TTC Board Meeting October 16, 2017 (Updated)

The TTC Board will meet on October 16. Among items of interest on the agenda are:

Continue reading

Where is the Extra TTC Service?

When John Tory became Mayor of Toronto in the 2014 election, he quickly discovered that his transit briefings as a candidate were far from the truth. Service deteriorated under Mayor Ford, and although Tory opposed a plan to increase bus and streetcar service during the campaign, he had a change of heart. The then-new Mayor supported the purchase of more buses and a partial restoration of service standards, notably those affecting off peak service.

What has happened between the start of 2015 and fall 2017?

Over the past three years, TTC ridership growth has levelled off and is dangerously close to slipping into decline. Many factors are cited, including the “do nothing” option of “every other city is losing riders so we’re no different” approach. That is not exactly a call to restore riders’ faith in the TTC.

An important issue is the question of actual service levels. Looking at the service budget and actual operations, one might get the impression that service is growing.

The budgeted service in 2017 is higher than in 2015, but the details are important:

  • The budgeted regular service in 2017 is only slightly higher than in 2015. For September, the budget is 2.1% higher in 2017. However, not all of the planned increase was actually operated in response to the ridership numbers and budget pressures.
  • More than half of the total increase in service hours is due to construction, not to actual improvements.

One can argue that without the extra construction service, riders would have even worse service than they do, but this does not change the fact that regular service has improved very little over three years.

This can be seen in detail in the chart linked below which compares peak headways (the time between vehicles) for the January 2015 and September 2017 schedules.

201501_201709_Comparison

The bus changes summarized in pie form show the breakdown.

 

The total number of buses in service during peak periods is up from 2015, but this is entirely due to substitution of buses for streetcars either for construction projects or to offset the late delivery of Flexity cars from Bombardier.

When the total count of buses in service is adjusted for vehicles running on streetcar lines, the AM peak value is down slightly, and the PM peak is up only by 20 vehicles.

This does not account for extra vehicles dealing with congestion delays on the TYSSE project in 2015, nor the Crosstown project in 2017.

On a route by route basis, there are more routes with service cuts in 2017 relative to 2015, or with no change in peak service at all, than there are routes with better headways. Most streetcar routes have service cuts in both peaks.

This is hardly a system that is actively trying to attract ridership with better service and more capacity.

In coming months, the TTC expects to publish an updated bus fleet plan (October 2017) and a Ridership Growth Strategy (November 2017). A fleet plan requires not just buses, but also garages, operators and maintenance workers. Both of these reports will be vital inputs to the Capital and Operating Budgets. Good intentions alone will not bring more service to TTC routes.

An important component of the fleet plan will be the target spare ratio for vehicles. The TTC has increased this value in past years to ensure that its fleet gets the pro-active maintenance required to field revenue service every day. However, adding to spares makes the fleet larger without changing scheduled service.

Absent the details of coming budgets, it is unclear how much provision, if any, has been made for substantial service improvements. Capital subsidy programs such as Ottawa’s PTIF (Public Transit Infrastructure Fund) may let Toronto buy more buses, but if Toronto Council will not fund the operation and maintenance of the added vehicles, then this is nothing more than a make-work program for the bus builders and a way for the TTC to retire much of its fleet sooner than planned.

There is an inherent dishonesty in many transit funding announcements because they address only the initial cost of provisioning vehicles and building infrastructure, not actually running and maintaining them. The TYSSE alone will add about $30 million (net of new revenue) in costs to the TTC’s budget, and it is far from clear whether they will receive added subsidy, or be forced to trim service elsewhere to handle this. (Part of the increase was already digested in the 2017 budget as startup costs for the new subway, but most of the cost hits in 2018.)

Meanwhile, Bombardier is supposed to ramp up streetcar deliveries this fall reaching 7.5 cars/month by October. As of September 24, the second “September” car, 4446, is about to leave Thunder Bay, and it is clear that Bombardier will not make the target of four cars for this month (ending with 4448). It is hard to believe that they will actually ramp up to a much higher production rate, and this threatens service quality on both the streetcar and bus networks. Even if Bombardier does begin to deliver at the planned rate, there remains the question of whether the TTC will restore streetcar operations on all routes, or continue to retire older cars.

The TTC faces several difficult years in a political environment where added spending on transit operations will be hard to achieve in the face of “tax fighting” politicians. The one product they are selling is service. Without real improvements, not just marketing flim-flam, the TTC will lose political influence and, with that, the funding they so badly require.

Sources:

  • TTC Scheduled Service Summaries (copies available on this site)
  • TTC Service Planning Board Period Memos

TTC Service Changes Effective October 15, 2017

The service changes for the October-November period are relatively minor.

On the streetcar network, provisions for construction diversions on 505 Dundas and 506 Carlton end, but for 501 Queen and related routes a new project at McCaul plus the ongoing streetcar shortage leave service in that corridor with temporary schedules:

  • 501 Queen streetcars will operate from Neville to Roncesvalles, but they will divert via Church, King and Spadina for about three weeks during the reconstruction of the intersection at McCaul.
  • 501L Long Branch and 501M Marine Parade bus service continues as it was in September.
  • Bus trippers have been shifted from King to Queen where they will operate between Woodbine and Sunnyside.
  • Kingston Road services 502 and 503 continue to operate with buses.

The 511 Bathurst route continues to operate with buses although the introduction of Flexitys on 512 St. Clair is releasing CLRVs that could be redeployed elsewhere. How quickly streetcars will return to temporarily bused routes remains to be seen.

Although schedules for routes at the Renforth Station (formerly “Gateway”) were altered in September, this facility is not yet complete and routes continue to operate to their former termini.

2017.10.15_Service_Changes

TTC Service Quality Update for September 2017

As part of the September CEO’s report to the TTC Board, management made two presentations on the evolution of service quality on the surface and subway networks.

For the surface routes, Rick Leary (Deputy CEO and Chief Service Officer) ran through an overview of two sets of statistics: on time performance and the numbers of short turns.

For the subway, Jim Ross (Deputy Chief Operating Officer) reviewed schedule versus actual capacity provided on Line 1 YUS as well as a breakdown of the types of incidents that produce most delays.

Continue reading

510 Spadina Platform Delays

From time to time, readers mention in comments the common delays at Spadina Station caused by the manner in which streetcars unload and load at the platform.

Before the introduction of the new Flexitys, cars would enter the station, unload at the east end of the platform, then pull ahead to load at the west end. There was even enough room that three cars could be on the platform at once with one ready to move forward into the loading zone as soon as room was available.

However, with the longer Flexitys, two cars will not quite fit on the platform, and although there are still separate unloading and loading stops, in practice only one car can be on the platform at a time. When this is compounded with delays for crew changes and with the siestas some cars take while loading, delays to passengers waiting to get off of arriving cars are chronic and lengthy.

I asked the TTC’s Brad Ross why cars did not make better use of the space.

An option that comes up is for the loading car to pull slightly beyond the platform and not open its front-most door. This would allow the following car to come fully onto the platform and unload. The front of the loading car would not be in the tunnel itself, but adjacent to an unused area of the platform at the west end.

The reply from the TTC is:

  1. It is not possible to isolate the front doors of the LFLRV.  If the streetcar were to go past the glass barrier to allow a second streetcar to access the platform, we would not be able to open the remaining 3 doors, while keeping the first door closed.  All doors would closed but enabled, and would require our customers to push the door button to gain access to the streetcar.  In addition, we would not be able to prevent anyone from opening the first set of doors from the inside of the streetcar (or even from outside, should a customer go around the glass wall and push the button).  This is a huge safety issue, as the front doors are not flush with the platform at that position, and customers would have to step down to the rail bed.
  2. One obvious solution would be to extend the platform so that it does sit flush with the front doors, but that is not an option at this point.  The issue actually lies with the structural pillars that support the station; they are too close to the streetcar to allow proper egress.  This is where I am seeking clarification from the Construction group.  I am unsure of which pillar is in the way and what the actual requirement is, but have heard that it would cost millions to relocate the pillar to allow us to extend the platform.  That is why that section of the platform is not being used and the glass partition is in place.

This begs the rather obvious question of why it is not possible for an operator to selectively open doors on a car. It is not unknown for vehicles to be in locations where a physical barrier would prevent use of all doors. In the specific case of loading at Spadina Station, the loss of the comparatively narrow first door would be a good tradeoff for simply getting a car far enough into the loading area that its follower could unload behind it.

I await further feedback on the matter of the cost and practicality of modifying the station, but in the meantime, it is useful to look at this problem in the manner we normally see for evaluations of expensive rapid transit projects: the value of riders’ time. Billions in spending on rapid transit has been justified by the premise that people will move more quickly and thereby save time, time that has a value against which the capital investment can be offset.

(This is a dubious proposition because the public investment is “balanced” against a private saving in “money” that can never be recaptured, but stay with me for the purpose of the exercise.)

  • There are 15 cars per hour attempting to serve Spadina Station. If we assume that each car is delayed by an average of 2 minutes, and that it has an average of 50 passengers, this translates to 1,500 minutes of passenger delay per hour.
  • If this condition persists on average for 8 hours per day, that means there are 12,000 passenger minutes of delay, or 200 passenger hours.
  • The value of riders’ time is often quoted at about $30/hour, and this means a value of lost time of $6,000 per day. Scaling up to a year with a 300:1 factor (counting weekends as one day) gives us an annual lost time value of $1,800,000.

It is self evident that any of the variables used here can be tweaked up or down, but this gives the general idea of how the calculation would come out. Of course the City cannot “spend” that $1.8 million to offset reconstruction at the station because it is not real money, as I’m sure we would be told by the financial boffins.

In addition to any notional saving that riders might obtain, there is the real cost of, probably, one extra car on the Spadina route thanks to the extra running time needed to serve Spadina Station Loop.

There is a real need for the TTC to sort out operations at Spadina Station to minimize delays. This should include both figuring out how to use the loading area with a car projecting slightly beyond the platform and ensuring that crew changes happen as briskly as possible.

I will update this article when/if more information becomes available.

Updated Aug. 14, 2017 at 4:05 pm

Based on Twitter feedback, a few comments are in order.

In this article, I did not refer to explicit methods of crewing, but used the term “drop back” on Twitter. Some took umbrage saying that the term is “step back”. In fact both terms have been used over the years. It is the mechanism that counts, not the name. The idea is that operators get a break without the car having to sit there while they take it. To do this, there are more operators than cars, and an operator from car “n” “steps back” to car “n+2” (or whatever) so that they get a break of, in theory, two headways (the number can vary depending on how long a break is desired). This works fine as long as the operator who is supposed to take over an incoming car is actually available when it shows up.

At a location like Spadina Station, there is no stacking space to accommodate late crew changes whether they are part of a step back system or a regular shift change, and the problem can be compounded when the home division for the line is a long way away, and operators have to travel to pick up their cars in service.

Delays of cars getting on to the platform because of the loading techniques just make this worse by holding cars in the tunnel.

My aim in writing this article was twofold:

  • Everyone involved – TTC management, planners, line supervisors and operators – need to work together to find a way to improve operations at this busy station which have deteriorated noticeably since the new cars were introduced.
  • The principle of “value of riders’ time” is often used to justify big ticket capital projects, but it does not have the same clout in day to day operations.

 

TTC Service Changes Effective September 3, 2017

The September 2017 schedule changes primarily involve the reversal of summer service cuts to many routes with only minimal service improvements. This continues the TTC’s policy for 2017 of constraining service growth in the face of lower than budgeted ridership, as well as the shortage of vehicles.

Construction projects continue to affect route 501 Queen and will do so for many months to come:

  • Streetcar service is restored between Russell Carhouse (at Connaught) and Sunnyside Loop.
    • This will be affected in October when the intersection of McCaul and Queen is rebuilt requiring a diversion.
  • A bus shuttle will operate from River to Neville Park due to the reconstruction of the intersection at Coxwell and Queen.
    • This will also require the continued operation of buses on the 502/503 services on Kingston Road.
    • Through streetcar service to Neville Park will resume with the October schedules.
  • A bus shuttle to Long Branch will operate from Dufferin Loop, and Marine Parade will be served by its own local shuttle to Windermere.
    • Construction on The Queensway will prevent streetcars from operating to Humber Loop until the end of the year.
    • Streetcars will not operate west of Humber Loop to Long Branch until mid-2018.

With the return of ALRVs to the Queen route, 504 King will operate primarily with CLRVs, and the peak period trippers will mainly be buses, not streetcars. The effective capacity of the route will fall because of the lower capacity of CLRVs and buses versus the streetcars that have been used over the summer of 2017. This will be minimally offset by a small reduction in headways during all operating periods thanks to trimming of the running time. King cars now enjoy extended layovers leading to queues of vehicles at terminals thanks to an overly-generous schedule. The number of streetcars in service remains the same, but on slightly shorter headways.

New low-floor Flexity streetcars will be deployed on 512 St. Clair starting in September, subject to availability. The schedule will be based on CLRVs until new car deliveries reach the point where the line can be scheduled as a Flexity route.

The TTC plans to begin using Flexitys on 504 King late in 2017 subject to availability.

Between them, the King and St. Clair routes require about 60 CLRVs at peak. Allowing for some capacity growth with Flexitys, this translates to about 45 of the new cars, plus spares. It will be some time before both routes are converted, assuming Bombardier achieves their ramped up delivery rate in fall 2017. They are already slightly behind schedule with only two of three planned cars for July 2017 out the door in Thunder Bay, and they have not yet implemented the additional shifts/workforce to produce cars at a higher rate effective October 2017.

The northbound stop at Broadview & Danforth will be removed allegedly in the aid of transit priority signalling. In fact, this is a location where the substantial green time afforded to east-west traffic on Danforth makes the idea of “priority” for transit movements difficult to swallow. There is already an advanced green northbound for left turning motor traffic. Given the layovers now enjoyed by streetcars at Broadview Station, it is not clear just what this priority will achieve, but removing the stop will annoy the many riders who now use it. The southbound stop remains in service.

Other construction projects include:

  • 54 Lawrence East: Water main construction west of Victoria Park has completed.
  • Renforth Station opens: 32 Eglinton West and 112 West Mall are rerouted into the new regional terminal.
  • Kennedy Station: The schedule change to accommodate Crosstown construction is implemented for weekend service on 86 Scarborough.
  • Long Branch Loop: All buses will loop via the streetcar loop during reconstruction of the bus roadway.
  • 123 Shorncliffe: Additional running time to accommodate a City paving project.
  • 506 Carlton: The only remaining construction area/diversion is on College between Bathurst and Lansdowne. This will end in October.
  • 505 Dundas: The diversion between Bay and Church will end in late September or early October depending on progress of the road works east of Yonge.

The 400 Lawrence Manor and 404 East York Community Bus services will be extended. For details, see the TTC’s July 2017 update on these services.

2017.09.03_Service_Changes

514 Cherry Update re King & Sumach Noise

At a community meeting on June 27, 2017, the TTC presented updated information about their work on reducing the noise level from streetcars at King & Sumach. In response to complaints after the 514 Cherry route began operating in 2016, the TTC changed the 514 so that late evenings and early mornings it operates to Broadview & Queen (looping back via Dundas and Parliament just like a short turning 504 King car). During these periods, a Wheel Trans bus provided a shuttle service on Sumach and Cherry to Distillery Loop.

The TTC presented updated noise readings for this location showing the combined improvement of the full changeover to Flexity cars from CLRVs and of changes to the rail profile that were made to complement slower operation around the curves.

The chart above shows results for the tightest curve at King & Sumach, the east to south. The data plotted here summarize readings taken over a four-hour period, and so they reflect the contribution of whatever type of vehicles showed up. For the most recent reading on May 4, 2017, when the service should have been largely or completely run with Flexitys, the levels from the middle to the high end of the spectrum are markedly lower than they were in the fall.

From the vehicle tracking data for 514 Cherry, I can confirm that the vehicles in service on that date were:

  • CLRV 4071 from 8:28 to 9:55 am
  • CLRV 4049 from 2:56 to 5:45 pm
  • Flexitys 4402, 4404, 4406, 4409, 4412, 4414, 4423, 4425, 4428 and 4432

Depending on when the measurements were taken, there was at most one CLRV in service on the route, and none for most of the day.

By contrast, on Aug. 10, 2016, all but one car on the route was a CLRV with only a single Flexity in service, 4418, between 5:30 am and 2:07 am the following day.

For the north to west turn, the data show less of an improvement. Oddly, the readings for the gentler left turn curve are higher than for the eastbound right turn, but this could be a factor of the measurement location which is closer to the westbound turn.

As a matter of comparison, the TTC also presented readings from two intersections with comparable curve radii, Queen & Broadview and Bathurst & Fleet.

Note that this chart presents maximum values rather than a four hour average. The higher values for the comparator intersections are almost certainly due to the noise caused by CLRVs or ALRVs which have (a) inherently more squeal and (b) less car design factors to limit noise transmission.

Bathurst & Fleet would have had service only on 509 Harbourfront on May 4 as this predates the return of streetcars to 511 Bathurst. I do not have the tracking data for the 509 on that date, and so cannot comment on the proportion of service provided by each vehicle type. Harbourfront is supposed to be all Flexity, but routinely has a few CLRVs on it. It would take only one noisy CLRV to set the maximum values shown above.

The chart is also unclear about which turn was measured at each location, only that this was done from 8 metres away.

Future work of this type should be more careful in identification of the vehicle type and location specifics for any readings and charts. If nothing else, this will improve credibility with members of the public by showing the improvements new cars bring.

Based on the improvements recorded at King & Sumach, the TTC plans to return full streetcar service to Distillery Loop on a date to be announced in July.

This decision provoked something of a pitched battle between residents at various locations on the route. The high points (if they can be called that) included:

  • Wheel squeal at King and Sumach prevented some nearby residents from getting a full night’s sleep, and the respite with no cars making turns was 3 to 3.5 hours. (It was unclear whether the residents have ever had a Flexity-only late night or early morning service as a reference point because service was cut last November before the route conversion was completed.)
  • Squeal is worst after rain because the normal film of grease on the track (both from natural causes and from wheel greasers) washes away. Wet track actually is very quiet because the water acts as a lubricant, but track that is drying out can be extremely noisy. This also happens during periods of high humidity. The TTC was criticized for taking noise measurements only under ideal conditions.
  • Residents at King/Sumach who predate the installation of the intersection were used to quieter streetcar operation, and enjoyed a long period of no streetcars at all while the King leg of the Don Bridge was closed.
  • The Wheel Trans shuttle bus is utterly unreliable running on a schedule unknown to riders and with unpredictable headways that can be considerably longer than the round trip route would imply. Operators often bypass waiting passengers. There are safety issues for the large number of disabled transit users living in this neighbourhood if they are forced to make a transfer to an unreliable, infrequent service.
  • Residents along the Cherry Street portion of the route complained that they effectively lost service because the bus was so unreliable, and in any event, its wide headways and forced transfer at King Street added to travel times. They also noted that the change was implemented without notice to the wider community. (There were also complaints about poor publicity for the June 27 meeting.)
  • Aggrieved King/Sumach residents proposed that the 514 Cherry route be completely converted to bus operation during the hours when the shuttle runs now to eliminate the transfer connection and improve service to the Distillery. This option was rejected by the TTC and by some users of the 514 who noted that streetcars can be very crowded at late evenings downtown where the route is supposed to provide supplementary service on King.
  • Early morning trips from Leslie Barns to Distillery Loop make the west to south turn for which no automatic greasing is provided.
  • Not all who attended from King/Sumach objected to the streetcars, but as this was a small meeting, it is not clear what the balance of opinion in the neighbourhood might be.
  • Notable by its absence from any comments were complaints about noise from eastbound streetcars clattering through the trailing switch of the north to east curve. The slow order at this location appears to have dealt with this issue.

In addition to operating the 514 Cherry route with only Flexitys, the TTC is working on a design of a noise absorbing ring that will damp the high frequency vibrations. Wheel sets for two cars are now being manufactured, and they will be installed on test cars in the fall.

Further noise readings will be taken through the summer and fall to track conditions as they evolve, and the level of grease application will be increased. (There is a trackside greaser southbound at Distillery Loop, and the Flexitys have on board greasers that are triggered by GPS information to activate where lubrication is required.)

In a separate article, I will turn to the general unreliability of service at Distillery Loop on the 514 streetcars. The TTC puts this down to the usual problems of mixed traffic operation on King, but there are also issues with uneven headways departing from both the Distillery and Dufferin terminals following layovers that can be fairly long. Line management, as elsewhere on the system, is a problem for this service.

See the TTC’s King-Sumach page for complete information.

 

A Slow Trip to Express Buses (Updated)

Updated June 19, 2017 at 3:30 pm: The TTC has clarified that the hourly costs shown for various routes are net costs, not gross costs, and this addresses my concern that some of these values were understated. The text of this article has been updated where appropriate.

Updated June 20, 2017 at 10:30 am: A section has been added on gross operating costs (the TTC study includes only net costs) to illustrate how these vary from route to route.

The TTC Board recently approved an Express Bus Network Study that proposes several new and enhanced express routes in Toronto. The premise of the study – improving the bus network’s attractiveness and convenience to riders – is a good one addressing the basic function of any transit system. However, thanks to the TTC’s severe constraints on capital and operating funding, the actual implementation of these proposals drags out for the better part of a decade. Bus planning in Toronto is converging with new subways for a lengthy gestation period.

There are three types of “express” route in Toronto:

  • The (mainly) peak period services usually signified by an “E” suffix on the route number. Typically, buses will run express over part of a route stopping only at major transfer points or destinations, and will continue as local service on the outer part of the route. These services are useful for riders who would otherwise face a long stop-and-go trip on a local bus for their entire journey. By carrying the long-distance riders express for part of the trip, the cost of route operations is reduced from that of an all-local service and provides a more attractive service overall.
  • The Rocket services (route numbers in the 18x-19x series) operate for most of the day on weekdays and weekends, and provide a more limited stop service, end-to-end, than the “E” branches. Some are designed around major endpoints such as the 192 Airport Rocket from Kipling Station to Pearson Airport, while others more closely resemble the stopping pattern of “E” services. Unlike the “E” branches, the Rockets do not necessarily duplicate the route of a local service.
  • The five Downtown Premium Express services (route numbers 14x) charge an extra fare for the privilege of avoiding the crowded Yonge subway and the 501 Queen car.

The TTC proposes an interim classification of the first two of these as Tier 1 (Rockets) and Tier 2 (“E” branches) with the intent of coming up with some sort of branding that could be used to market them. Some cities have special bus services with their own names such as Hamilton’s B Line Express and Vancouver’s 99 B-Line. Given that there already is an established name for Tier 1 with a strong Toronto reference, it is hard to understand why a new brand is required. As for the Tier 2 services, riders are well acquainted with the “E” convention (broken in rare cases such as 60F Steeles West).

The first recommendation of this study is that a marketing effort is required to brand these services. That says a lot about where the TTC’s focus has been in recent years – selling the “pizzazz”, to quote a former TTC Chair, while the system gradually declines thanks to penny-pinching by two successive City administrations.

Summary

This is a long article. Here are the highlights:

  • Growth of express services is limited as much by the political question of transit funding as it is by planning and resource constraints within the TTC.
  • The TTC’s bus fleet plan should be thoroughly reviewed to determine how more buses can be made available for service sooner than 2019/20 when McNicoll Garage opens.
  • TTC budgets should reflect a return to full streetcar service on the streetcar lines in 2018 and the redeployment of replacement buses back to the bus network.
  • Express bus routes that were added in 2016 have performed better than expected showing that these are popular services and should be expanded as soon as possible.
  • New Rocket and express routes are proposed in two waves, one for 2019-21 and the second with a tentative date of 2026.
  • Costs and revenues allocated to existing and proposed routes should be verified.
    • Update (text deleted): The costs shown for some routes appear to be in error if the methodology for costing used by the TTC is to be believed. (This issue has been referred to the TTC for comment.)
    • Revenue allocation in a flat fare system can distort the benefit of a route, and the measure of value should be based on usage not on a misleading allocation of fare revenue.
  • Express buses provide a means of carrying riders on a route with a mix of short and long haul demand more efficiently and attractively than an all-local service.
    • Riders on express services travel further than on local services taking advantage of the faster trip between major points on the route. There are fewer riders per bus kilometre because there is less turnover of the passenger load on express services.
    • Express buses cost more per ride than local buses because of their lower turnover (i.e. more riders), but the overall route cost is lower with a mix of services.
    • A proposal by one member of the TTC Board to charge extra for express routes would be counterproductive.
  • The Premium Express buses to downtown operate at a very high cost per passenger, although this needs to be verified in light of the issue with cost calculations. Demand on these routes is relatively light, and they contribute only trivially to reducing demand on parallel services. The TTC proposes to leave them in place at least until 2021, but this should be reviewed even though removing the services will be politically challenging.
  • Transit Signal Priority is not just an issue for an Express network, but for transit in general. It should be pursued on major routes whether or not they include express operations.
  • Route supervision will be essential to maintaining reliable service not just on express routes, but on the transit system overall.
  • The staff proposal to “brand” services continues the TTC’s focus on marketing when what is needed is service. The “Rocket” name is already well-established as a service type in Toronto, as are the “E” express branches of various routes. In a few cases, the “wrong” name is associated with a service (some “rockets” are really just “E” services in terms of their service patterns), but this does not justify a complete rebranding exercise.

Continue reading

The Evolution of Streetcar Service from 1980 to 2016

Transit service on many of Toronto’s streetcar lines has declined over past decades and, with it, riders’ faith in and love for this mode. Unreliable, crowded service is considered the norm for streetcar routes, and this leads to calls to “improve” service with buses.

The historical context for this decline is worth repeating in the context of current debates over how Toronto should provide transit service to the growing population in its dense “old” city where most of the streetcar lines run.

When the TTC decided in late 1972, at the urging of City Council, to reverse its long-standing plans to eliminate streetcars by 1980 (when the Queen Subway would take over as the trunk route through the core), the level of service on streetcar lines was substantially better than it is on most routes today. Any comparison of streetcars versus buses faced the prospect of a very large fleet of buses on very frequent headways roaring back and forth on all major streets.

Service in 1980 (when the system was originally planned for conversion) was substantially the same as in 1972, and for the purpose of this article, that date is our starting point.

Ten years later, in 1990, little had changed, but the City’s transit demand was about to fall off a cliff thanks to a recession. During this period, TTC lost much riding on its network including the subway with annual travel dropping by 20% overall. It would take a decade to climb back from that, but various factors permanently “reset” the quality of service on streetcar routes:

  • During the recession, service was cut across the board, and this led to a reduction in the size of fleet required to serve the network.
  • In anticipation of the 510 Spadina line opening, the TTC had rebuilt a group of PCC streetcars, but these were not actually needed for the 509/510 Harbourfront/Spadina services by the time Spadina opened. “Surplus” cars thanks to the recession-era service cuts were available to operate the new routes.
  • Since 1996, any service changes have been  made within the available fleet, a situation compounded by declining reliability of the old cars and the anticipation of a new fleet “soon”.
  • By 2016, the fleet was not large enough to serve all routes, and bus substitutions became common.

Some of the decline in demand on streetcar routes came from changing demographics and shifting job locations. Old industrial areas transformed into residential clusters, and the traffic formerly attracted to them by jobs disappeared. Meanwhile, the city’s population density fell in areas where gentrification brought smaller families to the houses.

The city’s population is now growing again, although the rate is not equal for all areas. Liberty Village and the St. Lawrence neighbourhood are well known, visible growth areas, but growth is now spreading out from both the King Street corridor and moving further away from the subway lines. This creates pressure on the surface routes in what the City’s Planners call the “shoulders” of downtown.

As the population and transit demand have rebounded, the TTC has not kept pace.

The changes in service levels are summarized in the following spreadsheet:

Streetcar_Services_1980_To_2016 [pdf]

510 Bathurst: In 1980, this route had 24 cars/hour during the AM peak period, but by 2006 this had dropped by 50% to 12. In November 2016, with buses on the route, there were 20 vehicles per hour, and with the recent reintroduction of streetcars, the peak service was 10.6 ALRVs/hour, equivalent to about 16 CLRVs. Current service is about 1/3 less than it was in 1980.

506 Carlton: In 1980, this route  had 20 streetcars/hour at peak, but by 2016 this was down to 13.8.

505 Dundas: In 1980, service on this route had two branches, one of which terminated at Church after City Hall Loop was replaced by the Eaton Centre. On the western portion of the route, there were 27 cars per hour, while to the east there were 15 (services on the two branches were not at the same level). By 2016, this was down to 10.3. [Corrected]

504 King: This route, thanks to the developments along its length, has managed to retain its service over the years at the expense of other routes. In 1980, there were 25.2 cars per hour over the full route between Broadview and Dundas West Stations with a few trippers that came east only to Church Street. Despite budget cuts in 1996 that reduced service to 16.4 cars/hour at peak, the route came back to 30 cars/hour by 2006. Service is now provided by a mixture of King cars on the full route (15/hour), 514 Cherry cars between Sumach and Dufferin (7.5/hour), and some trippers between Roncesvalles and Broadview. Some 504 King runs operate with ALRVs and most 514 Cherry cars are Flexitys.

501 Queen/507 Long Branch: In 1980, the Queen and Long Branch services operated separately with 24.5 cars/hour on Queen and 8.9 cars/hour on Long Branch at peak. By 1990, the Queen service had been converted to operate with ALRVs and a peak service of 16.1 cars/hour, roughly an equivalent scheduled capacity to the CLRV service in 1980. By 1996, Queen service was down to 12 ALRVs/hour of which 6/hour ran through to Long Branch. Headways have stayed roughly at that level ever since. The Long Branch route was split off from Queen to save on ALRVs, and as of November 2016 6.3 CLRVs/hour ran on this part of the route. Bus replacement services are operating in 2017 due to many construction projects conflicting with streetcar operation.

502 Downtowner/503 Kingston Road Tripper: In 1980, these routes provided 15.6 cars/hour, but by 2016 this had declined to 10/hour.

512 St. Clair: In 1980, the St. Clair car operated with a scheduled short turn at Earlscourt Loop. East of Lansdowne, there were 33.3 cars/hour on St. Clair. By 1996 this was down to 20.6 cars/hour. The next decade saw an extended period of reconstruction for the streetcar right-of-way, and service during this period was irregular, to be generous. By 2016, the service has improved to 21.2 cars/hour, but this is still well below the level of 1980.

What is quite clear here is that the budget and service cuts of the early 1990s substantially reduced the level of service on streetcar routes, and even as the city recovered, the TTC was slow to restore service, if at all. The unknown question with current service levels is the degree to which demand was lost to demographic changes and to what extent the poor service fundamentally weakened the attractiveness of transit on these routes. The TTC has stated that some routes today are operating over capacity, but even those numbers are limited by the difference between crowding standards (which dictate design capacity) and the actual number of riders who can fit on the available service. It is much harder to count those who never board.

In a fiscal environment where any service improvement is viewed negatively because it will increase operating costs, the challenge is to turn around Council’s attitude to transit service. This is an issue across the city and many suburban bus routes suffer from capacity challenge and vehicle shortages just like the streetcar routes downtown.

The bus fleet remains constrained by actions of Mayor Ford in delaying construction of the McNicoll Garage with the result that that the TTC has no place to store and maintain a larger fleet even if they were given the money to buy and operate it. Years of making do with what we have and concentrating expansion funding on a few rapid transit projects has boxed in the TTC throughout its network.

Transit will not be “the better way” again until there are substantial investments in surface fleets and much-improved service.

The Cost of Running the Queen Car

Update: Minor changes were made to add some details to the costings presented here at about 10:10 am on June 14.]

The debate over which type of transit vehicle should operate on Queen Street, and by implication on the wider streetcar network, will inevitably get into the question of the cost of streetcar operations. The TTC has cited large ongoing costs of the bus operation:

This summer, the TTC is spending an extra $1 million per month to run buses on the route, according to TTC spokesperson Brad Ross. It also takes 60 buses to provide similar service to the 501 Queen’s usual 45 streetcars.

“Queen is a good example of a route where streetcars make good sense because of the capacity that they offer you in the downtown to reduce congestion,” Ross said, adding that Toronto’s streetcars produce lower emissions than buses.

[From CBC News Toronto]

The ratio of buses to streetcars in this quotation is somewhat misleading for a few reasons:

  • The bus service is scheduled with extra running time in anticipation of construction delays, although the actual construction has not yet begun. This is responsible both for the accumulation of large numbers of buses at terminals.
  • The replacement ratio of 1.3:1 is well below values the TTC normally uses in comparing transit modes, and in their own crowding standards. The design capacities of vehicles for service planning is 51 for a standard bus, 74 for a standard-length streetcar (CLRV), 108 for a two-section articulated streetcar (ALRV) and 130 for the new low floor Flexitys. This implies a replacement ratio of 1.45:1 for CLRVs, 2.12 for ALRVs and 2.55 for Flexitys. These numbers would be adjusted downward to compensate for faster operating speeds with buses, if any, although that adjustment would vary by time of day and route segment as shown in my analyses of operations on the route.
  • The capacity of scheduled bus service is less than the scheduled capacity of streetcars at the beginning of 2017. Service for 501 Queen is based on the capacity of ALRVs.
  • The actual streetcar service on Queen before buses began taking over was scheduled to use 33 ALRVs and 7 CLRVs (November 2016 service). The CLRVs were dedicated to the service between Humber and Long Branch Loops.

The TTC’s methodology for allocating operating costs to routes is based on three variables:

  • Vehicle hours (primarily the cost of drivers and related management and overhead costs)
  • Vehicle kilometres (part of the day-to-day cost of running and maintaining buses including fuel)
  • Vehicles (part of day-to-day costs for work such as dispatching, routine inspections and maintenance, cleaning)

The cost of routine streetcar track maintenance is included in the vehicle kilometre cost. This does not include major projects such as the replacement of track which are funded from the Capital Budget.

The factors for the two modes as of 2015 were:

                   Per Hour      Per Km        Per Vehicle
                                                 per Day
     Buses         $ 92.30       $  1.88       $  150
     Streetcars    $ 95.40       $  3.42       $  515

     [Source: TTC Service Planning via Stuart Green in TTC Media Relations]

As 2015 costs, these numbers contain almost no contribution from the new Flexity fleet, but they will be influenced by the cost of maintaining decades old CLRVs and ALRVs. The hourly component of streetcar costs is probably influenced by the relatively higher level of route supervision on that network than on the suburban bus routes.

The TTC’s most recently published detailed statistics for their network date from 2014. (The lack of timely data on route performance is an ongoing issue, but one that is separate from this article.) For 501 Queen, the daily factors for 2014 operation were:

     Vehicle Hours    595
     Vehicle Km     9,100
     Vehicles          36

The number of vehicles listed is lower than the peak requirement, and this will affect the calculated cost as discussed below.

When the streetcar costs are applied to these factors, the daily cost of the Queen car comes out to just over $100k (2015).

     Hourly costs     $ 56,763  53.3%
     Kilometre costs    31,122  29.2%
     Vehicle costs      18,540  17.4%
     Total            $106,425

Adjusting this for the higher number of streetcars actually shown in the schedules would add 4 vehicles (40 vs 36) at a daily cost of $2,060.

On an annual basis (taking one year as equivalent to 305 weekdays, the factor used by the TTC to account for lower demand on weekends and holidays), the Queen car costs about $32.5 million (2015) to operate.

Update: This does not include the cost of the 502 Downtowner nor the 503 Kingston Road Tripper cars. Annualizing the premium for bus service quoted by the TTC to $12m/year puts the relative cost by their estimation in context.

The important point here is that the hourly costs account for about half of the total, and so any calculation is most sensitive to the number of operators required to provide service. Larger vehicles have a strong advantage over smaller ones. Also, larger vehicles mean lower costs for vehicle distance travelled and per vehicle costs, but it is not certain that for a large-scale change in fleet composition that these cost factors would remain stable depending on just which cost components are allocated to each category. For example, a carhouse costs the same amount to operate whether it has 200 small cars or 100 large ones in it. Extrapolation to an all-Flexity environment should be done with care.

In the case of a bus operation, provided that the average speed could be increased during peak periods, this would reduce the total vehicle requirement and bus hours, but it would not change the bus kilometres in comparison to buses scheduled at the same speed as streetcars. (Fewer vehicles travelling at a higher speed run up the same mileage.) The big difference would come in vehicle (operator) hours because of the lower capacity of buses.

The problem of projecting a replacement cost then becomes one of “twirling the dials” of various factors to determine what the replacement service might look like. One obvious starting point is that this must be based on normal route conditions, not on the non-standard schedules now in use for the construction period. Possibilities include:

  • Using an ALRV:Bus replacement ratio of 2:1
  • Using a lower replacement ratio such as 1.5:1 (a sensitivity test to determine how costs would change with larger buses)
  • Using the 2:1 capacity ratio, but assuming a higher average speed for buses
  • Using the higher capacity of Flexitys

The results from these assumptions should be taken with considerable caution because it is far from certain that the cost factors can actually be relied upon across the different vehicle types and usage patterns.

  • On a 2:1 replacement ratio, the cost of bus operation is about 50% higher than for ALRVs. Costs allocated per vehicle are lower, even though there are more buses, but this is more than offset by higher costs for the hourly and mileage components.
  • On a 1.5:1 replacement ratio, the cost of buses is about 10% higher than for ALRVs.
  • On a 2:1 replacement ratio, but with a 10% increase in average speed, bus costs go down about 8%, but are still about 1/3 higher than the cost for ALRVs.
  • For Flexity operations, assuming cost factors are unchanged (valid for hourly costs, but mileage and vehicle costs are another matter), the replacement bus service would cost about 75% more than the streetcar service.
  • Flexity costs fall by 1/6 relative to ALRVs because of the larger Flexity design capacity. This is a comparatively small saving on Queen because the route is already scheduled (if not actually operated) as if it had the larger ALRVs on it. If we were looking at 504 King, for example, the schedule is based on CLRVs and so the replacement by buses would require many more vehicles proportionately than for the Queen route, and replacement by Flexitys would require many fewer vehicles to provide the same scheduled capacity.

[Note: I have deliberately not published exact numbers here because this is only a rough estimate subject to alteration as and when the TTC refines its cost base and the assumptions behind a comparative service design. Also, it is based on 2015 cost data and 2014 schedules.]

These costs do no include major capital projects including ongoing renovation of streetcar track, and one-time costs to bring infrastructure (notably the overhead power distribution system) up to modern standards.

The annual cost of surface track and special work (intersections) varies from year to year based on the scheduled work plans. The average for tangent track over 2017-26 is about $21 million/year although the amounts for 2017 and 2018 are particularly high due to the extent of planned work in those years. From the point where the TTC decided to retain streetcars in late 1972 until 1993, their track construction was not of a standard required for the long life expected of rail assets. Track was not welded, untreated wooden ties were used, and there was no mechanical isolation for vibration between the track and the concrete slab in which it was  laid. The result was that roadbeds fell apart quickly and the lifespan of the infrastructure was about 15 years.

Beginning in 1993, the TTC changed to a much more robust track structure using a new concrete base slab, steel ties, welded rail and rubber sleeves to isolate the track from the concrete around it. The structure is designed so that when track does need to be replaced, only the top layer, the depth of the track itself, needs to be removed. New track can be attached to the steel ties that are already in place. Conversion to this standard across the entire system is almost completed, and track reconstruction costs will drop due both to longer lifespan and simplified renewal work.

The average for special work over 2017-26 is about $14 million/year. Starting in 2003, the complex castings were set in a vibration-absorbent coating. Construction techniques have also advanced so that intersections are pre-assembled and welded off-site and then trucked to street locations for installation in large panels. The most recent intersection, Dundas and Parliament, went from initial demolition of the existing track to full assembly of the new intersection in one week. (Further work was required to complete other road upgrades, and new intersections are typically allowed to cure for a few weeks so that the concrete does not suffer vibration before it has properly set.) With a roughly 30-year cycle for special work replacement, the TTC is only about half way through rebuilding all of its intersections to the new standard.

Update: The Queen route represents about 28% of the track in the streetcar system, and so is responsible for about $10m of the annual capital work averaged over its lifetime. This is a relatively high proportion for one route, especially in relation to the amount of service operated there. 504 King, for example, is much shorter and has considerably more service than 501 Queen.

The cost of track replacement is essentially a fixed value that varies little with the level of transit service, although some of the lighter routes could turn out to have greater lifespans. This capital cost, therefore, represents an investment in the future of the streetcar system and the ridership growth that it could accommodate if only the TTC ran enough service. (The frequency of many routes is very much lower today than it was a few decades ago, and there is a lot of room for growth as residential density builds up along these routes.)

I will review the TTC’s Capital Budget for streetcar infrastructure in the next article in this series.

Any examination of streetcar replacement with buses must consider a variety of factors, but most importantly must look not at the streetcar system as it is today with service levels essentially frozen at or below the levels of two decades ago, but at what it can become as the backbone of travel in the growing “old” City.