Updated August 26, 2021 at 3:00 pm: The TTC has advised that the planned implementation of the 128 Stanley Greene bus has been deferred, contrary to information in the CEO’s Report. See the discussion under the 2021 Service Plan heading for more information.

Although there is no TTC Board meeting in August, the monthly CEO’s Report provides an update on TTC operations. Here are some points of interest and clarifications from the TTC on questions I sent about the report.

This article looks at ridership trends, the 2021 Service Plan status, service reliability and vehicle reliability.

Ridership, Revenue and Crowding

As Toronto reopens, ridership has started to grow, and as of the start of July had surpassed the level of early April when the stay-at-home order choked off the last attempted restart. The system as a whole reached 35 per cent of pre-covid levels. Buses still have the most boardings (44 per cent of pre-covid), but growth is stronger on the subway and streetcar networks in recent weeks. The TTC expects that this trend will continue through the summer and fall as in-person participation in school, office and other activities picks up.

Fare revenue is running well below historical levels thanks to the low ridership, and is also below the budgeted level because a stronger recovery was forecast in late 2020 when the budget projections were struck. The shortfall is part of the overall budgetary gap that the City faces in 2021 with pandemic-related costs.

As ridership picks up, so will bus crowding although the effect varies by route, location and time-of-day. The TTC does not break out this information in detail, but the data below show a clear trend into early July. An important consideration here is that the TTC’s recovery plan allows for greater crowding once the overall level of ridership crests 50 percent of historic values.

There simply are not enough buses and streetcars to accommodate twice the current riding level (i.e. a return to about 70 per cent overall) at current crowding levels. That said, the TTC’s fleet is substantially larger than its day-to-day requirement including provision for service and maintenance spares. There is room for growth in the total service operated provided that a way is found to pay for it. This will be an important issue going into the 2022 budget discussions.

In the chart below it is important to remember that these are all-day, all-system numbers. Many trips that are counter-peak, or offpeak, or on routes that tend not to accumulate large numbers of riders, are included in the total. A figure of 7 percent may not look like much, but the value is diluted by counting many trips that would never be crowded anyhow.

The issue, which the TTC does not report, is the proportion of trips on busy routes and times that are crowded. This results in a disconnect between rider complaints and reported average crowding levels. A basic aspect of transit is that when loads are not even, more people are riding on crowded buses than those that are nearly empty. The perceived level of crowding will always be higher than the average, but riders cannot board an “average” bus trip especially if that trip occurs on a route or at a time when they do not travel. Such is the inherent problem of reporting average values.

2021 Annual Service Plan

The CEO’s Report lists some of the service changes planned for October and November:

• 121 Fort York-Esplanade will be reorganized with service west of downtown to the Exhibition and Ontario Place removed. The east end of the route will be extended north via River Street from the Distillery District to Gerrard Street.
• 128 Stanley Greene will provide a new peak period service from Wilson Station.
• Weekday peak express service will be added on new routes 943 Kennedy and 968 Warden.
• New off peak service will be added to express routes 953 Steeles East and 960 Steeles West.

Updated August 26, 2021 at 3:00 pm: Since the publication of this article, the TTC has advised that implementation of 128 Stanley Greene is on hold.

Last week we decided to pause on implementing this change in November based on feedback from residents in the community who would like to explore alternative routing options (connection to Downsview Park Station). We are planning to do that work with the community in the fall. Any revision would be included in the 2022 ASP for approval. Based on these timelines, service to Stanley Greene will be implemented most likely in Q1/Q2 2022.

Email from Mark Mis, Head of Service Planning & Scheduling, August 26, 2021.

Readers of this blog often ask about the status of the planned Scarborough route reorganization. I asked the TTC about this, and here is their reply:

The following elements of the Scarborough East bus route changes, as described in the 2021 Annual Service Plan, have been or will be implemented by the end of this year:

• 905 Eglinton East Express weekday extension to Conlins Road began in May 2021
• New 938 Highland Creek Express in peak periods will begin service in September 2021
• 95 York Mills extension to Sheppard Avenue E

The final element of the plan is the extension of 54B Lawrence East to Kingston Road, and introducing new service on Brimorton Drive. These changes should be implemented together to maintain service coverage in the Orton Park neighbourhood. The status of these service changes are pending as we review with the local councillor.

Email from Stuart Green, TTC Corporate Communications, Aug. 25, 2021

Construction projects will continue to affect the Queen and King corridors for many months to come. The City recently announced that the segment of the Queen Street track and watermain project would begin west of McCaul on October 4. I asked the TTC whether the 501 Queen streetcar would be restored to that point with the October schedule change (typically on Thanksgiving weekend). They replied:

As originally planned, 501 Queen streetcar service would have been re-instated as far west as McCaul Loop as of the October 2021 board period, and replacement buses cut back to Jarvis. However, due to construction constraints within the broader context of the overhead upgrade program and other constructability concerns identified by the project team, streetcar service on Queen Street west of Parliament Street will not be restored until January 2022, upon completion of the full Bay to Fennings trackwork project. At that stage, cars will operate as far west as Wolseley Loop until overhead is restored in the KQQR intersection in May of 2022.

Email from Stuart Green, TTC Corporate Communications, Aug. 25, 2021

This also settles the question of when normal service through the King-Queen-Queensway-Roncesvalles intersection will resume. That project, as anyone who has visited it can tell, is running behind schedule.

Demand Responsive Service Plan

The TTC’s “demand responsive service plan” consists of operating extra buses to fill in on routes with crowding problems and react to ridership growth faster than schedules can be changed. The actual number of “Run As Directed” (RAD) buses that are available were the subject of a recent article. Because the number and usage of these buses is not well understood, I asked the TTC to confirm that my analysis was correct.

The TTC replied:

As we’ve said, approx. 120 to 140 buses a day, with three shifts of up to 55 buses providing coverage at any given time. Numbers will continue to fluctuate through the fall into the new year.

Email from Stuart Green, TTC Corporate Communications, Aug. 25, 2021

That is essentially what I reported, and it is important to note that the “up to 55” buses at any given time are far less than the 140 buses cited by some as if they were all on the road at once. This is important in the context of filling in other types of service replacement. These include shuttles for construction and replacement of scheduled runs that are cancelled either because an operator did not show up, or because they were never crewed in the first place due to a shortage of staff.

Yes, cancellation hours are tracked internally.

To help mitigate cancellations we utilize the spareboard, overtime and Run as Directed buses (RADs). There are instances when we would not be able to fill all the cancellations, especially last minute cancellations on the weekends when RADs are also helping with shuttles during subway closures.

Email from Stuart Green, TTC Corporate Communications, Aug. 25, 2021

Weekend cancellations are a particular problem because service is less frequent to begin with, and a missing bus can leave a big hole in the service. A combination of missing buses and lack of headway management can create very wide gaps in service. (See examples in the article on RAD buses.)

Service Reliability

The TTC measures service reliability by a combination of “on time” and headway-based metrics depending on the frequency of service. This has its pitfalls as I wrote recently. (See What Is “Reliable” Transit Service?) Although the margin for individual vehicles appears reasonably small, the swing in values between successive runs can produce much wider gaps than the standard implies.

For rapid transit lines, the reported values are based on this definition:

Headway (amount of time between train arrivals at a station) adherence of all service trains at end terminals. Data represents weekday service. To be on time a train must be within 1.5 times of its scheduled headway.

The “on time” values are averaged over entire reporting months hiding major disruptions. These are reported in accompanying text, but there is no formal standard to trigger exception reporting. The OTP values for all four rapid transit lines consistently sit above 90 percent.

Note that it is possible for trains to be missing and the service to still be measured as “on time”. For example, if the scheduled service is every 4 minutes, a 6 minute gap is within the 1.5 times metric. This means that an actual service of 10 trains/hour rather than 15 would still rank as “on time”.

The TTC tracks the proportion of peak service operated versus scheduled, but this has been moved from the main CEO’s report to an appendix. These charts do not show variations in the level of scheduled service that would reflect seasonal fluctuations, pandemic cutbacks or equipment availability problems (a notable issue for the SRT). From a rider’s point of view, 100 per cent of scheduled service is not the same as delivering the same service each month.

This is particularly striking on the SRT where only 75 percent of scheduled service was delivered during the latter part of 2020.

For surface modes, the standard against which service is tracked is different.

On-time performance measures vehicle departures from end terminals. Vehicles are considered on time if they depart within 59 seconds earlier or five minutes later than their scheduled departure time.

This is only one part of the TTC’s own Service Standards which also include headway-based targets for services running every 10 minutes or better.

The OTP value for streetcar routes in 2021 lies in the range from 60 to 80 percent overall, with poorer results on routes affected by construction which is a considerable part of the network. Bus OTP values lie in the 85-to-90 per cent range at least in part because so much of that network is not affected by construction.

• 

• 

The Service Standards state:

Passengers using high-frequency services are generally more interested in regular, even headways than in strict adherence to published timetables, whereas passengers on less frequent services expect arrivals/departures to occur as published.

TTC Service Standards, p. 11

The CEO’s Report depends on explanatory text to bridge the gap between the reported values and the targets, and does not address headway reliability at all. There is a naive belief that if service leaves terminals “on time” this will ensure reliable service downstream, but that is not how transit service actually works. Minor gaps from terminals get wider, and bunches get tighter.

There is no report of missed trips either because service was missing or because of short turns. This shows up indirectly in wider-than-standard headways, but it is a different and more serious problem than gapping.

The TTC claims that short turns have mostly been eliminated, but examination of vehicle tracking data (not to mention real-world experience) shows that this is not true. Because the goal is to have few or even no short turns officially, there is an incentive to under-report.

There is no report of cancelled service, a particular problem on weekends due to a relative shortage of resources to plug the gaps.

All service reliability metrics are averaged across all routes, all times-of-day, all month thereby hiding problem locations and times. This is a long-standing problem with the CEO’s Report.

Vehicle Reliability

Vehicle reliability is reported as the mean distance between failures (MDBF). This number varies from mode to mode for a variety of reasons including:

• A “failure” is defined as an event that creates a gap in service of five minutes or more for rail modes. For buses, it is an event that triggers a road call.
• Many types of failure are not related to distance travelled, but rather to time or duty cycles (e.g. door operation). Modes with higher average speeds will “fail” at a wider average distance simply because they accumulate more kilometres per hour than slower modes. In the case of the subway, a train can remain operational even if one car has a problem, and so a door problem on a train has less effect than on a streetcar.

• 

• 

• 

For some modes, the reported failure rate is capped even though actual values vary at some level above that cap. According to the TTC this represents situations where the values “exceed expectations”. This begs the question of whether those expectations should be raised.

It has been reported this way for years if you look back at past reports. Data is capped when it well exceeds expectations and reported in detail when it is below the target.

Email from Stuart Green, TTC Corporate Communications, Aug. 25, 2021

The TTC reports the percentage of scheduled service that is actually operated as a measure of fleet availability. This number is consistently high, sometimes over 100 per cent, because the scheduled service is well below the size of the fleet and the number of spares greatly exceeds the TTC’s 18 per cent target. (If 100 buses are scheduled, then a fleet of 118 is required to provide for spares.)

As service rises post-pandemic, there a basic question: can the TTC actually achieve an 18 per cent spare factor, or does it need a larger fleet due to reliability problems? When the spare factor is high, some problem buses might never leave the garage and pull down the reliability stats. On the streetcar network, fleet availability is reduced thanks to a major retrofit program that will continue into 2022. Service and budget discussions, at least in public, do not address the issues of overall fleet availability and the options for running better service.