Buses Vs Streetcars on 501 Lake Shore Service

The year 2017 brought a long-term shutdown of the 501 Queen street service west of Sunnyside Loop to permit various construction projects on The Queensway, at Humber Loop and on Lake Shore Boulevard. Although work will not be seriously underway until the weather improves, the year-long shutdown also gives the TTC some breathing room in its ongoing problem of streetcar fleet availability as it awaits the long-overdue Bombardier Flexity order.

City TV recently reported that riders on Lake Shore are very pleased with the replacement bus service. That’s little surprise considering how much more frequent the buses arrive compared to the streetcars. A well-known aspect of service evaluation is that riders are much more sensitive to waiting time, especially in bad weather, than to in-vehicle travel time, especially if riding conditions are moderately comfortable.

There is also the question of comparative travel speeds with bus operations versus streetcars. Do the buses make better time than streetcars and, if so, when, where and why?

Service Design

The 501 Queen car has been through a few changes in the past year or so, and riders on the west end of the route are understandably unhappy with constant changes.

Pre 2016 Service

The 501 Queen car operated as a continuous route from Neville Loop at its east end to Long Branch Loop at the west, with half of the service scheduled to short turn back from Humber Loop. This design was in place for many years after the TTC consolidated the 501 Queen and 507 Long Branch cars as a money-saving tactic in the mid 1990s, but this arrangment never worked as well, especially west of Humber Loop as the 501/507 split. The most common problems with the amalgamated services were:

  • One-way trips across the route were extremely long, and operators required substantial recovery breaks at the terminals, whether provision for these existed in the schedule or not.
  • Short turns were common at both ends of the route causing service beyond these points to operate at well above scheduled headways.
  • Integration of services from scheduled and unscheduled short turns was spotty, and pairs of cars commonly travelled across the city carrying a double headway.
  • Until 2016, the 501 Queen service west of Humber was not part of the TTC’s 10-minute network, and short turns of service bound for Long Branch could produce extremely wide gaps west of Humber Loop.

2016 Service

Effective January 2016, the services were split at Humber Loop with a separate streetcar operation from Humber to Long Branch operating very much like the former 507 Long Branch car. Through trips from Neville to Long Branch only operated very late in the evening during the bridge period to the all-night 301 Queen service over the full route.

Scheduled service west of Humber was improved to every 10 minutes or better. With cars captive to this part of the line, riders had a reasonable expectation that service would appear on a more timely basis than before. Short turns of westbound 501 Queen cars could still create problems, however, due to less reliable service making the connection at Humber Loop.

For much of 2016, the Queen service diverted both ways via King between Shaw and Spadina to allow water main construction on Queen Street. This added substantially to travel times and frayed the service quality, a problem compounded when schedules reverted to their “standard” configuration well before the construction and diversion were actually completed. (I will explore the service on the main part of the Queen route in a separate article.)

2017 Service

Effective January 2017, all Queen streetcars turns back at Sunnyside Loop (west of Roncesvalles). A bus service (501L) operates from Long Branch to Dufferin Loop with an overlap between streetcar and bus service to simplify transfers. An additional loop shuttle (501M) serves the condo district on Marine Parade.

501_2017

Bus service on the 501L to Long Branch is more frequent than the 501 streetcars it replaced:

               501 Streetcar    501L Bus
AM Peak            9'30"           5'15"
Midday            10'00"           6'00"
PM Peak            9'30"           5'30"
Early Evening      9'30"           6'00"
Late Evening       9'00"           7'30"
Wknd Early AM     10'00"           7'00"
Wknd Late AM      10'00"           6'00"
Wknd Afternoon    10'00"           5'30"
Wknd Early Eve    10'00"           6'30"
Wknd Late Eve     10'00"           7'30"

Travel Time Comparisons

In the following sections, comparative travel times are given for four months:

  • November 2015: Before streetcar route split at Humber Loop
  • November & December 2016: Split streetcar operation east and west of Humber. Late December provides an example of travel times during a period of lower demand and traffic congestion due to the Christmas break.
  • January 2017: Split operation with buses west of Sunnyside providing through service to Long Branch

For cases where the section of the route being compared spans Humber Loop, only data from November 2015 and January 2017 are included because most service during late 2016 did not run through from The Queensway to Lake Shore Boulevard.

Each set of charts contains six pages containing weekday, Saturday and Sunday data for eastbound and westbound travel times showing the average for each hour of the day. Times between points are in solid lines, while the standard deviations in values are dotted. The same colour scheme is used so that, for example, November 2015 data are in red on all charts.

Long Branch to Lake Shore at Humber Loop

Eastbound travel in the AM peak shows a much higher “hump” for November 2015 than for later data because of construction effects at that time. The streetcar running times are all quite consistent with the bus times being about 10% lower throughout the day. Bus times are also lower than streetcar times on weekends, although not by the same degree.

Westbound travel times are similar for all months until the early afternoon with buses showing a slight reduction relative to streetcar data from then on. Weekend times are similar for all dates. The hump in mid-afternoon Saturday times westbound in November 2016 was caused by an extended delay on November 4.

The charts below subdivide the route to isolate areas that could have different characteristics. Note that the scale on these charts is different from the charts covering the full distance from Long Branch to Humber.

501_Long Branch Lake Shore/Humber Streetcar vs Bus

Long Branch to Kipling

Bus running times are slightly better than streetcar times eastbound on weekdays, but otherwise lie in a similar range. The November 4, 2016 westbound delay shows up here as a spike in average travel times as do a few other smaller delays.

501_Long Branch Kipling Streetcar vs Bus

Kipling to Royal York

As with the Long Branch to Kipling segment, weekday travel times eastbound for buses are slightly better between Kipling and Royal York.

Westbound running times show a hump for buses (January 2017) during the hours 9 to 10 am, and noon to 1 pm. Examination of the detailed data reveals that this corresponds to layovers at Kipling that are probably for crew changes. The result is that some trips have unusually long travel times, several of more than 15 minutes.

501_Kipling Royal York Streetcar vs Bus

Royal York to Legion Road

Travel times during January 2017 (buses) for eastbound trips on weekdays are slightly faster by bus than by streetcar.

501_Royal York Legion Road Streetcar vs Bus

Legion Road to Lake Shore at Humber Loop

There is less difference between buses and streetcars in this segment than the others above. The construction-induced spike in AM peak travel times in November 2015 eastbound to Park Lawn shows up stongly here.

The peak for buses on Saturdays at 10 am was caused by one vehicle on January 28 that took a long layover west of Humber Loop. (This is an example of how the relatively low number of trips within an hour, and the fewer weekend days compared to weekdays, causes an outlier in the data to have a noticeable effect on the average.)

501_Legion Road Lake Shore/Humber Streetcar vs Bus

Lake Shore at Humber Loop to Roncesvalles

These charts compare streetcars in November 2015 with buses in January 2017. The bus trips are often faster, but this is due in part to their taking a different route that excludes serving Humber Loop. Moreover, the buses are not as constrained on The Queensway by slow order and delays at “transit priority signals” compared to the streetcars.

501_Lake Shore/Humber Roncesvalles Streetcar vs Bus

Long Branch to Roncesvalles

These charts compare streetcars in November 2015 with buses in January 2017 showing the collective effect of shorter travel times over the route west of Roncesvalles.

501_Long Branch Roncesvalles Streetcar vs Bus

Speed Comparisons: Long Branch to Lake Shore at Humber Loop

Another way of comparing the two modes is to look at fine grained vehicle speeds along the route. The charts below include:

  • Streetcar data from November 1-4, 7-11 (the 9th was omitted because of a major delay during the early evening)
  • Bus data from January 9-13, 16-20

Each set of charts contains 19 pages showing the average speed of vehicles along the route. Each page contains data for one hour of service with streetcar data in blue and bus data in orange. Where the average speeds are similar, these lines will lie more or less on top of each other. Where there is a difference, the lines are separated. Notches in the charts correspond to locations where most vehicles stop. The data are presented east-to-west so that Humber is at the left side of the page and Long Branch at the left.

An interesting contrast between the eastbound and westbound charts is the behaviour at stops. There is a consistent profile to speeds at stops with a gradual drop on the approach, but a quick acceleration on leaving. This shows up most obviously in the way the shape of the dips reverses on the east and westbound charts.

However, another difference is that for eastbound trips, the dip for streetcars tends to be deeper than for buses. My hypothesis for this is that streetcars have to wait longer at stops on Lake Shore because it is a wide street and passengers must walk out to the stopped car. This takes longer than on a narrow street, and certainly longer than for a bus stopping at the curb. Other factors will include traffic signal behaviour (if a red signal traps the transit vehicle for one cycle, the mode has no effect on travel time). Locations with safety islands should not exhibit the same behaviour as those where riders have to cross lanes to reach streetcars. Finally, of course, inbound trips have more loading than unloading, and so the effect is not as visible westbound.

501_201701_WB_Streetcar vs Bus

501_201701_EB_Streetcar vs Bus

Speed Comparisons: Windermere to Dufferin

Between Dufferin (where the Queen buses turn off the route to loop south to the loop at the CNE) and Windermere, both modes share the same route but their speed profiles are quite different. Of particular interest is that the streetcars run slower than the buses even where the streetcars are on their own right-of-way. This is clearly a question of operating practice, not of traffic congestion, and is probably related to a long evolution of “safety” related speed restrictions on this section of the route. This is particularly visible west of Roncesvalles where streetcars tip-toe through the special work at the carhouse and at Sunnyside Loop.

Also visible is the effect of nearside versus farside stop placements where the streetcars can be held twice, one for a green signal, and again to serve their stops. Buses stopping nearside stop only once.

501_201701_WB Streetcar vs Bus (Dufferin to Windermere)

501_201701_EB Streetcar vs Bus (Windermere to Dufferin)

If Toronto is to have a “transit first” attitude to moving vehicles, one important change must be the understanding of how streetcars can be delayed by traffic design and by operating practices that bring on slower operation.

Once the full bus replacement of the 501 Queen route begins in May 2017, I will compare operating speeds over the entire line for the two modes.

20 thoughts on “Buses Vs Streetcars on 501 Lake Shore Service

  1. I liked the route split on 501 Queen so much. It improves service west of Humber Loop and reduces delays. Splitting the route should go for many long and busy routes. They should keep the route split at Humber Loop forever.

    Like

  2. Having 65 buses replacing 36 streetcars means it will be expensive, coming out of the TTC operating budget. Now if they replaced the 65 buses with 65 streetcars after September.

    Steve: The city is paying the TTC an extra subsidy this year to cover costs triggered by construction projects. This will not cover the whole thing, however, as some of upheavals are due to TTC work, some to the city, some to third parties.

    Like

  3. One additional reason streetcars can be slower, which I didn’t notice in your analysis, is that with their longer headways, there is a greater chance of having to stop at more of the lightly-used stops on Lake Shore.

    Steve: That’s a good point, although it will require fine-grained analysis of the data to verify (individual vehicles rather than averages).

    There is also operating practice and supervision. Many streetcars used to travel at nowhere near the speed that it was easily and safely possible to them to go. The difference was evident when one caught the rare ambitious operator who was running late. Particularly on The Queensway, streetcars could keep pace with cars, or even gain on them, but it rarely happened (the rails made for a pretty bouncy experience).

    Of course the “streetcars 10 km/h through intersection” means big drops in speed approaching Colborne Lodge, where during winter or after dark it’s very unlikely that there will be any need for stopping. Oh, and the Parkside Drive stop is out of service during the bustitution (there’s no safe way to arrange a bus stop at that location) which gives the buses a touch more speed advantage.

    Steve: When the rail is rebuilt in concrete, the “bounciness” should only be a memory. It will be interesting to see how operators take to this. As you say, however, the “safety” impositions make for a much slower trip. It will be interesting to see whether the Parkside stop even returns as it is not accessible, and expensive to adapt for that.

    It seems to me that the new LFLRVs have more aggressive initial acceleration, more like a PCC than an ALRV. The ALRVs can move right along, but they seem to take their time getting there. Buses, on the other hand, are almost always pulling away from stops at full acceleration, since they have to pull out into fairly fast traffic on Lake Shore.

    Like

  4. Is it expected that once the track work is complete along the Queensway that that should resolve some of the slow orders at TSP locations?

    Steve: I am willing to bet that nothing will change. The fact that the speed profiles clearly show cars are stopping both near and farside indicates that “TSP” is not ensuring that streetcars can cross to their stop before the signal changes. Given how light the traffic is on the north-south cross streets and the comparative infrequency of streetcars, there is no excuse for this arrangement.

    Like

  5. Thanks for this analysis Steve.

    Would you consider supplementing it with a comparison of operating and maintenance costs for this route? Or comparing the TTC’s actual operational costs during bus replacement if that data is made public?

    It’s an important part of the holistic view and those who favour buses for simple reasons don’t always understand the related costs. It would be great to have an analysis to point to especially if it’s attached to a particular route or route segment rather than an abstract comparison of two transportation technologies.

    Steve: That sort of comparison is challenging because the TTC does not publish breakdowns of the component costs for each mode’s operation. Counting operators is easy, but more complex is the cost per vehicle, kilometre and hour of bus operation and maintenance.

    Also, as an aside, some of the 501L bus stop implementations are horrendous. Eastbound at St Joseph’s drops passengers off between two hydro poles on the southwest part of the intersection, where there is no sidewalk, no ramp, and no crosswalk. The traffic is navigating a fairly sharp, righthand ascending turn just before that light so the visibility can be very poor. Someone is going to be seriously injured or killed.

    Steve: Sounds like I need to visit and document this foolishness. The TTC always talks about “safety first”.

    At Long Branch the westbound bus pulls in well east of the intersection, so the additional time to walk to the light, wait for it to change, and then continue on to a transfer point for MiWay or GO cancels out any gain in travel time. Eastbound, the drivers pull in from the intersection so briefly that if you are waiting under the canopy due to weather, as I have, they pull out before you even have time to walk to the stop.

    Cheers,
    Dave

    Like

  6. The split 501 route is terrible. It adds a significant amount of travel time even if your streetcar is there. To have all passengers disembark, walk over, then board takes a long time. Last time I did it it added 7 mins. to the journey and that was with the next streetcar there and waiting.

    The City of Toronto’s Transportation Planning Department’s 2012 Southeast Etobicoke Travel Survey of 3,307 residents found that 88% of those who used the 501 used it to travel downtown. It is ludicrous to characterize a service disruption that worsens travel time for the majority of riders as an improvement.

    The solution is to improve service for all 501 riders, not a small minority.

    Steve: The last riding survey I saw showed that peak period travel was heavily downtown oriented, but that over half of the off-peak travel was local. For those riders, having a separate route ensures that it will have reliable service. When the routes are reunited with bus service, it will be interesting to see how the route overall behaves compared with the split operation.

    Like

  7. I agree Steve. Splitting the 501 Queen makes a reliable service along Lake Shore. Same story should still apply when 501 uses only buses with one branch from Long Branch to Dufferin Gate and the other from Park Lawn to Neville Park.

    Like

  8. Yes, there may be a higher percentage of local trips off-peak, but what percentage of total ridership and fare revenue is it? And are those trips as time sensitive as AM commutes? There is nothing worse than giving yourself ‘extra’ time in the morning to get in for an early meeting and arriving late anyway.

    What is the feasibility of headway-based service guarantees west of Ronces? Instead of focussing on the schedule, use gps route management data to check next arrival westbound at Roncesvalles. If it will exceed the 10 min headway, the car goes through. If not, it can be short turned if necessary. Would that not help both distant and local riders? Your western headways would still be met but you would not be splitting every ride.

    And if that is too complicated, then why not simply have peak trips run through and off-peak trips split?

    Steve: Service guarantees anywhere on the TTC simply will not happen. There is more to managing a long route like Queen than a service guarantee on one part of the line. Once upon a time, the Long Branch car did run into downtown during peak period only, then that became some Lake Shore trippers, then a few Queen cars. As the TTC cuts rather than adds service, things like that have fallen off of the table.

    Like

  9. The split of 501 streetcars was good for EB trips. There seemed to always be a streetcar waiting at Humber loop, and the transfer was quick. WB trips, however, were much different. Often there were longer wait times at Humber loop for the next long branch streetcar.

    Any chance of seeing the 508 return?

    I wonder how much peak ridership might be diverted to the 145 express if double fares didn’t exist?

    Steve: The 508 was cancelled allegedly because of a shortage of streetcars. When/if the TTC will get around to restoring it, I don’t know.

    Like

  10. The slow order for intersections on the Queensway is to try and prevent collisions from vehicles turning left in front of the streetcars. The rebuilt ROW won’t make a difference for that.

    Things the might fix it are a red arrow for the left turn signal, but that would require changes to the highway traffic act, or a sign that warns when a streetcar is approaching or even just a mirror.

    Liked by 1 person

  11. @doconnor

    The US Manual of Uniform Traffic Control Devices, that is signals and signs, says that red signals for left turns should be a left facing red arrow. The green signal head over the left hand through lane next to a left turn lane should be a straight through green arrow. If right turns are permitted, but not protected for, on green signals then the green signal over the right lane should be a green ball. If right turns are protected then there should also be a green right turn arrow to indicate this. The red signal for both of these should be a red ball.

    It is amazing how normal people can come up with obvious fixes to these problems whereas the MoT refuse to look at what is accepted practice in the rest of the world.

    The Transportation Research Board in the US did studies of accidents between LRTs on their own rights of way and automobiles and found that these simple fixes reduced the chance of collisions dramatically. I need to look for the address or the report and post it.

    Liked by 1 person

  12. Steve,
    Have there been any updates to streetcar operator safety practices and training? It seems that these are becoming a hindrance to operations and need to be rectify. I’m surprised there hasn’t been any uproar over the band-aid solutions that the TTC continually applies to the streetcar network. European car operators would have a good laugh at the TTC’s antiquated operations and at the irrational thinking towards so called safety practices if they were given a chance to ride and observe.

    Overall though, I feel like the TTC has zero confidence in its streetcar network and operations. This is not good when the system carries such high daily ridership. Is there someone who one can email or contact regarding the current timid approach to operation practices? It seems the only way to fix this would be to light a fire under operation management and the head of the streetcar division.

    Steve: Leaving the response for the anonymous TTC operators who read this site.

    Like

  13. Re: the red left turn arrows:

    I always thought that the use of red right and left turn arrows would have been a good solution for Queens Quay, to clear up ambiguity over the difference from the through signals. Others have previously commented frequently on the inability to use European or American style transit signals. This is a case where the layouts that are permitted under the HTA and developed in the 60s or 70s do not apply very well to alternative, complex intersection configurations, and instead of pushing for changes or exemptions that would better suit a new configuration, the designers tried to make the HTA-approved configurations work in a way that would be reasonably applicable to the intersection design, in some cases more successfully than in others. (There are other issues around placement of the signal heads, again partly due to interpreting the HTA requirements to the letter even where it may not make sense for an unconventional intersection.)

    If it was an intentional decision, I can understand why the QQ designers might have been reluctant to go the province for HTA revisions or exemptions, as it would probably have been an onerous and time-consuming process. But it’s unfortunate, as it would have been a useful tool for other future similar projects (including Cherry), whereas now we are probably stuck with the current approach. The use of a red left turn arrow would have been useful even at more conventional intersections, including on streets with ROWs like Spadina but also at non-ROW intersections, by cutting down on sign clutter at the intersection (those LEFT TURN SIGNAL signs would no longer be required).

    (My interpretation of the US MUTCD is that red right arrows are valid for signals applying to fully protected right turns, but the red ball is to be used where right turns are permitted on red while the red arrow is to be used where right turns are always prohibited on red. Hence this would have been a good tool for the right turn signal on Queens Quay east of Simcoe.)

    Like

  14. Only 2 new streetcars delivered so far this year, I think that the TTC should reduce the size of the Bombardier order so that another manufacturer can produce the rest of the streetcars in parallel. When will the TTC wake up and realise that Bombardier is chronically incompetent and incapable? Metrolinx seems to have gotten the message and it’s time for the TTC to wake up also.

    Like

  15. On a broader PofView, the challenges to both King and Queen could have been alleviated or almost fixed by pursuing a Front St. transitway, especially about 15 years ago instead of that road folly, where all three governments had lined up to support a car-based ‘fix’ to congestion. The intent is to advantage transit in this obviously high-demand corridor from the pinch point at the base of High Park in to the core, almost the same distance as a current one-stop subway folly in relative sprawl to a single owner, vs the core of the City.

    With the obduracy of pursuit of this Suspect Subway Extension, and draining of planning time and resource and political will etc., we are building shut a very rare chance for a surface corridor and cheaper fix by allowing condo buildings east of Strachan, and instead the City is OK with building a road between Strachan and Dufferin.

    I’m quite willing to compromise with a reversible one-way transitway a la Jarvis, just to have SOMETHING happen, and have whatever vehicles are used return via King/Queen. Also, it’d be smart to have GO buses be able to use what transitway might arise starting at Dufferin.

    Don’t we need new transit in new corridors more than condos and a subway in sprawl??

    Like

  16. Concerning the open tracks west of Roncesvalles, is there a specific problem in this area that makes the maintenance issues different from the open tracks on the subway? Or regular railways for that matter?

    Will the replacement be open tracks on a concrete pad like the RT, or will the tracks be completely embedded in concrete like the street tracks? This would make a big difference in the ease of replacing rails.

    This question is also applicable to the construction to be used on the open LRT tracks on Eglinton and other LRT systems.

    Steve: The new Queensway track is to have concrete to the railhead, just like in street trackage. The claim is that this will save on maintenance costs, although I expect the same benefit could be achieved with open track on a concrete slab. Existing track (as on parts of the subway and mainline railways) does not have a concrete base. The problems near Davisville arise from the fact that the existing ballast and underlying structure needs complete replacement because of bad drainage. The ballast gets fouled with dirt and silt that binds it together, and it loses its resilience (trackbed like this is supposed to be a bit springy, not solid).

    To some extent I think that the TTC is trading higher capital cost on The Queensway against reduced future maintenance, but has gone overboard.

    Like

  17. The Streetcar Division of Brookville restores, rebuilds, and modernises old streetcars including but not limited to PCCs. My suggestion would be to cancel the Bombardier order which has no chance of finishing before 2050 and instead ask Brookville to modernise our old streetcar fleet which will give us more than enough time to replace Bombardier with another streetcar provider.

    Like

  18. @Prash;

    The problem with this is that the motor controllers would need to be totally replaced as the old ones are to antiquated to make repairs worth while. The other problem is that the AODA requires all vehicles to be accessible within the next few years so the cars would need to be replaced almost as soon a they are finished.

    I have ridden Bombardier Flexitys in many cities and they are a good car. I don’t know what has happened to Bombardier and their management in the past few years to result in such poor performance.

    Like

  19. That is my route out by Royal York to King & Dufferin. All that really matters is frequent service first and speedy service second. With smaller more frequent vehicles which is buses on this route for now, it would be better service. Bigger vehicles at less frequency is worse service. It doesn’t matter if it is streetcar or bus or hovercraft or automated self driving unit of the future, whatever does the job. I do prefer the streetcar as a ride, but time and job is money. I have since needed to move closer to work and am now a pedestrian fighting against transit authorities and private automobilia all winter. I don’t want my tax money used for it. Keep the sidewalks clean.

    Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s