It makes sense to get streetcar platforms at stations to 60M wherever practical.
But to get bunching under control, they need to take excess bloat out of trip times across the network.
Main Stn never used to have more than 1 bus sitting on its opposite side due to arriving too early.
Now, 6-8 buses is standard with buses bunching up all over the station, impeding each other's ability to get in and out.
Removing running time does not resolve bunching, it resolves congestion in terminals. Those are opposites to each other.
More running time results in more congestion in terminals but less bunching;
Less running time results in more bunching but less congestion in terminals.
That said, I do agree that there is too much time in the schedules. Part of the reason that so much time appears to be needed (according to the GPS traces of the streetcars) is that many of the Streetcar policies require operators to operate in an exceedingly leisurely pace - even when behind schedule. For example, they're expected wait until the doors close on their own, leading to pointless delays at busier stops while people keep pressing the button to come in, and pointless delays at minor stops where the doors haven't yet figured out that everyone is already onboard. This is in contrast to the subway where guards anticipate the moment when everyone will have boarded, and start the door chimes a few seconds beforehand, such that there is no time wasted before departure. The Flexities have cameras and door chimes making this possible, but operators aren't allowed to use them.
Then there's the issue that the Streetcar Operations group has a tendency to just introduce a new speed restriction as a knee-jerk response to every incident.
Newsflash: Other railways don't achieve safety by expecting operators to be able to stop the train in time to avoid collisions. It's not like VIA train drivers can see level crossings 2km down the line while they're travelling 160 km/h. Instead the focus is on alerting people to approaching trains, and encouraging them to use common sense by obeying signals and/or looking both ways before crossing the tracks.
People often think of streetcars as buses on rails, but they can't stop nearly as quickly. They should really be treated more as little street-running trains, just as they are in virtually every other city. One little change I think could help is to swap the current horns which sound like car horns, to a sound which resembles a train horn. Many light rail systems in North America use train-like horns on their light rail vehicles, so there are plenty of off-the-shelf products available.
It's easy to make fun of the TTC for having horrendously unreliable streetcar service, but when you think about it, it's actually quite difficult to run regular headways on most of Toronto's streetcar lines. In any other type of transit service, there are certain stops where the vehicle sits when it is ahead of schedule, but on many of the mixed-traffic streetcars routes, there isn't anywhere practical to sit. There's also little opportunity to "hurry" to make up time, since so much of the streetcar's time is spent at stops rather than moving. The amount of time saved by driving faster won't come anywhere near enough to offset the extra time spent at stops due to the additional people who are waiting there due to the streetcar being late.
I think that introducing headway-based signal priority (
as planned) will help matters somewhat by creating some kind of force nudging vehicles forward or back toward the desired headway. Of course that alone won't solve the bunching issues - line managment and operator discipline are also key - at least it increases the number of practical headway management tools from 0 to 1.
Beyond that, they need to move strictly to headway management, not schedule;
The TTC is reluctant to use strictly headway management because it results in operators being in unpredictable locations at the time of their breaks and at the end of their shifts, thereby requiring a lot of annoying crew swapping between streetcars opposite directions to get them to the location where the next crew is waiting. Note that it's a lot more awkward to swap crews between streetcars in the middle of the street than it is to swap subway crews in stations where they can just walk across the platform.
Each time the crew is swapped represents a couple minutes delay, which is a big deal on lines scheduled as frequently as every 3 minutes. Furthermore, one or the other of the two streetcars will probably be sitting right before a signalized intersection, maxing out the signal priority there for a few minutes. The latter doesn't really affect the TTC much, but it increases the resentment that the City has against the signal priority system by creating a large and pointless delay to other types of traffic.
I think a decent compromise would be to dispatch from the terminals based on schedule (when service is running relatively close to schedule), but then always manage spacing along the line based on headways.
and to step-back crewing so that the operator takes a break, not the vehicle.
Streetcar schedules are already written entirely with step-back crewing for breaks. The problem is that some operators act like the recovery time scheduled in terminals is an additional break they're entitled to, which it is not. When the schedule includes 5 minutes sitting at a terminal, an operator who arrives more than 4.5 minutes late is expected to depart immediately.
