Recent articles on this site looked in great detail at the 501 Queen car and the problems with its service. Often, when people talk about Queen, they miss the streetcar routes that are, in effect, branch operations of the Queen line serving Kingston Road in The Beach:
- 502 Downtowner (formerly Kingston Road) operates between Bingham Loop (at Victoria Park & Kingston Road) and McCaul Loop sharing trackage with 501 Queen west of Woodbine Loop (which is actually at Kingston Road, and is named for the old racetrack, which itself became “Greenwood” when “New Woodbine” opened in northern Etobicoke). This route operates weekdays until the end of the PM peak. Evening and weekend service is provided by the 22a Coxwell bus.
- 503 Kingston Road Tripper operates rush hours only between Bingham Loop and York Street running into the core via King from the Don Bridge, and looping downtown via Church, Wellington and York.
This service design has been in place, with only a few changes, since 1948:
- 1954: Streetcar service cut back from Birchmount Loop to Bingham Loop.
- 1966: Coxwell bus replaced Coxwell streetcar and evening/weekend service on Kingston Rd./Coxwell (same as the 22A today).
The route name “Downtowner” arose from an ill-advised proposal to provide “relief” to the downtown subway by extending Kingston Road cars from McCaul Loop west and north to Bathurst Station in 1973. This didn’t last long. A year later the extended service became a peak-only operation, and that remained, on paper at least, until 1984. We have the name as a memento of that extension now 30-years in the past. The basic problem was that very little of the service actually reached Bathurst Station with many cars short turning either at Wolseley Loop (Queen & Bathurst) or at McCaul Loop.
The situation is not unlike what we see today because the 502 Downtowner schedule does not provide enough running time, and short turning is a chronic problem. This is particularly troubling because the short turns defeat the purpose of the route’s existence:
- A short turn eastbound at Woodbine Loop removes service from the street which the route is intended to serve.
- A short turn westbound at Church (looping via Richmond and Victoria) sends a car east without serving the major stops downtown from Yonge to University.
- A short turn westbound at Parliament (looping via Dundas and Broadview) removes a car even more from downtown, and not even a clever rider walking a block east from Yonge (an “illegal” move with a regular transfer) can take advantage of the service.
This is compounded by extremely erratic headways that are far worse than I have seen on any other route I have analyzed. According to TTC route performance stats, the 502 is “on time” (that is to say, within ±3 minutes of the scheduled headway) 30% of the time. As we will see later, even that claim is a stretch.
As for the 503 Kingston Road Tripper, service on that route is supposed to be blended with the 502, and during AM peaks it can work out, sort of, there is a vaguely reliable headway of alternating 502/503 cars on Kingston Road. But it’s a hit-and-miss situation, and very large gaps in 503 service are quite common.
Anyone attempting to use transit on or to Kingston Road is well advised to get on the first thing that shows up and be prepared to transfer. This appalling situation is a mockery of what the TTC claims is its “customer service”.
Service on Kingston Road was substantially better in past decades, and it is no wonder that ridership and scheduled service levels have fallen given the unpredictable nature of these routes. Recently, there has been some improvement. In April 2013, off-peak headways of 502 Downtowner improved from 20 to 16 minutes, and in June 2015, from 16 to 10 minutes. However, the fundamental problem of headway reliability undoes much semblance of “improvement”.
The following table summarizes the scheduled service levels during the periods included in this analysis.
2013 was a trying time for Kingston Road riders as the evolving schedules (and actual operations) show:
- May 2013: Routes operating “normally” to McCaul and to York downtown, and to Bingham Loop in the east end. Midday service had just been improved in April 2013 from every 20′ to every 16′ on the schedule.
- July 2013: Reconstruction of Kingston Road caused all streetcar service to be cut back at Woodbine Loop. Meanwhile, construction on York required both routes to turn back at Church/Victoria.
- July 29, 2013: Service west to McCaul Loop resumed on 502 Downtowner, but the 503 Trippers could not use York Street for their loop because it was still under construction. This situation continued until December.
- September 2, 2013: Service was scheduled to resume to Bingham Loop, but reconstruction of the street was not yet complete and streetcars continued to turn at Woodbine Loop. This arrangement continued into the late fall.
- December 23, 2013: Service was planned to resume to Bingham Loop, but the massive ice storm of the preceding weekend made Kingston Road impassible for a few days.
During the extended construction period, the 502 and 503 cars were treated more as unscheduled extras on the Queen and King lines, and supervisors filled many gaps to Neville Loop with 502s that had nothing else to do.
In 2014, the service operated with the same peak schedule as it had before the construction projects including the new 16′ midday headway. This continued until June 21, 2015 when the midday headway improved to 10′ and Kingston Road became part of the new “Ten Minute Network”.
The 503 Trippers were also affected by the long-running Queen Don Bridge diversion that sent all King Street service via Queen and Parliament to route. This ended in early June 2015.
A Note About File Names
The various charts use a similar naming convention to the one in the recent 501 Queen articles. “523” is the dummy route number for the combined 502/503 service.
- For headway charts, the number after the route name gives the direction (“1” = westbound, “2” = eastbound) and location (“01” is the first point on the route counting from the east where I have measured headways).
- For link time charts, the number after the route name gives the direction (“1” or “2” as above) and the pair of points on the route between which running times are measured.
This convention sorts files in geographic rather than alphabetic street name order.
Service at Bingham Loop (Victoria Park)
This chart shows, for the periods that streetcar service did operate to Bingham Loop, the actual headways westbound on Kingston Road just west of the loop. (Using this point avoids variations in GPS data depending on how cars circle the loop and lay over.) Average and Standard Deviation values are shown for each week with breakdowns by hour. A headway is counted within an hour based on when a vehicle passes the measurement point. Therefore, a gap spanning from 6:50 am to 7:10 am is counted as a 20′ headway within the 7:00 hour.
During the peak periods, the data include both the 502 and 503 services. This is evident from the much larger average values through midday. During the peak, the averages should be 6 minutes except for late June 2015 when the summer service at 7’30” operated. However, the averages are generally a bit higher and, much more troubling, the standard deviation values lie above 3 minutes, particularly in the PM peak.
Midday service should average 16′, but the values are generally higher, and the SD have a wide range rarely below 5′. This shows that not all of the service reached Bingham, and what did make it to the end of the line ran on an erratic headway.
Service at Coxwell Avenue Westbound
At Coxwell, any service that short turned at Woodbine Loop is back in the mix, and this improves the averages somewhat, although the peaks still run above the expected 6′ mark. Part of the problem is that the 503 service, being trippers, does not always make it out if there is a shortage of operators or vehicles. As at Bingham, the standard deviation values are high, and by the PM peak they are comparable to the headways showing severe bunching.
Midday service at Coxwell westbound is atrocious. The average values lie generally around the scheduled 16′ headway, dropping to 10′ with the schedule change in June 2015, but the SD values are generally over half a headway showing that there is a wide range of typical headways even on the part of the route that would have benefitted from short turns. The schedule may say 16′, but what riders actually get is often much different.
The service during summer 2013 was not operating strictly to schedule, but rather as if the cars were extras to be used wherever they could help out on the main routes.
The pattern at Coxwell is mirrored at points further west along the line showing that there is no effort so impose regular headways on the inbound service.
Service at Yonge and Queen Westbound
At Yonge and Queen, only the 502 Downtowner service is present. It should appear on a 12′ peak and a 16′ offpeak headway. For the AM peak, this average generally holds, albeit with a standard deviation in the 5′ range indicating uneven service. Midday, the average is routinely at or above 20′, and this is particularly bad in 2015 with midday averages in the 25-35 minute range. This is a direct result of the amount of service short turned westbound at Church and even at Parliament that does not reach the core area to serve outbound trips. The PM peak averages run well above the scheduled value, and the SD numbers are routinely at or above the scheduled headway.
Service at University Eastbound
University Avenue is a short distance east of McCaul Loop, and so measurements here show the eastbound service before most effects such as “congestion” might take their toll on reliability.
The AM peak is reasonably well-behaved in the early hours with average headways at roughly the scheduled level and low SD values showing regular departures from the loop. However, by midday, the average headway values start to bounce around and are generally above 20′ even though the schedule says 16′. Most of the SD values are above 10′ showing very irregular service. The PM peak values are also well above the scheduled numbers, particularly in 2015.
As at Yonge westbound, this shows the effect of the amount of service that is short turned before reaching the area where most outbound riders would be waiting.
Service at Parliament and at Broadview Eastbound
Service eastbound at Parliament and Broadview makes an interesting comparison for several reasons.
Peak period service at Parliament includes only the 502 Downtowner cars because the 503s are either on King Street, or are joining Queen at Parliament (on diversion) and are not counted as “crossing” Parliament in these charts. Comparing the charts at Broadview shows the difference in the average peak headways with and without the 503 trippers included.
Midday service eastbound at Parliament includes cars that short-turned westbound at Church, but not cars that short turn at Parliament. Average headways are closer to, but still generally above the target values of 16′, dropping to 10′ in late June 2015.
Service at Woodbine Avenue Eastbound
As with the westbound charts for Bingham, this set only shows periods when service ran up Kingston Road. Service here mirrors the pattern at Bingham and shows the effect of short-turns at Woodbine Loop. Midday average headways are almost always above the 16′ scheduled value, and SD values indicate that the values vary widely, certainly well beyond the TTC’s ±3 minute target.
Comparing Actual and Scheduled Running Times
A fundamental problem with the 502 Downtowner service is that the time provided in the schedule is inadequate, except early in the morning, for actual conditions on the route. As a result, cars are almost always late, and the result is chronic short-turning. Moreover, headway regulation takes a back seat to simply keeping operators on time, and chaotic service is the accepted norm on this route.
These charts track the time required for streetcars to operate between points near the eastern and western termini. The values are quite consistent with low SD values showing that the 502 Downtowner service is not subject (at least for the months represented here) to the type of delays that plague its sister route 501 Queen. There is a slight increase in the 2015 values, but things have been consistent for a few years.
The scheduled one-way trip from Bingham to McCaul has been 42 minutes peak, 40 minutes midday, for a long time. Only in late June 2015 was the peak time extended with 3 minutes recovery time. This is more about making the trip time an even multiple of the headway for the summer schedule rather than to match actual conditions, and the recovery time will almost certainly vanish again in the fall schedules.
Actual times from west of Bingham to University lie at or above the scheduled times for the slightly longer trips including turnaround time at the terminals. Clearly, it is impossible for streetcars to achieve these times.
These charts cover the shorter segment of the route for which there is data during 2013 when streetcars all turned back at Woodbine Loop. The scheduled running times from Woodbine to McCaul Loop were 30′ (AM peak), 32′ (Midday) and 33′ (PM peak). The progression in values mirrors but does not match the actual increases seen in the data where there is a wider range between the AM and PM peaks. As with the data for trips to/from Bingham, the amount of time required for the segment excluding the termini is close to or greater than the scheduled value over the entire route.
There is one point with a disruption in service in early August 2013 that extended the averages. This was caused by severe congestion eastbound from University to Yonge during the PM peak.
503 Kingston Road Tripper Service
Monitoring the 503 on its downtown loop is challenging because GPS errors are common among the many tall buildings. Cars routinely report that they are much further north than their actual location, or even that they are nowhere near downtown. To avoid this problem, I have used Sherbourne Street as a reference point. This picks up all service to and from downtown regardless of whether it is bound for York or only for Church.
Service westbound into downtown is roughly at the scheduled 12′ headway early in the AM peak, but the average gets worse in later hours. Moreover, the SD values are high indicating that the cars are not regularly spaced. Eastbound from downtown in the PM peak, average headways are only close to the 12′ schedule for the 4:00 pm hour, and by 5:00 pm, the averages are much worse, especially in 2015. As with the AM peak, the SD values are high indicating irregular service.
June 2015 In Detail
The following charts show the detail of headways by day and week at three locations: westbound from Bingham Loop, eastbound at University (502 service only), and westbound at Sherbourne (503 service only). (Note that Wednesday, June 10 has no data available.)
For the 502 Downtowner service, the trend lines show broadly how the headways behave near the termini with widely-ranging values some of which are so high that the service is for all purposes completely absent. In the middle part of the route, east of Parliament and west of Coxwell, the service may be “on time”, but it is doing little more than operating as extras on 501 Queen.
Equally with the 503 service, the headways vary widely and bear no resemblance to the advertised service (12′ headways except in late June 2015 when they move to 15′).
Looking at Individual Days
The layout of these charts is similar to previous time-distance charts I have published except that it contains a separate segment for the portion of the 503 Kingston Road Tripper route that splits off from 502 Downtowner. This segment is placed at the top of the chart and shows the portion of the line west of Parliament Street. Lines moving diagonally up to the right are westbound cars, downward are eastbound. A horizontal line is a stationary vehicle.
Tuesday, June 2, 2015
June 2 was a typical day, to the extent one can make this statement, given stats showing widespread gapping and bunching as an ongoing problem.
The first page covers the early morning up to 7:00 am. Cars pull into service from Russell Carhouse, operate east to Bingham Loop, wait for their departure time, and leave. The first of the 503 trippers leaves Bingham just after 6:25, operates westbound to Parliament (the Don Bridge diversion was still in effect) and then shifts from the “502” part of the chart to the “503” section. Two cars are held westbound at Pape just before 7:00, and they will be held again west of Broadview shortly later. This is probably an equipment failure on the first car as it disappears from the data stream near McCaul Loop at about 7:25 and is not seen again.
The second page covers the AM peak from 7:00 to 10:00. The 503 runs are easy to spot because of the way they vanish at Parliament (reappearing on the 503 segment), but what this also shows is that, in general, they are not running half way between 502 cars. This problem gets worse as time passes. The somewhat irregular headway of the 503s is visible in the top part of the chart. Note that missing data in the eastbound trips is due to GPS problems on King Street in the “canyon”, although the phenomenon extends east of Church for some cars. (This illustrates why I used Sherbourne Street as a reference point above.) All of the 503s run in to Russell Carhouse eastbound.
Service to McCaul is somewhat irregular with a large gap at 8:40.
- The car that had been delayed twice around 7:00 am (“turquoise”) continues to make full McCaul-Bingham trips but it is closely shadowed by the following car (“light green”) for a round trip until that car short turns westbound at Church. “Turquoise” runs in eastbound to Russell.
- One car (“amber”) shows a propensity for running close behind its leader starting eastbound from McCaul at 8:10, and this continues until about 10:50 when it is short-turned at Church.
The third page covers the midday from 10:00 am to 1:00 pm. The horizontal lines east of Greenwood are caused by data from cars that have run in to Russell Yard, but whose operators did not “log off” the tracking system. Events of note:
- No cars short turn eastbound at Woodbine, but service to Bingham is irregular, especially for a 20′ headway, and there are some gaps of over 30 minutes.
- Three cars short turn westbound at Church, and one at Broadview.
- Short turns leave a gap of roughly an hour eastbound through the core starting at about 11:15.
- One car is short turned twice at Church and is accompanied for much of a round trip by another car.
There was no e-Alert giving any indication of troubles on the route and the short turns appear to simply be a case of trying, badly, to keep cars on time.
Page four covers the afternoon from 1:00 to 4:00 pm.
- Four cars short turn westbound at Broadview, and one at Church causing very large gaps in service from Yonge to McCaul Loop.
Page five covers the PM peak from 4:00 to 7:00 pm.
- The 503 is supposed to be operating every 12 minutes (5 cars/hour), but only 9 trips show up in the 2.5 hour period when service would be expected.
- One 503 car (“pink”) is rerouted via Church to Queen to fill a gap on the 502 just before 6:00, but as it operates east from Church, it does not serve the primary stop eastbound at Yonge.
- Six of the westbound 502 trips are short turned at Broadview returning east from Parliament. It is unclear just what service these provide to people waiting downtown for cars to Kingston Road.
- During the three-hour period, only five cars reach McCaul Loop on what should be a 12′ headway.
Thursday, June 11, 2015
June 11 starts out well, but things fall apart as the day progresses.
For the early morning, service builds up normally, although one 503 car (“grey”) runs from Russell to Woodbine Loop, takes a long layover, and then enters service westbound on Queen just after 7:30.
For the AM peak, the 502 service operates normally between Bingham and McCaul, but the 503 service has large gaps probably due to missing cars.
For the midday, the 502 service operates almost entirely without short turns, but there is one “problem” car (“mauve”). On the 10:15 trip westbound from Bingham, it is already closer than scheduled (the 20′ midday headway) to its leader (“dark blue”) which appears to be falling behind schedule due to a widening gap. By 11:30, the two cars are running as a pair for much of a round trip until “mauve” short turns eastbound at Woodbine Loop. Meanwhile, another pair has formed westbound from Bingham at 12:10.
In the afternoon, the effect of pairing and short turns becomes more evident:
- “Mauve” and “turquoise” make much of a round trip together leaving McCaul eastbound at about 1:20 pm, and they are followed by a one-hour gap.
- “Dark blue” makes two trips where it short turns westbound at Broadview.
Also visible during this period is the onset of congestion between University and Yonge.
In the PM peak, congestion west of Yonge becomes severe with some cars taking close to an hour to make the trip westbound. This was the day of an intense storm that would later close the east end of the Queen route due to a fallen tree, and flood Museum Station. There was no e-Alert issued for either the Queen or Downtowner route, and it is unclear what was happening around City Hall at the time. (The Gardiner East debate was in progress inside the chamber.)
The 503 trippers managed to operate almost all of their trips, but with some of the second outbound departures delayed and bunched.
Wednesday, June 24, 2015
June 24 was a fairly typical day for service on the 502 and 503 after the introduction of the more frequent 10′ headways for midday service on 502 Downtowner. Peak headways were on the summer frequency of 15′ on each of the two routes.
Early morning service is unremarkable. Its line vanishes at around 6:50, the point where it turns off of Queen onto King at the Don Bridge, and reappears at Berkeley Street, one block west of Parliament, shortly later.
The AM peak shows the same pattern seen on other days with the 503 cars failing to provide an evenly split headway westbound from Bingham as the morning progresses. The GPS problem on King is so severe for one car (blue line leaving York eastbound about 8:20) that it actually appears briefly as if it were on Queen crossing Yonge Street, then disappears back to King Street.
Meanwhile, the service to McCaul Loop has been well-behaved with only a few oddities:
- Delays eastbound at Woodbine caused by traffic signals favouring westbound cars turning south are evident.
- There are delays at Pape (the almost horizontal sections for several cars between Greenwood and Broadview, mainly westbound.
- One car (turquoise) appears out of nowhere eastbound at Coxwell at 9:10. This car will have an interesting progress for the next few hours as we will see.
For the midday:
- There is one delay eastbound just east of Broadview that holds a car for half an hour starting at 10:50. During this period, two cars divert via Parliament, Dundas and Broadview, and only lose a little time in the process (note how the breaks in their graphs have an offset because it takes longer to cover this distance via Dundas than via Queen). The delayed car short turns eastbound at Woodbine Loop, and then westbound at Broadview finally getting back into place eastbound from Parliament at about noon.
- Our “mystery” car (turquoise) runs west to McCaul, then east only to Woodbine Loop. On its next eastbound trip, it runs through to Bingham, but shadows another car most of the way. On their westbound trips (about 12:05 from Bingham), our “turquose” car short turns at Church/Victoria, while its cohort (“light blue”) short turns at Broadview returning east from Parliament.
- Two cars are short turned at Church/Victoria around 11:30, but their eastbound re-entry is closely followed by another car (“pink”) which short-turns eastbound at Woodbine Loop.
For the afternoon period:
- Five cars short turn westbound at Church/Victoria, and two cars short turn westbound at Broadview (re-entering eastbound from Parliament). With a scheduled 10′ headway, this means that 7 of 18 cars, almost 40% of the service, never reached Yonge Street. Headways through the core suffered accordingly with two gaps of over 30 minutes eastbound from McCaul.
- Four cars short turn eastbound at Woodbine Loop.
- One car (“blue”) has a particularly interesting journey. Eastbound just after 1:00 pm it short turns at Woodbine Loop and immediately follows another car (“brown”) westbound. “Brown” turns back from Church and neatly splits an eastbound gap. Meanwhile “blue” turns again at Woodbine Loop and closely follows “brown” westbound to Church and short turns, only to be closely followed east by another car (“light blue”) coming east from McCaul. “Blue” finally morphs into a 503 car for its westbound trip around 3:30 pm.
- Meanwhile, or friend “turquoise” comes east to Bingham at 1:20 pm, and is closely followed west by “pink”. “Turquoise” short turns at Church only to become part of a pair with “light blue” eastbound. Eventually, thanks to everything else being short turned, “turquoise” makes a trip to McCaul as a solitary car trailed by a 40 minute wide gap.
For the PM peak:
- Service on the 503 looks as if cars missing, but they are just badly spaced. In the course of two hours there are 8 trips, what one would expect on a 15′ headway, but some of them run much more closely to each other than the scheduled interval.
- During this period, 5 of the 502 cars short turn at Broadview and return east from Parliament, and 1 car short turns at Church.
- A delay at Queen and Bay held service between roughly 5:30 and 5:50 pm. During this period most of the 502s were not even getting to Church, let alone Bay, and one car at McCaul Loop (“brown”) disappears from the chart while it loops north to Dundas, east to Victoria and finally south to Queen rejoining the line at about 6:00 pm. There was almost no 502 service eastbound through the core during the main part of the PM peak.
Service on this date may not look very good, but it is not unusual.
The thread running through all of this is that a relatively minor streetcar line can be extremely frustrating to riders through inattention to adequate scheduling and through line management that values short turns and being on time to provision of service where riders actually might be located.
Making 502 Downtowner part of the “ten minute network” should improve the lot of riders on Kingston Road, but this will only actually happen if schedules are adjusted to provide adequate running time when it is needed, if short turns are rare and used only for the need to restore service after a bona fide delay.
The “blended” 503 Kingston Road service is a special challenge and at the very least, it should be managed during the AM peak when demand is primarily inbound to give reliable service westbound from Bingham Loop. Whether the two routes should be consolidated as one is a separate debate, and that should occur in the context of much better line management in general. Consolidation will send more peak cars to McCaul Loop, but they could be just as disorganized (not to mention short turned) as the 502 service is today.