Route 7 is a frequent line that runs from St. Laurent to Carleton University on the northern segment of St. Laurent, the Hemlock-Beechwood corridor, and Bank Street, an arrangement which has not changed substantially since the 1980s. This is one of two Bank Street routes, which is currently being contested amidst a proposal to convert two lanes of parking to bus lanes, as well as a plan to introduce bus lanes to St. Laurent.
Route 7 is among the busiest of OC Transpo routes, but suffers from the usual problems of a high rate of cancellations and buses that do not appear on-schedule or on-headway.
The example of Route 7 is a continual reminder of how poorly schedules create erratic service, both the pitfalls and benefits of route management, and how reliability can differ on the same route at different times of day.
As usual, the data for this analysis comes from Transsee.
Cancellations
The overall cancellation rate was 5.9%, with St. Laurent-bound service seeing more cancellations than the Carleton-bound service. The cancellation rate for St. Laurent buses is 6.9%, and the rate for Carleton buses is 4.9%. This is a consistent pattern over the entire route - eastbound service is consistently more delayed, has a greater gap between scheduled and actual runtimes, and has a higher rate of bunching.
Schedule Adherence
These charts show the on-time performance of the route across the entire four weeks. On weekdays, performance is passable at the origin, but delays grow along the route, suggesting that the route is not well-scheduled. This growth is more pronounced in the eastbound direction, which matches the travel time analysis.
While OC Transpo does not measure on-time performance on frequent routes like Route 7, this remains an important metric thanks to random interlining. As buses are not assigned to one particular route, delays on other routes may "bleed" into Route 7 and vice versa; what's more, some trips see cancellations (certain interlines are more prone to this than others) if they are delayed at the terminal and need to be somewhere else.
In both directions, the worst chokepoint is the heavily trafficked section from Parliament Station to Rideau Station. Its combination of variability and low travel speeds is schedule murder on a half dozen well-used downtown bus routes. Circulation management and transit priority measures (including closing Rideau to cars) are desperately needed here.
In the morning peak period, both directions see an average on-time performance.
The need for bus lanes on Bank Street in the AM Peak is not particularly evident, as buses generally remain on-schedule through the Glebe. However, they could contribute to a higher average speed in this time period.
On-time performance in both directions is abysmal. There is a clear pattern of trips that progressively get delayed heading east on St. Laurent; even then, the locations of the two proposed bus priority locations are not delay hotspots. Heading towards Carleton, the problem is one of poorly scheduled interlines and/or EB trips that arrive late. The westbound service is an example of how non-native problems can still affect a transit route, and how this is compounded by random interlining.
Evening on-time performance is still quite bad heading eastbound, while westbound service is more heavily affected by late departures from Carleton.
On-time performance in both directions on weekends is worse than on weekdays. Service is not allocated adequate travel time, which causes heavy delays on a large percentage of service. Chronic lack of runtime and bad interlining practices are not limited to weekdays, and are practiced every day of the week.
Headway Adherence
These charts show the headway, or gap, between arriving vehicles. The solid line shows the average headway, while the dotted line shows the additional wait time that bunching creates.
There are high "peaks" of bunching, as well as a noticeable spike in cancellations at certain times of day (often from bad interlines). The "peaks" get smoothed down heading eastbound, but the baseline level of bunching increases - the nature of the eastbound 7 is such that buses will bunch without intervention, with the added wait time reaching 50% of the scheduled headway at St. Laurent/Montreal.
Westbound service shows more variation in trip cancellations across the analysis period. This presents the challenge of ensuring reliable service despite differing bus availability rates.
The spikes in headway (that is, cancellations) become more pronounced going westwards, and serious bunching appears along the length of the route. The "baseline" change is less serious than in the eastbound service, but trips leaving St. Laurent have a poorer on-time performance, as seen in the schedule adherence charts.
The string charts show the progress of every trip across the day, and are useful for service troubleshooting. The string charts here reflect the final week of analysis from 20 to 26 October; the PDF for the weekday set can be seen here and the weekend set here. The charts for Wednesday, 22 October can be seen below.
On this particular day, bunching is not a big problem in the AM peak, but begins in earnest after 0900, and grows until 45-minute gaps appear at various points in the afternoon peak. There seems to be no more than a cursory attempt at controlling bunching, and the attempts to do so are not effective.
There are several chokepoints on the route. The most obvious is the "central corridor" of buses in central Ottawa, on Queen, Elgin, and Rideau. A solution to that should be found, which would reduce operating costs and improve reliability and speed for thousands of transit riders in the urban area. Two other notable chokepoints are the NB segment of St. Laurent coming out of the station and Beechwood right east of the Rideau River. A speedup campaign would focus on these areas to deliver the best results for transit service speed and reliability.
Travel Times
In order to control layover times, reduce variability, and cut bunching, travel times should have low variability, and scheduled times should match the reality on the ground.
Please note that these charts include only weekdays; moreover, they exclude 22-24 October due to Transsee data wackiness.
What's more, the medium-high levels of travel time variation means that extra service hours (and therefore cost) needs to be spent on long layovers, or risk spiraling service into an unreliable mess - the latter choice has been OC Transpo's and City Council's. Reducing the variability of travel times would reduce operating costs and improve service for riders, but this is an efficiency that City Council is reluctant to take advantage of.
Westbound service is, outside of the AM Peak and evening period, much better behaved. Not only does the scheduled travel time match the actual travel time, sometimes even going with too much padding, but it also has much lower variability, which theoretically makes scheduling much simpler.
The dual sore spots of the evening period and the AM Peak suggest that work needs to be done in those periods. However, the City's proposed Bank Street bus lanes are only active in the PM Peak for this direction - clearly, the City has not put much effort into analyzing delays on its own bus network, or they discarded the results of such an analysis.
On the whole, a travel time analysis with these results, instead of the eastbound service, is much easier to schedule and operate. If only Transit Committee were willing to tackle the dual devils of random interlines and inadequate run times.
Looking Forwards
Multiple factors contribute to OC Transpo's erratic bus service: variable travel times, which itself has many contributing factors; a scheduling system which includes inappropriate interlining and inadequate runtimes; and a City which puts more and more money into public transportation, apparently with no curiosity as to why their return on investment is so poor.
Route 7 shows the effect of poor travel times causing bunching and delays heading eastbound; many westbound delays can then be attributed to these delays ricocheting back. Interlining causes cancellations on a small percentage of trips, but the following trips bunch and create a widespread problem. Meanwhile, buses crawl through the central corridor of the downtown bus network, and there is barely any attempt to understand this problem, let alone in finding solutions.
No amount of City Council's transit priority hot air talk will solve this problem without them understanding the root causes of why buses start their trips late, how they become late on the route, and why they cannot recover that lost time at terminals. As long as we're not even understanding the basic reasons why the service is unreliable, then we will not be able to make progress solving problems on the network.
Until next time.
No comments:
Post a Comment