In many past articles, I have reviewed the quality of service on various routes from the point of view of headway regularity, travel times and short turns. While these analyses can show that disordered service is commonplace, they do not address a more basic question: what is the actual capacity of service offered, how consistently does the TTC actually provide room for passengers to ride?
This article uses vehicle tracking data in a different manner.
Service at Bathurst & King picks up almost all transit vehicles in both directions without effects of short turns. However, raw vehicle counts do not directly represent “capacity” because this must be adjusted for vehicle size. For example, twenty streetcars in one hour can carry more passengers than twenty buses. The capacity per hour is affected by the combination of two factors: how many vehicles of each size actually passed the intersection, and how many vehicles even operated in the peak period. The latter factor is important because “missed trips” don’t just arise from erratic service, but also from a failure to field all of the scheduled vehicles.
Hour-by-Hour Capacity
The following sets of charts show the capacity of service passing Bathurst Street eastbound during the AM peak (0600 to 1000) and westbound during the PM peak (1500 to 1900). Data are shown for all weekdays from September 2013 to July 2016 except for February 2016 (because I do not have data for that month).
There are three types of chart in each set:
- Vehicle counts by type by hour
- Capacity of the vehicles observed by hour
- Total capacity over the four-hour period
Visible in these charts is the fluctuation from time to time in the proportion of service provided with standard CLRV streetcars, articulated ALRV cars, buses and new Flexity LFLRVs (since June 20, 2016). Very low values on individual days correspond to situations where a diversion to much or all of the service away from the intersection, typically to Queen Street. Another factor is that occasionally information for the time in question does not exist in the data provided by the TTC.
Over the length of an hour, a small day-to-day variation might be expected in vehicle counts through minor service irregularity. However, the mix of vehicles affects overall capacity. Late 2014 saw the onset of bus replacements of streetcars, and the drop in streetcar counts came primarily from ALRVs which are much larger vehicles. The number of bus trips required to replace ALRV trips was substantially larger as the charts show. However, there remains a considerable fluctuation from day-to-day in the number of each type of vehicle.
Capacities are calculated from the TTC’s service design values:
- CLRV: 74
- ALRV: 108
- LFLRV: 130
- Bus: 51
Note that vehicles can carry more people than these numbers suggest under crush conditions, but service cannot be designed based on crush loads on every trip.
What is quite striking about the capacity charts is the fluctuation over a range of about 500 passengers per hour, roughly 25% of the typical value, during the busiest periods. In other words, even if the vehicles arrived evenly spaced and, therefore, evenly loaded, there would be a considerable change day-to-day in the quality of service experienced by riders.
This begs the question of whether the days with lower capacity arise from “traffic congestion”, the TTC’s favourite explanation for erratic service, or if another factor could be at work.
Vehicles in Service
Another way to look at the data is to simply count the number of vehicles in service on the route. The next charts report on the number of vehicles of each type observed on the central part of King (Jarvis to Jameson) from 0700 to 0900, and from 1600 to 1800.
What one would expect to see if the schedules are to be believed is that the vehicle counts would stay more or less the same for a series of weeks corresponding to one or more “board periods”, the five-to-six week periods for which set of schedules remains in effect. However, what we see, particularly for streetcar modes, is a substantial change day-to-day in the number of vehicles the TTC actually fielded. The total vehicle count has fluctuated quite a lot (a range of about 10 vehicles on a likely scheduled total in the high 50s in the AM peak).
The numbers here do not translate directly to capacity, but they are linked:
- More streetcars and fewer buses can lead to higher capacity depending on the replacement ratio, but this can be offset by the degree to which larger streetcars (ALRVs) are used.
- Schedule changes to increase running times, even at similar headways, result in more vehicles in service (because the round trip time is longer) but not in more capacity (the vehicles/hour counts are unchanged).
The TTC speaks of the bus trippers on King as “expanding capacity”. As the charts clearly show, they do not achieve this effect because they are only replacing streetcars, and not necessarily on an equal capacity basis. The buses are a response to a shortage of streetcars, not a service improvement, except in the sense that without them the remaining streetcars would be even less able to cope with demand.
One factor which the vehicle-based data cannot reveal is the degree to which runs are cancelled because there are not enough operators available to drive them. This is not just a manpower issue, but one of schedule design and the degree to which all runs are part of regular crews, and how many depend on spare or overtime operators whose availability fluctuates. Trippers that operate for a short period are especially vulnerable to this because they are short pieces of work more likely to be crewed as extras.
The TTC has many challenges on King Street, and the City is now studying ways to redesign the street to aid transit operations, among other goals. However, fielding a consistent level of service is an essential part of delivering a consistent quality to riders on the line, and the TTC does not achieve this.
There has been some growth in capacity of service provided on the central part of the route, notably with the service redesign for 514 Cherry, but how long this will last in an era of trimmed budgets remains to be seen.
Note:
The counts and capacities shown here include service on 514 Cherry starting in late June 2016, but do not include service on 508 Lake Shore. This service was suspended due to the streetcar shortage in January 2015 (PM peak), and June 2015 (AM peak). It contributed a small amount of added capacity when it operated, but this was quite erratic because the arrival times and numbers of vehicles assigned varied considerably. I do not have 508 tracking data for many of the months included in these charts, and so have omitted the route from the analysis.
Steve Munro 
Steve I gather that Bombardier is still struggling to get more than one vehicle per month delivered, so the question must be asked, if there was actually 20 or so of the new cars on King, how far would that go to resolve the issue? (assuming that this was a straight up swap new car and remove a bus from the route). Also to what degree does having to have buses here, damage service in the balance of the city?
Steve: At this moment, there are no buses on King, only streetcars, but the buses will be back in the fall. Total peak vehicle requirements are about 50 including trippers and the 514 Cherry cars. 20 Flexitys would help, but wouldn’t cover the whole need. Also, current fleet plans do no include enough Flexitys to allow 1:1 replacement of existing fleet. That’s what the extra 60 cars proposed will do.
LikeLike
If only the King Streetcar would Not have to compete with car traffic.
LikeLike
I wonder from a technical perspective if the TTC even really has the tools for determining if they are running the correct amount and types of vehicles each day…this will obviously be made easier when we eventually get to the point when all vehicles are the same capacity…
From a management perspective I continue to have concerns about the stats that are being presented in the CEO report…one great example is that they show the number of hires per month, but not the number of people who left, or the number of people they targeted to hire…
It would be great if they could present a target service level report with more details include: target number of passengers per section of route, actual passengers per section fielded based on vehicle size, and the reasons the others were not fielded (confusion, lack of vehicles, lack of staff, planning, etc.)
Over time the lack of improvement in the analysis (and public reporting) area is becoming a very large gap in the system.
Steve: There are many stats that TTC management should report, but do not, and I am getting more than a little tired of all the back-patting that goes on for the KPIs they do choose to report.
LikeLike
I was actually simply thinking in terms of leaving all other vehicles in place, and replacing the bus trippers with Flexitys. That is removing the smallest vehicle, and replacing with largest, and otherwise leaving, the others in place. I am focused on where the slow roll out of the cars, is causing the most damage, and where replacing a relatively small number of vehicles will have the most impact. I cannot help but think if Flexitys were one vehicle in three, and they were arranged that way – ie a CLRV followed by and ALRV followed by a Flexity, and repeat, it would hugely reduce the frequency of longer no board conditions. The Flexity offering a reset at the stops as it came through.
Having a CLRV followed by a bus, followed by a CLRV when a CLRV does not have the capacity to begin with, will result in a repeated no board condition (ie a continued line at a stop, that is not cleared by multiple vehicles).
LikeLike
I have noticed on more than one occasion at Dundas West Station around 11AM-Noon (in other words well beyond ned of AM rush) a lineup of King Cars. The platform filled with 3 cars, 2 cars sitting out on Dundas waiting to get in (and passengers unable to get off and into station to transfer) and yet another car waiting at stop south side of Bloor Street.
Looks like the complaint about rotten service on Roncy has been solved. At least at that time of day. Would a slight increase in 514 Cherry cars going to/from Dufferin Loop ease that over- supply of King cars at Dundas West Stn?
Steve: We have the same problem at Broadview. When the weather and traffic co-operate, the cars have too much running time and they pile up at the end of the line.
LikeLike
Dundas West Station is to be rebuilt/expanded but it looks constrained for present or near term requirements let alone future growth. I cannot find the drawing showing surface loading platforms but I recall they are to be oriented on a northeast angle as opposed to the present straight west angle. It looks like 2 streetcars at two platforms. Not sure if this is Flexity capacity or just CLRV size. I forget if there is to be a spare track to hold excess cars something missing presently. Of course there should not be “storage” capacity at a loop in the sense that cars ought to be on the street providing service.
Steve: That design is the work of Metrolinx and their fantasy about a “Mobility Hub” at Dundas West. There is no sign that this will actually occur, and there is nothing in the TTC’s budget or plans for the rebuild. If anything, the new loop is smaller than the existing one.
One wonders if there should be room for the Carlton Car to be run in there as well. Ridership along Howard Park Avenue to Parkside Drive and into High Park Loop cannot be very high considering the low amount housing (and zero high rise) that exists. Connection with GO and UP would also be possible at Dundas West.
Alternatively, run Carlton Cars up Parkside Drive to an expanded Keele Station. This ought to take riders off 89 Weston Road and 41 Keele routes as well as Line 2 riders.
Yet another possibility might be to run Carlton Cars north on Lansdowne to an expanded Lansdowne Station with better bus facilities for loading/unloading and more importantly waiting shelter.
Either way having a second subway station to transfer to/from streetcar routes would provide an alternative diversion in the event of a problem at one subway terminal or on-street.
LikeLike
I see that once again the City are allowing TIFF to close King Street and the already very poorly managed 504/514 will be disrupted yet again. The Star says:
When a group like TIFF forces a TTC diversion and hiring of “Ambassadors” and police to do they pay the extra costs?
Steve: I will pass this on to Brad Ross for comment. By the way, originally there was not going to be a diversion, but Tiff has friends in high places.
LikeLike
The King streetcar was the reason I stopped buying a Metropass. Packed to the doors at 6:30 am and 3:00 pm. And I’d get really antsy and angry not knowing if I’d make it to work in time as a half dozen non-King streetcars sailed past at 6 am from wherever they’re stored. And as soon as I had no Metropass, I stopped buying tokens, because $3something is way too much to pay for anything I can walk in an hour or less. So if the TTC wonders why the declining ridership, they should look at people like me. I’ll pay $3 to go to Scarborough – great deal! – but forget about shelling that out to ride up to Wellesley and have to fight my way out the doors past all the long-distance commuters who seem surprised that anyone would get out anywhere other than Bloor.
LikeLike
In an another example of the TTC’s seemingly random dispatching, today (Sunday Sept 4) the LFLRV count is 12 on 510, 2 on 509, 2 on 514. This is the first time I’ve seen 2 on the 514. Why the “extra” isn’t assigned to a 509 run during the air show weekend is anybody’s guess. Yesterday there were 12 on 510, 5 on 509, and 1 on 514.
I’d like to see them manage a roll out where they have 2 assigned to each route, if that is indeed the plan.
Steve: Yes, vehicle assignment has always been a mystery, for as long as I can remember.
LikeLike
The hard part of this, is that there should be a set of variables, that is both communicated, and clear that would make this clear as day. Peak load, for instance, loading times, maximum number of vehicles that can effectively navigate, given signal timing.
LikeLike