On Sunday, June 19, 2016, the TTC eliminated several carstops on the central portion of the 504 King route with the alleged purpose of speeding up service. This is an idea that sounds good in theory – fewer stops means faster trips – but that’s not necessarily what happened.
The stops in question were located at:
- Victoria Street westbound
- York Street both ways
- Simcoe Street both ways
- Brant Street both ways
Another stop outside of the core that was removed was on Broadview north of Queen, northbound. This was originally created for the Don Mills night bus when it served Broadview, but had become a popular location for riders making the transfer from westbound Queen to northbound King cars without having to make a double-crossing of the intersection. That convenience is gone in the name of faster service.
Readers of this blog will know that a common thread through many of my service analyses is that some of the worst ongoing delays come from poor or non-existent transit priority including badly timed traffic signals or the lack of transit-only green time for turns. Other major delays arise when there is anything unusual such as utility construction with no corresponding change in traffic/parking regulations to retain as much of the street’s capacity as possible.
When the TTC floated the idea of eliminating minor stops at various locations, the idea sounded plausible. Did the change actually achieve anything?
Travel Times Between Jarvis and Bathurst
One way of looking at the situation is to examine travel times for June 2016 to see how they varied before and after the change. In the sets of charts below there are seven pages for each direction:
- Pages 1-5 give the individual travel times (points) and trend lines through these points for each week of the month, with each day’s data shown in a different colour.
- Pages 6-7 give the same view, but for weekend data on Saturdays and Sundays
Looking at the westbound data, the most obvious point is that the values are spread over a wide range and vary from day to day even at the same time period. Any improvement in overall numbers will have to be quite substantial just to be visible among the “noise” of hour-to-hour and day-by-day variations. Also fairly evident is that the data for pre- and post-elimination lie generally in the same area (more about this later). The weekend of June 11-12 has some of the highest numbers (yellow), but of the “post” days only June 25 lies convincingly below other Sundays. Eastbound data have similar patterns.
Another way to look at the numbers is to compare operating speeds along the route. The following charts show average speeds hour-by-hour through the day for the section of 504 King from Gerrard & Broadview to Shaw & King which brackets the affected area. Step through the charts page by page to see the evolution of streetcar speeds through the day. Note that the eastbound charts should be read right-to-left as that is the direction of travel.
All of the stop locations are visible as notches in the speed charts because these are places where transit vehicles slow and stop. The depth of the notch gives an indication of whether all cars stop, or only selected vehicles when there is a passenger to board or alight. The “pre” period (weekdays June 6-17) is in blue while the “post” period (June 20-30) is in orange.
For westbound service, the visible effects are:
- At Victoria, there is a slightly better speed mainly during the late afternoon when this stop doubles for Yonge Street.
- At York, there is little difference in speed because this is a signalled intersection and streetcars are often caught by a red light. Eliminating the stop without also implementing a green-extension transit priority capability achieves little.
- At Simcoe, there is some benefit during parts of the day. This location has a traffic signal but it also has transit priority to extend the green phase for King.
- At Brant, there are improvements primarily during the AM peak. What is quite striking is the slowdown in transit speeds after 10 pm between Spadina and Portland for the both the pre- and post- periods.
For eastbound service, the visible effects are:
- At Brant, there is some improvement, but unlike westbound, this is visible for much of the day. The late evening slowdown is mainly between Bathurst and Portland.
- At Spadina, although the pre/post times are the same, the effect of delays caused by 501 Queen cars turning north with no priority is evident in the slower speeds approaching Spadina from the west.
- At Simcoe, there is some change through the day, but the biggest effects are in the PM peak and early evening.
- At York, there is small improvement during some periods.
The data for the post period could be affected by the fact that many streetcar operators continued to serve the stops because people continued to wait at them. The TTC’s half-baked notices originally consisted of small sheets of paper, some of which disappeard within days of the change. A transit shelter is a powerful advertisement that this is a transit stop, and removals in the future should be co-ordinated with eliminating the “shelter”component of the street furniture. At the very least, large signs of the size typically used for subway shutdowns should be deployed on shelters at discontinued stops.
Another issue could be that the time “saved” by skipping a stop was simply eaten up again when the streetcars caught up with the traffic in which they were travelling, particularly for stops that were not at a traffic signal that would penalize the act of stop service with a delay of one cycle. This would be somewhat like the impatient motorist who speeds ahead only to catch up with traffic sooner than those who get there by driving normally.
Finally, here are the link times for May and June 2016 broken down into individual weeks and half-hourly increments through the day. These show that there is little difference in the data for the last two weeks in June compared with the preceding period. Worth noting here is that the downturn in values for June Week 5 can be seasonal as there is typically a light traffic day before a long holiday weekend (Friday July 1 onward).
When data from July and August are available, I will update this article.