Introduction
This article discusses service quality on Route 5 between 28 July and 24 August. Route 5 was created as part of New Ways to Bus earlier this year; it runs frequently between Rideau and Elmvale on Elgin, Main, and Smyth. Previously, the "5" designation belonged to a local route between Rideau and Billings Bridge, and the current service span was covered by a smorgasbord of local and frequent routes.
Route 5 is fairly well-behaved most times of the day. However, problems appear at certain times of the day, and cancellations remain a scourge on service quality.
Route 5 is relatively well-behaved in most times of the day, with problems being concentrated to certain times of day ... and to the cancellation of service. Internal analysis like this can identify problem spots on the network, to be targeted with something more specific than redesigning the scheduling system or similar interventions.
Data, as always, comes from Transsee.
Cancellations
In the analysis period, the northbound cancellation rate was 4.3%, the southbound cancellation rate was 6.8%, and the overall cancellation rate was 5.6%.
The southbound cancellation rate was 50% higher than northbound cancellations. I believe many northbound trips are at the beginning of a shift, while southbound trips are often the last trip in a driver's shift, and therefore more likely to be delayed to the point of cancellation.
Headway Adherence
Headway adherence - that is, bunching - is fairly good on Route 5 throughout most of the day. Unlike, say, Route 80, bunching does not substantially increase along the route, which suggests that problems mostly originate at the trip's origin (interlining) rather than from Route 5 itself (inadequate runtimes).
However, in the PM Peak, bunching is a problem in both directions. Also worth noting is the worsened headway adherence in weeks 3/4 compared to weeks 1/2.
Southbound service quality is worse than northbound, with bunching being a more serious problem outside the PM Peak. Additionally, bunching increases a bit going along the route, implying a less robust schedule. As with the northbound service, headway adherence is worse in weeks 3/4.
On-Time Performance
The next two sections reflect service quality in week 4, between 18-22 August.
On-time performance is generally good, except during the PM Peak. Outside that period, trips in both directions generally see about 80-90% on-time performance, and more importantly, they remain at that level at the terminus, which signals that the route is well-scheduled.
In the AM Peak, on-time performance never falls below 70% in both directions. The difference in performance between the origin and terminus is zero going northbound, while on-time performance falls from 90% to 80% southbound.
During the midday, almost all trips depart on time. However, trips in both directions, especially going northbound, see delays along the route, with a small fraction late by ten minutes or more, indicating that the route needs to add a few minutes to runtime and/or recovery time during this time period.
In the same pattern as observed in the headway charts, on-time performance falls during the PM Peak. Only 60% of trips in either direction depart on-time, and about 5-10% are over twenty minutes (one full headway) late. Even with this kind of start, on-time-edness remains even on the route, a sign that the schedule is robust enough to handle delays without spiraling into bunching and delay propagation.
In the evening, on-time performance is good in both directions, though northbound service generates delays on Elgin, which return to normal at Rideau.
String Diagrams
The string diagrams shown reflect Monday, 18 August. The strings for the entirety of the week can be found in this PDF.
One observes that bus spacing is fairly even throughout most of the daytime. From the AM Peak until about 1430 or whereabouts, and after 1830, buses arrive every 15 minutes in both directions. However, bunching and large gaps appear in the PM Peak, right when customer demand is highest, leading to the perception that OC Transpo is unreliable.
Some chokepoints on Route 5 ("slowpoints"?) include the detour through the TOH campus, the northbound detour around the blocked stop at Elgin/Argyle (I say just remove the stop!), slowdowns in both directions over the Pretoria Bridge, and generally slow speeds on Elgin. Fixing these would speed up the route, reducing the operating cost of Route 5 and improving travel times for customers.
Travel Time
Travel time during peak periods is consistent, with few observed anomalies that might greatly impact reliability. This can be seen in the on-time performance charts, where with the exception of interlined services in the PM Peak, on-time performance is good.
Travel time is also generous in the PM Peak. In both directions, the route is scheduled with enough travel time; and the top quadrant of travel time does not extend too far out from the median.
Conclusion
Route 5's on-time performance is good, which stands in contrast to many frequent routes which run downtown. If every bus route had the performance of Route 5, OC Transpo would be a very reliable service and I would not need to write Transit Ottawa anymore.
By analyzing a route whose on-time performance is good, positive (and negative) features begin to appear which make best practices clearer. On Route 5, we see consistent travel times with no large jumps, schedules whose travel time reflects the reality on the ground, and timely departures from the origin (most of the time), which ensure that the route sees reliable service.
Route 5 does also have lower reliability, in the form of a high cancellation rate (partly the result of interlined service, and partly due to a systemwide bus shortage), and in the PM Peak, which comes from delays which originate on other routes and are propagated to Route 5 through interlining. An end to random interlining would likely improve service quality at the PM Peak as well.
All of this comes with one caveat: this analysis was performed over the summer, when traffic is low. After returning to school, reliability will not be as good without seasonal adjustments for travel time, which should be done anyways.
The example of Route 5 strengthens the case for scheduling reform in Ottawa. With work, reform, learning from other cities, and possibly a boost to service hours, most routes in Ottawa could have this kind of reliability.
Until next time.
No comments:
Post a Comment