Transit on Commercial Drive…or the cost of a bike lane

November 3, 2014

Some numbers extracted from the Translink GTFS feed [4] (for the day of Sept 5th, 2014), for the 2km segment between Hasting and Broadway. The current average speed is ~11.5km/h, could be increased to ~15km/h with a bus lane…or reduced to ~9km/h according to the tradeoff done to implement bike paths

  • number of #20 runs: 304 (but I counted only 276 between Broadway and Hasting) requiring a minimum of 19 vehicles in revenue service [3]
  • time and speed between Broadway and Hasting [6]:
  • Min time Average time Max time
    10.3mn 5.3mn 12.4mn
    Max speed Average speed Min speed
    22.6km/h 11.65km/h 9.65km/h
  • ~15,700 annual operating hours meaning $1.57 millions in annual operating cost (at $100/hr, in line with [5])

bus lane Impact on Commercial Drive
We are considering the previously presented Commercial Drive proposal as illustrated below

4.5meters wide bike+bus lanes, with bus keeping in its lane at bus stop. It features transit signal priority and right turn specific signal to protect both transit and cyclists – Transit average speed is estimated at 15km/h

  • This bus lane, featuring clearly marked corridors (protected in one direction) and transit priority signal, suggests that average speed typical of BRT or urban LRT could be achieved: that is ~20km/h.
  • That said, noticeabily because the stop are closely spaced, an average speed of 15km/h could be more realisticaly and conservatively achieved:
    • That is roughly the average speed of the bus 20 outside the Commercial Drive segment.

Annual operating cost

average speed Average time Annual operating cost
9km/h 13.3mn $1.9M
11.5km/h 10.5mn $1.5M
15km/h 8mn $1.2M
20km/h 6mn $0.9M

The potential operating cost saving is in the tune of of $300,000 to $600,000/year.

On the opposite, a configuration of Commercial Drive with a single lane of traffic per direction to preserve parking [2], negatively impacts the speed of the bus, as we have seen before:

Commercial street redesigned as per StreetForeveryone group

Commercial street redesigned as per StreetForeveryone group – Transit average speed is estimated at 9km/h

Similar configurations, be on Davie or Robson, suggest a reduction of the average speed to ~9km/h; That could increase the route 20 operating cost by $400,000/year:

    the bus+bike lanes proposal is conductive of $1 Million in operating cost saving versus a proposal favoring street parking over transit.

A bus lane + traffic signal priority, allows an increase in the bus schedule reliability: lay over can be reduced accordingly, increasing the operating saving

Operating cost is only part of the picture:

Capital cost

the slower a bus route is, the more buses are required at same frequency/seat capacity:

The steeper the slope of a line, the faster the travel, and the sooner a vehicle return to its orgin, ready to do another run. the number of starting lines in between represent the required number of vehicle – credit Melbourne on Transit

The bus requirement is compounded by two conflating issues:

  • Demand is at its greatest at peak hour, but
  • transit speed is also at its slowest at peak hour
  • .

    On the route 20, afternoon peak hour traffic cost ~4 buses:

    number of vehicle in service on route 20 according to the time of the day (graph for friday Sept 5th, 2014)

    A bus lane, making transit more immune to traffic congestion, allows to reduce drastically the peak hour buses requirement (in our example, the average speed maintained at ~15km/h, vs 9.5km/h currently in peak hour)

    Adding a peak hour bus is a very expensive proposition: it means (to preserve spare ratio, and other contingency)

    • the Purchase of an additional bus
    • Adding storage capacity for this bus (even if in use 20mn a day)
    • Adding maintenance cost
    • adding a driver on payroll and all ancilliairy cost (training, administration)

    According to a conversation with a former Toronto Transit Commission employee, the TTC is costing an additional peak hour bus at $100,000 a year (that is for a 40footer, typically sold a ~$300,000)

    It is worth to note that Translink is in very short supply of articulated trolleybus, estimated each at $1M


    It is no secret that the faster a transit service is, the more ridership it will attract. That has been again recently verified in Seattle, with a quasi linear relationship:

    • an increase of 20% in speed is conductive of a similar increase in the ridership, which de facto increase the bus operator revenue[1]

    This coumpounded to lower operating cost makes Transit much more financially sustainable.


    When all the effects are combined, it is relatively conservative to estimate that a bike lane, done at the expense of transit on Commerical, could end up to cost more than $1 million/year to Translink, when compared to a solution improving both

    …and here we have analyzed only the direct cost for Translink…

    [1] New markings aim to keep drivers out of Battery Street bus lane, Aubrey Cohen, SeattlePi- Tuesday, October 21, 2014.

    [2] We refers here to the “Street for Everyone” proposal we have previouslly discussed, which has also been discussed on the pricetag blog.

    [3] That makes the route 20 the 4th most frequent bus route of the network, behind route 99,9 and 41.

    [4] See our reference spreadsheet (which has been updated with the 2014 data) for further detail.

    [5] We use here the hourly operating cost as stated in the 2013 Bus Service Performance Review (see Annex A): it is worth to note that this hourly operating cost doesn’t include neither bus lay over and dead end trips. It doesn’t differentiate artics buses from standard ones too: the $100 mark is a very significant under estimate of the real operating cost of a route. A $180 per customer hour service could be closer to reality as we have seen before.

    [6] It seems that the average speed of the route 20 is decreasing year over year, almost 10% reduction in the last 7 years according to our spreadsheet [4] (which also depends of the Translink data quality): A probable consequence of the city council inaction on Transit front


    13 Responses to “Transit on Commercial Drive…or the cost of a bike lane”

    1. Matt Taylor Says:

      Here is my very quick analysis of something that needs to be taken into consideration:

      On the corridor being discussed there might be about 200 metered parking spaces. The parking spaces I checked are $2 per hour. These would generate $500,000 in revenue if they are occupied an average of 3.5 hours a day, and $1,000,000 if they are occupied for 7. It is just a guess but it would not surprise me if it landed somewhere between there.

      • Voony Says:

        You are absolutely right that parking revenues are important to consider

        2 points:
        -The total street parking revenue for the entire city is $40M/year
        -According to, there are ~105 park meters on Commercial between Hasting and Broadway, some already used by parklets (+ ~66 off Commercial).

        That suggests you are probably erring on the optimistic side regarding the parking revenue on Commercial. That said, this parking vs bus lane issue raises 2 other questions:

        • Should we really delay the ~27,0000 daily transit rider because some consider they can’t live both without this ~100 parking stalls and a segregated bike lane?

        Notice, that our presented proposal preserve parking on one side…and because there is no bus bay on the parking side, the presented proposal retains probably 80% of on street parking (on commercial+off street)?

        And the other question arise of who bear the cost of street allocation choice?

        • The parking revenues go in the city coffer, and due to a tax loophole are exempt of Translink tax (revenue lost for Translink = $8 Millions!)
        • but

        • The additional operating costs (due to the fact parking is priotirized over Transit efficiency) are supported by Translink (hence the regional taxpayer)

        That is conductive of bad transportation policy

        • Matt Taylor Says:

          The significant majority of those parking meters are of the heart shaped variety, so those 105 parking meters represent closer to 200 parking stalls. There is 1 parklet in the area. At 6:30 tonight the metered parking spaces were 70-80% occupied which looked to my eyes like standard evening usage.

          We can prioritize through movements by improving traffic flow, or we can prioritize local traffic by providing greater access and a more comfortable environment. In terms of being able to transport more people more quickly, your proposal does a better job – but the latter may be more important in this situation.

          Also, you are applying your average speed analysis to the 2km of Commercial Drive between Broadway and Hastings. The street section narrows to 4 lanes north of 1st, and parking lanes are already used as through lanes in the peak direction. I’m not sure what changes you or Streets for Everyone would be proposing north of 1st, but I wouldn’t expect the differences in average transit speed to be that significant. The average speed analysis along the 800m section between Broadway and 1st (using 9 km/h and 15 km/h) gives 2 minutes of transit delay. This is associated with $400,000 worth of additional transit operating costs according to your analysis.

      • Voony Says:

        yes, I stand to be corrected on the number of metered space (I have overlooked the “heart shaped meters” đŸ˜‰

        I have also over looked the narrower section North of 1st Avenue, and our proposal effectively can’t apply north of first (or rather north of Graveley…between Graveley and First, the street section is ~75feet).

        But then there is only ~45 “heart shaped” meters between First and Broadway…

        and you still have very significant operating cost difference on the section South fo first as you have noticed (I will try to refine the number on thsi section).

        That said, North of first, the street section could be on each side, sidewalk (4 meters wide) a raised bike lane (1.5meter) + one general traffic lane (3meter)-leaving a 3 meters wide middle lane which can either be used as a left turn lane, or a queue jumper lane for buses. That is, all street parking go away: street parking is a major source of congestion which also occur off peak on Commercial.

        The key driver, is that Transit is prioritized over parking.

        If you are not willing to remove parking South of first, why you could be ready to remove it North of it?

        On thru movement vs greater access: I agree with you.
        Faster transit to the heart of Commercial Drive is also improving access.

        That said, we have also to see the region at large:
        the bus route 20, is the privilegied one (and practically the only reasonnable one) giving access to the East Hasting corridor from either the Expo or the Millenium line …And that is anchored by the Commercial/Broadway skytrain station (leaving practically no other option than Commercial for Transit)

        The East hasting corridor is a very important transit destination, because many jobs are concentrated here:

        picture is small, but I guess we get the idea on job concentration location,… and the reason why bus 7 (Nanaimo) or 16 (Renfrew) can’t be good alternative for most

        • Matt Taylor Says:

          I will have to apologize for just realizing now that you are preserving a lane of parking. As a concept then it improves or mostly preserves the level of access for each mode – but it is definitely very tight.

          You are showing raised bike lanes and a landscaped buffer separating the bus lane on the east. Would there be room for these? Enough buffer space will have to be provided for people to stand on and plug the parking meters.The bus on the east side will have a very narrow canyon to operate in. Getting to and from a parked car will be daunting, as will loading and unloading the bus directly into a bike lane.

          Two thing that it rules out are wider sidewalks and parklets. The value you place on these things is somewhat subjective, but they are something I would personally like to see.

        • Matt Taylor Says:

          I suppose you would keep the parking meters at their current location. It’s less strange but sill a somewhat awkward.

        • mike0123 Says:

          South of 1st, Commercial is 17 metres wide, divided mostly like 2.5 m/3.0 m/3.0 m/3.0 m/3.0 m/2.5 m to create 2 parking lanes and four travel lanes.

          It could be divided like 2.0 m/3.0 m/3.5 m/3.5 m/3.0 m/2.0 m to create bike lanes, bus platforms, and travel lanes, all at least 0.2 m wider than the minimum usually required for each.

          It could be divided like 4.5 m/2.0 m/3.0 m/3.0 m/4.5 m with 4.5 m bus/bike lanes, except that 2.0 m is not wide enough for a parking lane. It could be divided like 3.5 m/0.5 m/2.5 m/3.0 m/3.0 m/4.5 m, with bikes and buses unable to pass on the first (mostly downhill) side and a buffer so that more doors stay on cars.

          The 16 also connects the Expo and Millennium Line to Hastings on Renfrew.

        • Voony Says:

          I have considered 18m between side walk south of first (the total street RoW is 80feet):

          And for sure, that is a tight arrangement as I have mentioned in a previous post: Effectively, one meter less put the parking lane in jeopardy

          The illustration below is the Mike0123 proposal
          ( 1.5m m/3.0 m/3.25 m/3.25 m/3.0 m/1.5m to optimize sidewalk width… there is clearly room for wider bike lane…).
          ( I didn’t move the tree, since they are roughly at their current location, and mature tree are hard to move …)

          I like it (the idea of a 3.5 meter wide only bus+bike lane: not so much, since it is not wide enough for take-over, so leading to frustration and stress on both side)

          yes 16 connect with the expo line, but it means a way much longer trip: that is a geometry issue – the shortest transit trip to East Hasting from the Expo line pass thru Commercial (for the Millenium line, Clark could be an option)

        • mike0123 Says:

          I don’t think it’s immediately obvious that your proposal requires taking 1 m or more from the sidewalks (and cutting down all the trees on at least one side of the street) in order to maintain just one lane’s worth of parking.

          Even after taking 1 m from a sidewalk, the bus lanes at 3 m wide are less than the minimum that TransLink would usually require for a single-lane bus route, which is 3.3 m. The buses often take both lanes on Commercial because the existing lane widths are substandard.

        • Voony Says:

          The issue seems to be whether the actual space between the sidealk is 17m or 18 meter. As suggested by the rendering I have used googlemap (and vanmap for the total right of way): I didn’t go with my tape on the actual street.

          bike+bus lane at 4.5 meters are generally accepted good practice (example here in french), now also adopted in Montreal(does that allow raised bike lane between bus stop? may be not, In Richmond they did it on 4.75 width)

          if the lane is next to a curb lane parking, that create a “tunnel effect” slowing down traffic (as well, like for cyclist, “you need a “safety” buffer between the parked lane and the traffic lane) : so yes in those conditions 3meters could be too narrow.

    2. Rico Says:

      Great post.

    3. […] Several ways to address the bike+bus interaction exist, as noticed by Jarret Walker. As him, we also prefer a “table” or shared space solutions for the bike lane that alert the cyclist to yield to peds in this situation, as we have seen before: […]

    Leave a Reply

    Fill in your details below or click an icon to log in: Logo

    You are commenting using your account. Log Out / Change )

    Twitter picture

    You are commenting using your Twitter account. Log Out / Change )

    Facebook photo

    You are commenting using your Facebook account. Log Out / Change )

    Google+ photo

    You are commenting using your Google+ account. Log Out / Change )

    Connecting to %s

    %d bloggers like this: