The agenda for the TTC Board meeting of July 10 was not extensive, and the major issue of the day – the pending layoffs at Bombardier – was not debated, at least not in the public session.

Items of interest included:

• The CEO’s Report
• An update on the Line 2 wheel flats issue

CEO’s Report

Ridership

Ridership continues to be an issue for the TTC with results to June 1, 2019 being down 1.2% compared to 2018. This is attributed to two factors:

• Severe winter weather affecting results in January and February
• Migration of former Metropass users to pay-as-you-go Presto rather than loading a pass onto their card

The question of how Metropass users were counted for “ridership” purposes has affected TTC results for a few years. Based on survey data, the average Metropass holder took 71 trips/month although obviously there would be a range of values from very frequent riders down to those who bought a pass for convenience and the unlimited transfers it offered.

With the shift to Presto, Metropass users at the low end of the scale could obtain the convenience of a card for payment including two-hour transfer privileges without actually buying a pass. These riders now count only for the trips they make, not for the average usage of all passholders, and so there is a drop in trips allocated to them.

The TTC is still wrestling with restating old ridership statistics in light of possible over-reporting of Metropass trips.

Completely missing from the CEO’s Report is any discussion of fare revenue. This information is now reported quarterly by the CFO, a position that recently changed hands with the unexpected departure of Dan Wright. There was no financial update from the new acting CFO.

Regardless of how the TTC counts riders and trips, revenue and expenses are the real year-over-year measure. We will not see an update on this until the next board meeting in September.

GO-TTC Double Discount Fare

The Double Discount fare provides a reduction of up to $1.50 for GO+TTC trips using Presto with subsidy coming from the Province of up to $18.4 million annually. This was implemented by a three-year agreement in 2017 that is due to expire in March 2020. However, the uptake of this is so strong that the allocated subsidy will be exhausted by October 2019.

Recent changes to the GO/UPX tariff eliminated this discount for “local” UPX+TTC trips that do not go to Pearson Airport thereby undoing some of the benefit of frequent UPX service in the Weston corridor.

Phil Verster, President and CEO at Metrolinx, is proposing a “sustainable strategy for the continuation of this fare integration initiative, one that does not use a subsidy from the provincial government.” He proposes GO Transit and TTC continue to offer the DDF reduced fare without Provincial subsidy starting in October 2019. [pp 11-12]

It is unclear just what a “sustainable strategy” for a fare subsidy might be as new sources for operating funds are not thick on the ground. If this requires diversion of revenue from other sources, then what do we give up to offset the provincial download? At an annual cost of about $30 million, is this where Metrolinx and Toronto wish to spend their money? To what extent should Toronto be expected to fund what was a provincial program to encourage cross-system ridership?

While this could be portrayed as another in a long list of downloads from the Ford government to Toronto, the real issue is that Metrolinx has been utterly unable to produce a coherent regional fare strategy for years. The fundamental problem is that any fare consolidation and boundary elimination will reduce fares for some riders be they 905-416 cross-boundary travellers (e.g. YRT+TTC) or riders who use GO for part of their journey. Metrolinx wasted a lot of time attempting to create a “zero sum” tariff that would not require additional subsidy, but this inevitably created a scheme that would rob Toronto riders to pay for cross-border journeys. At a time there should have been a robust discussion of the cost and subsidy arrangements for flattening fare barriers, the Province continued with ad hoc fixes including the Double Discount fare leaving the larger problem for another day.

Staff will report back to the Board in September with an analysis of the financial effect if the TTC and Toronto continue this program at their own cost.

Service Quality

The CEO’s Report contains many charts reporting on service quality. I will not review these in detail because many of them have a basic flaw – data are reported on an average basis over the entire operating day without regard to quality during key periods. Many stats reflect conditions at terminals, but the average riders experiences service along routes where conditions may be very different.

Subway service is presented with two quite different metrics: “on time performance” and “percentage of capacity operated”. The difference is quite striking for Line 1 stats.

For a train to be “on time”, it must depart from a terminal within 1.5 times the scheduled headway. For example, if service is intended to run every 4 minutes, a train can leave in a gap of up to 6 minutes and still be “on time”. The chart below shows that about 90% of service met this criterion, but it is important to note two critical factors:

• The data are for all-day operations where off-peak hours dominate in the overall contribution, and where 1.5 headways is a wider allowable target than in the peak.
• There is no measure of the number of trains that actually operated. In a worst case situation, one third of the service could be missing while what remained still achieved the 1.5 headway target. (For example, 10 trains in an hour on 6 minute headways are “on time” even though the scheduled service might be 15 trains on a 4 minute headway.)

The second chart shows capacity actually delivered at a selection of points along the route. This number was very low in January because of problems getting the AM peak trains out of Wilson Yard on time. Operational changes at Wilson have improved things considerably. By the PM peak, it is easier to have all service in place and so the capacity figures are better.

Note that these are percentages of scheduled capacity. The PM peak percentage is higher than AM, but this is against a wider scheduled value (25.5 trains/hour AM, 23.1 trains/hour PM).

The TTC uses “gap trains” (also known as “run as directed” trains) to fill holes in service caused by the inevitable minor delays which will occur. These are not inserted at terminals, and so are not subject to the headway limitations of existing terminal operations. In the AM peak, the trains can fill gaps southbound on Yonge to clear out waiting demand, particularly at Bloor. This allows the line capacity along the route to get back to scheduled level even if there is a gap from the terminal. The ability to do this on the subway is limited by both the number of gap trains scheduled and the locations where it is practical to insert them into service. In past years, there was a cutback in gap trains so that service could be improved at relatively low cost with the tradeoff of lost flexibility. Now the pendulum has swung back to recognizing that a gap train, even though it may appear to “do nothing” a lot of the time, is vital to maintaining service quality.

Current schedules provide 4 AM and 2 PM gap trains on Line 1, with 1 AM and 1 PM gap trains on Line 2. The August schedules will add a second gap train to each period for Line 2.

Surface routes are reported on the basis of terminal departures with “on time” defined as a window of +1 (early) to -5 (late) minutes to the scheduled time. This only shows all day figures, and does not indicate whether all of the scheduled service actually operated, or how much of it was running in packs. Surface route metrics are due for an update later in 2019, and I look forward to more meaningful information including crowding data.

Another oft-cited metric of surface operations is the decline in short turns. As I have already written (Zero Short Turns Does Not Equal Better Service), simply driving the short turn count down does not automatically mean that service overall will be better. I will leave further discussion until later in the year when the TTC produces its updated Service Plan.

Equipment Reliability

The TTC reports the Mean Distance Between Failure (MDBF) where a “failure” is defined as an event that delays service for five minutes or more. The T1 fleet on Line 2 is running well above its target level and the TR trains on Line 1 are doing even better.

These numbers bounce around a lot because there are few failures each month (4 on Line 1, 6 on Line 2 in May 2019). Therefore, a change of only ±1 event can have a big effect on the MDBF. It is the overall trend, not the individual monthly values which is most important as a “good” or “bad” month may just be a case of timing.

The bus fleet continues to operate with an MDBF of 20,000 km or better thanks to the retirement of the most failure-prone of older vehicles. In the short term, these stats will look good, but the real challenge will be to maintain them as a large fleet acquired in a short period ages.

The streetcar fleet’s performance improved in May with the Flexitys hitting just under 28k MDBF, their highest ever value. The contract target is 35k and the TTC hopes to reach this eventually despite a large backlog of outstanding fixes needed on this fleet. I asked TTC staff whether this number is being sustained into June and beyond, and they replied that the number will stay high, but will move up and down because there are so many outstanding fixes to be applied to the fleet.

The ALRV fleet is for all practical purposes retired, although a few cars remain available, in theory, as extras if required. The CLRVs continue to disappear as new cars arrive, although the reliability of those that remain is rising as the worst cars are retired.

The challenge for the TTC will come when the fleet is 100% Flexity and the size of spare ratio that is needed to sustain ongoing repair work. For example, if the peak service is 160 cars (June 2019 schedules), this would mean a spare factor of over 25% which is very high.

Line 2 Wheel Flats: Report and Presentation

From October 2018, the TTC received a large increase in complaints about noise from trains on the Bloor-Danforth subway, Line 2. The problem was not confined to one part of the line showing that this was not due to local changes (for example rising groundwater changing ground-borne vibration patterns) but to a problem with the trains. This co-incided with a very steep rise in the number of trains with flat wheels.

The typical proportion of trains with flats is about 4% of the fleet rising to 8% in the fall when rail conditions, mainly from falling leaves, increase the likelihood of trains sliding when they stop. However, in October 2019, 90% of the Line 2 fleet had “moderate to severe” flats producing higher than usual noise levels from almost all trains.

With so many flats, the TTC simply could not keep up with the work needed to restore wheels by machining out the flats, and noisy trains were inevitable along the line. Wilson Yard, which normally maintains Line 1’s TR fleet, took on some of the work on the Line 2 T1 fleet to supplement the capacity at Greenwood Shops.

Flats are created when a train that is stopping loses traction between the wheels and the brakes keep the wheel locked in position while it slides rather than gradually slowing. This can be triggered by various factors:

• Greasy rail: Under normal circumstances a stop on clean dry rail might be possible, but this will not work if the rail is greasy.
• Emergency braking: During an emergency stop, the brake shoes are applied sooner than in a normal stop, and this can trigger sliding depending on rail conditions.

The basics of traction between rails and wheels are not unlike what motorists or cyclists experience with various conditions. Both acceleration and braking depend on wheels turning without slipping. Although most of the braking effort on subway cars is done with regenerative braking through the motors (converting the train’s kinetic energy back into electricity), the last stage is done with brake shoes bearing on the wheels. Normally, the wheels are slowed by the brake shoes, but if there is poor traction between the wheel and the track, the wheel locks up and slides.

The TTC has a wheel flat detection system on both Lines 1 and 2 so that it can spot problem cars and take the worst offenders out of service for repairs. In the illustration below, note how the condition of the same train is much worse in a space of under 24 hours showing that there was some event that produced a large crop of flats.

The main report includes a detailed chronology of events from fall 2018 through spring 2019. What is quite evident is that there is no single cause behind the rise in flat wheels, and indeed, the TTC has still not settled on an individual “culprit” among the issues that were investigated.

• At some locations, an excessive buildup of grease from lubricators was found, in part due to old units that have since been replaced.
• A new method of controlling wheel squeal, a “top of rail friction modifier”, was installed on the curves west of Islington which are the source of constant noise complaints from nearby buildings when the lubrication system is turned off or fails.
• Many emergency brake applications were triggered by the Speed Control System (SCS) which monitors train speed against posted limits and triggers an emergency stop. There is a question of whether some of these events were errors, and changes are planned to SCS to reduce these events. Meanwhile, speed restrictions were implemented in some of the open cut areas to avoid triggering emergency stops on areas which, seasonally, have poor rail traction.
• A batch of wheels produced metal flakes that dropped to track level causing signal problems. Conversion to wheels from another manufacturer was already underway since 2016.
• A new type of brake shoe was installed to reduce wheel flaking, but these shoes may have contributed brake dust that added to problems with rail traction.

The main report contains substantially more detail on these issues for readers who are interested.

In January 2019, the TTC retained the National Research Council of Canada as part of a team from several departments to investigate the problem, and more recently Network Rail Consulting has been retained to follow up on the NRCC’s recommendations.

The rate of wheel flat production is, as of June 2019, down to typical levels, but the real test will come with the fall weather.

In preparing this article, I posed a number of questions to the TTC to clarify what they are doing:

Q1: Reading the report, one gets the impression that the old wheel greasing system seems to have gotten out of hand and was doing its job too well. Replacement with new units is part of the overall fix. Is this is a permanent change, and will all old units be changed out?

A1: All old units are planned to be changed over the next few years. The lubrication machine is estimated to have life cycle of 10-15 years and most of our machines are older than that.

Q2: There is reference to cleaning the track. Was this a one-time effort to get rid of a backlog of excessive grease, or a standing program that will occur seasonally when there is buildup of material such as leaves?

A2: Cleaning of track to remove flammable trash is a year round continuous Track Maintenance activity. The rail surface is either cleaned by spraying water with wash car, or manually with degreasing agent in response to reports of rail contamination around lubricators and in select locations in the Fall when leafs are causing slippery layer on rails.

Q3: What is the top of rail friction modifier and how is it dispensed so that it is spread along the track surface?

A3: More technical information on chemical composition and operation of this product can be found here.

Top of rail friction modifier (TOR FM) used at the TTC subway is a water-based solution mainly containing water, Propylene Glycol and Molybdenum Disulphide. After application to the top of rail and after the water evaporates, a paint-like film is formed between the wheel and rail which helps adjust the coefficient of friction to an optimum level. This optimum level of friction has many advantages including reduction of the noise emission from the wheel-rail interface while ensuring that the traction and braking of the vehicle is not compromised. It is dispensed with specially designed spreader bar to the the surface of the rail and is dispersed and carried by the wheel thread. Pictures of the dispenser bar can also be found on the site above.

Q4: There are references to changes in both the wheels and brake shoes, and that some of these changes have been backed out. Is it correct to say that the fleet is now operating with wheels and shoes of the type employed before the flats became a problem?

A4: No. The current wheels are in the process of being replaced with wheels from a different manufacturer for the last few years. Only a very small portion of the fleet (30 cars) were reverted to the older brake shoes for investigation into the wheel flats issue. Older brake shoes would be installed across the fleet after a full conversion from the old wheels to the new wheels in about 2-3 years. Wheels are replaced by attrition as they reach their life expectancy.

Q5: Is there any sense of the degree to which the wheel and brake shoe changes interacted with the excess of track grease to create a compound problem?

A5: This investigation is still in progress by Network Rail Consulting and TTC engineering, however at the moment there is no agreement that excessive grease contributed to the spike of wheel flats.

Thanks to the TTC’s Stuart Green for this information.