A Big Announcement, or a Transit Three Card Monte?

On October 16, the governments at Queen’s Park and Toronto City Hall announced a deal to sort out competing transit plans for the city. The current provincial priority projects are the Ontario Line (Don Mills/Eglinton to Exhibition), Scarborough Subway extension from Kennedy Station to Sheppard, Yonge Subway extension from Finch to Richmond Hill, and the Eglinton West LRT extension from Mount Dennis to Renforth.

The main City of Toronto report will be discussed at Executive Committee on October 23, and then at Council on October 29-30.

This article reviews that report with reference to a few parts of its many attachments. I will turn to the technical attachments in a second article. To focus material on each subject for readers, I have grouped related items together or re-sequenced things for emphasis. There are extensive quotations of key material so that readers hear not just my “voice” but that of the report’s authors.

Despite the importance politicians at both levels place on the proposals, the fundamental problem remains that many of the details are cloudy, to be kind. Specifically:

  • The City of Toronto retains ownership of the existing transit system avoiding a complex realignment of responsibilities and governance, but with this comes total responsibility for funding the ongoing state of good repair.
  • A large gap remains between the amount of funding needed to maintain and expand Toronto’s transit system relative to the amounts actually available and committed in budgets at various levels of government.
  • Ontario will build four key projects substantially with its own money, but continued support for transit beyond this is uncertain.
  • Toronto will redirect funding originally earmarked for its share of the key projects to other priorities, notably the TTC’s repair backlog. However, much of that “funding” does not exist as allocations in existing budgets and new money is required from Toronto to pay its share.
  • Cost estimates for the key projects are based on preliminary estimates that could change substantially as the design process unfolds. These estimates are in 2019 dollars and make no provision for inflation. The reports are silent on how the proportion of total spending by each contributor might change over the decade or more of construction.
  • A substantial total of project costs will be born by private sector partners through a “P3” financing mechanism. These arrangement will require future payments for what will be, in effect, a capital lease, but the mechanism for funding this from three levels of government is unclear. The reports are silent on the split between short term borrowing to pay for construction as opposed to long term payments to the P3 financier.
  • Project details as they are known today will change in response to design work and the need to keep costs within the projected level. This will affect alignments and stations, and what we think we are buying could be quite different from what we actually get.

The challenge in all of this is, as always, the question of money. We can watch the hands of politicians and managers at all levels as they shuffle cards on the table. We hope to “find the Queen”, to win in the subway sweeps rather than being taken for suckers who will cheer any plan, but lose every game. It is far from clear whether the proposal is a “good deal” for Toronto, and there are huge future transit costs that are barely addressed.

The whole exercise is a political deal to bring peace, comparatively speaking, to the transit file which was needlessly fouled by Doug Ford’s insistence that he knows more about transit in Toronto than anyone else. Does Toronto take this as its last best chance to preserve some semblance of control over its transit future, or do we keep fighting for a better deal?

There are a lot of holes in this plan and severe implications both for the City’s finances and the future of Toronto’s transit system. Many questions need to be asked and answered, even if the result will be a whole new plan after provincial and municipal elections in 2022.

Continue reading

How Reliable Are TTC Statistics?

Ben Spurr in the Toronto Star published an article on September 30 about the mis-reporting of vehicle reliability for the fleet of Bombardier Flexity streetcars.

In brief, the TTC reports defects for the new cars on a different basis than for the old ones (the CLRVs and the recently retired ALRVs), and this has two effects:

  • The reliability of the old cars looks worse by comparison to the new ones, and this supports the argument that the old cars should be retired as soon as possible.
  • The new cars have recently crested the performance specification from the Bombardier contract, but this is based on the way the failure rate is calculated.

The September 2019 CEO’s Report contains Mean Distance Between Failure (MDBF) charts for both types of streetcars still in active service. August 2019 saw the new fleet’s reliability go about the 35,000 km MDBF, and CEO Rick Leary reported at the TTC Board meeting of September 24 that the current number was running above 50,000 km.

By contrast, the CLRVs have failed roughly every 4,000 to 6,000 km for much of the past three years with problems more common during the cold months.

However, according to Spurr’s article, the basis of calculation is different for the two fleets. In the case of the new cars, only failures chargeable against Bombardier’s contracted reliability level are counted while for the old cars any failure counts. This makes a big difference when one considers how many of the in service failures were not included in the calculation for the new fleet.

Spurr writes:

The vehicle contract the TTC and Bombardier signed in 2009 set a MDBF target of 35,000 km. The cars were supposed to reach that figure by the time the 60th vehicle was delivered. That car arrived in January 2018, but the new fleet failed to hit the target then or in subsequent months.

That changed this summer. As Bombardier edged closer to completing its delivery of the 204-car fleet, and the TTC weighed the option of placing an order for additional streetcars with the company, the publicly reported reliability figures shot up.

They showed the cars had an MDBF of 36,500 km in July, and 51,500 km in August, the best the fleet has recorded since the early days of the order. CEO Leary cited that most recent figure at last Tuesday’s meeting as evidence the cars are “performing exceptionally well.”

However, over the same period the unpublished reliability figures didn’t improve. The “legacy” numbers showed an MDBF of just 16,400 km in August, which while much better than the early months of the year, was virtually unchanged from the mark set in May.

The unpublished “legacy” figures are consistently significantly worse than those the contractual numbers.

He goes on to write:

Internal TTC documents reviewed by the Star show that in [August] the new streetcars experienced dozens of delays related to faulty brakes, malfunctioning doors, broken HVAC units, and short circuit warnings. The agency tabulated 43 significant delays during that period, but only 15 were deemed Bombardier’s responsibility and included in the version of the stats that are made public.

That lower delay figure led to the contractual number of the cars running more than 51,500 km without a failure.

Readers can judge for themselves the type of delay that is omitted from the TTC’s reliability numbers from the following table which is compiled from TTC delay reports.

TTC_201908_LFLRV_DelaySummary

There are 56 items in the list and several patterns are immediately obvious:

  • Problems with the power collection system are common including pantograph failures, lost trolley poles or defective shoes, and dewirements snagging poles and/or damaging the overhead.
  • Brake problems
  • Mobility ramp problems
  • Failures early in the morning on cars that are probably just entering service

Many of the problems have nothing to do with Bombardier reliability stats and are not included in the calculation included in the CEO’s Report. If they were, the numbers would not look anywhere near as good.

Something that is evident in reported reliability stats is that there can be large variations in MDBF values from month to month. The TTC does not make huge changes in the mileage operated by its fleet each month, and so the large swings must be due to a relatively low number of incidents. For example, if there were typically 100 incidents per month and this swings up or down by 10%, then the MDBF would not change much. However, if there were typically only 20 incidents per month, a small change in the month-to-month numbers would produce a big swing in the MDBF. This is evident in the Flexity reliability values and in those cited for the subway fleet, notably the newer TR trains on Lines 1 and 4.

Even if all types of failure were counted, the service delay it causes must be five minutes or more. This is a standard adopted from the NOVA group of rapid transit operators and really is more appropriate for rapid transit lines.

An important distinction is that vehicles that run in trains have the capability of “getting home” even if one unit is disabled under most circumstances, and reliability stats for this type of operation will be higher than for single vehicles on a streetcar system. Also, rapid transit lines operate at higher average speeds, and failures that are affected more by hours of service than by mileage are spread over a larger distance operated. This is quite evident in TTC subway stats where the MDBF is much higher than for streetcars.

By contrast, it is difficult to imagine how a bus breakdown can cause a significant service delay except in comparatively rare circumstances, and the five minute delay screen for a chargeable delay makes no sense for the bus fleet.

The question of just how reliable various vehicle types might be is part of a larger issue with the selective, and possibly misleading reporting of statistics by TTC management.

Delays to service, especially on the subway, are caused not just by equipment failures, but by a raft of other subsystems and problems such as signals, track, power supply, fires, passenger assistance alarms and track level incidents. The TTC tracks the various types of delay, but reports on them only rarely in public. This means that sources of service delay that might be under the TTC’s control are not tracked in a report that is routinely seen by the TTC Board, nor is there any tracking of the effects of preventative maintenance or capital works to reduce this type of delay. One obvious example is the new Automatic Train Control system which is now operating on about half of Line 1 YUS, but we know nothing of service reliability on that section, Vaughan to St. Patrick, compared to the old signals still in use from St. Patrick to Finch.

Bus reliability is reported in the aggregate for a fleet that ranges in age from brand new to over twelve years old. The TTC used to keep buses for at least 18 years, but now chooses to replace rather than rebuild old vehicles. Retiring a large tranche of 12-20 year old buses in recent years has had three effects:

  • The average age of the fleet is now quite low, and it will continue to drop. Half of the fleet is less than five years old, but as the “bulge” of new buses ages, the fleet reliability will fall.
  • With many new buses coming on stream, the TTC can keep old buses in service and maintain a high ratio of spares to service requirements. The situation is very different for the streetcar fleet where with the retirement of old cars, the fleet is too small to provide service on all routes with an adequate number of spares for maintenance.
  • The large order of buses soaked up the then-available funding for transit infrastructure as it was the only way Toronto could spend its allocation within the short timeframe dictated by the federal government.

For reasons best known to the TTC, the chart above is clipped at 20,000 km rather than showing the actual variation, and this has been the case since early 2018. It is unclear whether the actual numbers are rising or falling over the past two years. Moreover, the values average the reliability for the entire fleet rather than showing subsets such as diesel and hybrid buses, or buses of varying ages or manufacture. This type of breakdown is vital in understanding fleet planning, not to mention tracking the benefits (or not) of technology changes such as the move to an all-electric fleet which is only just beginning.

The TTC fleet of buses is much larger than its requirement for service. In total, as of the September 2019 vehicle list (taken from the Scheduled Service Summary, last page) shows a total of 2,076 vehicles as compared to a peak service requirement of 1,626 (p. 63 of the same document). This is a generous spare factor of over 27%, or one spare bus for every four in service. It is easy to get very good performance from your fleet with such a high ratio, but this also means that, in effect, the TTC operates one garage worth of spares for every four garages worth of regular service. This is far higher than the target spare ratio for rail vehicles.

In a separate post, I will turn to the question of service reliability, scheduling and the way in which service quality is presented by management to the TTC Board. This is another area where there has been a lot of work to make the numbers “look good” but with detrimental effects on the system.

Questions for the TTC:

I have posed a series of questions to the TTC and await answers from them. This article will be updated when they reply.

1. Has the Flexity reliability number always been quoted on the basis of failures chargeable to Bombardier, or was there a change in the methodology somewhere along the way? To put it another way, was there a change in what counted as a failure that created an artificial improvement in the reported numbers?

2. What is the situation with subway delays and MDBF numbers? Are all failures counted (at least those producing a 5 minute or greater delay) or only those considered to be the manufacturer’s fault? Is the calculation done the same way for the TR and T1 fleets?

3. The NOVA metric which the TTC uses is based on the idea of a failure that causes a delay to service. This only makes sense in the case of rail modes where a car/train failure can block the line. For buses, only a rare and well-positioned failure could actually block service. How is a chargeable failure calculated for the bus fleet?

4. Are numbers available for subsets of the bus fleet (e.g. all buses from the same order, age, technology) so that reliability figures can be compared as they have been with rail modes?

5. The CEO’s Report includes only stats for delays caused by vehicle faults, not from other sources such as infrastructure failure. Why is this info not also tracked in the report so that the effects are clear on a proportionate basis? In particular, there is no tracking of signal failures on various parts of the subway with older and newer technologies.

Summary:

TTC management should report vehicle reliability numbers on a consistent basis for all types of vehicles.

The calculation of service interruption rates should reflect what riders experience, not simply numbers to establish contract performance for suppliers or to artificially enhance the reported performance of some vehicle types.

The reliability statistics for the bus fleet should be broken down by major vehicle groups (manufacturer, propulsion technology, age) to allow meaningful comparisons and to ensure tracking of maintenance/reliability as parts of the fleet age.

The very large spare ratio for the bus fleet should be reviewed to determine whether this size of fleet is actually required, or if more service could be operated if only the TTC would budget for the cost of its operation.

Delays caused by infrastructure issues and other interruptions should be tracked and reported so that their effect on service quality can be seen in comparison to vehicle related problems.

TTC Service Changes Effective Sunday, October 13, 2019

The TTC will make many changes to its services on Thanksgiving weekend, Sunday, October 13, 2019.

These include:

  • The formal introduction of new Flexity streetcars on the 506 Carlton route on a scheduled basis.
  • Implementation of the route changes flowing from the Junction Area Study.
  • Construction at Runnymede Station for the Easier Access Program trigger some route changes and interlining.
  • Many “service reliability improvements” intended to make scheduled trips better match actual operating conditions.

Of particular concern in this round of changes is that a policy adopted by the TTC Board in June 2019 allows schedule changes intended to improve reliability to take precedence over loading standards. Several routes where service is nominally “improved” will also have crowding beyond the Board approved standards. This is caused by a shortfall in the budget which does not provide enough resources, the TTC claims, to operate service at a level that meets the standards. There are no numeric values given and so the degree of overcrowding is not known. The affected routes and time periods are shown in the chart below.

On a related note, a few routes which are now part of the Ten Minute Network will slip beyond that target headway.

The intent is to correct these situations either through the 2020 budget or by reallocation of service from other routes.

(The two charts above are included in the PDF containing all of the schedule changes at the end of this article.)

In response to my query about these practices, Mark Mis, Manager of Service Planning replied:

Q1:

SM: … it was my understanding that the table of crowded routes was going to be published regularly, but nothing has appeared after the first iteration … [earlier this year].

MM: The TTC will be reporting on crowded periods of operation in the CEO Report later in the fall when we have a complete board period of service reliability data. We think it is important to report on both crowding and reliability at the same time.

Q2:

SM: … does this footnote mean that the affected services are being scheduled at over the Service Standards for loading?

MM: The footnote is an internal note in the Board Period memo for record keeping purposes. It indicates a period of operation that is scheduled over the crowding standard because of a change in running time as opposed to demand. This strategy to address service reliability was approved by the Board on June 12, 2019 in the Improvements to Surface Transit Schedules report. This is important to note when you take the response to Q1 into account.

The June 2019 report included the following observations:

An effective transit schedule is comprised of two key components:accurate estimates of demand and cycle time—the time for an operator/vehicle to complete a round-trip. If the cycle time is insufficient, a transit schedule can not be delivered as planned resulting in customers experiencing unreliable service and higher levels of crowding than intended.

Customers value consistent and dependable schedules because it improves their trip planning, reduces wait times and reduces trip durations. This was demonstrated by the King Street pilot. A key outcome of it was more consistent and less variable service which resulted in tremendous increases in ridership. [p. 3]

The TTC is leveraging new technology and applying a new approach to preparing reliable schedules which includes setting cycle time at the 95th percentile of observed travel time along a route in a given period of operation. This is meant to ensure that operators have sufficient time to deliver service as planned and advertised during most of the operating period. [p. 5]

The TTC 2019 Operating Budget includes some funding to improve service reliability and capacity. It is not possible to bring all schedules that actually operate outside of the tolerances of service standards in line with them within existing funding. The new schedules will reflect actual operating conditions and will result in more reliable service for customers. [p. 7]

On several routes that already have new schedules, a common observation is that vehicles cluster at terminals because they generally arrive early and/or have long scheduled layovers. Despite this, the vehicles do not manage to depart on a regular headway (time between vehicles) nor do they achieve the “on time performance” goal. The TTC is quite generous with a six-minute window in which a departure is considered “on time”. Service becomes progressively more bunched as cars and buses proceed along their route, and there is no measurement of service reliability other than at terminals.

This begs the question of rejigging the schedules in the first place when service can be so irregular. In almost all cases, the “improved” schedules offer less frequent service because existing vehicle spacings are stretched to give longer trip times. Longer trips may reduce short turns, but this is only one of several possible measures of service quality. See Zero Short Turns Does Not Equal Better Service.

Streetcar System Changes

The 506 Carlton route is now officially a low-floor streetcar service with Flexitys scheduled at all hours. Bus and CLRV trippers will supplement peak services. Further details are in a previous article.

The 505 Dundas route remains a bus operation, but the schedules will change to remove extra running time added for water main construction in the central part of the route. This may relieve the worst of the bunching of Dundas buses now seen at terminals. Riders on the 505 will see improved service, or at least the bunches of buses will come slightly more often than they do today.

Conversion of 505 Dundas to Flexity streetcar operation is planned for the winter/spring after conversion of the overhead on that route for pantograph operation.

Routes 510 Spadina and 509 Harbourfront will get new schedules with blended services during most off-peak periods at Union Station. Scheduled headways will be the same on each route with the intent that cars will alternate on a regular, reliable basis. The downside to this scheme is that service on 510 Spadina will run less frequently, and layover times, which already cause queuing problems at the terminals are increased. 509 Harbourfront will see less frequent service as a regular seasonal change.

Overnight service on 306 Carlton will change to every 20 minutes, and on 310 Spadina to every 15 minutes to push a few more cars out of the crowded carhouses. It is unclear how long this practice will actually last once the CLRV fleet is retired because capacity will be lost at Russell Carhouse for construction starting in 2020.

Bus Network Changes

Construction of the new mouth of the Don River in the Portlands has progressed to the point that on October 30, 2019, Commissioners Street will be closed between Cherry and Saulter for excavation of the new river. A new Commissioners Street road bridge will cross the river, but at this point there is no money in the Portlands project for a streetcar bridge although provision exists in the plans. This would allow the eventual connection of an extended Cherry Street service east to Leslie Barns via Commissioners.

The route changes from the Junction Area Study will go into effect concurrently with route modifications to accommodate elevator construction at Runnymede Station.

The new route structure, in brief:

  • Route 30 High Park will operate from High Park Station to Runnymede Loop taking over the route number formerly assigned to Lambton.
  • Route 40 Junction – Dundas West will operate from Dundas West Station to Kipling Station with a short turn service to Jane Street.
  • Route 71 Runnymede will keep only its branch to Industry Street in Mount Dennis.
  • Route 189 Stockyards will provide service from the subway at High Park and Keele Stations to St. Clair and Scarlett Road.

Due to construction, no buses will be able to loop at Runnymede Station triggering the following changes:

  • Routes 71 Runnymede and 77 Swansea will be interlined.
  • Route 79 Scarlett Road will loop via Annette, High Park and Bloor Street as shown below.

Seasonal service will end on:

  • Weekend evening service to the Zoo on 85 Sheppard East and 86 Scarborough
  • 121 Fort York – Esplanade to Ontario Place and Cherry Beach
  • 175 Bluffers Park

The details of all changes are in the file linked below.

2019.10.13_Service_Changes

Metrolinx Board Meeting September 12, 2019

The Metrolinx Board met on September 12, 2019, but there was not much of substance on the public agenda. Presentations consisted as much of rehashing old news (including he oft-announced service improvements on GO Transit), but almost no substantive policy discussions.

Links here to the Agenda and Video for the meeting.

Ontario Line Initial Business Case (Video at 23:35)

This report was presented by a team of four from Metrolinx:

  • Mathieu Goetzke, Chief Planning Officer (Acting)
  • Malcolm Mackay, Project Sponsor for the Ontario Line, and until 13 days ago an engineer at TTC with 13 years experience, now transferred to Metrolinx. His previous major TTC projects included the Relief Line and the Union Station second platform.
  • Duncan Law, Head Sponsor for the subway program
  • Becca Nagorsky, Director for Project Planning

As I have already reviewed the Initial Business Case (IBC) in some detail, I will not dwell on that here, but will flag comments during the presentation and discussion of particular interest.

There were two threads on which nobody remarked, but which were significant given the way that the Ontario Line was announced:

  • The project details are far less advanced than the bluster of the original announcement might have indicated, and Metrolinx acknowledges that significant technical challenges remain for the design.
  • Language implying the general incompetence of the TTC to build a “modern” rapid transit line is much reduced if not eliminated from the discussion.

These are welcome changes, but we now face the need to build something because the Premier announced it.

Mathieu Goetzke introduced the presentation saying that although the Initial Business Case (IBC) was published in July, they are now going into more details. The Preliminary Business Case (PBC), the next step in the process, must resolve some issues and Metrolinx needs to “activate all possible levels” to address project costs. (See video at 27:00.)

Duncan Law continued in this vein saying that it was important to recognize that the IBC is an early stage of the project. Both the Relief and Ontario Lines are underpinned by the recognition that more capacity is needed. With roughly 50 per cent of the Ontario Line being at or above grade, there would be cost savings. Moreover, with the line separate from the existing subway system, there is an opportunity for technology change that would not otherwise be possible. There is a big difference in this outlook from saying that the TTC uses out of date technology.

Becca Nagorsky echoed the remark that the IBC is a first phase saying that its purpose is to define the project’s goals that must be preserved through the life cycle of more detailed design. She continued what has become a standard Metrolinx comparison of the original Relief Line project to the Ontario Line considering only the Relief Line South. This works from the assumption that the Ontario Line’s technology change will save so much money that the Relief Line North, as a conventional subway, does not even come into the discussion. This precludes the possibility that future design work might discover that the Ontario Line could be more expensive than originally thought, but by then the idea of going back to a subway project will be difficult, if only for political reasons. (There are parallels with the now-entrenched concept of a Scarborough Subway.)

The capital cost projections include anticipated savings due to “risk transfer” to a private sector partner in a Design-Build-Finance-Maintain (DBFM) P3 arrangement, but as with so many P3 schemes, there is little explanation of how exactly this is achieved. In particular, there is always the possibility that circumstances and designs will change, and the private sector “partner” will not assume this risk.

Note that none of the Benefit-Cost ratios exceeds “1” indicating that any version of the project does not produce a “profit” within the Metrolinx benefit-cost methodology. This came up in discussion a bit later (see below). The important issue here is that a large project such as the Ontario or Relief Line has benefits (and possibly costs) that the methodology does not capture notably the value of increasing resiliency in the rapid transit network by provision of alternate routes, and the enabling of projects such as the Richmond Hill extension that would otherwise overload the system.

A new addition to the discussion is a map showing the supposed benefit of the Ontario Line to residents of low income areas. It is no surprise that the OL (and the full RL to Eglinton that preceded it) benefit low income areas such as Thorncliffe Park and Moss Park, but there is a bizarre problem with the map which shows a reduced access to low-income jobs for residents between the Spadina Subway and the Barrie GO corridor south of Eglinton, far from the Ontario Line, not to mention Flemingdon Park north of Eglinton. There is something wrong here with the underlying model, but nobody at the meeting picked up on this.

Duncan Law bravely observed that “we are trained to challenge how things have been done” and this will lead to cost avoidance in the design (video at 38:00). He noted that early works on the route would be accelerated, although this is a tactic already in place (after much political fighting with Mayor Tory who eventually embraced it) for the Relief Line. At this point, Metrolinx is considering what their options for the OL design are before they take them to the public for comment, and they are still at an early stage.

In other words, they have a line on a map, but even that may change, and they fear alarming the locals with designs that are not yet definitive.

Malcolm Mackay spoke about the early works and the importance of co-ordination with large programs in the corridor to “leverage” contracts and consultants for other projects such as delivery of the (now) six track structure from East Harbour to Gerrard (four GO tracks plus two for the OL). There is also the potential role of Transit Oriented Development (TOD) works where some OL work could be combined with private development. However, it is not clear whether the likely construction timeframes of East Harbour and other projects would mesh with the schedule for construction of the OL.

In a marvellous piece of bafflegab, the presentation notes:

To demonstrate visible progress and to de-risk the schedule, a progressive works program is being contemplated with a ground breaking target of 2020 – 2021. [p 20 of the PDF]

What this means is that if Metrolinx actually undertakes some work soon, there will be political benefits of “progress” (shovels in the ground) and would-be bidders for the larger project will see that it has progressed beyond a political slogan and a line on a map. It is unclear just how much will actually be achieved by 2020 when the requests for qualifications and proposals (see chart below) occur.

Among the potential early works is design work the Don Valley crossing and possible launch sites for tunnel boring machines (TBMs). The decision to place some of the OL at or above grade means that there are more transitions in and out of tunnels than would be the case with an all-underground line, and launch sites at the transitions are required. These have significant effects on their locations as recent experience on the Eglinton Crosstown shows.

Mathieu Goetzke observed that the Queen Street corridor has challenges, but of course that would also have applied to the RL at least as far west as Osgoode Station. There is the larger question of the choice by the RL project of Queen versus a route further south, and again that is both a technical and political decision that is now set in stone.

The Next Steps slide below contains the troubling observation that Metrolinx will work to “understand community engagement” as if somehow Metrolinx has been operating in the dark while the Relief Line project went through its assessment and consultation stages. In a telling, but common, misuse, Metrolinx describes what they will undertake as “fulsome” intending to imply “copious” or “substantial”, but the word can also mean “excessive and insincere”, the fawning behaviour of one who insults by being overly complimentary. Metrolinx and GO before them have a long history of insincere public participation.

Discussion by the Board raised various questions starting with one from Michael Kraljevic who asked how much of the work already done on the Relief Line can be used for the Ontario Line? Mathieu Goetzke replied that work on the Queen Street corridor “feeds in” to the OL project, and Queen is “incredibly complicated”. The northern branch of the OL was built in part on the Relief Line North study. Malcolm Mackay stated that all of the work done so far will still be used giving the example of an underground station where geological information would inform design for a nearby structure. Some strategies that were “not successful” will not be pursued for the OL.

Regarding the construction challenges listed in the IBC, Vice-Chair Bryan Davies asked about “showstoppers” in the project. Duncan Law stated that there are none in the project “at this stage”, and he claimed that the benefit is that we see the risks now. Undoubtedly there will be several challenges, but the team will work through them, he said. The objective is to get people to jobs and home. Integration with GO for local and regional travel is important, Low continued. Environmental Assessment amendments will be required including a review of technical options for the OL. There is a “significant mountain to climb”.

Consultants have been engaged for the EA process, but subways tend to be environmentally positive, said Mackay, Metrolinx claims, and challenges with the elevated sections will be overcome.

These comments really are a dodge of the main question about the ability to reuse work already completed. The basic fact is that the OL alignment diverges considerably from the RL in places, and the detailed RL work, including public consultation, does not reflect the OL plan as it now stands. Again, we hear that this is a complex, challenging project, a rather different characterization than the self-confidence of the line’s announcement.

Director Paul Tsparis asked how the OL project helps to alleviate pressure on Line 1, and how is the TTC helping on that front.

Malcolm Mackay trotted out the usual list of TTC efforts including:

  • Larger trains
  • Painted tiles on the platforms at St. George and Bloor-Yonge to channel waiting passengers
  • Automatic Train Control
  • Bloor-Yonge Station expansion

The reference to larger trains is getting tiresome whoever cites it because the “new” Toronto Rocket (TR) equipment has been operating for some time now, and their extra capacity was long-ago consumed by latent demand. The painted floor directions may have some effect, but the big problem at busy stations is that platforms become totally filled even with a slight delay and this prevents easy exchange of passengers with trains. As for ATC, Mackay despite his years working at TTC, was unsure of the dates when it will be implemented. He also neglected to mention that more service requires more trains and, eventually, more train storage when the TTC exhausts what it now has. If the province takes ownership of the subway, this problem will land in their lap.

This was capped off with an observation that going north to meet the Eglinton Crosstown is a “beautiful addition to relief” to Line 1. Well, yes, many advocates have been saying this for a very long time while others downplayed the importance of continuing north of Danforth. Even Metrolinx flagged the added relief of the northern extension, and this informed support for work on it by the previous provincial government.

Michael Kraljevic asked about the benefit cost ratio where the value is less than one, although the P3 arrangement is alleged to improve that factor. How does this line up against other subway projects?

Mathieu Goetzke replied that it is hard to get a ratio beyond 1 with brand new infrastructure. GO improvements have good numbers because they build on legacy infrastructure. Moreover other modelling techniques would pick up economic development issues that are not included in the Metrolinx model. Phil Verster explained that Metrolinx does not consider benefits outside of purely transit ones, and the wider economic benefits would make every transit business case a good one. A case will always be touch-and-go for tunnels to get to a ratio of 1. Goetzke added that the OL can enable other works [e.g. the Richmond Hill extension].

This is quite an admission for Metrolinx who have wrestled with their business case analysis for some time. In a political climate where projects must at least break even, the benefits that are balanced against costs have a huge influence on the results. This can include both the scope of benefits (how wide a net is cast to capture benefits) and what payback period is used in the calculation. If the scope is too wide, there is a risk that presumed benefits are not entirely due to the project itself. If the timeframe is too long (Metrolinx uses 60 years), there is a financial problem of substantial expenditure in the short term for savings that might or might not accrue over the very long haul. Moreover, a large proportion of the “benefits” do not capture revenue that can be used to pay off project debt, but rather accrue to transit riders in reduced travel time and increased mobility.

These approaches can be defended on the grounds of “city building” and the long term, cumulative effect of having more transit infrastructure. However, the attempt to make any one project “pay its way” can distort how it is evaluated for political reasons.

Footnote: The Metrolinx Blog includes an article which emphasizes that the Ontario Line will not be built on unproven technology. The ghost of the UTDC and the Scarborough RT still haunts provincial decisions.

Ridership Initiatives (Video at 1:13:19)

Metrolinx ridership for the second quarter of 2019 is up 4.1% over the same period in 2018. However, children did not ride free a year ago, and when they are removed from the “before” numbers, the remaining ridership rose by 5.2%. It is worth noting that the TTC attempts to count children even though they ride free on the system, and so ridership numbers for the two systems are not directly comparable, at least on that basis. There is also the challenge of defining a “ride” when trips can involve a series of transfers and the benefit of the two-hour fare on Presto. The whole question of reporting demand on the GTA network needs work, including granularity about when and where people actually travel.

Statistics reported by APTA show commuter rail up 2.1% among reporting agencies, and bus ridership down 1.0%. GO bus ridership is up 4.5%.

A detailed map (PDF) shows the ridership by station across the rail network. There is no data for the bus network where there has been some controversy about which services should be maintained and which should be cut. Moreover the rail network counts do not distinguish by time of day to break out growth, if any, in off-peak travel as GO moves beyond peak period, peak direction commuting service to downtown Toronto.

Ridership growth varies by corridor and station. The high roller is Barrie with a 10.6% growth in the corridor. Kitchener us up 5.6% and Lakeshore East is up 4.8%.

GO is doing a lot of online marketing which they report as being quite successful, and there is an uptake of e-Tickets as a way of purchasing fares. Many of the promotions are for attractions in the off-peak period and for casual users who would be new to the system.

GO regards riders in Toronto as “transit natives”, people familiar with what transit can offer, and markets to them differently than to potential riders in the 905. Whether this is valid all the way out to the 416/905 boundary is hard to say.

The lower base GO fares are driving ridership within Toronto, but there was no discussion of the effect that will occur if the GO-TTC co-fare arrangement ends thanks to lack of funding from Queen’s Park. This has already affected riders on the UPX who no longer get a UPX+TTC discount. All the marketing efforts in the world can be undone by fiscal policies that affect fares and service.

Presto Quarterly Report (Video at 1:28:08)

Director Janet Ecker asked about efforts to minimize TTC criticisms of Presto. From what she is reading, criticisms of the system way off base. She asked how Metrolinx is trying to deal with this.

Phil Verster replied that it is illuminating to see how some comments get headline status and do not reflect what’s happening on the ground. Things are challenging, he said, and Metrolinx continues to work closely with TTC. There are claims that Metrolinx feels are not valid, and they have encouraged the TTC “to put this behind us”, not to go to dispute resolution.

Annaliese Czerny, Executive VP, PRESTO, felt that it was a shame the story is not about new products and better experiences. Verster was optimistic about making progress with TTC to move to a better future – new devices, open payments – and that this will be the story rather than problems. Czerny noted that TTC is an active and positive partner in the process for future developments.

This was the usual positive Metrolinx spin on PRESTO, but it was undone by the TTC when they released their agenda for the Board Meeting to be held on September 24. In it, the CEO’s Report is quite clear that the TTC will pursue arbitration under their PRESTO contract in an attempt to obtain payment of lost revenue due to non-delivery of a working fare system.

With PRESTO readers on every bus, streetcar and fare gate, and with PRESTO fare vending machines and self-serve reload machines at every station, the provincially-led fare card system has given our customers many benefits, but also many challenges.

Over the summer, I met with Metrolinx President and CEO, Phil Verster, to discuss the outstanding claims between the parties and the status of the outstanding deliverables of the contract for the implementation of PRESTO on the TTC. It is clear from our discussions that Metrolinx considers the contract deliverables complete.

So, while these discussions were informative about the positions of each organization, we were not able to reach a common understanding and agreement. We did agree that the next step is to proceed with arbitration, which is the dispute resolution process provided in the contract.

We are working with external counsel to review the process and finalize material and submissions. As we outlined in our report to the Board in June, the TTC does not consider the contract closed. Rather, there are significant deliverables outstanding, including open payment and account-based technology (which includes equipment), equipment to provide PRESTO Tickets on buses and streetcars, an acceptable third-party distribution network and Service Level Agreements for all equipment.

[TTC CEO’s Report at p. 14]

A major problem with Metrolinx’ perception of a “working” system is that they assume that any rider who encounters a fare machine that is out of service will use an alternate just as they would at a GO station. However, on crowded transit vehicles getting to another reader, let alone another fare machine for tokens or cash, can be very difficult and many riders do not bother to try. If they have a Metropass on their card (or a two-hour fare from a previous tap), this really does not matter because they have already paid, but for other riders this represents lost revenue to the TTC. Credit card holders cannot pay at all because this function rarely if ever worked.

Regular riders are familiar with the situation and just shrug when their tap does not register. I personally encounter this problem at least once a week, and see others having the same problems with unresponsive machines even more often. Things may be improving, but perfection is some distance off and Metrolinx has a lot to answer for from the earlier days of their PRESTO implementation.

Metrolinx tests the availability of fare equipment by “pinging” each device (sending a signal to a machine that elicits a response simply saying “I am here”). However, that function takes place at a low level within the hardware and the application software could be hung even though the “ping” gets a positive response.

Metrolinx measures of PRESTO access are likely too rosy because of assumptions about how easily riders can access the machines and about what constitutes a “working” box when tested remotely.

Service Changes Coming to 506 Carlton in October 2019

With the shift to the larger low-floor streetcars, the TTC will begin schedule changes on Sunday October 13 (Thanksgiving weekend) for route 506 Carlton. In many periods of operation, the headways will widen in recognition of the capacity of the new cars, although the change will also bring a capacity increase.

The schedules are designed on the basis that most service will be provided by Flexitys, especially in the off-peak period. This could lead to problems such as those seen on other routes where CLRVs operate on headways designed for larger cars and are badly overloaded as a result. A lot depends on there being enough new cars to fully populate the route.

On the schedules, the only CLRVs remaining in operation will be the 506 Carlton trippers and 511 Bathurst (which is planned to start conversion late in 2019).

The TTC Service Standards set design capacities for vehicles which in theory dictate the level of service a route received. The standard is more generous for off-peak than for peak service.

Peak Off-Peak
CLRV 74 42
Flexity 130 70
Ratio Flexity:CLRV 1.76 1.67
Ratio CLRV:Flexity 0.57 0.60

If service were replaced purely on the basis of scheduled capacity, then there would be about 60% as many Flexitys/hour as on a schedule for CLRVs.

The new service design (click to enlarge, or retrieve PDF version) is shown below.

The change in AM peak and midday headways is from 5’40” (340 seconds, 10.6 cars/hour) to 7’50” (470 seconds, 7.7 cars/hour), a decrease in cars per hour to 72% of current service. This will be partly offset by the larger capacity of the new cars plus the benefit of trippers scheduled during the height of the peak period.

The change in PM peak headways is from 6’00” (360 seconds, 10 cars/hour) to 8’30” (510 seconds, 7.1 cars/hour), a decrease in cars/hour of to 71% of current service.

This change comes on top of wider headways introduced in August in an attempt to improve route reliability by giving streetcars more running time and longer terminal layovers in a bid to cut down on short turns.

The capacity goes up, but the service frequency worsens. This will almost certainly be compounded by the TTC’s inability to maintain even spacing of vehicles as they move across the route. Even though they might leave their terminals somewhat “on time”, this situation quickly deteriorates into bunches and gaps.

There is always a challenge when the TTC substitutes larger vehicles on a route and widens the headway. Laissez-faire route management that might work tolerably on shorter headways falls apart when headways are wider. If two vehicles nominally ten minutes apart actually operate as a pair, then there is a twenty minute gap and the service might as well not exist at all. Equally, the short-turn stats might look wonderful because both cars have time to reach their terminal, but riders see the same gap they would have if one of the cars short-turned.

The TTC has a huge challenge in this regard, but shows little sign of trying to fix the problem preferring to assume that more running time and longer layovers will do the trick.

Continue reading

TTC CEO’s Report: August 2019

Although the TTC Board takes a long siesta through the summer, the CEO produces a monthly report even in months when the Board does not meet. The August 2019 edition was recently posted on the TTC’s site.

This report continues a format established some time ago by CEO Rick Leary in which the focus is on measures of system performance. There is no financial information here, and only summary ridership numbers with no sense of the associated revenue. All of the financial reporting was hived off of the CEO’s report into a quarterly CFO’s report, but we have not seen one of those since April 2019, and that covered the year ending December 2018. One might have expected an update at the July Board meeting, but the abrupt and unexpected departure of CFO Dan Wright in early June appears to have iced the short-lived report.

I spoke with Wright at the Audit & Risk Management Committee meeting the day before he left the TTC, and he gave no indication of his impending departure. We spoke about the problems of counting “rides” in an environment where there is only a tenuous link between fare payment and the actual number of trips (or trip segments) taken.

The TTC has been wrestling with the possibility that ridership has been over-reported for some time, and the situation is further complicated by Metropasses, the move to Presto and the two-hour fare. Just what is a “ride” for statistical purposes? I had planned to follow this up with Wright for an article here, but alas he vanished like so many other senior TTC staff in the past year. The problem is summarized in the CEO’s Report:

Higher PRESTO adoption appears to have affected measured ridership in two ways. First, we now have more precise ridership data compared to counting tokens and weighing paper tickets. Second, more than 25% of our former monthly pass customers have converted to PRESTO pay-as-you-go e-purse each month in 2019, likely to take advantage of the two hour transfer and for some, the TTC/GO discounted co-fare. This would affect measured ridership to the extent that these customers may ride less often than the monthly average of 71 rides per adult monthly pass. [p 25]

Josie La Vita, the Executive Director of Financial Planning for the City of Toronto, replaced Wright on an acting basis pending recruitment of a new CFO, a process expected to take 12-24 months.

The CEO no longer reports financial data on operations or capital projects, and the absence of a CFO’s report leaves a major gap in information available to the TTC Board and Council on the system’s actual performance.

Moreover, riding counts (i.e. vehicle occupancy) are only rarely reported at a route level, and a “crowding report” has not recently seen the light of day. When crowding data are released, they are averaged and this gives no indication of the effects of bunching and gapping on individual vehicle loads. Performance metrics that do appear in the CEO’s Report do not fully describe the service quality actually experienced by riders particularly on surface routes.

Ridership for 2019 to the end of June is reported as 267.8 million as against a target of 271.9m and a 2018 figure of 270.3m (down 1.5% and 0.9% respectively). For the month of June itself, ridership is on target. As the chart below shows, the shortfall in 2019 came primarily in the winter months which were unusually cold. June’s ridership was boosted by the Toronto Raptors Championship Parade without which the number would have been down 0.6% compared to 2018.

Weekend ridership is down, and this is thought to be due to various factors including the number of subway shutdowns. These are not going to end any time soon with the ongoing signalling projects and other infrastructure upgrades, and at some point the TTC cannot treat their effects as unexpected. A follow-on problem for the TTC is the perception by riders that “the subway is always closed” and this can affect ridership as much as the actual shutdowns.

Presto fare card usage continues to increase and by June 2019 the adoption rate reached about 80%. Presto ridership for the first half of 2019 was 214.2 million out of the total reported ridership of 267.8m.

Vehicle Reliability

The fleet benefits from improved preventative maintenance and from the retirement of older vehicles. Before the onset of hot weather, there was a concerted effort to get air conditioning systems in good order, and AC failures were rare even with the particularly hot weather in 2019. (Personally, I never encountered a “hot car” on the subway, a first for several years running.)

On the streetcar fleet, the decline in the proportion of service provided by the nearly 40-year old CLRVs and the disappearance of their ALRV cousins (of which a few are still officially active but never seen in revenue service) contributes to a reduction of in service failures. The Flexity fleet has its ups and downs for reliability, but even running below its target, these cars are much more reliable than those they replace.

On the bus fleet, the retirement of old buses and the recent purchases of hundreds of new vehicles has substantially lowered the average age of a bus and increased the proportion of buses that are “spare” relative to service needs. The fleet is over 2,000 vehicles, but the peak requirement in June was 1,641 including buses used on streetcar lines. This generous spare ratio has benefits in service reliability but also a cost in both capital and in garaging.

Continue reading

TIFF 2019 To Demolish Downtown Transit Service, Again

Updated Aug. 22, 2019 at 8:05 pm: The TTC has advised that temporary stops for the diversions are still to be finalized. Also, there will be Paid Duty Officers to manage traffic at Spadina, York and Church Streets. Thanks to Stuart Green for the update.

The City of Toronto and TTC have announced various road closures and service diversions associated with the 2019 edition of the Toronto International Film Festival (TIFF).

As in past years, King Street will be completely closed to traffic including transit operations from University Avenue to Spadina Avenue. This will begin at 5 am on Thursday, September 5 and continue through to 5 am on Monday, September 9.

Additional ad hoc diversions might occur on both Monday and Tuesday, September 9-10, for “red carpet events” from 3:30 pm onward.

Service arrangements this year are somewhat different from 2018 because there is now the split 504A/504B King service and the soon-to-be-restored 508 Lake Shore.

The primary diversion for almost all services will be:

  • From King to Queen on Spadina, both ways
  • From King to Queen on York, northbound/westbound
  • From Queen to King on Church, southbound/eastbound

Riders wishing to transfer from the subway to an eastbound King car to Broadview Station or to the Distillery should do so at Queen, not at King, or they will have to walk from Yonge to Church from King Station. There will be no eastbound King cars at Osgoode Station. It is not yet confirmed whether the TTC will establish a stop at Queen and York for a walking transfer.

Riders wishing to transfer from the subway to a westbound King car must do so at King or Osgoode Stations. There will be no streetcar service at St. Andrew Station, although there will be plenty of King cars nearby at York Street. Whether a temporary stop will be created at King and York is still to be confirmed.

This service design will see ALL of the King Street services, both ways, operating northbound on York Street giving an extremely frequent service, not to mention the potential for a total bottleneck making the turns east and west at Queen Street.

504A Dundas West to Distillery Service

This service will be broken into two segments:

  • From the west, 504A cars to/from Dundas West will operate downtown via Spadina and Queen east to Church, then loop via Church, King and York.
  • From the east, 504A cars to/from the Distillery will loop downtown via King, York, Queen and Church.

504B Broadview Station Service

The Broadview Station service will use the same loop downtown as the 504A Distillery cars:

  • Westbound on King to York, then north to Queen, east to Church and south to King.

There will be no replacement bus service parallel to King as has been attempted in some past years. Anyone destined for the area between University and Spadina on King will have to walk in from the bounds of the closed area or south from Queen Street. This is of particular concern for anyone going to screenings at TIFF Bell Lightbox which will have no transit service during the diversions.

508 Lake Shore Service

The new 508 Lake Shore trippers will loop downtown via:

  • Eastbound via Spadina, Queen and Church
  • Returning westbound via Richmond and Victoria to Queen, then west to Spadina

304 Night Service

The 304 night cars will divert via Spadina, Queen, Church/York both ways.

A shuttle night bus will operate between Wolseley Loop and Parliament/King bypassing the TIFF district via Adelaide and Richmond Streets.

503 Kingston Road Bus

Not mentioned in the TTC’s announcement is the 503 bus service which consolidates the 502/503 Kingston Road services in September. These buses are supposed to loop via York, Richmond, and University to King including a layover point on York north of King. That area will be thick with streetcar service. It is ironic that the only service that will stop eastbound at St. Andrew Station will be the 503 bus on what is sure to be a “now and then” schedule.

I am a TIFF supporter as a member and donor, and have attended the festival for over three decades. That said, I am disgusted by the gorilla-like behaviour of TIFF in elbowing aside vital transit services on weekdays in Toronto.

These diversions produce severe effects on service not just downtown, but on parts of the King and Queen routes far from the TIFF district. Riders across the city suffer so that TIFF can have its street fair.

Every year we hear that “next year will be different”, but nothing happens.

From the TTC’s diversion announcement:

We encourage you to plan your trip in advance. We thank you for your patience during this important event benefitting Toronto’s economy and international reputation as a world-class city.

A “world class city” would figure out how to integrate its transit service into a major cultural festival.

TTC Service Changes Effective Sunday, September 1, 2019

On September 1, TTC routes that had summer service cuts in May and June will revert to their regular service levels.

Several routes will have “reliability improvements”, a catch-all term for a variety of changes.

  • Most commonly, the buses assigned to a route are stretched further apart on wider headways with additional driving and/or recovery time in the hope that the service will stay more on time than it is now. Whether this actually happens is a matter of conjecture because the TTC does not regular report on route performance at a granular enough level to judge the effect of schedule changes.
  • In a few cases, running times are trimmed in recognition that the current schedules are excessive. This usually frees up buses that are either assigned to another branch of the same route, or which go into the pool for improvements on other routes.
  • In a few cases, there are adjustments between the local and express services on a route.

The TTC has not published crowding statistics since March 2019, although this was supposed to happen quarterly.

There will be major changes on the streetcar network, notably on 501 Queen and related routes, due to construction at Kingston Road and Queen, as well as the conversion of service west of Humber Loop to 100% low-floor cars. See my previous article for more information. The service plan for the restructured routes is included in the spreadsheet linked below.

Other changes of note:

  • Construction changes to 34 Eglinton East, 51 Leslie, 54 Lawrence East and 91 Woodbine related to the Crosstown project will end and these routes will resume their normal configuration.
  • Running times on 36 Finch West are extended to compensate for the effects of the Finch West LRT project.
  • Construction at Sheppard West Station will affect many routes there.
    • The 84/984 Sheppard services will not enter the station, but will serve on-street stops.
    • 101 Downsview Park, 106 Sentinel, 107 St. Regis, 108 Driftwood and the Wheel-Trans services will move to new bus bays as the work progresses.
    • A temporary loop will be provided in the passenger pick-up area for use by 104 Faywood, 105 Dufferin North, 117 Alness-Chesswood and YRT services.
  • With the return to fall service levels on Line 2 Bloor-Danforth, the second gap train recently introduced on that line will be removed.

Details of the service plans are in the spreadsheet here:

2019.09.01_Service_Changes

Various construction projects that have required buses to be added to several routes are scheduled to wind down in the coming months. Whether the buses released from construction service will actually shift to service improvements on the network as a whole remains to be seen. A basic budgetary problem for the TTC is that at least some of the extra construction service is paid for in capital project budgets, but this money is not available on an ongoing basis to run the same vehicles in day-to-day operations.

TTC Service Changes Effective Sunday, August 4, 2019

The TTC will implement several service changes on August 4, 2019. Normally this is a quiet time of the year for schedule work, but TTC Service Planning is working through “reliability enhancements” for many routes with a batch in the August schedules and more to come in future months.

Unfortunately, the changes almost always involve running the same number of vehicles on wider headways so that there is more driving time, but most importantly, more recovery time at terminals. However, nothing has changed in the TTC’s standard that “on time” has a range of +1 to -5 minutes, and there is no indication that service reliability will actually improve without active management of headways.

The following routes have reliability enhancements that result in less frequent service:

  • 11 Bayview
  • 61 Avenue Road North
  • 62 Mortimer
  • 63 Ossington
  • 83 Jones
  • 88 South Leaside
  • 506 Carlton headways will widen at many times. Late Sunday evenings, Carlton will go to a 10’30” headway thereby removing this route, technically speaking, from the 10 Minute Network. This condition will be corrected in the fall when Carlton gets a new schedule for low floor car operation.
  • 512 St. Clair (weekends)

The TTC has not published crowding stats since March 2019, and so there is no indication of the effect of the wider scheduled headways on crowding levels. The real test will come in September when demand rises with the end of the vacation period.

One construction project, at Royal York Station, is now complete and the route structure here will be restored to its normal arrangement.

  • The 15 Evans and 48 Rathburn routes will no longer interline during weekday daytime periods.
  • The 73 Royal York and 76 Royal York South routes will no longer interline.
  • The 315 Evans night bus will no longer divert to Islington Station as its northern terminus.

Additional running time which had been added to 33 Forest Hill to compensate for construction on the Crosstown will be removed, and the route will revert to a 30′ headway at all times.

Even more running time will be added to 505 Dundas to compensate for the effects of watermain construction downtown, and scheduled headways will be wider during most periods. Whether the TTC will do anything about the tendency of buses on this route to operate in herds of two or more vehicles remains to be seen. The change is expected to be in place until the October schedules.

Planned improvements include:

  • One additional “gap” train will be added on 2 Bloor-Danforth in both the AM and PM peak periods.
  • Late evening service on 39 Finch will be extended to Old Finch & Morningview.
  • Early evening service on 64 Main will be improved from every 20′ to every 12′.
  • The 176 Mimico GO shuttle will be changed to loop via Marine Parade rather than at Park Lawn Loop to improve its reach in the Humber Bay Shores area. The afternoon schedule will be adjusted to match the observed (usually late) operation of the GO service at Mimico.

Details of these and other changes are in the spreadsheet linked below.

2019.08.04_Service_Changes

TTC to Presto: Deliver What You Promised

At its meeting of June 12, 2019, the TTC Board considered a report and presentation setting out the many shortcomings of the Presto fare card system.

Presto implementation has been underway since a 2012 master agreement between the TTC and Metrolinx, and they are now at the half-way mark in a 15-year contract. In spite of this, some functionality included in that contract has not yet been provided, and there are no service level agreements (SLAs) in place setting out basic issues for system performance and financial penalties to Metrolinx for failure. Indeed, although the TTC has claimed revenue losses thanks to problems with Presto, Metrolinx has not accepted that it owes the TTC anything.

Until quite recently, Presto, like so much that comes from Metrolinx and Queen’s Park, was a “good news story” where TTC riders switching to the fare card drove fast growth for that system. The convenience of Presto versus tokens and the disappearance of the Metropass shifted over two-thirds of TTC riding onto Presto. The chart below shows the breakdown for April 2019. The Presto share is expected to reach 95% when legacy media are withdrawn.

However, the good news hides problems with the system as it exists both for Presto users and those who might shift to that system in the future.

The heart of Toronto’s problem with Presto lies in the way the system was imposed. TTC was prepared to go with an alternative vendor for a new fare card system, but were told by Queen’s Park that failure to adopt Presto would put all of the provincial subsidy programs at risk. This was a very big, heavy stick for the government to use, and it shows just how desperate they were that Ontario’s largest transit system be a client (with associated revenue and prestige) of the Presto system.

During the ongoing discussions that began in the fall 2010, the Province and Metrolinx maintained that the common PRESTO Farecard system was their recommended approach to achieving the Plan’s inter-regional policy objective of increasing cross boundary travel. The Province and Metrolinx committed to upgrade the PRESTO system to meet the TTC’s distinct and forward-looking customer, business and financial needs, including advances in fare payment technologies using open payments. The Province and Metrolinx indicated that billions of dollars of funding for some existing TTC programs which had been promised publicly (Provincial gas tax, new streetcars, Eglinton and Scarborough transit initiatives) could be in jeopardy without PRESTO. The adoption of PRESTO was thus approved in June 2011, subject to developing acceptable operating and financial agreements and confirming the funding necessary to proceed. [p 7]

Despite a competing proposal from Xerox subsidiary ACS that would cost over $300 million less than Presto, the TTC was forced to accept the provincial system.

Presto did not perform as it needed to, and contracted functionality is still not available.

The TTC entered into the Agreement with Metrolinx on November 12, 2012. The base term is 15 years (2027), with an option to extend for one additional year at Metrolinx’s discretion and an additional four years by mutual agreement. Key elements of the Agreement include:

  • Metrolinx to make modifications and enhancements to the PRESTO system to allow for an e-fare account based payment system with an open architecture using industry standards to accommodate open loop financial cards, mobile applications and future technological innovations (“PRESTO NG”)
  • Metrolinx to finance, implement and operate the PRESTO NG system and provide a wide range of “managed services” (e.g. back office operations; customer services; revenue collection and maintenance of all system field equipment)
  • In return, TTC to pay a fee of 5.25% of TTC fare revenues processed through the PRESTO system [pp 7-8]

The TTC contemplated a variety of payment mechanisms years ago.

The TTC’s Business Requirements specified both open payment and an account-based architecture, which was to have been built alongside the existing PRESTO card-based architecture. In such a system, customer convenience and flexibility would be maximized. A customer could choose to have a PRESTO card with all its fare policy benefits (e.g. fare concession, 2-hour transfer), or to get some of those benefits using a non-registered open-payment card (e.g. Visa, Mastercard), or to get all fare policy benefits with a registered open-payment card. These concepts were an essential component of the Agreement and were fundamental to the TTC’s agreement in 2012 to accept the PRESTO system versus a competitive market-based solution. [p 11]

Two key capabilities – fares for travel between transit agencies such as York Region Transit and TTC, and the move to “open payments” and an account based system – have yet to be implemented. Regional integration is hoped for later in 2019, but there is no firm date for a change to the payment mechanism.

What Are Account Based Fares?

Account based fares and open payments are closely related, and they are fundamentally different from how Presto works today.

With Presto, fare calculations and validation occur between card readers and the cards themselves. Information about the account balance and any bulk purchase such as a pass is stored on the card. This allows a transaction to occur without any link back to a central system, an arrangement dating back to Presto’s early days when wireless links from their roving bus fleet were not reliably available. However, this forces all of the payment logic into the architecture of the readers and cards constraining the functions they can support.

This arrangement is responsible for the lag between an online account update and the appearance of new “money” on your Presto card. Until you tap onto a reader that “knows” you loaded more money online, you have money in the central system, but cannot use it because the fare machines don’t “see” it.

Imagine if software had to be loaded into every phone so that it could calculate the cost of a call and maintain your account balance. That, in effect, is the complexity Presto imposes, something that should have been designed out of the system years ago.

In the case of a credit or debit card, Presto cannot “write” information onto the card, and so distance-based functions such as GO fare calculations cannot be handled. Only a flat, standard charge is possible such as an adult TTC fare, and without a two-hour transfer privilege.

In an account based system, a rider has a transit account just as they would have one for their mobile phone or credit card, and activity (trips, transfers, border crossings) is accumulated as it occurs. At the end of a billing period, the cost of these trips is calculated with any appropriate discounts such as loyalty programs, time-of-day fares or special event promotions. A Presto card, any other smart card, or a smartphone app can be registered as the rider’s transit identification, but in all cases processing happens in the “back end” with monthly billing to a bank account or credit card. Riders who have not set up accounts simply use their credit cards for pay-as-you-go billing.

The important distinction is that the system needs only track a rider’s travel, not compute the fare in real time based on “today’s rules”. Those rules can be updated in the central billing system rather than having to be integrated in the reader software and pushed out network wide. There is no need for an electronic purse or “e-purse” on the card which must hold enough money (or some form of pass) to pay for any trip a rider might take.

Presto tickets (sometimes referred to as limited use media or “LUMs”) would still exist, but Presto readers would only have to verify when the ticket expires or how many trips are left on it.

Metrolinx plans to make account based fares available sometime in 2020 with open payments in 2021, but there are no firm dates, nor is there a specification of just what functionality Metrolinx will support.

Gaps in the Presto TTC System

Other gaps between the TTC’s Presto contract and current capabilities include:

  • Flexible and dynamic fare policies/products
  • Support for other than standard fares and environments such as
  • Presto tickets dispensed as receipts for cash fares on surface vehicles
  • Device availability/reliability
  • Service Level Agreements (SLAs)
  • Third-party sales network

Continue reading