Route 88 service analysis: October 28th to November 24th

 This article analyses the reliability and behaviour of OC Transpo's route 88 over a period of four weeks, from Monday, October 28th, to Sunday, November 24th. Data from Transsee, as always. 

Cancellations

• October 28th: 8 EB, 10 WB
• October 29th: 9 EB, 11 WB
• October 30th: 7 EB, 12 WB
• October 31st: 0 EB, 0 WB
• November 1st: 8 EB, 11 WB
• November 2nd: 1 EB, 5 WB
• November 3rd: 3 EB, 6 WB
• November 4th: 7 EB, 10 WB
• November 5th: 5 EB, 8 WB
• November 6th: 3 EB, 5 WB
• November 7th: 4 EB, 11 WB
• November 8th: 0 EB, 0 WB
• November 9th: 1 EB, 6 WB
• November 10th: 0 EB, 4 WB
• November 11th: 4 EB, 9 WB
• November 12th: 5 EB, 7 WB
• November 13th: 5 EB, 8 WB
• November 14th: 0 EB, 0 WB
• November 15th: 5 EB, 12 WB
• November 16th: 4 EB, 7 WB
• November 17th: 3 EB, 4 WB
• November 18th: 2 EB, 7 WB
• November 19th: 7 EB, 5 WB
• November 20th: 6 EB, 3 WB
• November 21st: 3 EB, 0 WB
• November 22nd: 1 EB, 7 WB
• November 23rd: 1 EB, 0 WB
• November 24th: 0 EB, 0 WB

This route runs: 

• EB weekday: 90 trips
• EB Saturday: 63 trips
• EB Sunday: 43 trips
• WB weekday: 82 trips
• WB Saturday: 61 trips
• WB Sunday: 39 trips

This amounts to a cancellation rate of 5.0% eastbound and 7.6% westbound. 

Headway Deviation

As always, we can look at (weekday) statistics to see how wide the gaps between buses are. Note - week 1 is October 28th to November 1st, week 2 is November 4th to November 8th, week 3 is November 11th to November 15th, and week 4 is November 18th to November 22nd. 

Oh my. Oh dear. The gap starts quite large at Greenbank, having already run a long route from Kanata; many times, the deviation is already reaching half of the gap, meaning every bus is, on average, bunched. During the PM Peak, just as frequencies decrease, the deviation becomes worse. Schedule adherence is not better, as we'll see in a minute. 

This gets worse over the day; at Navaho, the AM Peak is more noticeable, and the schedule deviation is reaching the average headway in the PM Peak. At Billings Bridge, the same is true of both the AM and PM peaks; the evenings are quite bad as well. Clearly, the schedule on this route is out of whack, as the buses bunch the further along they are. Let's look at westbound behaviour. 

WB shows buses better behaved, outside the PM Peak, at Billings Bridge. However, the schedule breaks down as the route runs west, and the peaks are noticeably taller and troughs shallower by Navaho, with little change to Greenbank. 

On-time performance

We can also look at on-time performance for the route over a certain week. Please note that these charts do not include short-turn trips that terminate or originate at Baseline or Bells Corners, only the full Terry Fox to Hurdman runs. No peak hour charts, because the short turns need their separate graph. 

On-time performance is bad in both directions, with westbound trips being worse than eastbound trips. In both directions, the percentage of trips that are late increases further along the route - clearly, the schedule needs to be adjusted, as the buses seem consistent in either falling behind, or running early, the bunching that is seen in the headway gap charts. 

Weekend performance isn't better - in fact, it's worse. 

I cannot emphasize enough that nearly two-thirds of westbound trips on a weekend are delayed at the terminus (and really, east of Greenbank). OTP is barely above half in the eastbound direction. 

I don't even know what to do with this information, other than curse at former and current mayors. 

String Diagrams

These charts show a Friday, November 8th. 

The routes start well spaced in the early morning, but the introduction of short turn trips also introduces bunching. By 10:00 AM, the bunching extends to Terry Fox trips as well; despite the presence of "paddle shifts" (buses which operate on one route, rather than deadheading all over the place due to the very worst efforts of "efficiency" at OC Transpo), those buses still bunch up with no effort made to correct that; additionally, many trips bunch halfway on the route, indicating a poorly designed schedule. 

By the afternoon peak, short turns are departing close to each other and to full-route runs, leaving large gaps in between. This persists until 9:00 PM, but there are still some bunches in the late night. 

As always, we can identify problem spots with these charts: 

• The loop-the-loop around Hazeldean Mall is a problem in both directions. The easy, or not so easy solution, is to cut a door through the Good Life Fitness and eliminate the loop altogether. This obviously imposes a cost on the mall though, and the ridership at that stop is, anecdotally, not zero. 
• The intersection with Greenbank is a problem eastbound, especially in the AM peak. Frustratingly, the intersection rebuild, due to break ground in spring, includes through-bus lanes going westbound, but the approach is a shared right-turn lane due to a narrow right of way. (westbound is a lesser problem compared to eastbound) 

• The Algonquin College detour is a problem in both directions, especially the left turn onto Navaho going westbound. 
• The westbound approach to Merivale is also a problem; it's quite variable eastbound. Clyde is a lesser problem that appears occasionally. 
  
• The section between Fisher and Baseline is slow in both directions. 

Travel times

We can also look at the variability of travel times to see whether some things can be (ahem) improved. Weekdays only. 

The variability from Terry Fox to Baseline is low, the maximum landing within 5-7 minutes of the median, which is acceptable on a forty minute run. The variability is worse to Billings Bridge, with the maximum running double the median travel time; this is no way to write a schedule and run it - this is something that OC Transpo is not good at. 

The variability going west is very high, reaching a maximum travel time of ninety six minutes for Billings Bridge to Navaho. Even if it is a single trip, that is 4.5 km/h. I can walk faster than that. And I don't think that's a single trip, since weeks one and three also show high variability, with median travel times in the mid-twenties, but maximums in the forties. If we could cap travel times at twenty five minutes, 17.5 km/h is acceptable for a rush hour route. 

Maybe when we get Baseline BRT, which is fraction of the cost of Stage 3, in the 2040s. /sigh

The travel times from Baseline to Terry Fox are less variable. 

Here are the figures for the PM period. 

Eastbound variability is worse in the eastbound, even though median travel times are similar.

Westbound travel times are high going westbound from Billings Bridge to Navaho, averaging around 14 km/h. Variability is not so high, which surprised me; almost all trips remain under forty minutes. The trips to Kanata show more variability, but there are no crazy outliers. 

That wraps up another article on OC Transpo, this time about Route 88. The route suffers from slow trips and a number of problem spots. Along its high ridership, I believe that the number of chokepoints and a general "problem stretch" from Fisher to Baseline contributes to the 88's reputation for unreliability. Stay tuned for another article soon.