T3G
Senior Member
There is no reason to assume this will be a regular occurrence, though.
|
|
|
nfitz
Superstar
The SRT replacement was being planned long before it was even mooted that the Eglinton Line would go east of Allen - let alone Laird. Besides, extending Line 2 north of Kennedy will not remove riders from Line 5.
There's only the vaguest hint in the Sheppard Subway studies of going east of McCowan (Shepperd East station). There is no pricing, let alone plan. And again, I don't see how Sheppard east of McCowan relieves Eglinton between Laird and Kennedy.
You are new here, I'd suggest reading up on the full thread before making up stories.
TTC has been bunching non-grade separated rail services for over a century. And they were doing so on Finch since day 1. There's every reason to assume that it will continue, and nothing to suggest otherwise.
T3G
Senior Member
I didn't observe much by way of bunching on day 1 - if anything, the opposite problem existed, with the headways being stretched out far more than they should have been.
And the mechanics of why the bunching occurs is extremely important - if the bunching occurs because of incompetent dispatchers, it's a lot different than if it was because of the route design being problematic. One can be solved with the will to improve services (and has been known to occur from time to time on the subway, too), the other is a handicap inherent to the line design. Look back through the whole conversation again, this whole discussion was triggered by someone who unequivocally stated that the eastern portion of the 5 will be rebuilt because of the bunching that it will create, two definitive statements based on assumptions that are in no way reasonable to make.
Again, if someone can explain to me what is so special about the transition point on line 5 that it will cause bunching, but it doesn't happen at Bloor-Yonge or Eglinton West, I'm all ears. Hell, if this was remotely a reasonable concern, how come the 900 bus doesn't see bunching southbound at the 427 at Dundas, when they all slow way down? If Urban Toronto science applied here, you'd have the entire stock of the line running together in one big pack.
Last edited:
Rainforest
Senior Member
I think both factors can contribute, but the quality of traffic control is more important. Some route designs may be more prone to bunching than others, and thus require more efforts from the dispatchers to avoid bunching. However, a good dispatch + cooperation from the traffic signal control should be able to keep bunching to the minimum.
Unless we are looking at a really desperate situation, such as a mixed-traffic route on a street that is constantly clogged; then the dispatchers can't do much. But this is not the case for the new LRT lines that have 100% exclusive lanes.
Westey
New Member
I was replying to the post immediately above - that I quoted.
For Sheppard to be a Crosstown - it would have to go beyond McCowan.
I assumed the assumption of that post was that the terrible design for Eglinton, between Laird and Kennedy, has already been locked in so we need to find another option to go "crosstown".
nfitz
Superstar
You don't see the entire stock on any streetcar line running in one big pack (except for very extreme circumstances - and then they are normally parked in one big pack).
But you do see bunching on routes like 512. The mechanism for Line 6, and Line 5, would be the same. And there's been reports of bunching on Line 6; perhaps TTC has solved that by running so few vehicles that they can't ever bunch; tough to say much without analytics.
What is special is traffic lights. I'm not aware of any vehicular or pedestrian crossings at Bloor-Yonge.
I'm not sure why you think scheduled speed might be a factor.
But more to the point - you don't see bunching on the 900? I've certainly seen bunching.
Just looking at the last hour of 900 data you can see southbound arrivals at East Mall and Dundas for 900 are anything but regular! There's supposed to be one every 8 minutes. But instead you had 2 come after only 2 minutes, and 2 more after only 4 minutes! And 20 and 25-minute gaps.
Surely the 900 data, even in the last hour, doesn't support much.
I'm simply responding to your claim. You said "Now we are spending $10B to extend the B-D line and another $8B to extend Sheppard Subway (beyond McCowan) - all to make up for the mistakes of not making Eglinton the Crosstown".
For example, can you provide information on the extension and cost of the Sheppard subway east of McCowan? I think you are incorrect on this; let alone the assertation that either of these projects have any relation to the technology selection on Line 5.
ARG1
Senior Member
nfitz
Superstar
Wrong photo?
Do the next vehicle displays show more than one train coming, like the buses or streetcars, or just the next train, like Lines 1 to 4? That would make it clear.
T3G
Senior Member
Yes.
1) Where do you see bunching on this photo?
2) Where did I say that bunching could never occur? I specified multiple times in this thread that it can happen as a result of delays or just plain bad management, but the idea that a slow down halfway along the line, which is what we are discussing, would cause bunching is impossible, because that's not how physics works (again, see my example about running from point A to point B, and then walking to point C, or my example about Line 1 at Bloor-Yonge and Eglinton West).
Some weird doublespeak going on with this whole bunching thing. On the one hand, LRT is argued to be bad because it's slow. On the other hand, people are claiming there is going to be an epidemic of Schrodinger's LRVs, whereby the hypothetical lead car is supposed to be slow, but somehow all the cars behind it are magically going to be so fast they're going to be able to close the 4-5+ minute gap in front of them. And those fast LRVs will never stop for lights or stations, they'll just keep chasing the mysterious slow LRV the whole time. What part of this makes any sense?
Because that's literally the entire basis on which this discussion is founded.
Again, people are arguing that there is going to be bunching at the transition between subway and tramway, because of the scheduled slowdown in speeds between the two sections. So, yes, scheduled speed is a rather significant factor in the whole discussion!
Traffic lights that will cause one vehicle to be stuck for 4-5+ minutes, but absolutely none of the vehicles behind it?
For the millionth time, I never said bunching can't occur, period. I'm arguing against a hyper specific criticism that people are levelling against the line, arguing that there is going to be bunching at the transition between subway and tramway, because of the scheduled slowdown in speeds. Thus, using examples like the 900, which also see a scheduled slowdown in speeds, is relevant. The fact that this argument doesn't make any sense doesn't mean that bunching can't occur for other reasons, such as breakdowns, driver breaks, drivers misbehaving, passengers misbehaving, traffic along the route, etc. These are all random acts of chance and have nothing to do with how the service - any service - is designed.
Westey
New Member
McCowan to Morningside. 7.5km @ $1B/km = $7.5B. Rounded to $8B.
B-D extension. current cost estimate $10B.
Once again, because Eglinton did not serve the Crosstown role (left off too much of Scarborough from good transit), they needed to connect STC. Also, following the statement of micheal_can, if Sheppard were to take over this Crosstown role, it would have to go much farther East.
ARG1
Senior Member
There's a train arriving at Stevenson, and you can see the lights for the next train at Mount Olive waiting for the Green Light. They are less than 1 minute apart.
urbanclient
Active Member
I made an edit earlier "[...]Dwell times are often longer and less consistent than metros due to bottlenecked boarding and alighting inherent in tram design."
We're not saying bunching occurs because someone fell asleep at the wheel and decided to idle at one stop for 4 minutes. We're saying bunching occurs due to cascading, compounding delays. For example, a driver that is just 1 minute slower at the halfway point, a green light that is missed by a few seconds leading to a 60 second red light, a handicapped passenger taking longer to board, a rowdy passenger refusing to pay their fare, a call from dispatch to hold 30 seconds longer at one stop, higher passenger traffic at one stop etc... etc...
In theory, the 510 Spadina shouldn't bunch with 7'45" headways on Saturday afternoons, but literally every time I'm on it at that time I see bunching. 506 Carlton had 9 minute headways today, and yet I could see two 506s within metres of each other going westbound near Yonge:
Longer headways only help prevent bunching in the same manner that excess schedule padding prevents bunching (see TTC schedules 24/7 365). It doesn't solve the underlying deficiencies inherent in design. Here's proof of Line 6 Finch West bunching, the train behind going eastbound is only 6 masts west of another train arriving at the Driftwood eastbound stop, or about 150 metres away:
@T3G I think you've misinterpreted at least some of us. When you hear hooves think horses not zebras. I'm not sure many are claiming the mathematically impossible here. I'm pretty sure they are claiming the inherent deficiencies of a surface ROW will lead to bunching, not the difference in average speeds between tunnel and surface. Maybe some others worded their posts ambiguously, sure. But the way you have replied to posts (Paris T9 etc.) it does seem like you were claiming surface tram lines should not have bunching in general:
This post as well:
In all seriousness: for the headways to be stretched out on day 1, logically bunching would occur the moment they resumed regularly scheduled departures at the terminus, which is likely exactly what happened. There was a lot of bunching observed on day 1.
We're not saying bunching occurs because someone fell asleep at the wheel and decided to idle at one stop for 4 minutes. We're saying bunching occurs due to cascading, compounding delays. For example, a driver that is just 1 minute slower at the halfway point, a green light that is missed by a few seconds leading to a 60 second red light, a handicapped passenger taking longer to board, a rowdy passenger refusing to pay their fare, a call from dispatch to hold 30 seconds longer at one stop, higher passenger traffic at one stop etc... etc...
In theory, the 510 Spadina shouldn't bunch with 7'45" headways on Saturday afternoons, but literally every time I'm on it at that time I see bunching. 506 Carlton had 9 minute headways today, and yet I could see two 506s within metres of each other going westbound near Yonge:
Longer headways only help prevent bunching in the same manner that excess schedule padding prevents bunching (see TTC schedules 24/7 365). It doesn't solve the underlying deficiencies inherent in design. Here's proof of Line 6 Finch West bunching, the train behind going eastbound is only 6 masts west of another train arriving at the Driftwood eastbound stop, or about 150 metres away:
When responding to this, I and many others took this to mean "bunching will not occur on the surface section" period. It wasn't clear you were aware of any ways bunching could occur on a regular basis.
Of course there being a transition from underground to surface in itself is not going to cause bunching. There being a manually operated surface section though, will cause bunching. A subtle difference, but as @nfitz said, it doesn't make much difference at the end of the day. How many people actually believe it's the transition from underground to surface that will cause bunching and not merely the surface section itself? There will be bunching.
@T3G I think you've misinterpreted at least some of us. When you hear hooves think horses not zebras. I'm not sure many are claiming the mathematically impossible here. I'm pretty sure they are claiming the inherent deficiencies of a surface ROW will lead to bunching, not the difference in average speeds between tunnel and surface. Maybe some others worded their posts ambiguously, sure. But the way you have replied to posts (Paris T9 etc.) it does seem like you were claiming surface tram lines should not have bunching in general:
From this post, it appears you're not aware of many contributing factors to bunching, and how frequently bunching happens on transit systems in general. Because the signal interaction duration variability often doesn't 'average out', that's why bunching can occur on Line 6, even with 12 minute headways.
This post as well:
Dwell time variability is just one of many contributors to bunching. It's not one excess dwell at one station that causes bunching, it's a confluence of small, cumulative deviations in dwell time, acceleration, signal interaction etc., as @lastcommodore explains:
There are countless reasons to know bunching will almost certainly be a regular occurrence. Seriously, have you ever been on a streetcar in Toronto besides the occasional visit downtown? Been on Line 6 besides day 1?
Ok, pack it up everyone. @T3G didn't see bunching on day 1. Bunching has been solved. Case closed
In all seriousness: for the headways to be stretched out on day 1, logically bunching would occur the moment they resumed regularly scheduled departures at the terminus, which is likely exactly what happened. There was a lot of bunching observed on day 1.
@T3G , there is lots of bunching on a daily basis on surface bus routes in Toronto, including the 900. Virtually every time I touch a bus there is bunching. It's the degree of severity that is up for debate: are buses 5 minutes apart considered bunching for a route with 15 minute headways? Are 3 minute headways instead of the scheduled 10 minute headway considered bunching? When it's somewhere between two buses literally back-to-back and two buses exactly as scheduled it becomes very subjective.
Last edited:
Bojaxs
Senior Member
I don't get Metrolinx, man. Can anyone explain their logic, rationale here? On the surface portion of Line 5 they didn't construct "Two stage crossing" for every intersection. Some intersections have it, and some don't. This makes no sense. So some intersections might get aggressive TSP and some won't? I just don't get why you wouldn't construct "two stage crossing" at every intersection along the line and then have the ability to implement aggressive TSP on the entire line. Particularly when you need the surface portion of the line to keep pace with the underground portion of the line.
I can't be bothered to post images of every intersection, but here are some examples.
"Two stage crossing"
Eglinton & Leslie .
Eglinton & Sloane
No 'two stage crossing"
Eglinton & Victoria Park
Eglinton & Swift
Eglinton & Pharmacy
I can't be bothered to post images of every intersection, but here are some examples.
"Two stage crossing"
Eglinton & Leslie .
Eglinton & Sloane
No 'two stage crossing"
Eglinton & Victoria Park
Eglinton & Swift
Eglinton & Pharmacy
Last edited:
lastcommodore
New Member
My boring, middle-ground thoughts:
- Bunching is a natural occurrence when a vehicle is, on average, slower than the vehicles behind it.
- There's a lot of reasons for this to happen. There is no silver bullet that magically ends bunching forever.
- There is a component of infrastructure & situation that makes bunching more or less likely to happen.
Ex: Mixed vs. Segregated traffic, intersections vs. complete separation, TSP, PSD, heavy ridership, etc.
- There is a component of operational skill that can mitigate bunching incidents, or make them more likely.
- The TTC is very bad at managing bunching, even in ideal scenarios.
- Bunching can and does occur in ideal scenarios, especially with poor operational skill, like the TTC has.
On the topic of bunching as it relates to at-grade intersections:
Theoretically over a long enough time, each LRV will hit the same number of reds and greens. However bunching occurs at a much smaller timeframe, thus it is totally possible for a single LRV to hit several unlucky red lights which now causes bunching. Effectively, bunching is caused by chance, and thus to reduce bunching you must reduce the variations of chance.
A "stronger" TSP which makes LRVs, on average 50% faster, but only 30% of the time is such a case where bunching would increase.
Theoretically, an absolute TSP system that activates always would reduce the effect of cycle chance to zero, effectively making the intersection have no effect on chance of bunching (Save for incidents like pedestrian/car/bike block the tram)
tl;dr anything short of absolute TSP introduces cycle luck, and cycle luck means potential unluck, and unluck means bunching incident. there's a lot of sources of unluck, and good operations can resolve unlucky incidents, while bad operations can make unlucky situations worse.
- Bunching is a natural occurrence when a vehicle is, on average, slower than the vehicles behind it.
- There's a lot of reasons for this to happen. There is no silver bullet that magically ends bunching forever.
- There is a component of infrastructure & situation that makes bunching more or less likely to happen.
Ex: Mixed vs. Segregated traffic, intersections vs. complete separation, TSP, PSD, heavy ridership, etc.
- There is a component of operational skill that can mitigate bunching incidents, or make them more likely.
- The TTC is very bad at managing bunching, even in ideal scenarios.
- Bunching can and does occur in ideal scenarios, especially with poor operational skill, like the TTC has.
On the topic of bunching as it relates to at-grade intersections:
Theoretically over a long enough time, each LRV will hit the same number of reds and greens. However bunching occurs at a much smaller timeframe, thus it is totally possible for a single LRV to hit several unlucky red lights which now causes bunching. Effectively, bunching is caused by chance, and thus to reduce bunching you must reduce the variations of chance.
A "stronger" TSP which makes LRVs, on average 50% faster, but only 30% of the time is such a case where bunching would increase.
Theoretically, an absolute TSP system that activates always would reduce the effect of cycle chance to zero, effectively making the intersection have no effect on chance of bunching (Save for incidents like pedestrian/car/bike block the tram)
tl;dr anything short of absolute TSP introduces cycle luck, and cycle luck means potential unluck, and unluck means bunching incident. there's a lot of sources of unluck, and good operations can resolve unlucky incidents, while bad operations can make unlucky situations worse.
Voltz
Senior Member
Two stage crossings aren't needed, pedestrians wouldn't follow them anyway, the signals just need to detect approaching LRV's from far enough away to have the signals give them a green light.I don't get Metrolinx, man. Can anyone explain their logic, rationale here? On the surface portion of Line 5 they didn't construct "Two stage crossing" for every intersection. Some intersections have it, and some don't. This makes no sense. So some intersections might get aggressive TSP and some won't? I just don't get why you wouldn't construct "two stage crossing" at every intersection along the line and then have the ability to implement aggressive TSP on the entire line. Particularly when you need the surface portion of the line to keep pace with the underground portion of the line.
I can't be bothered to post images of every intersection, but here are some examples.
"Two stage crossing"
Eglinton & Leslie .
Eglinton & Sloane
No 'two stage crossing"
Eglinton & Victoria Park
Eglinton & Swift
Eglinton & Pharmacy