Measuring and Illustrating Headway Reliability

In recent articles, I have reviewed a pervasive problem on the TTC’s network with uneven headways (the time between vehicles). This is annoying enough in “normal” times, but with the concerns about crowding during the pandemic, anything that contributes to crowding is more than just an annoyance.

The CEO’s Report presents only a few measures of service quality:

  • the proportion of service that is “on time” leaving terminals,
  • the number of short turns, and
  • the amount of service fielded compared to what is scheduled.

These metrics are utterly inadequate to showing how the system is behaving at the level seen day-by-day, hour-by-hour by riders for several reasons:

  • Most riders do not board vehicles at or near terminals, and measures that look only at terminal performance do not reflect most of each route.
  • “On time” is a meaningless concept for most routes because service is (or should be) frequent enough that riders do not time their arrivals at stops to meet their bus.
  • A count of short turns only indicates that most vehicles reached their terminals, not whether the absence of short-turns contributed to poor service quality. Moreover, there is good reason to believe that the number of short turns is under-reported.
  • Statistics are averaged on an all-day, all month basis and do not reflect route-by-route or hour-by-hour variations.

Riders to not ride “average” buses and streetcars – they take what shows up when they try to use the system. Indeed, if the TTC actually provided “average” service, a lot of problems would be solved because the service would be much more reliable.

At the recent TTC Board meeting, CEO Rick Leary noted that the metrics in his monthly report will be changed in the near future. This is long overdue, but it remains to be seen just how informative these will be.

This article includes displays of headway reliability in a new type of chart I have been working on recently. Reader feedback on this is welcome.

Continue reading

The Problem of Scheduled Service Irregularity: Update

In a recent article, I reviewed the scheduled service for several routes and the potential problems associated with uneven headways and crowding. As noted in that article, several routes had their schedules modified in recent months to reduce or eliminate these problems.

In my review, I only looked at a subset of routes across the system. The TTC has provided a full list of the routes with new schedules and their plans for coming months.

We have been actively working to reduce uneven headways created by the cancellation of crews in the May 2020 board period. 

In most cases, we utilized the built-in recovery time at the end of the route, and by shifting departure times, we were able to minimize significant gaps in the schedule. This approach allowed us to improve on most of the last-minute changes implemented in the spring while continuing to develop multiple service level scenarios through the summer and fall. Unfortunately, due to time constraints (working well past normal deadlines), full schedule re-writes were not possible and we continued to provide suboptimal headways on some routes across the system.

The shifting of schedules was implemented on the following routes in the September board: 7, 11, 25, 29, 32, 34, 51, 61, 63, 68 (gaps remain), 72 (southbound only), 74, 75, 79, 86, 88, 89 , 90, 98, 102 (northbound only), 105, 116, 124, 160 and 163.

We are currently working on the January boards. Most routes will return to pre-pandemic levels and scheduling anomalies resulting from cancelled crews and extra trippers will be eliminated.

Mark Mis
Head, Service Planning & Scheduling

When the January updates are announced, I will publish the details as usual. Stay tuned.

TTC Proposes Massive Fleet Plan

At its meeting on October 22, the TTC Board will consider a report setting out plans to purchase new buses, streetcars, subway trains and Wheel-Trans vans in coming years.

TTC Fleet Procurement Strategy and Plan

In an important departure from typical practice, the City is setting out its position including what can be achieved with already-committed City funding without waiting for confirmation of contributions from other governments. Both the provincial and federal governments will face voters sometime in the next few years, and this, in effect says “come to the table”.

The plan has many strong points although some important details are missing. Key to this plan is that it is a system plan, not a scheme for one tiny chunk of the network nor a flavour-of-the-day announcement from one politician.

Overview

The TTC proposes acquisition of hundreds of new and replacement vehicles over the coming years:

  • From 13 to 60 new streetcars from Bombardier to be delivered between 2023 and 2025.
  • Approximately 300 hybrid-electric buses for one or both of the two qualified suppliers to be delivered between 2022 and 2023.
  • Pending outcome of technical evaluation and product comparison work now underway, approximately 300 all-electric long-range buses in 2023 to 2025.
  • 70 Wheel-Trans buses for delivery in 2022 and 2023.
  • 80 subway trains to replace the existing fleet now used on Line 2 and to provide for future service improvement with ATC (automatic train control).

That list is only part of a larger scheme shown in the table below.

The “ask” for funding on these projects is based on the full quantity of vehicles (column 2 above) as opposed to what the TTC can achieve with only the City’s contribution (column 3).

A political problem for the TTC is that they are seeking funding for the ten year plan within the next few years even though some of the spending is in the latter part of the decade.

For example, the buses are unlikely to be contracted on one big purchase that would lock in a single supplier, and a new contract would be tendered two or three times during the decade. Similarly, the quantity of Wheel-Trans buses represents far more than one fleet replacement (as of June 30 there were about 280 WT buses). Part of this funding would not be required until late in the decade when the next purchases would be at end-of-life.

Commitments that far off are unlikely to be made by either the provincial or federal governments both of which would face at least one if not more elections in the meantime.

A further issue is that there are many more projects in the TTC’s long-range capital plan than the ones listed here, and there is no sense of relative priority for things like ongoing infrastructure maintenance. If the vehicles program soaks up all available funding, other projects could find that the cupboard is bare.

Missing from this report is an overview of the cash flow requirements for each project and the point at which money for each component must be secured. Projects with long timelines such as ATC installation need early commitment even though they would not finish until late in this decade or possibly longer. The same does not apply to the cyclic renewal of the bus fleets and some of the associated infrastructure.

TTC footnote 1: Estimated vehicle procurement quantities are based on Class 4 cost estimates. Given the need exceeds the funding currently available, TTC will seek to maximize the final number of vehicles to be procured through negotiation of contract unit pricing.

To support the electric vehicle purchases, the TTC together with Toronto Hydro and Ontario Power Generation (OPG) are working on plans for the charging infrastructure that will be required to move to a zero emissions fleet by 2040 in regular buses, Wheel-Trans and non-revenue vehicles.

The subway train order will likely grow because Metrolinx would piggy-back the needs of the Yonge North extension to Richmond Hill and the Scarborough extension to Sheppard for economies of scale and consistency of fleets on the two major rapid transit lines. However, the cost will be on Metrolinx’ account because these are now provincial projects. There is a danger that if future provincial funding is constrained, the provincial projects could elbow aside requests for local projects.

The committed and required funding amounts are set out below.

TTC footnotes:
1: Number of Vehicles reflects the current fleet plan as described under the Comments section of the report.
2: Estimated Total Costs includes the following: (1) vendor contract payments for vehicle design, production, delivery and commissioning of vehicles; and (2) delivery costs including procurement, project management, engineering, quality assurance, and project contingency
3. Total Estimated Cost has been revised from $5.84 billion (Class 5) to $6.17 billion (Class 4).

The City’s share is provided by the City Building Fund, a supplementary property tax introduced in the 2020 budget, together with funding that had been allocated to a planned rejuvenation of the Line 2 subway fleet for an additional decade of service. Now that those trains will be replaced, the money set aside to refresh the old fleet is available for this project.

City Building Fund Project$ millions
Bloor-Yonge Station Expansion$500
Line 1 Capacity Enhancement$1,490
Line 2 Capacity Enhancement$817
Line 2 Automatic Train Control $623
Other Critical Subway State of Good Repair (Note 1)$160
New Vehicles and eBus Charging Systems$1,140
Total City Building Fund$4,730
Note 1: These values do not exactly match numbers cited in the TTC report due to rounding.

The vehicle procurements are funded on the City side by a combination of CBF monies (see above) and the previous allocation for renovation of the Line 2 fleet of T1 trains.

Project$ millions
80 New Subway Trains$ 623
T1 Overhaul and Maintenance to 2030$ 74
Procurement of Buses$ 686
eBus Charging Infrastructure$ 64
Wheel-Trans Buses$ 22
New Streetcars$ 140
Total$1,609
Existing Approved Funding (T1 Life Extension)$ 474
City Building Fund$1,140
Total$1,614

Combining the $1.61 billion above with the Line 2 ATC funding brings the City’s total to about $2.2 billion. The TTC and City invite their partners at the provincial and federal levels to make up the difference of just under $4 billion between City allocations and the total required for this portion of the overall capital plan.

The City’s strategy is to start spending its $2.2 billion and hope that the other governments will come in for their share. There are elections at both levels that could provide some leverage, but there are also problems with Toronto’s appetite for capital compared to other parts of Ontario and Canada.

Continue reading

How Much Service Can The TTC Run?

A lot of ink (some real, but much virtual) has been spilled in past weeks on the topic of crowding on TTC buses. This was compounded by less-than sympathetic responses from various places, including the TTC inself, suggesting the speakers/writers have no sense of the real world that transit riders inhabit.

Although there has been a recent plateau, ridership has been building through the summer and photos of crowded buses now appear commonly on social media. This co-incides with a second wave of infections and a selective move back to “stage 2” protocols. Just when concerns about personal safety are rising, the TTC gives the impression of shrugging its shoulders. This is not a recipe for confidence in rebuilding transit ridership.

That said, I must confess that I fall midway between camps that cry out for vast increases in transit (possibly with lower fares) and those who wring their hands wondering how we will even pay for what operates today. Trying to take the middle road is not easy.

Here I will attempt to put various issues and options on the table. Whether those responsible for the TTC act on them is out of my hands.

How Full is Full?

During the early days of the Covid recovery period (remember only a few months ago when we thought this was all going away soon?), the TTC produced charts showing demand levels and crowding configurations for its vehicles.

The degree to which they might achieve these levels depended on the overall level of ridership. This chart shows what happens if the TTC attempted to provide distanced service even though demand was low. At a 30% level, the service would have to match pre-pandemic levels to leave enough room for passengers to spread out.

This is not feasible from either a budgetary or operational point of view. The cost is very high, and there is no headroom for growth beyond that 30 per cent line.

Another variation of this chart foresaw a combination of increased crowding and service as ridership grew. On a system-wide basis, we are now at Level 2 with overall ridership at around 40 per cent and this translates to a crowding standard of 25 per bus, about three quarters of a seated load (see chart above).

The problem, of course, is that there is a big difference between system-wide averages and the conditions from route to route, location to location and time of day.

The TTC reported that as of early September demand on the bus system had reached almost half of pre-covid levels. About one quarter of all trips were running at or above the Level 1 standard of 30 per cent capacity, and almost a tenth were running at or above the Level 2 standard of 50 per cent.

The oft-cited figure that 92 per cent of TTC trips are not crowded comes from this chart. What it really means is that only 8 per cent are over the Level 2 standard, but it does not mean that the balance fall at or below Level 1. This distinction has been lost in translation by those who seek to put the best possible spin on the situation.

A major source of confusion has been the question of what would trigger resumption of full service. The chart above shows that 50 per cent demand would require 100 per cent service to provide some degree of distancing, albeit at Level 2, not Level 1 standards.

This was taken by TTC management to mean 50 per cent overall, but as we well know the recovery on many bus routes outpaces the system average. Many bus routes are likely to be well beyond this level before the system as a whole hits that target.

At the September 24 TTC Board Meeting, management made this point, that bus ridership was growing faster and would have to be dealt with but there is little real progress on this so far. The TTC does run some unscheduled extra service (about which more later) and claims that its effect shows up in the slight downturn in early September in the proportion of trips cresting the 50 per cent capacity line.

Continue reading

The TTC 2021 Service Plan

The TTC Service Plan for 2021 is still in draft, but the TTC wants public feedback on their proposals. The deadline for comments is October 9, 2020.

See:

In earlier stages of public participation, the focus was on implementation of the RapidTO bus lanes, notably the one on Eglinton-Kingston-Morningside that has just been installed.

Now the TTC has added material about other proposals and there is a 24-minute video overview on the presentation page linked above.

Continue reading

TTC Covid Service Recovery Plans: Part II

This article continues comments written before the TTC Board meeting of September 24, 2020. The staff presentation had more information than in the report linked from the agenda.

The following chart was included in my first article on the TTC’s Covid response. The red and green lines in this chart marks the number of trips that accumulated more than a 30% and 50% load respectively. The 30% load has been the target for service through the summer, but this is challenging to achieve for a few reasons.

Demand on the bus network has recovered more quickly than on the rest of the system and now stands at almost 50%. This is not evenly distributed by route or time of day, and there will be trips that routinely face higher demand than can be handled.

However, on some routes, notably those that formerly had express services, the revised all-local service is well below the average of 85% of pre-covid service the TTC commonly cites. In a few cases, routes have only half of their previous service unless unscheduled extras are added to compensate.

TTC management note that both the green and red curves turned downward slightly in early September, and attribute this to the operation of more unscheduled buses than in earlier months thanks to operators who have been recalled from layoff.

The more severe challenge, however, is that there simply are not enough buses and operators even at full service to provide generous spacing with demand at 50% of pre-covid levels, let alone higher proportions.

More operators will be recalled in October 2020 and this will add to on-demand service for school travel, particularly in the midday which now will have peaks that did not previously exist as half-day attendees switch over from the morning to the afternoon panel.

The success of using demand responsive service will be seen in how these stats behave in coming weeks.

Continue reading

TTC Covid Service Recovery Plans

At its September 24, 2020 meeting, the TTC Board will consider a report from management on the current status of the “Restart and Recovery” program.

This report contains a variety of information including:

  • A review of the physical changes the TTC has been implementing and investigating to improve health and safety on its system.
  • Demand and service plans for the coming months.
  • Financial projections
  • A current survey of rider attitudes to travel on the TTC

This article deals only with the service plans. I will turn to the other components after the TTC meeting.

Conventional System Service

Ridership on the main part of the transit system (excluding Wheel-Trans which I will discuss later) has been growing slowly from the low point in April 2020.

The return of demand varies by mode and this is strongly influenced by the higher proportion of trips to downtown jobs and academic institutions that have been replaced by work-from-home arrangements.

The bus system is back at about half of its normal demand although the subway is at only about a third. The TTC does not plan to restore full service until the system as a whole reaches 50%, but is running extra unscheduled buses on routes where crowding is a problem. As I have previously reported, the TTC does not list where and when these extras are used, nor does it routinely publish crowding statistics.

A big problem with the 49% level for bus boardings compared to pre-covid conditions is that this is an average. It blends data from all routes, all times of day, and all parts of the city. There will be places and times of lower demand and places where routes are as crowded as pre-covid. This is borne out by photos posted on social media by riders.

Certain types of trips, indeed even the potential for some trips, will not return to the system until “normal” times especially for special events and entertainment. School travel at all levels will run below “normal” while classes are provided via distance learning for the safety of students.

This will dilute the returning numbers of workers who continue to use the TTC and bring down the overall percentages even though some routes will see demand well over the 50% level, but with less than 100% of the pre-covid service.

Overall crowding levels are shown in the chart below which is driven by Automatic Passenger Counter (APC) data. As ridership grows, the proportion of trips running above 30% or 50% of capacity (reference points for acceptable crowding in stages of the recovery plan) have also grown. Note that this includes the benefits of whatever unscheduled extra service the TTC has operated.

What is badly needed is a better public understanding of where crowding is a problem at the route and time of day level, not simply on an average basis across the system. Only in this context can the addition or redeployment of resources — drivers and buses, not to mention funding — make sense.

The TTC plans some service improvements in coming months:

  • Starting in September, the TTC is dispatching extra buses to deal with the rise in school trips as elementary and secondary schools re-open and some of their population returns as TTC riders.
  • In October, the TTC will make improvements on Eglinton East to take advantage of the recently installed bus lanes. This will include restoration of 905 Eglinton East and 986 Scarborough express services.
  • In November, more express services will return and local services will be improved by using some of the buses and drivers now operating on demand-based standby service.

The October changes will be accomplished by reallocating service from poor-performing routes, and it will be interesting to see which routes are on the chopping block. Indeed, it would be useful to see a full list of routes and service periods that are under threat of service cuts considering that full funding for 2021 is far from certain. For 2021, any new services will have to compete with existing routes for resources.

An important distinction in service design that will be with us for some time will be the spreading of the “peak” period, or more accurately a lower demand at the height of the peaks resulting in more uniform demand and service levels. This has been part of the schedule design for recent months when “trippers” that ordinarily would be in service for only a few hours in each peak operated for 6-7 hours/day with only a brief mid-day window between the “AM” and “PM” peak service levels.

Another issue related to core area demand will be the degree to which non-core oriented trips take a proportionately larger role on the TTC and how network and service designs react to this.

Travel between suburbs has always been more difficult than travel to and from downtown because so much service is oriented to feeding the subway network. Coming months provide a chance to address this problem with a stronger focus on the surface network and the areas it serves.

For 2021, various options are part of TTC planning:

  • Continuation of some demand-responsive service to allow quick adjustments as travel patterns evolve.
  • The potential effect of a “second wave” that could affect both ridership and the level of distancing required on transit vehicles.
  • Adjusting service patterns and network structure, as discussed above, to react to and support new patterns of demand.
  • Improve service reliability (a subject of several recent articles on this site).
  • Continue the RapidTO plan of creating more exclusive bus lanes.
  • Cross-border service and fare integration with systems in the 905.
  • Micro-transit and automated shuttle services (*).

(*) The last bullet is quite clearly a sop to the political forces at Queen’s Park where there are assumptions that large savings can be had by using self-driving mini-buses and/or some form of alternate transit to replace lightly used routes. The big problems on the TTC are on routes where these options are hopelessly inadequate, and yet there will be pressure to focus on them as some sort of magic bullet for transit budget problems.

Public consultation regarding the 2021 Service Plan is now in progress with the intent of reporting to the Board late in 2020. The big unknown for next year is the state of the TTC Budget and the possibility that the system will face arbitrary cutbacks to fit within a City and Provincial budget envelope.

Wheel-Trans

Although Wheel-Trans is a much smaller operation than the “conventional” system, ridership has been returning to it roughly in line with the rest of the surface system. Demand is expected to be close to 50% of pre-covid levels by the end of 2020.

As demand rises, the ability to distance on WT services will begin to be difficult and some ride-sharing will have to occur starting in late September.

WT ridership is a particular challenge for the TTC on three counts:

  • The ratio of a fully-loaded vehicle to one operating at covid-based distancing and capacity is much higher for WT than for regular TTC vehicles. This means that distancing problems kick in much sooner as demand grows on the WT network compared with the “conventional” one.
  • The TTC has attempted to shift WT riders to its “family of services” model where trips are taken part by a WT ride and partly on the conventional system. However, crowding issues, already challenging for some WT users under “normal” circumstances, are more of a problem with limited capacity on buses.
  • The WT user community has a higher proportion of riders with risk factors for covid exposure and their concerns about behaviour by other riders are heightened.

Coming Soon …

I will update this article with additional information, if any, from the TTC Board meeting.

TTC Recalls More Laid Off Workers

The TTC has announced that it will recall 132 more workers effective October 2, 2020. Of the 450 who were laid off in April, 150 were recalled in September. This leaves 168 still awaiting recall.

The TTC intends to use these drivers for “demand responsive service” with buses (and a few streetcars) that are dispatched as needed on routes where scheduled service cannot handle demand. The return of students to in-school classes adds to system load, and more service is needed.

The layoffs are a temporary measure with all operators to be recalled when the TTC reaches 50 per cent of pre-pandemic ridership levels on all vehicle modes (before the pandemic the TTC was carrying 1.7 million rides on a typical weekday day). At the lowest point of the lockdown, the TTC was moving roughly 15-20 per cent of pre-pandemic ridership. Currently, the TTC is seeing daily ridership in the 35-40 per cent range (or more than 630,000 customers each weekday).

TTC Press Release Sept. 17, 2020

This statement clarifies a confusing element in past announcements where it was unclear whether the 50% criterion would apply to selected portions of the network where demand was strong or to the network as a whole. Because demand into the core area, particularly on the subway, is much lighter than usual, the bus system will have to reach a level well over 50% of pre-covid riding before overall system demand will trigger full service restoration.

In previous articles, I have been tracking the level of scheduled service on many routes (the final article will appear soon) as well as the irregularity of headways (the time between buses) both in the schedules and in actual operation.

What is missing is any report of where the standby buses the TTC operates are used. With an increasing proportion of service provided by unscheduled vehicles, knowing how they are allocated is important as well as how this correlates to on-board crowding.

There is roughly a two month lead time for changes in the scheduled service, and the next two sets of schedules will come into effect at Thanksgiving and in late November. Details of proposed changes are not yet public, but at this point it is fair to assume that the October schedules are locked down and the November changes are well along in draft state. These lead times are required to give time for workforce and fleet planning, and to organize sign-ups for crews.

Using standby or “run as directed” service allows the TTC to have a pool of vehicles and drivers that are not pre-allocated to specific times and locations, and this gets around the lead time problem for scheduled service. The downside is that systems that depend on schedules including public information and vehicle tracking don’t “see” the extra buses.

One member of the TTC Board, Commissioner/Councillor Shelley Carroll, plans to move a motion at the September 24, 2020 Board Meeting that full service be restored. The earliest that is likely to happen would be year-end on a formal basis, and if all remaining drivers are recalled they would likely be used for standby buses until then.

The agenda for that meeting will be published soon, and we will see whether management has any specific proposals to address this.

Still at issue is the question of service standards both to deal with crowding levels and for “poor performing” routes where the Ontario Government wants transit systems to consider alternative means to provide service.

Crowding and Service on the TTC: September 2020

A regular feature of Twitter and other platforms are complaints about TTC crowding and photos of packed vehicles and stations. Far be it for me to quote BlogTO as a source, but Becky Robertson put together a collection of Tweets on this subject in an article published on September 9. I have included the top of it here to illustrate a few points.

This view is at St. Clair West Station on the northbound platform seen from the mezzanine level. The original post is timestamped 5:02 pm on September 8, 2020.

This stop is not usually this busy, and what has almost certainly happened is that a train short-turned into the pocket track north of the station and dumped its passengers.

Looking at the service alerts, there was a fire that resulted in suspended service north of this point. The original alert was issued at 4:30 pm, and an all clear went out at 4:56. However, there was a power failure in the same area for which the notice went out at 5:11 and the all clear at 5:20.

Crowded platforms, not to mention crowded shuttle buses, are going to be a fact of life in this type of circumstance, but it would be too easy to let the TTC off the hook because this was an emergency. They happen every day, and not just on the subway. Alerts with unspecified “operational problems” or “mechanical difficulties” are common, but the details are not routinely available, especially for surface routes. (Archived delay logs are available through the City’s Open Data Portal, but the surface route logs do not contain the same extensive set of explanations for delays as those for rapid transit routes.)

Line 1 Yonge-University is operating at a level of of 17.1 trains/hour (210 seconds between trains) compared to the pre-covid PM peak level of of 23.1 trains/hour. The pre-covid AM peak service was 25.5 trains/hour between Glencairn and Finch with half of these running north to Vaughan.

Line 2 Bloor-Danforth operates at a level of 16 trains/hour (225 seconds between trains) compared to the pre-covid peak level of 23.5 and 23.8 trains/hour (AM and PM respectively).

A large chunk of the subway’s capacity (based on the TTC Service Standards) is the substantial room for standees on subway trains. Proportionately this is higher for the subway than for surface modes with peak:offpeak ratios of 2 for the subway, 1.76 for the new streetcars, and 1.46 for a standard low-floor 12-metre bus.

During the early covid days, the standards were much lower with the intent of providing social distancing, but as demand builds up, this is less and less likely. Seats that are marked off for spacing are routinely occupied by passengers when routes get busy.

A previous article on this site TTC Preps For Covid Recovery includes diagrams showing the various levels of crowding the TTC is aiming for, at least on paper.

The basic problem with target crowding levels is that they are averages, and reality can be quite different. This is not simply a question of emergencies when they occur, but of irregularities in service.

In previous articles (Intro, Parts I, II and III), I reviewed the operation of several routes in the east, north and west of the bus network. (Part IV is in preparation and will deal with a few major east-west routes missed earlier as well as some smaller routes in the central city.)

There are common threads across the system:

  • Service is erratic on all routes although the degree varies by time of day and location.
  • Service is scheduled to be erratic in many cases either because branching routes do not have headways that blend, or because extra “tripper” service operates at a different headway from the basic route.

The TTC is operating “run as directed” buses (and a few streetcars) to help out where routes are crowded, but these are not tracked because they are not part of the schedule, and they run under a different internal route number than the line they actually serve. There is no standing report of how these vehicles are used, or whether there are standing assignments to specific routes and times to make up for schedule problems.

Irregularities in the schedule could be dealt with on an ad hoc basis although this causes problems for the TTC’s own service reliability metric which is based on “on time performance”. The buses cannot be both evenly spaced and on time.

The Service Standards provide leeway both for being on time, and for operating at a reliable headway (the time between buses), but on many routes these combine to make bunched service fit within the acceptable range.

“On Time” is measured only at terminals, and a vehicle is expected to be no more than one minute early or five minutes late. The goal is for 90% of departures and 60% of arrivals to be “on time”.

Reliability has a different metric depending on the headway:

  • For service scheduled at more than 10 minute intervals, the on time rule applies because riders expect vehicles to show up when they are scheduled.
  • For headways between 5 and 10 minutes, the metric is that the deviation from scheduled spacing is no more than 50% and that this is achieved 60% of the time. On an 8 minute headway, a range from 4 to 12 minutes is acceptable.
  • For headways under 5 minutes, the metric is that the deviation from scheduled spacing is no more than 75% and that this is achieved 60% of the time. On a 4 minute headway, a range from 1 to 7 minutes is acceptable.

These are all-day averages, and they leave considerable leeway for service to “meet” the standards while in practice being a complete mess. For starters, 4 out of 10 trips can lie outside of the standards, and the leeway for trips within standards is wide enough that bunched service qualifies as “ok”.

To call these “standards” is something of a joke because they say, in effect, “we’re not going to try too hard to provide reliable service, and we will do that less than 2/3 of the time”.

These were approved by the TTC Board, but there was no detailed discussion of the implication of the standards for the actual quality of service on the street. Management produces only superficial rolled-up stats for service quality and complaints about crowding, even in pre-covid times, were more likely to meet with “we have no budget/buses/drivers” as a stock response.

During the covid era, even vehicle spacing, and hence even vehicle loading, is even more important than ever. We do not know what the vehicle-to-vehicle crowding situation is because the TTC does not publish any breakdown of this information, only averages.

A fundamental issue with uneven headways is that the bus carrying the largest gap will probably have the most passengers. Instead of having three buses each with 30 riders, one could see 50 on the first bus, 30 on the second and 10 on the third. Same riders. Same average load. But the “average rider” sees a crowded bus (over half of the 90 are on the first, full bus), while only 10 of the 90 lounge in the comfort of a nearly-empty vehicle.

This problem is intimately linked with the question of whether there are enough buses on the road because if the TTC only looks at averages (and even worse if they do so over many hours), they will completely miss both spikes in demand and crowding problems from irregular headways. Indeed, we could well hear a familiar refrain of “we are monitoring service and everything is running within Board-approved standards”.

It is not enough for the TTC to say “we have standby buses we use to address problems”. They should report on where, when and how these vehicles are assigned with a view to integrating them into scheduled service.

A basic principle of transit operational planning is that the cheapest “new” capacity comes from properly managed and spaced service. Riders on many TTC routes suffer from unreliable service and have done so for years.

Fixing this is a job for everyone from the top to bottom of the organization: from a Board that approves standards without understanding how lax an operation these actually permit, through management who prefer easy metrics that make them look good, to supervisors who need to actively manage service. Finally, there are the drivers who are mainly very good, but some of them have only a passing sense of the importance of service reliability. Running “hot” to get a long recovery time at the terminal should not be an accepted practice.

If the TTC’s goal is to put off full service restoration as long as possible to reduce subsidy requirements, then it is their duty to make the best of the service they have on the street.

The scheme of holding off until ridership gets back to 50% of pre-covid levels sounds practical in theory, but it misses the basics about where the recovery is taking place. The subway will be the last to hit this target, but bus routes should not have to wait for better service.

As I write this, plans for fall 2020 and for the 2021 budget year have not been announced. The province is nibbling around the edges with talk of Microtransit even though this will not address the TTC’s fundamental problems because so little of the network is an applicable target for this type of operation.

Politicians love to avoid hard questions. We can expect to hear a lot of simplistic slogans about efficiency and belt-tightening. The net effect will not be a careful review, simply an edict to save some arbitrary amount. If anything, this will work counter to making transit more attractive especially if the problems of service reliability are not addressed.

The TTC is a good system, at least by North American standards, and it is getting more financial support than most. But the gap between its own view of service quality and what riders see is too wide, and this needs to be fixed as an integral part of restoring the TTC’s role in Toronto’s transportation network.

Postscript: A Message For People Documenting Crowding Problems

Take pictures and post them if you can. This is the only way we are going to hold the TTC’s feet to the fire. But be sure to include the date, time, route and location, and any info on service quality such as how long you waited for the bus.

TTC Begins To Recall Laid-Off Drivers

On August 27, the TTC announced that it will begin to recall workers who were laid off in April due to a dramatic drop in demand on the transit system.

450 drivers were laid off, and of these 150 will be recalled to provide extra resources for the demand school opening will bring in September.

TTC ridership had dropped to between 15 and 20 per cent of normal levels in the spring, but now sits in the 35 to 40 per cent range across the system. From previous announcements, we know that growth is strongest on the bus system, and crowding problems there have been reported often by riders even before any school traffic returns.

The TTC has not yet announced where service will be added as this depends on plans by school boards that have not yet been finalized.

As ridership continues to grow, more drivers will be recalled and the intent is to have all employees back to work when the system reaches 50 per cent of its pre-pandemic demand.

The TTC has enough buses to operate full service, and with the change in peak demand to a flatter service design, the fleet can go further to provide capacity than it did before. For off-peak service, the question is less whether TTC has the vehicles, but whether they have the drivers and budget headroom to use them.

Many routes currently operate with “trippers” that under normal circumstances would only run a few hours a day in the peaks, but now operate roughly seven hours at a time roughly between 6 AM and 1 PM, and again from 3 PM to 10 PM. The marginal cost of running longer “peak” service is actually less than it might seem because drivers spend more of their day driving on routes rather than to and from garages, and premiums associated with longer split shifts do not apply.

The challenge for 2020 and especially 2021 will be whether the funding level from various governments will allow the TTC to stay ahead of growing demand, and for how long a comparatively uncrowded transit experience will be possible.