This post is in the best journalistic tradition.  On those slow days and holiday weekends, all the material you’ve been saving up for months finally sees the light of day.  To readers who submitted comments and wondered where they went, take heart.  Today is for you!  This post collects comments loosely related to service quality and operations.

Back in March, I was writing about how transit services are analyzed.  We’re still waiting for the 2007 Service Plan to give us updated stats, but it’s time to rekindle the discussion here.

Sam wrote:

I found it comforting that the lines I’ve used all my life were not performing as poorly as I thought.  Then I scrolled down, and sure enough 104 and 96 made the cut. How embarrassing.

Steve:  A great peculiarity of the TTC figures is that the reported performance of routes considered together is worse than the system average.  There is something fundamentally wrong when major routes show “poor performance”.  The figures are almost certainly wrong, but they give ammunition to the TTC’s critics who say that they are wasting money.

Don Hamilton, who has been involved in local activism regarding the Graydon Hall bus, wrote:

I was introduced to the strange ways of the TTC by the series of events described on my website and here.

I only learned of your website recently and have to tell you how much I enjoy your approach even though much of it is over my head.

All I know is the microcosmic example available to me locally.  I am astounded by the degree to which the TTC ignores facts that they don’t like and just plain lie about ridership figures.  Are the TTC overall ridership figures ever audited?

Steve:  Overall ridership figures come from the folks in finance, and they are based on the actual sales of fare media.  From periodic surveys, the TTC knows what the average usage of monthly, weekly or day pass is, and they use this to convert pass sales into ridership numbers.  Similarly, most cash fares are adults, but there is some small proportion that are children.  The route by route counts are, with expections where there is a special study going on, taken at best once per year.  It is not unusual to see the reported counts in the annual Service Plans stay the same on a route like Queen for two or three years at a time.  This is a staffing issue, but it also means that the TTC is not getting up-to-date information at the route by route level when riding is changing, and figures purporting to give the “performance” of a route are out of whack even more than usual. 

TTC schedules seem to be driven by providing union members with 8 hours per day instead of a service driven by exhibited need.

Just a few observations from an ordinary but curious citizen.

Steve:  The 8-hour situation is a little more complex.  No route operates only 8 hours per day, and when garage time is taken into account (the time needed to get a bus into service in the morning an out of service in the evening), even a route with no evening or weekend service will clock up at least 70 hours per week per bus.  That’s two operator shifts with a bit left over.

There are special arrangements to deal with the fact that more vehicles (and operators) are in service during the peak than at other times.  (If there were no peak, then all crews could be set up more or less as straight-through shifts.)  Some operators work a few hours on a peak vehicle, and then move to a series of relief shifts through the day.  This involves taking over a vehicle for 20-30 minutes to give an operator with a straight shift a break.  Other operators may work part of the day on one route, and part on another.  A variety of other special provisions exist, but I’m not going into the gory details here.  The basic point is that the scheduling of operators is based on the overall service on all routes operated by each garage, not by individual routes.

Later, Don wrote:

I see that the experience related on my web site and blog with the TTC staff was not a special performance meant for me and my neighbours alone, but just everyday obfuscation and creation of facts to suit their already made decisions.

Since discovering your site I have made it number one on my daily routine even before checking my mail, please keep up the good work.

Steve:  Fan mail is always welcome!

And still later he wrote:

Wow!  I thought I was the only guy standing on a corner monitoring suspicious TTC information (and finding it totally spurious).

I stand in awe of your efforts and equally disappointed in the response from the TTC.

Steve:  So am I.  One project I am thinking of undertaking [TTC planners please note] is a Freedom of Information request to get the detailed vehicle tracking information for a selection of routes and days.  From this, I should be able to reconstruct the same sort of overview of route operations that we did manually two decades ago.  I understand that the TTC never quite got around to doing this internally.  Shows just how much priority the operations folks assign to actually understanding how their system works.  They would rather just bellyache about traffic congestion and rights-of-ways.

Gordon wrote:

It looks to me like the allocation to subway riders is about double that of the surface routes.  If the subway and RT were costed in proportion to their actual cost, how close to the chopping block would the Danforth subway be?

Steve:  Subway operations are very complex to cost and to tweak because so much of it comes from fixed costs like station operations, track and signals maintenance and route supervision.  There are actually times when the onboard crew are outnumbered:

On Sundays, the number of trains per line is:

• Bloor-Danforth: 20
• Yonge-University-Spadina:  23
• Sheppard:  4  [there are 4 trains on this line all the time]
• Scarborough RT:  4  [there are 4 trains on the SRT all the time except at peak when there are 6]

This gives us a total crew of 98 (2 per train except on the RT).  There are 69 stations, some of which have two separate fare control areas (Queen, Dundas, Bloor-Yonge).  Then you have janitors, escalator mechanics and so on as part of the overhead of stations.  The result is that cutting actual service does not make a huge dent in the cost of having a subway.  [I could get into a rant here about how you save most of this with an LRT line, but I will save that for a separate post.]

The flip side of this is that service design on the subway lines is completely different from the surface system where the TTC trims everything they can.  If we were to hold the rapid transit network to the same standards, we would close the Sheppard line most evenings, and weekends would be dodgy.  The RT has often been replaced by a bus and does not run after midnight to allow for ongoing power rail maintenance. 

When people say “I want a subway” what they are often meaning, unconciously, is “I want a line where the service is good all the time whether there are any riders or not”.  Those who have written to me about “value for money” audits should ask why we have a double-standard for the rapid transit network and equally what would happen if we took a less draconian approach to surface service standards.

Geoff wrote:

Having been stuck waiting in enough streetcars for some boob to turn left, I’ve often wondered if Torontonians could somehow learn the Melbourne hook turn … on second thought probably not.  [This is a right turn onto a sidestreet followed by a U-turn.  Of course in Melbourne it would be the other way around because they drive on the left.]

Sort of off the topic of streetcars, the other thing I’ve wondered about while trying to cross the Danforth in the Pape or Coxwell bus, is if the city and TTC could change the signals.  After the lights East-West turn red, if there could be north-south right turn advance plus TTC advance light (white bar like at Broadview) to clear the right lane and avoid pedestrians, along with getting the bus ahead of the traffic.  Then the north-south lights could turn green.

Steve:  Very slowly, the TTC is rolling out signal priority on its bus routes.  However, major intersections like Pape and Danforth will always be ruled by the Danforth phase as it has much more traffic.  Your idea of a separate transit phase would be great of the road engineers in the Works Department would let us have it, but it presents a special problem for transit vehicles that don’t have their own lane.  Say you are a bus stopped south of Danforth but behind two or three cars.  If you get your white-bar transit signal, the cars are not going to get out of your way.

I received a long comment from David Seto that I’m not going to post here because it has a huge amount of detail about the physical geometry of the streetcar track layout and intersections.  The nub of his argument is that many of the diversions that are implemented either for emergencies or for construction projects are more “out of the way” than they need to be due to limitations of the track layout. 

As a simple example, if the King car needs to divert around King and Bay, the westbound diversion is simple:  Church-Wellington-York.  However, the eastbound diversion is much longer:  Spadina-Queen-Church.  The King car is much further away from its route and for a longer time.  There are many other examples of this problem.

I have mixed feelings about this.  Toronto is blessed with a wealth of extra trackage that other streetcar systems don’t have.  This is a holdover from the days before the subway system when many more routes operated in the city.  Track on Ossington and Shaw is left over from the Dovercourt car.  Track on Church and Parliament is left over from those carlines.  Track on Adelaide, Richmond, York, McCaul and Victoria is a remnant of many old lines that looped through downtown.

The TTC has added a few curves during intersection rebuilding projects, but has missed a few opportunities.  One in particular was the chance to add an east-to-north at King and York a few years ago.  The missing quadrant at Queen and Dufferin was not added partly for budget reasons but also for concerns about installing curves on a grade.  The real mystery is whether the eastbound track on Adelaide from Charlotte Street (the Spadina car’s loop at King) to Victoria will ever be re-activated as a Queen/King diversion.