Several items have languished unreported here for a while, and it’s time to push them all out of the door in preparation for the deluge of budget information and the new Service Plan that will come in December. My apologies for not keeping you as up to date as I might have.

The items covered here are:

• Ridership and revenue
• Vehicle reliability
• Service quality (briefly)
• Automatic train control
• eBuses (electric buses)
• Fare evasion

Ridership and Revenue

For the three months July-September 2019, ridership numbers ran slightly above budget levels. Year-to-date numbers are still down, but the trend looks optimistic. However, this must be taken with a grain of salt because the “ridership” is a derived number with several components:

• Actual sales and usage of tickets and tokens, plus cash revenue.
  • The proportion of fares collected in this manner has been dropping as riders migrate onto Presto. Sales of legacy media will end on November 30.
• Presto counts
  • User “taps” need to be converted to the conventional version of “trips” allowing for the change to a two-hour transfer rule.

A regular commentator and long-time friend, Robert Wightman, left me a question a while back about how riding is reported by the TTC:

The APTA data for North American ridership for the second quarter of 2019 is available. Take a look at the last page for the TTC’s increase in light rail (street car) ridership growth.

The data in question are shown here:

Note that APTA reports “unlinked trips” and so every transfer connection (except between subway lines) counts as a new “trip” for their counts which are much higher than published TTC numbers. However, there is a huge jump in “light rail” ridership (streetcar), and an 8.38% increase in overall trips, a much larger number than we see in general TTC ridership stats.

I sent a query to the TTC asking about this, and their reply from Mark Mis [email of October 4, 2019] explains what is going on:

I asked: “The number for LRT is way off of the scale but the other modes so substantial shifts.”

Mark Mis replied:

Three reasons:

1. with more new Streetcars available, the use of Buses on the Streetcar network is less, reversing the prior years’ decline,
2. a change in methodology (previously modal allocation of rides relied on two modal analysis by Revenue Operations, now we allocate modal rides based on PRESTO tap data),
3. new transfer rates are calculated annually by mode by Service Planning.

I asked: “The total unlinked trips show an increase of 8.38%, and this is very different from “ridership” as reported by management. Can you find out how the numbers supplied to APTA are created and if there is any sense of why the magnitude of changes is so different from the numbers reported by the TTC itself?”

Mark Mis replied:

The TTC’s reported revenue passenger rides are based on a the number of customer trips/journeys made whereas APTA figures reflect all boardings as a new ride.

In 2018, after a review of TTC’s transfer rates calculation, it was noticed that TTC did not include subway interchange stations in the transfer rates. As a result, beginning in 2018, TTC started including Subway interchange transfers in their transfer rate calculations. Same for other multi-modal stations (e.g. Kennedy).

In 2018 and 2019, the TTC was in transition of changing its fare collection system from a Legacy fare system to a PRESTO fare system. The changeover has created many challenges in reporting Revenue Passenger rides.

Despite seeing Revenue passengers decline in 2018, Service Planning’s measures indicated overcrowding.

The new transfer rates have been increasing (in 2017 was 1.59 system wide “without interchange station transfers’, in 2018 was 1.81 system wide (with inclusion of interchange station transfers) and in 2019 was 1.92 system wide (with inclusion of interchange station transfers).

APTA 2019 results vs. 2018 year-to-date as at June are correct.

In other words, actual ridership on the TTC is going up faster according to Service Planning’s world view than might be implied by the ridership numbers charted above that are based on fares. Indeed, what is even more important from a financial point of view is the total revenue, but this is only reported in the quarterly financial reports of which the most recent gives the data up to mid-year.

As of June, revenue was down from budget, but so was the projected expenditure. For the details of these changes, please refer to the September 2019 budget update [pp 4-5].

A comparison with 2018 actuals shows where the changes occur in 2019.

These changes are explained in the report [p 6]. Of particular concern is that the extra revenue from the April 1 fare increase is less than half of the total of new revenue which depends on a one-time draw from a reserve that will be depleted this year. That is new money that will have to be found for the 2020 budget before any other changes are considered.

At a time when there are widespread calls for lower fares and better service, not to mention the need for billions to fund state of good repair capital programs, the operating budget will be a difficult debate for the TTC Board and for Council facing a tough budget year in 2020. The TTC Budgets will be considered at a Board meeting on December 16, 2019.

The TTC’s 5 Year Service Plan and 10 Year Outlook will be considered at the Board meeting on December 12, 2019. I participated in two public working group meetings for the development of this plan, but will not report on details until the final version is out because the ideas presented to the working group were only in draft form.

Streetcar Reliability

As I reported in How Reliable are TTC Statistics, there was a controversy over how the stats are reported for new cars versus the “legacy” fleet of CLRVs and ALRVs. Until quite recently, only failures that were chargeable to the manufacturer were counted on new cars for warranty purposes, while all failures were counted on old cars. This substantially increased the gap between the reported mean distance between failures (MDBF) for the two fleets.

In the November 2019 CEO’s Report, the two metrics for the Flexity report are shown separately:

• The moving annual average MDBF on a contracted basis is 20,852km as compared to the contracted target of 35,000km which must be reached by one year following the delivery of the last car in the order.
• The operational MDBF calculated on the same basis as the numbers for older cars was 15,035km.

The month by month values bounce around a lot because the number of failures is small, and so a small rise or fall in the monthly count has a big effect on the monthly calculation. The chart below shows the values for individual months, not the moving average which is reported in the accompanying text. Although the value for September fell from the high in August, the overall trend is still upward for contract-related failures.

A chart including the operational failure rate will be included in a future CEO’s report. It would have been interesting to see these numbers during the period when the old cars were pushed out the door as quickly as possible because the gap between the two fleets, especially for the best of the old cars, was probably much less than what was presented.

The legacy fleet’s reliability continues to drop as maintenance of these cars ends and cars are withdrawn from service when they fail.

CLRV operation is expected to end on the last weekend of December with arrangements to celebrate, if that is the correct term, the event still to be announced. CLRVs will operate as extras on 501 Queen on weekends from Sunday, November 24 to Saturday, December 21 as I previously described in the November service update.

The overall rate of failures on the streetcar fleet is reflected in the “Road Calls and Change Offs” statistics which have been dropping steadily as older cars are retired. The target is to get this number down to about 3 per month or 1.5% of the peak scheduled service (based on September 2019).

Bus and Subway Reliability

The bus and subway fleets continue to perform well.

Because the TTC has a very large fleet compared to service requirements, and a good proportion of that fleet is only a few years old, vehicle reliability is very high. With a spare ratio of over 20% (2,000+ vehicles versus a scheduled service of 1,655 buses in the AM peak), maintenance can stay ahead of failures and any “bad actors” in the fleet simply don’t get out of the garage.

On the subway, the T1 trains on Line 2 are performing well above their MDBF target of 300,000km, while the TR fleet on Lines 1 and 4 has difficulty achieving its 600,000km target. Indeed, for September 2019, the actual MDBF numbers for both fleets were similar. As with the Flexity stats cited above, these numbers bounce around a lot because the number of failures per month is small, and only a little change in this count produces a big change in the MDBF value.

Service Quality

The TTC continues to measure service on the basis of scheduled departures at terminals, and counts of short-turns. As I have shown in many articles, the times leaving terminals give a best case view of the variation in headways (gaps between vehicles) along a route, and most riders do not experience the comparatively reliable service at the ends of lines. Gaps get wider and bunching gets worse within a few kilometres of terminals, and it is common for riders to have longer than scheduled waits followed by a pair of vehicles.

There are fewer short turns, but this comes at the expense of scheduling for “service reliability” by stretching headways (the times between vehicles) so that the same number of vehicles provide less frequent service over a route with more running and recovery time. However, the lack of headway management on most TTC services simply means that scheduled service gets worse while the scatter in headway values remains. Short turn rates might improve, but all riders suffer just to drive down the short turn count. (See the analysis of 41 Keele for examples.)

Automatic Train Control (ATC)

The ATC project on Line 1 continues around the downtown “U” and is projected to be expanded in four segments to eventual completion in 2022:

• Q1 2020: St. Patrick through Union to Queen
• Q4 2020: Queen to Rosedale
• Q4 2021: Rosedale to Eglinton including Davisville Yard
• Q3 2022: Eglinton to Finch

These dates are somewhat later than the original target for this project in part due to complexity (the addition of Davisville Yard to the scope), waiting for the Eglinton Station reconfiguration associated with the Crosstown LRT project, and track and tunnel work needed north from Eglinton.

The TTC has used some fairly aggressive numbers in the past when speaking of ATC capability, most recently in the plans for the Relief Line operating at 33 trains/hour, or 33,000 passengers based on a service design load of 1,100 per six-car train. However, at the November 12 Board Meeting, CEO Rick Leary talked about a rate of only 31 trains/hour, and it is unclear just what target the TTC is aiming for.

In any event, ATC would allow “gap trains” to be inserted southbound from Davisville on a closer headway than is now possible improving the short-term capacity through Bloor Station southbound once ATC extends at least to Eglinton.

As for Line 2, the date for ATC is off in the future thanks to the TTC’s decision to rebuild rather than replace its fleet of T1 trains. This gives a short-term saving in the capital budget by delaying:

• The ATC conversion of Line 2 itself,
• The replacement of the T1 fleet, and
• The construction of a new carhouse at Kipling (Obico Yard, already purchased by the TTC).

While the existing signals and trains might be kept working for another decade or more, this policy raises two major issues:

• Even with the platform and circulation space expansion planned for Bloor-Yonge station, there will be no added capacity to take riders away from this key transfer point in the PM peak. Much study and hand-wringing concentrates on southbound congestion in the AM peak and the benefits of both ATC and any “relief” line that might be built. However, this ignores the outbound capacity constraint in the evening.
• The Line 2 extension to STC and beyond to Sheppard should be built with ATC, unless the TTC and Metrolinx plan to replicate the screw-up with the Vaughan extension that was originally planned to use conventional signals and converted while the line was under construction. In turn a move to ATC on the L2EE will accelerate the need for ATC capable trains on Line 2 even if the existing portion does not go “live” with ATC until sometime after the extension opens.

The three-four year delay in the opening date for the L2EE gives the TTC more breathing space in its schedule, but a decision about improved signal technology, trains, carhouse and line capacity must be made soon allowing for lead times on these major projects.

eBuses

The TTC has 60 eBuses on order, 25 each from suppliers Proterra and New Flyer, and 10 from BYD. As I write this on November 14, the number of buses tracking on the TTC’s vehicle location system for each vendor are:

• New Flyer: 10 buses (3700-3709)
• Proterra: 9 buses (3725-3731, 3733-3734)
• BYD: None

These buses are not all in revenue service, and no BYD vehicles have been delivered yet although they are expected by the end of 2019 (as of a September 2019 update).

This is an ironic situation considering that BYD were responsible (through a heavy-handed lobbying attempt and vendor presentation effectively sponsored by the Deputy Mayor) for getting the green bus purchase in the TTC’s agenda.

During discussion of the CEO’s Report at the November meeting, an issue came up about vehicle charges which are proprietary technology for BYD, but not for the other vendors. This means that either BYD buses will be confined to specific garages (assuming they play any significant role in a future fleet), a constraint that other vendors will not face.

TTC plans to reach a zero-emission fleet by 2020 [corrected] 2040, but management have yet to produce a detailed financial plan showing the cost of converting existing garages or the extra capital cost of vehicles and garage infrastructure (power feeds and charging systems) compared with a hybrid diesel-electric fleet that is supported by current infrastructure. Any premium cost for the vehicles and infrastructure must be balanced against operating and maintenance savings, yet to be determined, just as with rail transit modes.

There is not yet enough operational experience with the new buses for MDBF stats to show up in the CEO’s report, and the fleets that are in Toronto have to go through winter testing in coming months.

Fare Evasion

In February 2019, the City’s Auditor General reported on TTC fare evasion. See Fare Evasion on the TTC: The Auditor General’s Report.

At its September 2019 meeting, the Board received a response from TTC management that included the results of some field work done both to check on fare evasion levels, and to determine how well Presto data matched actual observations. The Presto data lined up with field observations at 93%.

Three reports were considered by the Board:

• Revenue Control Strategy – Phase One: Response to the Auditor General Revenue Operations Phase One – Fare Evasion and Fare Inspection Report
• Presto Tapping Observations on Streetcars
• Revenue Control Strategy – Presentation

Tapping behaviour was monitored in the AM and PM peaks on the King and Queen routes. The over 6,000 observations were reconciled to Presto data at 93% accuracy.

There was a difference in observed behaviour in the second phase of the field work when uniformed fare inspectors were present. More people tapped with inspectors present, and some would-be riders were observed to change their minds about boarding when they saw that they would be checked for a valid fare. It is worth noting that even the surveyors were unable to check everyone because of vehicle crowding. (In the chart below, “SRVM” refers to the Single Ride Vending Machines.)

The rate of tapping behaviour varies with location. It is 64% downtown on surface routes (with “downtown” defined as the planning district bounded by Bathurst, the CP Rail corridor, the Don Valley and Lake Ontario), 76% outside of “downtown” (although this contains much of the old City of Toronto), but only 59% in the subway where riders could only enter without tapping either because there was no barrier, or through efforts such as “tail gating” through a fare gate. Considering the relative proportion of ridership represented by the subway system, this is a major problem compared with what is seen on the surface network. (See p 4 of the Tapping Observations presentation.)

The locations where riders were least likely to tap are almost all close to the core area where trips are likely to be short and the chance of being inspected small.

• On King: Portland, Spadina, Peter, John and University
• On Queen: Ossington, Yonge, Church, Jarvis and Parliament

More riders tapped when boarding at the first and second doors on Flexity streetcars (74-75%), but this rate falls at the third and fourth doors (64% and 54%). Considering that operators perform no fare inspection at all, this behaviour may be left over from legacy car operations where rear door entry avoided the operator’s gaze and intervention.

The review found that riders make a greater effort to pay either with a Presto tap or at an SRVM when inspectors are present, but even then “vehicle crowding was still an impediment to payment” (p 8). Moreover, in 10% of instances one of the two SRVMs on a car was not working. (It should be noted that these machines are near the second and third doors, and on a crowded car are unreachable.)

Although no children tapping Presto cards were observed, nine unique “child” Presto cards were recorded. This shows the importance of having distinct cards and visual/audible displays on readers to show when an adult is fraudulently using a child’s card to appear to be making a payment.

When questioned, riders who did not have a valid fare gave a variety of explanations, several of which are quite credible:

• Will reload Presto card at the connecting subway station.
• Did not know that the two-hour transfer only applies to Presto cards and not to paper transfers.
• Time lag after reloading Presto card online.
• Presto reader blocked by other customers.
• Presto reader or SRVM out of service.
• Carrying multiple Presto cards and used the wrong card. [p 9]

Revenue losses are estimated to be $64 million annually of which $61 million is due to various forms of evasion. To put this in context, TTC’s fare revenue for 2019 will be over $1.2 billion, and so the loss is approximately 5%. The budgeted fare inspection complement for 2019 was increased by 70 (45 fare inspectors, 22 enforcement officers, 3 staff), and these positions will be filled by year end. Eventually the goal is to have 13 2-person teams at a minimum during peak periods, and 8 teams on weekends.

Inspection of the use of children’s cards for free rides in March 2019 yielded 291 charges over a four week project with 7,140 cards checked. With the increased use of concession cards, deployment of inspectors will be based on location and time of day where there is heavy use. Also, Presto data can be cross-referenced with Automatic Passenger Counter (APC) data to determine the ratio of boardings to taps and identify problem locations.

There will be an effort to have Shoppers Drug Mart be more selective in issuing children’s Presto cards, but this is unlikely to have much effect as a “parent” can legitimately say they are getting one for the children without them actually being present. The TTC wants Metrolinx to change both the physical colour of these cards and the sound/display on a Presto reader when one is used, but Metrolinx is dragging its feet on these requests, even though they are included in the master agreement with TTC as required features.

During discussion of the reports, the Board asked whether there is a plan to reduce evasion and what the target will be. Management replied that benchmarking is in progress with targets to come in 1Q2020. Evasion rates will be reported out publicly at least annually based on studies mimicking the Auditor General’s report. Commissioner Lalonde felt that it will be very helpful to have fare evasion data in the CEO’s report, but than at a minimum it should be reported quarterly starting as soon as practical. It is unclear what resources this level of monitoring will require over and above routine enforcement.

Commissioner Carroll observed that the biggest finding is that transit fare inspectors work and asked if this will be be part of TTC strategy. Management replied that they plan to figure out the appropriate mix of fare inspectors and enforcement officers.

Commissioner Bradford asked how a progressive enforcement model would work. How many warnings would an evader get before being charged? What about those who can’t pay? Management replied that fare inspectors are expected to exercise discretion, and a model for how this will work is yet to come.

Bradford asked whether there is any indication that Metrolinx will help the TTC implement the needed changes. CEO Rick Leary replied that Metrolinx will “look at” distinct cards, and that a lot has changed since the Auditor General’s report. Bradford observed that he did not have a lot of confidence in the Metrolinx relationship.

Commissioner Osborne compared the cost of a fare increase to the lost revenue to fare evasion. However, any offset would be a one-time change (recovery of the current 5% loss) and would require ongoing enforcement to ensure that the evasion remained at a new, lower rate.

Although the discussion focused on Presto, there is also the question of legacy transfer forms issued on surface vehicles and in the subway.

With the retirement of the legacy streetcar fleet, conventional transfers on these routes will vanish, although there will still be paper receipts from the SRVMs for cash and token fares, and senior/student tickets will have a timestamp (that may or may not be accurate).

In the subway, once tokens and tickets are no longer in circulation, the need for transfer machines as we know them will disappear. Riders paying cash will do so at a fare machine and get a single ride Presto ticket. It remains to be seen, with proposed changes in station staffing, how the TTC expects people to enter subway stations with a fare receipt/transfer that is not machine readable.

Only the bus fleet will remain as a location where operators have to issue paper transfers because Presto does not yet have any mechanism to issue receipts or single ride tickets on these vehicles.

All that said, the degree to which anyone actually inspects the remaining paper transfers will affect how much transfer abuse contributes to TTC fare losses.