This article is part of a continuing series tracking the behaviour of transit service on King Street during the pilot implementation of pedestrian and transit priority measures. The last update was in May, and I skipped June because there was little new to report.
Although we are now into the summer when conflicts with pedestrians and space constraints from recent takeovers of curb lanes with a variety of artworks and seating areas, travel times on King have not been affected. In fact, thanks to the re-activation of Transit Priority Signalling (TSP) at various locations on July 7, travel times have actually dropped during some periods.
Peak Travel Times
Continuing the tradition of these articles, here is the travel time chart for the 50th (median) and 85th percentile values westbound from Jarvis to Bathurst from September 2017 to the end of July 2018.
The collection of charts for five periods during the day for the two directions are linked here:
The following service disruptions show up in the charts above for the June-July period:
- June 14, 20 and 26 eastbound: Congestion eastbound to University Avenue from 5-6pm (typically this is caused by north-south traffic blocking the intersection)
- June 26 eastbound: Service held at Church Street just before 2pm (and therefore counting in the 1-2pm travel time stats) by a fallen overhead wire.
- June 26 westbound: Service held at Peter Street at about 10:40 pm by a collision.
- July 10 westbound: Service held east of Bathurst Street at about 1:15 pm by a collision.
- July 25 westbound: Service held near Church Street just before 9 am. Reason unknown (no TTC eAlert was issued).
The change in schedules effective in the last week of June reorganized service on King Street:
- 504 King cars, a mix of CLRVs and Flexitys, operated between Dundas West and Broadview Stations.
- 514 Cherry cars, all Flexitys, operated between Dufferin Loop and Distillery Loop.
- 503 Kingston Road cars, all CLRVs, operated between Spadina and Bingham Loop (Victoria Park and Kingston Road) weekdays from 6am to 7pm.
- 504 King cars, all Flexitys, operated between Dundas West Station and Distillery Loop.
- 504 King cars, usually CLRVs, but often with some Flexitys, operated between Dufferin Loop and Woodbine or Bingham Loops depending on the time of day.
- No service was scheduled to operate as a 503 or 514 car.
- A shuttle bus operated from King and Parliament to Broadview Station. This is outside of the pilot area.
The effect on line capacity varied depending on where one looks. Notably, service in Parkdale and beyond had greater capacity because all cars west of Dufferin were scheduled to be Flexitys. There was a small decline in capacity westbound at Yonge because the former 503 service to Spadina had been rolled into the 504 King service to Dufferin.
The continued migration of service to the new fleet also contributed to capacity growth.
Here are the full chart sets:
A few samples:
Inbound AM peak capacity at Jameson Avenue in Parkdale shows the effect of almost all trips operating with larger Flexity cars.
The revised service design reduced the number of cars scheduled eastbound at Bathurst in the AM peak, but this was partly offset by a higher portion running with Flexitys instead of CLRVs.
As noted above, the consolidation of the 503 service to Spadina into the 504 service to Dufferin reduced the capacity westbound at Yonge. However, the 503 cars to Spadina were only actually useful to riders bound a comparatively short distance in the core and so the higher apparent capacity before the schedule change was not actually available to a majority of riders.
Detailed Breakdowns of Travel Times and Headways
I have omitted these charts from this update because there is little change in the character of the data from earlier reports. In October, when the effects of TIFF (the film festival) can be seen in September data, I will return to this part of the analysis.
Travel Speed Comparisons
With the activation of Transit Signal Priority at several intersections, obvious question to ask is where are the effects most noticeable and at what times of the day. The format of the charts that follow will be familiar to those who have read earlier articles comparing the speed of bus and streetcar operations on various routes. Two sets of average speeds, in this case one for two weeks in mid-June 2018 and one for two weeks in mid-July, are plotted together to show their behaviour along the route. The granularity of the plotted data is 10m, and this allows one to see speed variations at intersections and at stops.
There are a few important caveats here:
- Speed at a point is not the same thing as travel speed across the pilot area. For example, if a car is stationary at a stop or traffic signal, its speed is zero for however long it sits. A high speed between stops is of less value if it cannot be sustained and/or is offset by long stop service times.
- Where TSP is active, streetcar operators approaching an intersection with a farside stop will stay at a higher speed on the approach because they are facing a green signal they can expect to stay clear. “Double stops” are reduced, and even when they do occur, they can be shortened by pre-emption of the cross street’s green time when a transit vehicle is waiting.
- Although the vehicle tracking data updates the position (and hence speed) every 20 seconds, and this can cover multiple 10m increments in the chart, the data presented here are the accumulation of values for one-hour periods over ten days and represent the combined behaviour of many streetcars. A single car would give a very choppy view of speeds with a snapshot only every 20 seconds, but the contribution of a few hundred observations gives a more granular view of speeds.
For example, the chart below compares westbound speeds during the 8-9am period in the area from Parliament to Shaw. The pilot lies between Jarvis and Bathurst, but behaviour on the shoulders of this area is worth seeing for comparison.
The blue line shows June 2018 data while the green line shows July. The red dotted line is an interpolated trend line for June data, while the yellow line is for July. The westbound approach to Jarvis stays at a higher speed and does not exhibit as much effect from double-stops in July (green) as in June (blue). There is a similar effect at Church and at York. Travel speeds are also better at other locations by varying amounts.
The data for eastbound trips is shown below. The chart should be read from right to left because that is the direction of travel.
Some of the locations with higher speeds are not adjacent to signals where TSP was activated (eastbound at Bay has no TSP), but improvements in nearside delays are evident at Simcoe, Church and Jarvis.
These data are complicated by the fact that not all TSP locations were working properly on every day included for the July comparison. Further analysis will have to await more data, especially beyond the summer season.
The full charts with data in hourly increments are linked here:
These charts are most interestingly view by stepping back and form from page to page and watching the change in the data at specific locations. This shows the evolution of patterns over the day and how they differ along the route. I leave that detailed review to readers who will all have their own points of interest to look at.
Note that the charts become more “spiky” late in the evening when the number of observations included in the averages is smaller as headways widen to the late night service.
Speed Comparisons vs the Pre-Pilot Era
The charts above compare June and July 2018, but equally interesting is a comparison of July 2017, before the pilot but with comparable seasonal effects, with July 2018. The charts below have the same format, and the only change is that the June 2018 values have been replaced with July 2017.
Within the pilot area, the shift from nearside to farside stops is evident in the position of the lowest values from Jarvis to Spadina (the last farside westbound stop). Note also the improvement on the approach to Jarvis westbound probably due to better signal priority at George and Jarvis, as well as the shift to a farside stop.
Eastbound, there is again the effect of the shift to farside stops. In both cases, there are faster speeds between stops probably due to the reduction in competing traffic. The effect varies by location and direction over the day.
Here are the full day sets of charts comparing July 2017 to 2018.