com's discussion forum is one of the world's most active community discussion portals for skyscraper, urbanism and building enthusiasts. Featuring coverage of the world's projects, there are topics that will appeal to the skyscraper enthusiast in everyone.

forum.skyscraperpage.com

This $1.21 figure is a telling example for how the CDPQ and their many supporters frame facts and figures: By just comparing it with the $0.72 per passenger-km subsidy the REM's promoters will receive, one could think that the REM transports passengers at 59.5% ($0.72 divided by $1.21) of the operating costs incurred by AMT/RTM/Exo's (it's difficult to keep track of all the name changes) commuter rail network, thus saving the taxpayer $0.49 ($1.21 minus $0.72) per passenger-km. For whatever reason, the EXO's Annual Reports no longer specify cost and revenue figures per line and we'll have to resort to the 2016 AMT Annual Report, which features the following tables:

Table 1: Performance indicators of AMT network (2016)
View attachment 194584
Source: AMT 2016 Annual Report (p.43)
Note: the figures I will use in the following tables will be "Achalandage" (ridership), "Trajet (km)" (route length), "couts d'exploitation" (operating expenses) and "recettes usagers" (revenues).

To compare, the performance indicators provided by the EXO 2018 Annual Report only include the passenger counts, which hinders any cost comparison of the current commuter rail network with the future REM:

Table 2: Performance indicators indicators of Exo network (2018)
View attachment 194586
Source: EXO 2018 Annual Report (p.31)


At this point, it is important to note that the REM network will only replace entirely one single line (the Deux-Montagne line), while it will truncate two other services (the Saint-Jerome and Mascouche lines, which will terminate at the projected A40 station, where passengers will be forced to transfer onto the REM for the final miles into downtown Montreal) and partly overlap with the catchment areas of the remaining EXO lines: Candiac, Mont-Saint-Hillaire and most importantly: Vaudreuil-Hudson.

Therefore, to determine whether it is more cost-effective for the taxpayer to transport the existing ridership on the Deux-Montagnes lines with the existing commuter rail service or with the REM, we need to somehow estimate a per-passenger-km figure for that line. Unfortunately, the AMT Annual Report does not provide any passenger mileage data, which means we'll need to get creative, but let's start by assuming that $1.21 is the per passenger-mile for the entire AMT network:

Table 3: Estimation of passenger-km figures by assuming per passenger-km subsidy of $1.21
View attachment 194590
Compiled from: AMT 2016 Annual Report (p.43) with "Passenger-km" figures derived by using the $1.21 per passenger-km figure (Annual subsidy divided by per passenger-km subsidy equals annual passenger-kilometre count).
Note: line length provided for the Vaudreuil-Hudson line excludes the mileage between Vaudreuil and Hudson, as it is only served by one train per day.

However, dividing the passenger-km figures we derive for the various lines by their respective ridership counts yields improbably short average trip lengths (compare second-last column in table above with the table below: for instance, 2.0 km on the Deux-Montagnes line doesn't even cover half of the distance to the first station), if we look at the station spacing of the various routes:

Table 4: Distance between the downtown termini and the subsequent stations for the various commuter rail lines in Montreal
View attachment 194589
Measured with: Google Earth

Given that there is only one line for which the estimated average trip length equals more than 1 station from its terminus, it becomes clear that the Mascouche line is the line with the worst per passenger-km figure (as also indicated by it having the lowest cost-recovery rate, as shown in Tables 1 and 3) and that that figure must be significantly smaller (than $1.21) for the other lines. If we assume that the average trip length as percentage of the respective line's route length of the Mascouche line is representative for all other lines, the per passenger-km subsidy on the Deux-Montagnes line decreases from $1.21 to $0.44 (i.e. by 64%), at which point transporting a single passenger on the Deux-Montagnes becomes pricier rather than cheaper with the REM (as we will see in more detail in Table 7):

Table 5: Estimation of passenger-km figures by assuming an average trip distance of 18.8% of total line length
View attachment 194605
Compiled from: AMT 2016 Annual Report (p.43) with "Passenger-km" figures derived by using the $1.21 per passenger-km figure (Annual subsidy divided by per passenger-km subsidy equals annual passenger-kilometre count) for the Mascouche line and assuming the same ratio between average trip length and total line length as on the Mascouche line (i.e. 18.8%) on all other lines.
Note: line length provided for the Vaudreuil-Hudson line excludes the mileage between Vaudreuil and Hudson, as it is only served by one train per day.

However, the average trip length is still slightly inferior than the distance to Canora as the first station when traveling from Gare Centrale (5.6 km vs. 5.7 km). If we instead assume that the average distance travelled on the Mascouche line (i.e. 9.4 km) is representative for all other lines, the per passenger-km subsidy on the Deux-Montagnes line decreases even further (to $0.26, or less than one-quarter of the $1.21), at which point the REM's cost disadvantage (for the taxpayer) of transporting a single passenger on the Deux-Montagnes line increases even further:

Table 6: Estimation of passenger-km figures by assuming an average trip length of 9.4 km
View attachment 194610
Compiled from: AMT 2016 Annual Report (p.43) with "Passenger-km" figures derived by assuming that 9.4 km (i.e. the average distance travelled on the Mascouche line, given the $1.21 per passenger-mile subsidy figure) is also the average distance travelled on all other lines.
Note: line length provided for the Vaudreuil-Hudson line excludes the mileage between Vaudreuil and Hudson, as it is only served by one train per day.


So, in summary, what effect do the difference "subsidy per passenger-km" figures have on the Deux-Montagnes line? Most obviously, a decrease of the AMT's figure of $1.21 to $0.26 turns the REM's per passenger-km cost advantage of $0.49 to a disadvantage of $0.46. More importantly, as (given that we know the total subsidy) a lower subsidy per passenger-km figure results in a higher passenger-km count (or average trip length figure), the REM's per passenger cost advantage of $1.00 turns into a disadvantage of $4.33. This means that rather than reducing the operating costs of transporting the current ridership by $7.6 million (or 40%), it increases it by $32.8 million (or a shocking 176%). Note that an average trip length of 9.43 brings you to just past the 3rd (of 11) stations when travelling from Gare Centrale towards Deux-Montagnes, which means that it is quite likely that the AMT's per passenger-km subsidy figure is even lower than $0.44, which would mean that the REM's cost disadvantage is even worse. But even at $0.26, these figures are disastrous enough for a mode (rail) which is usually characterised by economies of scale:

Table 7: Financial effect of the various per passenger-km subsidy figures on the REM's relative cost-effectiveness
View attachment 194613
Compiled from: data provided or estimated in Tables 3, 5 and 6.


Having followed this project closely from its beginning, this selective and distorted use of facts and figures is very typical of the REM project and its promoters. Trying to justify an already obscene figure of $0.72 by waving around an even bigger figure ($1.21) to justify the removal of the most cost-effective commuter line, which has a much, much lower figure just demonstrates how an honest discussion about this project and its merits and costs is obscured. The fact that AMT/RMT/exo stopped publishing the figures one needs to somewhat compare the operating costs faced by the taxpayer with those of the current commuter rail network just at the time where the REM project was revealed (April 2016), just exemplifies how this project is shielded from any due diligence.

Talking about being selective and distorted with the use of facts. You omit one of the most important item. Infrastructure, trains, operation costs and financing are all included in the CDPQi number per km. It would cost nearly a billion to just to upgrade the Deux-Montagnes line to have two way service and a secured tunnel, and that's without two new stations. So if you want to compare real costs, compare real costs. If the CDPQi can be competitive with building a whole new system and operating it comparer to just operating costs, it's the public provider that's being expensive.

 

[rbt]

... whereas Toronto has all but given up on it's big lofty transit goals like Network 2011.

That's a good thing; that plan was based on the faulty assumption that nobody wanted to move or work downtown, and that suburban subway corridors could mimick the success of the nearby highway corridors. In fact I walked through a parking lot near York Mills that was set aside about the same time as Network 2011 was approved, for offices expected to be constructed in the early 90's (still no takers, plenty of interest for condos though). The sad bit is some of that plan did get built; and we pay for it every single year with reduced service city wide on every other corridor (read the TTC operating budget; those heavily subsidized short segments of subway could instead run hundreds of additional buses with more carrying capacity than the subway segments handle).

If Toronto followed through on all the bad plans over the last 100 years, then it would be both much more expensive and not a very nice place to live.

Frankly, ditching Network 2011 should be a point of pride alongside ditching the highway plans that preceded it.

 

[felix123]

Also saddened that some think that one must ignore the undeniably racist laws that are being passed to prevent people driving these trains from wearing turbans.

In general, no one is allowed to drive those trains, because they're automated.
Not to downplay this by the way. I think that any Quebecers who support it are total f-ing morons, even if that's the majority, as it won't help "save" Quebec's culture.
That said, I do think that the religious symbol ban should not be part of this thread since this is about RT in Montreal, not the Quebec government's awful new law.

 

[Urban Sky]

Montrealers daring to think and dream big is something that ought to be commended, not mocked.

You misunderstand the reasoning of the REM's critics: they don't object with capital costs in excess of $6 billion and the annual operating subsidy of $240 million per se - they simply don't agree that the project they are promised can justify investing the equivalent of one entire Metro line (in capital costs) and more than twice the annual budget of the entire commuter rail network ($98 million in 2016). If you maybe took a closer look at how well the Skytrain (i.e. basically the same technology) suits Vancouver, you may be better able to understand that they believe many questions remain unanswered:

Opinion: Canada Line is a model example of a poorly-designed, under-built toy train

@Urban Sky
An extremely interesting and well thought out analysis, I must say.

I can add to the average trip length question for the Deux-Montagnes line. The 2016 ridership for the following stations were:
Canora — 431,400 passengers
Mont-Royal — 600,000
Montpellier — 634,700
Du Ruisseau — 716,300
Bois-Franc — 827,500
Sunnybrooke — 791,500
Roxboro-Pierrefonds — 1,332,300
Île-Bigras — 161,600
Sainte-Dorothée — 963,400
Grand-Moulin — 506,100
Deux-Montagnes — 1,608,300

Central is difficult to figure out, since it combines all three lines into one ridership number — 10,018,800. The sum of the 3 EXO lines that serve it have a daily ridership of 11,500,100 passengers per day, meaning that approximately 14% of riders do not travel into Central (probably more on the DMT line since the line is electrified), meaning that any estimation is going to be a little higher than assumed, but for the sake of argument, we'll say that all of the DMT line's passengers travel into Central station.

Under this argument and assuming some reduction in ridership numbers, the passenger*km distance travelled by the line users would be about 140,300,000 km/year, and that the average trip length would be 18.8 km, close to double what you assumed. Even when we consider a margin of error of 5 km, it's still almost 5 km higher than the 9.4 km assumed.

Thank you for making me aware that these figures are actually provided on the following page of the 2016 Annual Report, including the number of passengers boarding/detraining the Deux-Montagnes line at Gare Centrale (6,590,000):

Table 1: 2016 ridership figures provided by the AMT for each station of the Deux-Montagnes line

Source: AMT Annual Report 2016 (p. 44)

We can now divide the passengers of the Deux-Montagnes line into 2 distinctive groups:

1. Passengers travelling between Gare Centrale and any other stations
2. Passengers travelling between any two stations other than Gare Centrale

Thanks to above table, we know the size of group 1 (the 6,590,000 passengers counted at Gare Centrale), which means that the balance to the Deux-Montagnes total ridership must be the size of group 2 (7,581,600 minus 6,590,000 equals 991,400).

If we now take the passenger figures of all stations other than Gare Centrale and pretend that all these passengers travel to Gare Centrale, we indeed get an average distance of 18.8 km:

Table 2: Passenger and passenger mile count of all passengers counted at stations other than Gare Centrale (and assuming they all fall into Group 1)

Source: compiled with figures from Table 1

Obviously, the total of above passengers is overstating the known number of Group 1 passengers, as it has to match the passenger count at Gare Centrale. Therefore, we have to remove 991,400 trips (or 1,982,800 passenger counts) from above table. In absence of an origin-destination matrix with the exact passenger counts for every possible O-D pair, we will have to arbitrarily remove 1,982,800 passenger counts from above table. However, even though it's impossible to determine the exact passenger-km count for Group 1 passengers to calculate the average trip length, it is easy to define the upper and lower bounds for the latter, as we can remove the passengers from either end of the line:

For the upper-bound of the average trip length, we have to add up the passenger counts starting with the station nearest to Gare Centrale until we've reached the passenger count for Group 2 (i.e. 1,982,800). Then we have to figure out how to remove that number from the 2,382,400 passengers which were counted at either Canora, Mont-Royal, Montpellier and Du Ruisseau, while minimizing the number of passenger miles: as it just happens, the first two stations (Canora and Mont-Royal) happen to have the shortest spacing between any 2 stations along the line (0.9 km), while the next two stations (Montpellier and Du Ruisseau) are not much further apart (1.5 km), which means that our 991,700 Group 2 trips will be the 431,400 passengers counted at Canora (assumed to travel between Canora and Mont-Royal as "Subgroup A"), while the balance will be assumed to be trips between Montpellier and Du Ruisseau (as "Subgroup B"). This results in an upper-bound average trip length of 19.3 km on the Deux-Montagnes line, derived from a passenger-mileage figure of 146.5 million:

Table 3: Calculation of the upper-bound for the average trip length on the Deux-Montagnes line (in 2016)

Adapted from: Table 2
Note: Above tables excludes the passenger counts at Gare Centrale, which are a double-count of the passenger count of Passenger Group 1, while the passenger counts for Group 2 are adjusted (by dividing them by 2), in order to avoid a double count.

Conversely, for the lower-bound of the average trip length, we have to add up the passenger counts starting with the station furthest away from Gare Centrale until we've reached the passenger count for Group 2 (i.e. 1,982,800). However, even though Grand-Moulin and Deux-Montagnes already share more than 2 million passenger counts between them, we also have to include Sainte-Dorothée as there are not more than 506,100 trips which can be removed from Grand-Moulin. Therefore, we have to figure out how to remove 1,982,200 passenger counts from the 3,077,800 passengers which were counted at these three stations, while maximizing the number of passenger miles: we therefore treat the 506,100 trips counted at Grand-Moulin as "Subgroup A" trips between Grand-Moulin and Deux-Montagnes and the balance as "Subgroup B" trips between Sainte-Dorothée and Deux-Montagnes. This results in a lower-bound average trip length of 14.4 km on the Deux-Montagnes line, derived from a passenger-mileage figure of 108.9 million:

Table 4: Calculation of the lower-bound for the average trip length on the Deux-Montagnes line (in 2016)

Adapted from: Table 2
Note: Above tables excludes the passenger counts at Gare Centrale, which are a double-count of the passenger count of Passenger Group 1, while the passenger counts for Group 2 are adjusted (by dividing them by 2), in order to avoid a double count.

Given that I travelled between Grand-Moulin and Mont-Royal at least once in 2016, we know that both, the upper and the lower bounds, do not accurately reflect the actual average trip length. However, we do know that the actual figure has to be between these values and with the upper-bound being only just over one-third (34.6%) higher than the lower-bound, we have a good idea of where that figure actually lies. If we update Table 7 of my previous post, the current operating subsidy per passenger-km figure is neither $0.44 or $0.26, but rather $0.13-$0.17:

Table 5: Financial effect of the various per passenger-km subsidy figures on the REM's relative cost-effectiveness

Adapted from: Table 7 of my previous post with the average trip length figures from Tables 3 and 4 in this post

One should note that the REM's $0.72 figure is a marginal figure, whereas the AMT's operating subsidy of $0.13-$0.17 is an average figure, which means that attracting one additional rider travelling the Deux-Montagnes line end-to-end will increase the REM's operating subsidy by $21.23, whereas it is probably only a few cents (i.e. whatever marginal costs transporting an additional ~100 kg over 29.9 km causes) with the AMT. Let that sink for a moment...

 

[CityStay]

Absolutely - but why do you ridicule people because you weren't aware that planning for both is about 20 years old?

Please don't call people posers.

Champlain Bridge is a Federal matter (it crosses the seaway, international waters) . The feds finally decided to replace the old bridge in 2015 because they had to - it was literally falling apart and they could no longer avoid replacing it. There was no serious plan that could go from the South Shore until the bridge was replaced, obviously.

I don't consider pre-election napkin plans to be actual plans - we've seen dozens of those and they are what they are, napkin plans. Pretty lines in a Powerpoint presentation. Just because something that is actually being built happens to somewhat correspond to a previous 'napkin plan' doesn't mean that it's a culmination of years of planning - It's not, it's an actual plan that is being carried out irregardless of whatever 'napkin planning' preceded it.

And poser = question. I'm surprised that a proud Anglo such as yourself didn't understand that!

 

[p_xavier]

People always talk about the Canada line like it was underbuilt and yet it will still not be fully used for many decades. Maybe it was built the right size with the correct ridership estimates for once? Comparing to the TTC Yonge line is laughable with the maximum PPHPD as Vanvouver is a century from such numbers and surely it's better to build many more metro lines with smaller stations (where the expenses are) than do the Toronto way and divert its users from other lines on line 1.

What's seems like good design for an urbanist is mostly bad design for a transportation engineer.

 

[Streety McCarface]

People always talk about the Canada line like it was underbuilt and yet it will still not be fully used for many decades. Maybe it was built the right size with the correct estimations for once? Comparing to the TTC Yonge line is laughable with the maximum PPHPD as Vanvouver is a century from such numbers and surely it's better to build many more metro lines with smaller stations (where the expenses are) than do the Toronto way and divert its users from other lines on line 1.

What's seems like good design for an urbanist is mostly bad design for a transportation engineer.

The Canada Line hasn't even been around for a decade and they already have to extend the platforms. That's not a result of good transportation engineering, that's a result of being very conservative with ridership estimations, and not accounting for future ridership growth. Also, any good transportation engineer will know that the maximum PPHPD is not, and should never be an indicator of what to build in a rapid transit line. When you reach that peak capacity, theoretically, no one can enter the train if no one gets off, meaning that a person could be stuck at a station indefinitely. Take the Bloor Line as an example, it's fairly common to have to wait for 1-4 trains to go by before you can get on a train during peak periods at the inner stations, but the maximum capacity of 28K passengers per hour per direction maximum has not yet been met (It's closer to 25K PPHPD during the peak periods). The line isn't "technically" overcrowded but experience has you waiting 10-20 minutes to get on a train. In any other country, the Bloor line relief would be an issue that would have a similar awareness level to planners and politicians as Yonge line crowding is today, not because it's necessarily as bad as Yonge line crowding, but it has the potential to be within a few years.

The PPHPD metric does not account for small fluctuations in ridership, nor does it give a clear enough picture for the average commuter. Lines should always be built with enough capacity (or at least the provision to expand within a few weeks if necessary) to sustain future growth for a few decades (regardless if there's a PPP that's running the line and they have an interest to cheap out wherever they can), while not wasting resources on capacity the line will not require before the infrastructure's lifecycle ends. That is a much bigger challenge than building a line that will be at-capacity, or very close to capacity from day one.

The line doesn't need to be a full-fledged subway, but building it to run only 2 car trains was a joke, and everyone knew that. Canada line ridership has grown from 38 million passengers per year during its first full year of service (which was during the Olympics so ridership was definitely inflated by at least 2 million riders) to close to 49 million riders last year. The line has been seeing ridership increases of around 5.5-6% for the past 3 years.

 

[nfitz]

Champlain Bridge is a Federal matter (it crosses the seaway, international waters).

That's not the reason why it's federal - there are other bridges that cross the seaway that are entirely provincial.

The feds finally decided to replace the old bridge in 2015 because they had to ...

The decision was announced in years earlier, in 2011. And that was after years of study and consultation. Even the 2011 media reports mentioned transit on the bridge - and the 2007 report I referenced above is very detailed about LRT, using the current alignment. Things don't happen overnight, they take years of study ... though the two-station spur to Dorval and the 4-station spur to Baie D'Urfé might be the exception to that rule (or maybe not - perhaps there's a history there I'm unaware of ).

And poser = question. I'm surprised that a proud Anglo such as yourself didn't understand that!

Actually I'm not familiar with the word in that context - I'd wondered if you mis-spelled "poster" ... but the context didn't make sense.

I've never claimed to be an "Anglo" - isn't that considered a derogatory term in French Quebec? I lived in Eastern Ontario for years before ever going to Quebec, and am an immigrant to Canada. I simply said I was a Quebecker before I ever came to Toronto.

The Canada Line hasn't even been around for a decade and they already have to extend the platforms.

I hadn't heard about them extending the platforms - what is the time frame for this? There didn't seem to be any working going on last time I rode it in April.

Peak frequency is still only once every 3 minutes downtown. They should have been able to almost double capacity by buying more trainsets, before they needed to extend the platforms. It certainly was looking pretty busy at peak - though what surprised me is how full it was mid-day (but yes, peak PM out of downtown was worse).

Still, I do agree, they appear to have undersized it. The 40-metre platforms seem very small - not much longer than a TTC streetcar. And ultimate is only 60 metres? (do I have that right?) Only about 40% of a Montreal or Toronto platform, and even only 2/3 the length of a Line 5 or Line 6 platform. I wonder what the extension of the Millennium Line to Cambie will do to ridership.

Anecdotally, ridership did seem significantly higher when I rode it this year, compared to a few years ago the last time I rode it.

 

[Streety McCarface]

Peak frequency is still only once every 3 minutes downtown. They should have been able to almost double capacity by buying more trainsets, before they needed to extend the platforms. It certainly was looking pretty busy at peak - though what surprised me is how full it was mid-day (but yes, peak PM out of downtown was worse).

Still, I do agree, they appear to have undersized it. The 40-metre platforms seem very small - not much longer than a TTC streetcar. And ultimate is only 60 metres? (do I have that right?) Only about 40% of a Montreal or Toronto platform, and even only 2/3 the length of a Line 5 or Line 6 platform. I wonder what the extension of the Millennium Line to Cambie will do to ridership.

Anecdotally, ridership did seem significantly higher when I rode it this year, compared to a few years ago the last time I rode it.

Technically, you are correct in that additional trains can be added to improve service for the next decade.

Source: https://globalnews.ca/news/3793260/...-line-faces-tough-choice-as-demand-increases/

My understanding is that they have either already purchased or are in the process of ordering additional cars for the Canada line, however, as the article states, this will only be a viable option for an additional 10-15 years (written 2 years ago, so 8-13 years), meaning that platform extensions need to occur within the this timeframe. For the canada line, it's not a cheap and easy procedure, there are miles of tunnels and large grade-separated stations that don't necessarily have the available room for platform lengthening. This type of work requires at least 8 years of planning and construction to complete, meaning they have to get started soon assuming ridership doesn't grow much more in the next few years.

 

[nfitz]

Technically, you are correct in that additional trains can be added to improve service for the next decade.

Source: https://globalnews.ca/news/3793260/...-line-faces-tough-choice-as-demand-increases/

That's interesting! That should increase the fleet by 50%.

The article says that the new cars would be in place in 2018. But I thought they only had the 40 original cars still.

It's interesting, and a little sad, to see TransLink now saying, what we were all saying back then, that the capacity is too low, and that TransLink was short-sighted.

I'm curious as to why they are saying "adding more cars is just about all they can do", discounting extending the platforms. Were we misinformed that they'd designed them to be easily extended to 60 metres?

 

[p_xavier]

The Canada Line hasn't even been around for a decade and they already have to extend the platforms. That's not a result of good transportation engineering, that's a result of being very conservative with ridership estimations, and not accounting for future ridership growth. Also, any good transportation engineer will know that the maximum PPHPD is not, and should never be an indicator of what to build in a rapid transit line. When you reach that peak capacity, theoretically, no one can enter the train if no one gets off, meaning that a person could be stuck at a station indefinitely. Take the Bloor Line as an example, it's fairly common to have to wait for 1-4 trains to go by before you can get on a train during peak periods at the inner stations, but the maximum capacity of 28K passengers per hour per direction maximum has not yet been met (It's closer to 25K PPHPD during the peak periods). The line isn't "technically" overcrowded but experience has you waiting 10-20 minutes to get on a train. In any other country, the Bloor line relief would be an issue that would have a similar awareness level to planners and politicians as Yonge line crowding is today, not because it's necessarily as bad as Yonge line crowding, but it has the potential to be within a few years.

The PPHPD metric does not account for small fluctuations in ridership, nor does it give a clear enough picture for the average commuter. Lines should always be built with enough capacity (or at least the provision to expand within a few weeks if necessary) to sustain future growth for a few decades (regardless if there's a PPP that's running the line and they have an interest to cheap out wherever they can), while not wasting resources on capacity the line will not require before the infrastructure's lifecycle ends. That is a much bigger challenge than building a line that will be at-capacity, or very close to capacity from day one.

The line doesn't need to be a full-fledged subway, but building it to run only 2 car trains was a joke, and everyone knew that. Canada line ridership has grown from 38 million passengers per year during its first full year of service (which was during the Olympics so ridership was definitely inflated by at least 2 million riders) to close to 49 million riders last year. The line has been seeing ridership increases of around 5.5-6% for the past 3 years.

No they don't..

They recently acquired new trains from Hyundai-Rotem. Exactly why the engineers proposed a system for 15000 PPHPD, base on the studies by the city (and Translink). It's not the only factor but it's a big factor.

It's the same in Montréal for the Orange line, people want more space than what the technocrats refer as the standard. It's still a question of PPHPD, just that the accepted density in Canada is lower than Asia.

Still hearsay.

BTW the Azur trains have automatic counters for the amount of people in the trains, the REM will go further than that and will show on the platforms which car people should go in as more space is available.

 

[nfitz]

Exactly why the engineers proposed a system for 15000 PPHPD,

They aren't currently anywhere near that much. They only run 24 trains an hour - which would be about 10,000 PPHPD. Ultimately, they should be able to run about 40 trains an hour - about 16,000 with the 40-metre platforms.

Which means, if the original reports about being extended to 60 metres were true, they could get to 24,000 PPHPD. Quite frankly, if you exceed that, you should be looking at parallel lines - rather than having a single monster line. It's only in places like Manhattan where you already have multiple lines with very high ridership, that you need be pushing any higher.

Though if they really can't extend the platforms ... yeah, they may have limited themselves more than necessary.

It's still a question of PPHPD, just that the accepted density in Canada is lower than Asia.

I don't see how it's accepted density - you just can't physically squeeze more people on a Toronto subway train some days ... or Montreal, I've had to abandon someone I was travelling with recently in Montreal, because it was so crush loaded, that I could barely and very rudely squeeze in ... (I was trying to catch a VIA train at Central ... they weren't!)

Once you exceed design loading capacity (which is less than peak) then you reduce capacity of a very frequent system, because you significantly extend dwell times. You see this daily on some streetcar routes, and frequently on Line 1. The laws of physics are no different in Asia.

Personally, I've never seen transit as crush-loaded in Asia as I've seen it in Montreal or Toronto (though you don't see it in Montreal and Toronto every day) - not even in Seoul (I haven't ridden in Tokyo though).

I can only surmise that these odd pictures you see sometime, with the doors bursting (and just how aren't the door sensors tripping over this I don't know), are for lines that aren't very frequent. Because you can increase capacity by crush loading on less frequent (once every 6 to 10 minutes) services, because the increased dwell times, don't delay the following trains, and reduce the frequency - only the travel times).

 

[Streety McCarface]

No they don't..

They recently acquired new trains from Hyundai-Rotem. Exactly why the engineers proposed a system for 15000 PPHPD, base on the studies by the city (and Translink). It's not the only factor but it's a big factor.

It's the same in Montréal for the Orange line, people want more space than what the technocrats refer as the standard. It's still a question of PPHPD, just that the accepted density in Canada is lower than Asia.

Still hearsay.

Read the article, even with the new trainsets, the line is destined to be overcapacity in 10 years. Building a line (especially a rapid transit line!) that only lasts 20 years before needing a major reconstruction project completed is one of the worst transit planning errors in modern history. The canada line has special status because it needed to be completed before the 2010 olympics, but the fact that provisions for platform extensions weren't built into the contract is a huge problem.

The PPHPD factor has to be considered with the ridership growth rate over the lifecycle of the project. From the looks of it, they never really considered growth in the equation, and to ignore it was a huge mistake.

Also, the last statement is simply not true. I used to live in Tokyo and the only reason trains are so crowded is because there's no room to build much else. People in Tokyo have to live with 200% crush loads because there's simply no room for additional lines, above ground and underground. The trains themselves are built for the same number of people per unit area as similar rolling stock in North America. Take the Tokyo Metro 1500 series rolling stock, the train can carry 1,520 passengers at crush load (100% capacity), with car dimensions of 20.5 meters by 2.8 meters. There are typically 10 cars per trainset, meaning that there are approximately 2.65 passengers per square meter of space. Compare this to a Toronto Rocket, which can carry a crush-load of 1,458 passengers on a 6 car train with car length 23.2m and width 3.125m. (3.35 passengers per square meter). In other words, the Toronto Rocket technically carries more passengers per square meter than the Tokyo metro vehicle, and this doesn't even include the fact that the TR has a worse seating layout overall. Standards are not different, available space is.