Service Reliability on 47 Lansdowne: September 2021

This article continues a series reviewing service on relatively short routes within the TTC network in areas where one might expect service reliability to be easily achieved.

Overview

The southern part of the route lies between St. Clair and Queen Street. Every second bus continues north to Yorkdale Station on what was once a separate 18 Caledonia route. (These routes will be split apart at Caledonia Station in 2022 when Line 5 Crosstown opens.) During peak periods, the Yorkdale branch splits at the north end with an alternate route via Orfus to Yorkdale Station.

In addition to the type of problems shown in the previous articles on 22 Coxwell, 47 Lansdowne suffers from having what should be a blended frequent service on the common portion of the route. However, headways are not reliable either from the point where the branches merge at St. Clair southbound, nor at Queen northbound, the route’s southern terminus.

This schedule was in effect from September 1 to 4.

This is the schedule in effect from September 5 onward.

Change-offs, Breakdowns and Crew Changes

A common effect seen across the route is for a bus to disappear from service and, eventually, to be replaced by another in the same relative position.

  • In some cases, the same bus re-appears more or less where it disappeared indicating that it probably sat out of service awaiting a new operator.
  • In some cases, notably early in the day, a bus will remain in service for one trip or less, and then disappear. This implies that the bus was somehow faulty rather taken out of service because of a missed crew change.
  • In some cases, it is possible that a crew change is done by bringing a new bus into service from the garage replacing both the operator and the vehicle. This is not a scheduled event as can easily be seen by comparing operations on similar days (weekday to weekday, for example).

Where buses are missing for even a partial trip on a branch with a wide headway like the 47B Yorkdale service, the resulting gap can be very wide. These events are not reported as part of overall service quality and standards, just as bunching is not reported because it can often occur within the “approved standards” and their considerable margin for exceptions.

Northbound from Queen

Headways from Queen are measured north of Seaforth, the north end of the on-street loop. The averages and standard deviations of headways are well-behaved only for the first few hours of the day, and they deteriorate from 9am onward. There is considerable difference in values from week to week.

Note that because of the schedule change on the Labour Day weekend, the shape of the trendlines in week 1 of these charts are different than in weeks 2 to 5.

The clouds of data points span a range above 15 minutes for most of the day. This corresponds to the point where the standard deviation values rise after the am peak period. There are so many days with very short headways that I will not review each one in detail here, but give a sample. Days with very wide headways are of particular interest because this typically indicates either that buses are missing, or that bunching of more than two vehicles occurs.

This set of charts is particularly important because it shows the service at a terminal before the effects of passenger loads or congestion could disrupt service. This is also, as of September 5, a new set of schedules where any problems with the schedule itself should have been resolved. If anything, the service is worse in weeks 2 to 5 than in week 1, although this could be due to other factors.

Saturdays show bad bunching and gapping throughout the month. Sundays also show very erratic service except, oddly enough, on Labour Day (an honourary Sunday for these charts) when headways lie much closer around the trend line.

Southbound from Yorkdale Station at Bridgeland

The screenline for these charts is on Caledonia south of Bridgeland where route turns east to serve Yorkdale Station. As above, the shape of the week 1 chart differs from other weeks because of the schedule change.

Only half of the service reaches this point because of the 47A scheduled turnback at St. Clair. As at other locations, the SD values are high, but they are particularly so thanks to the wide scheduled headways and greater dispersion of data values. Weekend service is particularly unreliable with headways ranging over a wide span.

Note that some values go above the Y-axis cutoff, that is to say above half an hour.

Southbound from St. Clair

Service at St. Clair includes both the 47A buses originating at Earlscourt Loop and the 47B/C service from Yorkdale via Caledonia. The SD values are typical of a midroute location where branching services “merge” with little regard for each other. The clouds of headway values are spread over a wide range with values far from the average/scheduled service.

Friday, September 3

September 3 shows a few of the common problems on 47 Lansdowne.

  • At 7:20am there is a cluster of buses at Queen.
  • As the day continues buses on the 47A branch to St. Clair are often running close to 47B runs to Yorkdale.
  • The length of layovers at Queen, Earlscourt Loop (St. Clair) and Yorkdale (places where the plot of a bus is horizontal for some time) indicates that these buses are not stretched for running time.
  • Mid-morning service runs with only a few instances of bunching, but at 12:40 one bus, “purple”, goes out of service at Yorkdale and is replaced by “orange”. From that point, “orange” is running late, and makes a few short trips on the 47A before resuming 47B service at about 2:30pm. There is a gap in the Yorkdale service where it should have been.
  • “Light blue”, also on the 47B service, has an extended layover southbound at Davenport. It then operates on the St. Clair 47A service.
  • Two new runs enter service around 3pm: “dark green” from Yorkdale, and “brown” from Queen. Bunching is a problem particularly on the south end of the route.
  • Two runs leave service at about 5:20pm: “turquoise” south of Yorkdale, and “light green” near Queen.
  • Some bunching continues until around 8pm.
  • From 7:30pm, some buses disappear and reappear leaving gaps in various parts of the service.

Saturday, September 4

Although the scheduled service level on the south end of the route is 9 minutes during the daytime, the north end sees a bus only every 18 minutes, assuming the service is on time. When a bus is missing, the gap widens to 36 minutes or more.

  • Service begins fairly regularly, but just after 9am a bus, “yellow”, disappears from service. This creates a travelling gap that grows to 40 minutes on the Yorkdale branch.
  • This affects service until the bus re-enters service at 2:10pm.
  • Shortly later, at 2:50pm, “purple” goes out of service. It is replaced by “dark green” at 3:50pm.
  • After 6pm, “brown” and “turquoise” leave service in the normal transition to evening headways.
  • All buses remain in service through the evening, but problems with 47A and 47B buses running in pairs south of St. Clair remain.
  • “Turquoise” leaves service at 10:30 in a normal transition to late evening service.

Sunday, September 5

September 5 is the first day of the new schedule period. Running times and headways during some periods were adjusted from the August schedule (see tables above to compare them).

  • After 10am, one bus, “turquoise” consistently drops behind its place and by 11am is running as a pair with its follower “pink”. This continues until noon when “turquoise” short turns at Davenport, but running late continues for this bus until after 2pm.
  • A new bus, “yellow”, joins the route at about 2pm. This is a normal change for the afternoon, but the bus enters service late and a gap where it should have been in the preceding hours. “Yellow” gets into its proper place after 2:30pm.
  • Starting about 4pm, “brown” starts running late and pairs with other buses for a few trips.
  • At 9pm, “dark blue” misses a trip to Yorkdale by turning back from St. Clair, and at 9:20 there are three buses at Queen. Other pairings occur through the evening.
  • At 10pm, “purple” lays over for an extended period at Yorkdale and misses most of a round trip to Queen.

Wednesday, September 8

  • Through the am peak, bus pairings south of St. Clair are common. At 8am, “grey” leaves Queen Street early and runs in a pack of three to St. Clair, then as a pair with “yellow” north to Yorkdale. After a layover at the north end of the route, it returns south in its proper place.
  • The combined effect of bunching and a missing bus leads to a half-hour gap northbound from Queen just after 1pm.
  • As service builds up for the pm peak, bunching becomes more common with a trio of buses heading south to Queen just before 4pm.
  • The period from 4 to 7pm show buses running in pairs for extended periods (“brown/pink” and “purple/blue”) as well as some congestion effects notably southbound toward Dundas.
  • Three buses leave service northbound at St. Clair between 6:50 and 7:15pm, but one bus “dark blue” enters service shortly later southbound from Eglinton.
  • Through the evening, bunching continues south of St. Clair, and headways on the Yorkdale branch continue to be uneven.

Looking at the headways in greater detail reveals just how erratic the service is.

Seaforth, just north of Queen Street terminus, should be a location where buses pass at regular intervals. This occurs only for brief periods in the am and pm peaks. Far more commonly, a short headway is followed by a long one indicating that two buses are travelling together.

At Eglinton, with only the 47B Yorkdale service, the headway swings are wider because only half of the service comes this far north.

Southbound service at Lawrence is better-behaved until mid-afternoon when a gap caused by a missing bus creates a spike in the chart.

Southbound at St. Clair the service mirrors the northbound chart with little sense that the 47A and 47B services “blend” at all even though they are on a common headway all day.

Saturday, September 11

Saturday the 11th began poorly and continued that way all day thanks to missing buses. Bunching south of St. Clair is common, but this is not just southbound as 47A and 47B service fail to “blend” but also northbound from the terminal at Queen where, in theory, proper spacing should be restored.

  • “Pink” goes out of service northbound at St. Clair at 6:30am on its first trip and is replaced shortly later by “dark blue”.
  • Three buses operate together southbound from Yorkdale at 7:40am.
  • “Dark blue” runs as part of a pair from 10:40am to 12:40pm when it short turns at Bloor Street.
  • “Dark green” goes out of service at 1:15pm and is replaced by “purple” at 1:40pm.
  • After 4pm, “purple” and “ochre” form a pair that lasts until 7:20pm. They are joined by “dark blue” and “pink” resulting in a quartet of buses at Queen Street at 5:20pm.
  • There is a half-hour gap northbound from Queen between 5:30 and 6:00pm.
  • The effect of a missing bus is quite marked on the Yorkdale service when a 45 minute gap travels north from St. Clair. Although a new bus, “turquoise”, enters service southbound at 10:40pm, it runs as a pair with “grey” in a one hour gap.
  • Service continues to be erratic until after midnight.

Friday, September 17

September 17 is another day showing many problems.

  • Thanks to bunching, five (!) buses arrive southbound at Queen just after 8:40am and they travel northward in a pack. One of them, “purple”, on the 47A service turns back at St. Clair. “Dark grey” leaves service at that point, and “pink” follows at Lawrence. The remaining two buses travel together to Yorkdale.
  • Gaps and bunching continue into the late morning service including a 40 minute gap to Yorkdale from Bloor Street starting at 9:45am.
  • Service settles down through mid-day, although there is a brief period of congestion north of Dundas at about 2:30pm.
  • Congestion at St. Clair begins to develop after 3pm and continues until about 7pm.
  • Service to Yorkdale continues to be erratic with a trio of buses travelling north from Bloor at 4:15pm. Two of these buses, “grey/light orange” continue as a pair until early evening when a new partner, “blue” is added to the mix. Again there is a 40 minute gap in Yorkdale service.
  • A new bus, “turquoise”, joins the route at 9:30pm, but bunching and gaps continue through the evening. Three buses are at Queen together just before 8pm.

Monday, September 22

September 22 is, comparatively speaking, a “good” day with fewer disruptions than on many others.

  • Just before 7am, there is minor congestion south of Eglinton, but this is short-lived.
  • Through the am peak, there is some bunching between St. Clair and Queen, but it is not chronic.
  • Mid-day service is fairly well-behaved.
  • Starting at about 2:30pm, congestion develops north of Bloor, and this continues until about 4:30pm.
  • Service through the evening is fairly regular until “turquoise” goes out of service at Yorkdale just after 10pm. It is replaced by “dark green” at 11:30pm. This, plus some bunching, produces wide gaps between St. Clair and Queen.

Saturday, September 25

Another Saturday, and another catalog of problems on 47 Lansdowne right from the beginning of service.

  • “Pink” makes one round trip from Yorkdale to Queen, but then goes out of service at 7:20am. It is replaced by “ochre” at 8:10am.
  • “Dark blue” enters service at 7:40am at St. Clair, but goes out of service at 8:30am south of Yorkdale, reappearing at 9:10am replacing “blue” which leaves service just before.
  • “Dark blue” disappears again at 12:15pm south of Yorkdale creating a gap of about 35 minutes. This gap persists through the early afternoon.
  • Through the late afternoon, service southbound at Lawrence is quite erratic (this is shown in more detail in the next set of charts).
  • Gaps and bunching persist through the evening.

Looking at the headways in more detail shows just how badly they gyrate between very short and very wide intervals. Although the trend lines in these charts roughly follow the scheduled headway values (they will be higher when buses are missing), the range of actual headways is breathtakingly erratic.

Northbound at Seaforth, service should operate reasonably evenly as this is a terminal. In fact, alternating short and long headways typical of bunching are seen here,

Northbound at Eglinton, the 47A St. Clair service has dropped out, but the remaining 47B to Yorkdale arrives erratically. Some of the very wide headways of 30 minutes and more are caused by missing buses in what should be a 16 to 20 minute service.

The same pattern is seen in the 47B service southbound at Lawrence with some very wide gaps and occasionally buses operating close together.

Southbound at St. Clair, the 47A service is included, but the sawtooth pattern of short and long headways thanks to bunching predominates.

Sunday, September 26

September 26 was not a good day.

  • At 9:20am, “pink” disappears from service and is replaced shortly later by “dark blue”. The latter bus short turns at St. Clair to get back on time for its trip back to Yorkdale.
  • “Dark blue” and “purple” run as a pair for a round trip, and are joined by “yellow” south of St. Clair.
  • “Dark blue” disappears for a time around 11:15am southbound from Yorkdale, but reappears near Lawrence. This is probably a bus that “signed off” and forgot to “sign on” again, and therefore was not tracked. This bus was replaced by “brown” a trip later at 12:40pm.
  • Meanwhile, there was a 50 minute gap in the 47B service to Yorkdale northbound from Bloor starting at 11:20am.
  • The early afternoon sees more bunching with three buses, “purple/dark blue/green” running together and a 40-minute wide service gap on the 47B branch. Even with short turns, bunching remains.
  • Bunching continues through the late afternoon including a pack of four buses northbound just after 5:30pm.
  • Erratic headways, especially to Yorkdale, continue through the evening.

Tuesday, September 28

Like so many other days, September 28 has problems that persist through all service hours.

  • Just after 6:30am, a trio of buses heads north to Yorkdale from Queen. One of them goes out of service at Yorkdale, and the other two space out to reasonable headways. The missing run does not reappear.
  • Congestion appears southbound at Lawrence just before 8am and persists until early afternoon.
  • Bunching continues through the late morning including a trio of buses on the south end of the route between non and 12:30pm.
  • At 2:00pm “turquoise” goes out of service at St. Clair and is replaced one hour later as the pm peak buildup begins.
  • The afternoon peak is uneventful except for the usual bunching.
  • Come the early evening, one bus “dark orange” runs consistently late creating a wide gap. It can be seen regularly paired with other buses through into the late evening.

4 thoughts on “Service Reliability on 47 Lansdowne: September 2021

  1. Hi Steve, I live on this route at Dupont and St. Clair which is where services merge, and the St. Clair to Bloor section is the most crowded section in terms of people using the service. It is ridiculous how much bunching happens in the morning peak hours.

    What I don’t understand is how these buses travel together even from Queen where they have a chance to ‘reset’ or space out if they choose to do so. Is there no person within the TTC’s structure whose job is to look at service reliability and spacing? I am from Mumbai, a city of 20 million and we had a person called a ‘starter’ whose job was to keep the buses starting on time and spacing them at one end of the journey (whichever was most important ‘end’ of the route/ mostly a subway station as buses there are all about last mile connectivity usually).

    Steve: There is a certain irony that CEO Rick Leary is from Boston where they have Starters. But in Toronto the idea of actually dispatching/spacing service seems to be beyond our abilities.

    Like

  2. My 8 year old son says he wants to volunteer to be a ‘starter’ if the TTC is so worried about costs. He says all he needs is a phone with Google maps (which is what we use to track where the buses are) and he would be set. On a serious note, these kind of volunteers would serve better than having the ones at major subway stations giving directions.

    Like

  3. Steve, you should do an article on the 95, it’s been a ****show ever since they opened that temporary bus loop, its either bus bunching, or huge gaps between them. I regularly find myself to be waiting 15-20 minutes for a bus in the eastbound direction, but I see multiple pass me by westbound. Might have something to do with the 95A being extended too? There is also chronic overcrowding sometimes too.

    Steve: So many routes to review. This has been evolving from a manageable number of chronic routes to a system wide problem.

    Like

  4. The routes with poor performance should be contracted out to private sector players which has been the strategy used in many Asian cities with some of the most reliable transit service on the planet.

    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 )

Google photo

You are commenting using your Google 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 )

Connecting to %s