Ottawa: Moving on from the Busway

DSCN2036

Ottawa is moving forward with a plan to replace its partial busway network with light rail, including a new tunnel under downtown.  As usual, The Transport Politic provides a well-linked overview of the issue.  So this is probably my last relevant chance to talk about my tour and observations of the busway in 2006.  I took a particular interest in this busway because it is the conceptual ancestor of the busway network now being built in Brisbane, and the basis for the “Quickways” concept advocated in the US by Alan Hoffman.


DSCN2056

I have two really strong impressions of the Ottawa busway.  Both may be relevant to understanding why Ottawa is moving on with a conversion to rail, and why Brisbane shows no signs of doing so.

In this post, I’ll talk about the main structural problem with Ottawa’s busway.  In the next post tomorrow, I’ll discuss an aesthetic one.

The structural problem is simple: the downtown segment of the busway system was never built.  Buses flow through fully separated busways extending in three directions from the city, but all these buses are then dumped into the same one-way couplet, Albert and Slater Streets, when they get downtown.  More remarkably, all these buses stop at the same stops.  And when you tell around 100 buses per hour to stop at the same stop, you get this:

DSCN2078

DSCN1895
It needs to be noted that this situation was not an inevitable consequence of the busway, or of the decision not to build a bus tunnel under downtown, as Seattle and Brisbane did.  It was the consequence of a decision not to give the buses the space they would need to move across downtown efficiently.  When you are talking about well over 100 buses per hour, the only fully reliable on-street solution is a double-width exclusive bus lane with skip stops (different groups of routes stopping at different stops, which alternate as you go along the street).

Marq2-homeIt worked on the pre-rail Portland Mall, it works on Seattle’s 3rd Avenue, and it will soon be working in Minneapolis, as I discussed here.  Double-width exclusive lanes can comfortably move about 180 buses per hour, with no more than two buses piling up at any one stop.  The way they do this is simple:  they insure that buses never block each other; when a bus is stopped, other buses can pass it.

You also have to insist that the lanes not be shared with turning cars, as happens in Ottawa.  Most commonly — as in Portland, Seattle, and Minneapolis — there are no parallel car lanes, or if there are, turns across the bus lanes are not permitted.  That way you don’t get a pile of buses behind one right-turning SUV.

DSCN2075

A double bus lane is obviously a challenge to advocate, mostly because it commandeers a large share of the street.  Obviously, you make the case for it, if you can, by showing how many people will move through the bus lanes on buses, compared to how many will move through the car lanes in cars.

There is no reason to question Ottawa’s decision to move on to light rail.  Light rail will certainly deliver better outcomes in terms of emissions, energy efficiency, and a more efficient utilization of labor, all critical features to the modern transit system.

But I do object to using this case as a basis for anti-busway triumphalism.  When people claim that that Ottawa’s light rail decision proves the failure of busways, I’d reply that you can’t judge a busway network that was missing its most important link.  It would be as if you had chopped off the roots of a sapling before you planted it, then watched it die and said: “See?  What did I tell you?  Trees just don’t work!”

38 Responses to Ottawa: Moving on from the Busway

  1. Adam Parast May 18, 2010 at 6:20 am #

    I completely agree with you except on the last point. Transportation is an extremely political topic and can’t be removed from the discussion. If there isn’t political will to give buses the space or ROW they need in North American’s best developed BRT system I think that is something that can’t be ignored.
    I believe in BRT as a concept but this is the reality. The quality of BRT is easily compromised in ways that are not acceptable for LRT. For example if LRT in Ottawa was at street level like buses are now would they ever allow cars to drive on the track? Never. You have done a very good job of showing how Australia quickways are much better but they tend to be the exception not the rule.
    I would also like to point out that the only reason that 3rd Ave in Seattle was originally closed to cars was to mitigate the two year closure of the transit tunnel for LRT construction. After the closure SDOT and Metro agreed to keep it bus only during peak periods. So in some ways LRT in Seattle was an impetus for it.
    Also, one of the major reasons that the transit tunnels was built was to accommodate future LRT. This was before my time but I suspect that if conversion to LRT wasn’t part of the project when it was built it might not have happened.
    So again I completely agree with you on a technical level but I think the politics of transportation simply favor rail. This doesn’t mean it is better or worse, these examples just show that politicians bend over backwards to make LRT work but not BRT.

  2. David in Ottawa May 18, 2010 at 7:08 am #

    Jarrett,
    After all the posts I’ve put on here, I would have thought you’d have had the good sense not to make a claim of the sort that a solution comprising a “double-width exclusive bus lane with skip stops” would somehow solve fundamental problems with the concept.
    For one, you’re taking twice as many lanes as would LRT to achieve the “same” result. Except it isn’t the same result, because you’ve cut in half the number of stops that could have been made available to each user (or you’ve doubled the number of stops needed overall).
    And, well, it’s not a solution that will actually work.
    We have this arrangement at Tunney’s Pasture and at Hurdman, both of which are on the Transitway, and guess what? We get jams there too. Yes. On an exclusive busway with a four lane cross section we get bus jams at less than 10,000 pphpd. I’ve got video and pictures of it. If you were to do this downtown you’d definitely get bus jams because on either side of each station the buses have to change lanes. Now instead of just sending 180 buses (not 100 – 180 – I’ve mentioned this number numerous times before and it’s readily published in City of Ottawa Transit Committee documents on the City’s website so there’s absolutely no excuse not to know it) through a single lane, you’re now going to be having them weaving as well. What happens when the curb lane in one block is completely jammed up (which it will)? That means that the buses in the passing lane in the upstream block that are destined to enter that lane have nowhere to go. So they start jamming up the passing lane.
    Unless you’re going to treat downtown as an enormous toll plaza and use multiple streets for rapid transit, this isn’t going to work.
    I also point back to my point about your erroneous use of numbers. You’re making it out that we’re failing at around 100 but actually we’re failing around 180 (much less on snowy days) yet you claim your double lane system will work out to 180 – which is exactly where we already are. So what do we do to get above that? It seems that BRT people are too willing to keep throwing out increasingly outrageous suggestions for how to try to make a dysfunctional system function. John Bonsall, who is the “father” of the Transitway, the originator of Brisbane’s Busway and apparently Alan Hoffman’s idol, has even suggested that a bus tunnel be built as only two lanes (two single lane tunnels) to save on the cost of building a pair of two lane tunnels. I’ve got a PDF of his presentation to Transit Committee, so I’m not making this up. He argued that the problem in Ottawa was red lights and that single-lane stations could function. My video evidence of Tunney’s Pasture and that of others of Hurdman suggests otherwise, but what would I know? I only observe the system up close and collect data on it rather than dreaming up ideas from afar.
    On top of all that, on Yonah’s blog you’ve even posted that Ottawa should have devoted resources to a costly bus tunnel. It’s kind of hard to justify such an outlay for a system that at the time was losing ridership annually all the while the modern marvel of the Transitway was being built. As for your suggestion there that outer busway infrastructure could have been sacrificed to provide resources for the tunnel, I have to ask what outer busway infrastructure? Nothing much other than a few random pre-cursor stations was built after 1996 until the early 2000s when a ~ 3km busway across open farmland was built at a cost of less than $15M to allow the bus lanes they had been using to be converted to regular traffic lanes – so not exactly enough resources to finance a tunnel. The only other thing that was built was the O-Train at a cost of about $30M and was the only piece of transit infrastructure we’ve built that has actually demonstrably increased ridership. Indeed, the estimates of the cost of a bus tunnel in the late 1980s exceeded the entire cost of the system as built in 1996 (and that cost had inflation in it). So where were you going to get the funding for this tunnel?
    Meanwhile, Calgary’s surface LRT continues to carry more passengers across its downtown daily than could a busway, and they’ve still got the potential for a one third increase in capacity.
    The kicker of Ottawa is that the Transitway failed to increase ridership, but twenty years later when absolute ridership finally got above where it had started, the system promptly failed. In other words, had it succeeded it would have failed, but because it failed it appeared for awhile to succeed (it would be interesting I admit to see how the spatial and time-of-day distribution of ridership changed over time since the express buses were clearly bringing people in in increasing numbers at peak periods – my guess is that non-peak ridership and inner area ridership fell precipitously, but that’s just speculation to try to fit the facts). Ottawa was just the wrong city to put such a system into. Its ridership was already too high for a busway system and the snowy climate ensured that winter snarl-ups of buses with no snow tires available to them became commonplace.

  3. Alan Robinson May 18, 2010 at 7:45 am #

    I have a question for both David and Jarrett. What makes LRT more efficient at offloading people that buses? David claims that an LRT using a single track per direction will have greater capacity than two dedicated bus lanes. If this is just a factor with the small number of doors on standard transit buses, can this shortcoming not be corrected by increasing the number of loading doors on buses?
    I see that Adam beat me to my other point.

  4. Jarrett at HumanTransit.org May 18, 2010 at 8:21 am #

    @Alan. I’m not aware of any difference in boarding efficiency that is truly an intrinsic feature of light rail versus rail-like bus-rapid transit vehicles. Features that matter to boarding time are the width and number of doors, the configuration of seating, and of course eliminating all fare transactions at the boarding point. Having a car floor level with the platform is also helpful. Many of these things are popularly associated with LRT because most LRT provides them. But none of them are intrinsic features that you can only get by choosing LRT.
    @David. Bus capacity is about STOPS, not LANES. Freeway capacity is about lanes. Rail capacity is about stations. Busway capacity is about stops.
    The operations that I observed on Albert/Slater streets are the result of too many buses trying to serve the same STOP, not too many buses going down the street.
    I’m sure you have video of bus congestion at Tunney’s Pasture when too many buses are trying to get to the STOP at once. But it would be very hard to congest a through running bus LANE that was always separate from the stopping lane — except perhaps due to a backup of buses waiting to get into a stop.
    If you have separate bus running and stopping lanes, you can move 180 buses per hour each way even through a CBD grid with signals at every street. In exclusive busway, with stops not blocking the running lane, you can go much higher.
    That also explains why a bus tunnel doesn’t have to be double-width. You only need the width at the statons. So both Seattle and Brisbane built tunnels that are one lane in each direction, and two lanes only in the stations, since these are the only places buses stop. Both Seattle (pre-LRT) and Brisbane today also assign different routes to different stop along the underground platform, to distribute the volume and help passengers wait close to the point where their bus will board. This also reduces the likelihood of buses piling up at stops to a degree that could impact the through lane.

  5. Jarrett at HumanTransit.org May 18, 2010 at 8:23 am #

    @ Adam Parast. You write:
    So again I completely agree with you on a technical level but I think the politics of transportation simply favor rail.
    Yes, I know that. But I think it’s important for North Americans to be aware of what all of the options are, including options that are well-demonstrated only overseas.

  6. EngineerScotty May 18, 2010 at 8:48 am #

    Portland’s transit mall presently handles 180 busses per hour, PLUS twenty trains, with no difficulties at all. Its silly to suggest something isn’t possible when it counterexamples–working implementations–exist.
    Regarding Alan’s question–while the larger vehicle size helps (a single LRT vehicle holds about 3x the number of passengers as a 40′ bus, and can load and unload them faster due to its longer length providing more surface area for boarding doors)–the real advantage of a train is being a train. Were you to take a LRT line and abolish entraining–requiring each “train” consist entirely of one car–the capacity of the line would go WAY down, even if you exclude the issue of paying the driver, as each individual vehicle has to negotiate with the others for space.

  7. David in Ottawa May 18, 2010 at 9:16 am #

    @Alan
    The bus jams in downtown Ottawa occur more in the afternoon than the morning (we’ve started to get morning jams as well, especially when it snows, but we got them first in the afternoon). So clearly boarding is more of a constraint than alighting.
    One source of delay is that Ottawa’s buses are not prepaid. For passholders, this isn’t an issue since they just flash their pass and move on, but for those paying with tickets/cash, they have to wait a few seconds for the transfer to print. I’m sure this could be fixed by putting up machines on the sidewalk that issue transfers, but that would require a degree of creativity – and a transit agency that thought that BRT was a good idea in the first place doesn’t have a lot of that. But even with pre-paid fares, I don’t think it would solve the problem since there are other issues as well.
    Platform congestion and confusion is a major issue. With so many express routes there are lots of buses going past that people can’t take – they have to wait for theirs. So that means more platform congestion than would occur if people were getting on pretty much the next bus that came along (the longer the average wait, the more people are waiting, which results in more congestion). The congestion of course makes it harder to get to your bus when it arrives – which brings up the next issue: confusion. No one has any idea where their bus will stop at the platform, since the platforms hold about 3 buses. That means when a platoon arrives there’s a mad rush in all directions to get to your bus. And it gets worse still. Suppose you’re waiting at the upstream end of the platform and you see your bus pull in at 4th position – i.e. it’s not at the platform yet, it’s alongside some parked cars or an entrance or something. So you know that when the platoon clears, your bus will be at the opposite end of the platform since it will then be first in line. Some people “game” the system and simply walk out into the street and basically beg to be let on, and the drivers usually oblige (in part because if they don’t they’re just going to have to wait for you to get to the other end of a congested platform). But having gamed the system, someone else follows your lead and before long the platoon has cleared and that bus now has an open platform ahead of it but it’s taking on passengers. So it boards the last of the gamers, advances, and stops again to take on those who waited.
    Then there’s the boarding itself. It’s a step up. People simply cannot board up a step as fast as across a level platform as you would have with rail. It’s sometimes thought that pre-paid fare would deal with the boarding issue, and since most express bus users have a pass, we shouldn’t see queues to board express buses – but we do because stepping up takes time. Opening up the other doors helps, but those double doors are just narrow enough combined with stepping up to prevent people boarding two at a time.
    Once inside the buses have narrow aisles with few places to stand. Most have stairs at the back. Internal circulation on a fullish bus can slow down the rate at which people can get on. In this case, delays from issuing transfers or even stepping up cease to matter as they are no longer the constraint.
    Then there are the periodic (but increasingly frequent with an aging population) delays from the need to lower the bus or the ramp to allow those with mobility issues to board. The ramp in particular takes a lot of time because the first thing that has to happen is that the congested platform in front of the door has to be cleared of people, then the ramp has to come out, then the person in the wheelchair has to line themselves up to the ramp, go up the ramp, turn and hope that the wheelchair area is clear (by this time it usually is), and, well, I could go on, but I think it’s clear enough. The delay from a single ramp lowering could cascade in time for several cycles/platoons.
    All these things start to add up. With 180 buses per hour that’s one every 20 seconds on average, and about 60 with platoons. The traffic lights that falsely get the blame actually *help* by forcing platooning, since they run at about a minute.
    The level boarding of trains absolutely and unequivocally cuts down on the time taken for each passenger to board or alight. That’s where the savings really start to come in. Other savings come from the reduction in platform congestion and confusion – ones location on the platform doesn’t matter and people will self-regulate by going where there are few people waiting.
    Another thing about dwell time is that each separate transit vehicle experiences it. A 40′ bus has basically the same dwell time as a 60′ bus, but the latter carries more people. So the dwell of an articulated bus moves more people than that of a 40′ bus. And what’s true of an articulated bus is also true of a train. One train, 3 cars (=~ 6×40′ buses or 4×60′ artics) has one dwell time of 20-30s, but the corresponding number of buses has at least twice that if operating in platoons and potentially more if not or if the platooning breaks (which it can and does).
    This is Ottawa-specific (and goes to the “wisdom” of choosing BRT in the first place), but add some snow to the mix and things really get fun. Buses lose traction in snow (no one has invented a bus snow tire because snow tires are soft but buses require hard-walled tires since they’re always rubbing up against curbs), so they approach and leave stations slower, pushing up the average dwell time. Passengers take more care boarding, especially with snowbanks but also just “pushing off” when they step up since the movement has to be vertical to avoid slipping. More time. Passengers stand further back to avoid being splashed or run over by a wayward bus (it happens – hardened snow at the curb can become a ramp). So more platform congestion – and more time.
    At the beginning I mentioned snow and the effect on alighting in the morning. Snow really does make alighting from a bus a major PITA. The snow, especially if it was recent (i.e. within a couple of hours of morning peak, or indeed during it), gets piled up at the curb. This presents an obstacle to exiting the bus. But even without the furrow, snow (and slush, which is what it quickly becomes) on the platform/sidewalk slows people down because they take extra care not to slip when one foot lands (you can’t really walk forward out of the bus – you have to step down with all motion being vertical so as not to slip upon landing). Each and every person has to do this and the delays quickly accumulate. It gets to the point where you can outpace the bus – and I’m sure people just decide to get off early and walk. This might seem odd to people from warmer climes, but it’s never too cold when it’s snowing. It’s seldom below -10°C, which for Ottawans is mild, and it’s also seldom windy when it snows either. So the option of getting out and walking is a relatively attractive one if you’ve been cooped up in a filthy (windows are covered in muck) bus.
    Failure to consider the effects of snow on passenger alighting and boarding movements has got to be one of the single biggest screw-ups of the entire Transitway concept in Ottawa. Failure to account for the effects of snow on buses and on the infrastructure of the Transitway itself (well, the effect of the salt used to combat the snow) would have to rank not far behind.
    This is what happens when highway engineers are given or allowed to assume the responsibility of designing a transit system. They failed to factor in the humans who have to use it and even the conditions in which it would operate. A standing joke amongst transit advocates in Ottawa is that the Transitway would work great if it weren’t for all the people using it. By contrast, the O-Train suffers from few of these problems. Snow is never an issue for the O-Train and the level boarding solves all sorts of problems. The platforms can even be cleared of snow by just pushing the snow onto the tracks – whereas doing that with a busway doesn’t solve anything. People in wheelchairs can actually board the vehicle in dignity without an enormous attention-drawing scene being made out of it. When it takes upwards of 90 seconds to get a person in a wheelchair into a vehicle on a rapid transit system that requires that dwell times not exceed 60 seconds even in a platoon because the delay will cascade through subsequent platoons, that ought to be telling us something about the wisdom of such a system.
    Ottawa was just the wrong place to try such an experiment. It’s been a costly failure and it’ll be costly just to turn it into what should have been.

  8. Rob May 18, 2010 at 9:30 am #

    Seattle’s 3rd Avenue is only restricted to buses during peak periods, but I think it would make sense to either make the restriction 24 hours, or make it one-way and dedicate multiple lanes to bus traffic with the opposite-direction traffic on 2nd or 4th.
    Also, it’s interesting to note that the Portland Transit Mall handles mixed buses and trains without any trouble even after collecting fares on buses within the area formerly known as Fareless Square. This is especially interesting as Metro explores the costs and benefits of keeping the Ride Free Area, and the possible impacts traffic on 3rd Avenue and in the transit tunnel.

  9. Scott May 18, 2010 at 9:39 am #

    I’m sure the Minneapolis bus lanes are open — they were planning on starting to use them last December just before I moved.

  10. Brent Palmer May 18, 2010 at 9:54 am #

    @Jarrett: “Both Seattle (pre-LRT) and Brisbane today also assign different routes to different stop along the underground platform, to distribute the volume and help passengers wait close to the point where their bus will board.”
    So far, the only station with this arrangement is King George Square in the CBD. During afternoon peak at Cultural Centre, a bus might use any of four bays – subject to availability – with the stop it’s about to pull into announced over the PA. Rarely does a bus pull in without someone making a mad dash along the platform!
    Maybe other busway stations could have a segment of platform allocated for each selected route group the whole time, although I’m not sure how many bays could fit on each platofrm. 3? 4?

  11. Joseph E May 18, 2010 at 10:02 am #

    “I’m not aware of any difference in boarding efficiency that is truly an intrinsic feature of light rail versus rail-like bus-rapid transit vehicles. Features that matter to boarding time are the width and number of doors, the configuration of seating, and of course eliminating all fare transactions at the boarding point. Having a car floor level with the platform is also helpful. Many of these things are popularly associated with LRT because most LRT provides them. But none of them are intrinsic features that you can only get by choosing LRT.”
    What about level boarding? Can you get a wheelchair directly onto a bus without a “fixed guideway?”
    Light rail vehicles can pull up within 2 inchs of the platform edge while traveling 55 mph. Most low-floor, kneeling buses seem to leave a 6 inch gap despite the best efforts of the bus driver, which means you need a ramp to get a wheelchair on or off.
    That greatly increased boarding time when a disabled passenger wants to get on, no matter how low your bus is, and how many doors it has.
    I’m sure the problem would be improved by having wheelchair ramps at 4 doors on every articulated bus, but it would still lead to longer stops than you would get with a train that can get right next to the platform, reliably, every time.
    The other features (width and number of doors, the configuration of seating, proof-of-payment or fare gates) can be done for buses, but only if you buy all new buses and make a new fare system. And if the fare system is different, you either lose the advantage of being able to run BRT on many branches to the suburbs, or you have to redo the entire transit system. Not that that would be a bad thing, but it would be expensive.

  12. David in Ottawa May 18, 2010 at 10:14 am #

    @Jarrett
    “But it would be very hard to congest a through running bus LANE that was always separate from the stopping lane — except perhaps due to a backup of buses waiting to get into a stop.”
    That’s what I’ve been arguing all along – you will get a backup of buses. The dwell times at the stations are such that this will happen. You discounted the Tunney’s Pasture example, but there the station is split into two stopping places, each of which is long enough to hold two or three buses. We get the queues in part because the drivers don’t know whether it’s clear or not, so they don’t move until they know they can. This is exactly what we would get in a tunnel, even with a passing lane. At Hurdman, even with an enormous platform, the place still clogs up. Not because of the cross-overs, but because a driver won’t go until he knows he’s got a place to go, which he frequently doesn’t.
    And what of the effect on the drivers who are going to be expected to carry out all this weaving? OC Transpo has enough trouble with retention as it is without adding even more stress. Think especially about winter. Always always always think winter when proposing something for Ottawa. Slush and snow and poor traction and rear-driven articulated buses. You will get jackknifing. And then you’ll have both lanes blocked. Congratulations.
    John Bonsall’s presentation was for single-lane tunnels AND single-lane stations – which is a complete repudiation of his earlier beliefs in the four-lane station arrangement (the double lane tunnels proposal of the late 1980s were really a single lane tunnel with that ubiquitous shoulder lane for parking buses that break down or run out of fuel – it was to recreate the 2 lanes + 2 shoulders arrangement of the Transitway elsewhere). I had a double-take when I heard it it was so shocking. He was advocating using another sure-to-be-a-wintertime-winner, the fixed guideway of sunny Adelaide. Even assuming it was only in the tunnel, those guidewheels are going to get knocked off elsewhere and clogged up with frozen filthy slush in the winter (one of the entertainments of little boys in Canada is kicking these lumps of frozen slush off the mudflaps of parked cars). Of course if it was in a tunnel then you have to have covered areas where the buses enter and leave the guideway so as to avoid having the entry point to the guideway get snowed on and iced up. The entry point would ideally be at a station upon either entering or leaving it to minimize the effect of slowing down to enter the guideway, but whatever you do, expect another queue.
    The proposed solutions keep getting more outlandish just to avoid the obvious one: switch to rail. I think this is what really angers a lot of transit advocates in Ottawa. The BRT lobby just won’t give it up. They built a system people didn’t want, then proposed a giant tunnel costing more than the system they had built to keep it running and after that didn’t fly the same people came back with a guided-bus solution. Others in the same industry come up with ideas of turning Albert & Slater into above ground bus sewers to make them even more miserable than they already are.
    Just give it up. No bus-based solution will work. We’ve lost a decade against when we should have started conversion because of the BRT lobby’s rearguard action.
    @Engineer Scotty
    Is Portland’s transit mall trying to move 10,000 people per hour on those vehicles? So long as the dwell times stay below the average headway, a system could continue to work at 180/hour or even greater – but that condition is only going to occur if there are far fewer passenger movements.

  13. EngineerScotty May 18, 2010 at 11:10 am #

    I’m not sure what the current mall throughput is, but it is certainly in the 10k/hour range. TriMet does 300k boarding rides per day, and a significant fraction of these trips originate or end in the transit mall.
    And as others have noted, the busses in the Mall are 40′ models with street-level boarding, generally running local or express routes, with pay-as-you-board fare collection. In other words, as inefficient as you can get if your biggest concern is minimizing dwell times.
    Of course, we don’t get snow like Ottawa…

  14. dejv May 18, 2010 at 12:26 pm #

    Joseph E –

    What about level boarding? Can you get a wheelchair directly onto a bus without a “fixed guideway?”

    Sure, if the platforms are equipped with Kassel kerbs.

  15. Daniel Olson May 18, 2010 at 3:01 pm #

    Yes, as Scott noted the exclusive duel bus lanes on two parallel streets in downtown Minneapolis opened in December and they work great. Downtown went from being gridlocked with buses at rush hour to flowing freely every day. By concentrating most of the express buses on the two streets, congestion was relieved on other downtown streets including the beloved pedestrian mall. Conversion to LRT will not be an issue though, as the buses jump on various freeways on the edge of downtown and split off to park-n-rides in far flung, low-density suburbs in every direction.
    Minneapolis is building light rail in the shorter higher density corridors which seem to be a better comparison to the Ottawa corridor. Unfortunately though, we’re building LRT with even less grade-separation than the current Ottawa busway.
    Thanks for your work on the Access Minneapolis project Jarrett.

  16. Chad May 18, 2010 at 3:10 pm #

    David,
    Sheesh, we get it. You don’t like buses.
    From 29 May 2009, “82 per cent of [OC Transpo] riders indicate public transit service is good or very good.”
    Rarely have I seen such an effort to portray too much success as a failure.

  17. David in Ottawa May 18, 2010 at 6:05 pm #

    Chad,
    What I don’t like is the BRT lobby’s attempt to portray Ottawa as something it’s not. I didn’t even realize this was occurring until I left Ottawa for awhile. I would hear some so-called transit expert waxing poetic about how great Ottawa’s system was and what I was hearing didn’t correspond to what I knew from daily experience. How well Ottawa’s transit system functioned was not something I had given too much thought to until I left Ottawa. Heck, I didn’t even realize how rare a system the Transitway was – I guess I assumed most cities Ottawa’s size had them (I’d been on metros in major cities like Montreal, Toronto and London, but I never really visited a city similar in size to Ottawa). It was a really shocker to encounter a far more function light rail system in Calgary.
    Statements of the sort accusing me of conducting “an effort to portray too much success as a failure” is just the sort of absolute hogwash I’m talking about. What success? Ridership fell for a dozen years and took two decades to get back to where it started. Per capita ridership is still lower. Once daily ridership did get above where it started, it promptly failed, giving us bus queues. It collapses in a 10 cm snow fall. It hemorrhages cash. It requires distorted operator schedules to keep the multitude of express buses running to far flung suburbs, which puts pressure on the transit agency to demand workday schedules far out of line with the Canadian transit norm. We had a lengthy strike in the winter of 2009 in large part over that issue, and though the transit agency won on a number of other issues, it lost on the workday rules. Ottawans take no pride in the Transitway. It doesn’t enter the collective mental map of the city in the way rail does; for their part the media have a hard time just getting the terminology correct. For the neighbourhoods it goes through it’s a blight that they’d sooner not have. The transit stations are ignored by the development industry – beyond a brief spurt of integrated station development early on, there’s been a noticeable lack of any bona fide transit oriented development around stations. So again, where’s the success?
    The one success it has had is for the BRT promotion lobby in allowing them to dupe lots of other places into thinking it’s a success by quoting Ottawa’s high ridership and insinuating that the Transitway is responsible. It’s funny: the Transitway is apparently a marketing success everywhere but in Ottawa itself.

  18. David in Ottawa May 18, 2010 at 7:12 pm #

    Oh, the irony. Sometimes one just has to smile. Here’s what the “successful” Transitway looked like this afternoon (18th May):
    http://www.flickr.com/photos/ottawabusgallery/4620477288/
    The problem was somewhere downtown in the westbound direction.
    Better off walking:
    http://www.flickr.com/photos/ottawabusgallery/4619863105/
    http://www.flickr.com/photos/ottawabusgallery/4619864305/
    By coincidence, the Canadian Urban Transit Association is meeting in Ottawa right now. I don’t suppose they were taken out to see this particular aspect of Ottawa’s “successful” Transitway in operation.
    Less than a decade left of this kind of nonsense. Less than a decade…

  19. Chad May 18, 2010 at 7:17 pm #

    David,
    Ottawa is a success based on the primary goal all North American transit agencies strive for:
    Ottawa has the highest rate of transit ridership per capita of all North American cities of a comparable size.
    If that’s failure, then let’s have more of it.

  20. David in Ottawa May 18, 2010 at 7:48 pm #

    Chad,
    I am fully aware of Ottawa having the highest transit ridership per capita of all North American cities of a comparable size.
    The thing is, we had those levels *BEFORE* the Transitway was built and all the while it was being built ridership dropped, bottoming out in 1996. That’s not exactly a success, is it? We had such high ridership that we could lose tonnes of it and still be above everyone else. That’s great that we were so high and great that we still are, but it’s not a Transitway success story, either.
    The Transitway-is-a-success people are just using the high transit ridership of Ottawa to promote BRT without letting anyone know that there is no correlation let alone causation between the Transitway and high ridership.
    For the causes of Ottawa’s high ridership, you’ll have to look elsewhere than the Transitway. I would have to speculate because no one at the City has seen fit to do an honest study on the causes of our high ridership, probably because they know they’d have to explain why it is ridership actually dropped.
    As I said, this is speculation, but I would offer up parking policies of both the federal government in their own facilities and municipal policies generally. Road access into downtown Ottawa has never been too good and a number of projected expressway-type projects were cancelled or not carried out. We had an early bout of it from the Federal Government – perhaps we had this expressway building so early that people also ended up seeing better of it sooner than elsewhere as well. A civic-minded ethos may well be part of it too, and the fact that fortunately transit never quite became stigmatized in Ottawa in the post-streetcar era. Why that is is from long before my time and long before the Transitway’s as well. It just seems to be the case.
    But it isn’t the Transitway. Correlation (high ridership in a city with a busway) isn’t causation.

  21. dreww May 18, 2010 at 8:16 pm #

    David – The problem in Ottawa is the shared stops. Here in Minneapolis our dedicated dual bus lane system has been a total success. Routes are assigned one of four specified stop letters. Each block has two stops and buses are timed so as not arrive at the same stop simultaneously. Bus drivers have mastered the jog from the right lane to the left lane, across the next two intersections and back to the right. Real time information boards and detailed route timetables are at all but the most lightly used stops. Passengers pay their fare outside the downtown area and are permitted to board from the rear doors in the afternoon. Buses zip through downtown and quickly find their way to St. Paul, the southern reaches of Minneapolis and the burbs.
    Car and bus traffic on other streets has eased.
    BRT in Ottawa may be a failure but crummy downtown design is certainly to blame.

  22. Jarrett at HumanTransit.org May 18, 2010 at 8:28 pm #

    Thank you for all the comments so far.  As someone who spent a year helping the City of Minneapolis with its downtown bus circulation issues, and who first proposed the double bus lanes there, I especially appreciate the reports from Minneapolis on how well they're working. 
    Convincing our friend David, however, is probably not a realistic, or relevant, objective.

  23. Brent May 18, 2010 at 10:23 pm #

    I have also read some of the comments David makes elsewhere, particularly the comment about platooning through downtown. The system is set up to allow buses to proceed in platoons of three buses per cycle, stop at a platform on a red signal that has been precisely timed, and then depart when it turns green. In theory it should work, but because the number of buses is at capacity, any minor fluctuation has a ripple effect with no room for recovery.
    My understanding is that a lot of this has to do with the fact that there are so many routes traveling through downtown arriving at loading platforms in random position — if I am waiting for an express bus that runs every 20 or 30 minutes, do I wait at loading position 1, 2 or 3? You only find out when the buses arrive and then there is jockeying to get to the correct loading position, and additional potential for delays (and if one bus doesn’t make it through, it means that one bus in the subsequent platoon won’t have room at the platform).
    I suspect a fairly easy and quick “fix” pre-LRT would be to restructure the routes operating on the Transitway so that the central portion only consists of one, maybe two or three, trunk routes. All the feeder and express routes, many of which only run every 20 or 30 minutes, would be truncated to run either from a common major terminus (e.g. Hurdman) or from the first station accessed on the Transitway. This makes boarding downtown more straightforward because you can board at any position in the platform (the 3-bus platoons operate similar to a train), but it also might allow the number of buses through downtown to be consolidated and reduced, if not all of the feeder / express routes are operating with full buses on all trips.
    One issue I see here is whether this transfers a capacity problem elsewhere (e.g. does Hurdman have enough platform capacity to serve as a transfer point from central trunk to outlying branches? what if you divide the branches between several outlying stations?). I don’t know if this would be the case or not.
    However I suspect that the main issue is political. There is a powerful allure to the concept of a one-seat ride and that is one of the key selling points of BRT over LRT. Will riders be willing to board a central trunk operating every minute or two, and then wait for a less frequent feeder service at Hurdman instead of downtown on Slater? I would argue that they will have to do this eventually anyway, if/when LRT is built through the central portion — and there are lots of other systems where this is readily accepted (e.g. Toronto bus routes feeding into the subway network).
    There was an excellent point above about the Transitway being touted as a cause of high ridership, whereas the more important factor is transportation policy (in this case, severely limited parking availability downtown). This could be tested by checking modal split and ridership in downtown Ottawa vs. other employment centres in outlying areas (Kanata, Gloucester etc.) where there is plentiful free parking. I would be interested to see what the results are.

  24. Tessa May 19, 2010 at 12:04 am #

    I would like to share David’s frustration at the busway being given credit for Ottawa’s high transit ridership. I highly doubt that’s the case given, as he said, ridership fell when the busway was introduced. I’m not suggesting the busway caused ridership to fall, either (it was during a recession, after all), but the busway shouldn’t be given credit for Ottawa’s high ridership when the high ridership existed before the busway, and hasn’t really improved since then.
    I also have to agree with our friends at the transport politic, who point out that Ottawa really didn’t save any money with the busway. The LRT line seems about the same cost as a similar LRT line would cost if it were built from scratch, and yet the city has already spent a lot of money on the busway and a lot of money on operation expenditures that would be lower with a rail alternative. I honestly don’t think the busway has worked out that well for Ottawa. Yes, this is parly because of the failure in design, but I still don’t know if I see the benefit of a busway over a rail-based alternatives as the main transportation spine in most cities. That’s just my perspective, though.

  25. David in Ottawa May 19, 2010 at 5:18 am #

    Jarrett,
    Even if a 4-lane system were to work, you’re still taking twice as many lanes to achieve at best what 2 lanes are achieving in Calgary. There’s an opportunity cost to taking twice as much scarce downtown road space for use in one rapid transit corridor.
    Brent,
    In general OC Transpo is gradually consolidating express routes just to keep treading water.
    But there is an irony here: the one-seat ride that is the touted benefit of BRT is being curtailed at the same time.
    Another interesting phenomenon in Ottawa is the suburban park & ride and the use of express buses. In theory, people in the suburbs are supposed to (1) buy a premium express pass and take an express bus from their neighbourhood direct to downtown; (2) buy a regular fare pass, take a feeder bus to the Transitway and then a 90-series (trunk route) bus to downtown from there; or (3) as (2) but drive to a Park & Ride instead of taking a local bus. But what a fair number are doing is something they’re not supposed to be doing: they buy a premium fare express pass but drive to the Park & Ride and get on the first express bus that comes through, which, give the volumes, doesn’t take long. Significant numbers are eschewing the supposed benefit of the “express” system as configured in Ottawa: the neighbourhood circulator function. From the perspective of the system, this is an unexpected benefit: it helps to ensure that more buses are being better utilized because if one express bus arrives at the Park & Ride full people will just wait a couple minutes for the next one coming in from a different neighbourhood. Downtown in the afternoon it also means that people are waiting less on average since they only have to wait for a bus from their ‘series’ (a bus in the 20s, 30s, 60s, 70s) that goes to that Park & Ride and not their specific neighbourhood bus. On the other hand, many of these express buses are circulating suburban neighbourhoods practically empty.
    Another thing is despite the dominance of the single seat ride in the thought processes of OC Transpo, lots of people are transferring on a system that never gave much thought to it. Hurdman is a zoo. LeBreton sees massive numbers and it’s just a platform in a field. Transitway stations are designed with heated shelters on the Transitway portion where service is frequent but the upper levels serving infrequent local buses are cramped, seating limited and never have heating. Sure, it’s impressive to charter a bus and drive transit professionals from afar around the system showing them the heated shelters on the Transitway and then carry on without thinking about the local services.
    The Transitway’s designers were generally highway engineers and it really shows in the lack of consideration of the people who were to use the system. It was flawed before it began.

  26. Alan Robinson May 19, 2010 at 9:36 am #

    I imagine that the choice of using a busway vs. LRT may have also influenced the route chosen for the busway. With a busway, one is able to send vehicles off of the dedicated ROW to reach customers, whereas it is more critical for rail transit stations to be next to the intended users. A busway should be able to use the cheapest ROW possible, whereas rail should find it’s way into densified areas, where construction is far more expensive. As I have not yet lived in a city with extensive BRT or LRT, I’d like to hear other people’s take on this.

  27. anonymouse May 19, 2010 at 1:15 pm #

    On the topic of Ottawa and ridership: what about the O-Train? This 5 station line with 15 minute headways gets 12,000 riders per day. If I’m doing my math correctly, that’s an average of 85 riders per train throughout the whole service span (6:30 am to 11:45 pm), and they only have two trains in service at a time! It’s a pretty high level of ridership for such a short line that doesn’t even go to the city center. All of which tends to indicate to me that the high ridership on Ottawa’s buses has more to do with Ottawa than with the buses.

  28. dejv May 19, 2010 at 3:01 pm #

    Another interesting phenomenon in Ottawa is the suburban park & ride and the use of express buses. In theory, people in the suburbs are supposed to (1) buy a premium express pass and take an express bus from their neighbourhood direct to downtown; (2) buy a regular fare pass, take a feeder bus to the Transitway and then a 90-series (trunk route) bus to downtown from there; or (3) as (2) but drive to a Park & Ride instead of taking a local bus. But what a fair number are doing is something they’re not supposed to be doing: they buy a premium fare express pass but drive to the Park & Ride and get on the first express bus that comes through, which, give the volumes, doesn’t take long. Significant numbers are eschewing the supposed benefit of the “express” system as configured in Ottawa: the neighbourhood circulator function.

    That proves one of Jarrett’s older points: that frequency is one of most important features of transit. People even prefer it to one seat rides in this case.

  29. Brent May 19, 2010 at 3:11 pm #

    So, based on the above, truncate the express routes where they meet the transitway (essentially turning them into local feeder routes), and use the # of buses saved to beef up service on the feeder routes (and on the transitway trunk, if/where passenger loadings require).
    Again, need to deal with the “one-seat ride” arguments — show how the added transfer is offset by improvements in convenience elsewhere, and by the way, that transfer is not a significant issue because service on the transitway trunk is so frequent.

  30. jack horner May 19, 2010 at 4:07 pm #

    So that Ottawans won’t feel too embarrassed, I advise that in Sydney, Australia, similar queues of buses are seen as 150 buses in the peak hour try to enter the city from the Harbour Bridge.
    I have strolled beside buses taking 10 minutes to go half a mile in the congested section.
    It’s an insult to the intelligence to see this sort of thing in rich first world nations which you might have thought would have the intelligence and resources to get this stuff right.
    It screams either incompetent transport authorities and/or don’t-care politicians (Sydney has both of these in spades).
    Makes Bogota’s busways look pretty good, doesn’t it?

  31. David in Ottawa May 20, 2010 at 9:44 am #

    I think what really bothers me with commentary on the Transitway is how uncritical it is. I think this is in part due to just how high transit ridership is in Ottawa.
    People see the high ridership numbers and they’re told that Ottawa has a busway system and so they jump to the conclusion that the former is the result of the latter. Seldom does anyone actually ask if ridership in Ottawa has improved because of the Transitway. It’s just sort of assumed that it did – after all, the high ridership is such an anomaly and so is the Transitway, so it has to be the cause, right? BRT promoters from Ottawa are a major part of the problem. They’re either directly misleading people elsewhere and/or are allowing others to draw conclusions through implication and omission.
    But let me put forward an alternate theory on the Transitway.
    The relationship between the existence of the Transitway and high ridership doesn’t run between the former causing the latter but rather the other way around: the Transitway was built because of high ridership.
    In the late 1970s ridership in Ottawa was growing for the first time in decades and growing fast. There was widespread support for trying to shape growth around rapid transit and using transit to address transportation problems, since by then Ottawa had seen enough of the expressway projects. The Spadina issue in Toronto was fresh in people’s minds, as was a growing environmental awareness.
    If you’re at OC Transpo in this era (OC Transpo in that era being a transit commission), you’ve got a problem few other places have: you’ve got rising ridership, rising so fast that you’re running out of buses. And the buses you do have are getting stuck in traffic. What you want to improve your daily financial situation is a way to speed up the buses so you can send them back out to make another run. Increasing ridership isn’t a goal: it’s a problem to be addressed. A rail system might be nice to have, but what you want is a solution that can be implemented in a hurry.
    Meanwhile, the road engineers have just been deprived of work because all the expressway projects they were expecting to be planning and designing have been cancelled.
    So you have a transit commission with a transportation problem looking for solutions and road engineers looking for transportation problems to solve. The idea of a busway must have seemed a fine idea to both of them. What anyone else thought or what the planning implications were wouldn’t enter into it, since, after all, the public was taking transit in increasing numbers. Public and elected opposition to busways and support for light rail (including some impressive citizen proposals using images of German light rail vehicles superimposed on the still-existing tracks at Tunney’s Pasture of the rail corridor that became the West Transitway) was brushed aside with vague promises of convertibility and a largely unaccountable regional governance structure that was looking to wrest planning control from the Federal government’s National Capital Commission which had done most of Ottawa’s major postwar structural planning. The NCC had done its bit though by pulling out most of the railways, leaving behind tempting rail roadbeds and rights-of-way. The Province of Ontario offered to put up the cash.
    And so Ottawa got first BRT with exclusive lanes here and there and eventually busways. What it also got was a BRT lobby that went on to promote the concept elsewhere, and was so successful in that effort that it duped much of the rest of the world into believing that the Transitway itself was successful in creating Ottawa’s high transit ridership when, in all likelihood, the Transitway they were using to promote the concept actually owed its existence to Ottawa’s high ridership.

  32. Nathanael May 23, 2010 at 8:05 pm #

    Well, the interesting thing about this is that it shows a MAJOR advantage of rail which you haven’t discussed much: low real estate usage!
    The tight footprint of rail is another one of the reasons why rail ends up being superior to buses for high-capacity applications.

  33. Nathanael May 23, 2010 at 8:07 pm #

    “So, based on the above, truncate the express routes where they meet the transitway (essentially turning them into local feeder routes), and use the # of buses saved to beef up service on the feeder routes (and on the transitway trunk, if/where passenger loadings require).
    Again, need to deal with the “one-seat ride” arguments — show how the added transfer is offset by improvements in convenience elsewhere, and by the way, that transfer is not a significant issue because service on the transitway trunk is so frequent.”
    People prefer connections where one of the ends of the connection is a train, of course….
    In practice the light rail conversion will do exactly what you described, successfully.

  34. Nathanael May 23, 2010 at 8:13 pm #

    FYI, I’m not about to criticize the Minneapolis application. It has several features which make it very smart:
    (1) It took over *existing* general-purpose roadway capacity. This made installation cheap. The supposed cheapness of BRT is usually destroyed by refusal to take away lanes from cars. Converting car lanes to bus lanes where there are lots of buses is a *very very good idea*. How on Earth did you do it, politically?
    (2) It’s not considered a substitute for an effective local rail system. The busiest trunk corridors (Minneapolis-St. Paul and Minneapolis-Airport/Mall of America) got rail.
    (3) It actually connects properly to the rail as well as into downtown.

  35. Nathanael May 23, 2010 at 8:17 pm #

    “Sure, if the platforms are equipped with Kassel kerbs.”
    These should be installed universally. Though training the drivers to use them properly is another matter — I was disturbed when LA bus drivers just stop stuck halfway out in traffic.

  36. Nathanael May 23, 2010 at 8:23 pm #

    “@Alan. I’m not aware of any difference in boarding efficiency that is truly an intrinsic feature of light rail versus rail-like bus-rapid transit vehicles.”
    Wake up, Jarrett. There are no “rail-like bus-rapid transit vehicles”, and the missing factor is *trains*, which are made possible by *rails*. You can have 30 doors per side in a really long streetcar train. (See the “yellow caterpillars” in Budapest.) You simply cannot have that in a bus, because it would be too long and would fishtail too much on the back end. A whole bunch of buses doesn’t substitute because they need space between them, tight platooning slows them down, et cetera.
    Now, if you’ve designed your light rail track and stations to only allow one short streetcar, then you might as well have built a trolleybus. If you’ve designed them to be expanded to three very long vehicles (Minneapolis), or more, you have faster boarding.
    But perhaps you were only thinking of the *direct* comparison of one bus versus one short streetcar.

  37. Brisben June 14, 2010 at 12:38 am #

    I think the busway has served its job. It has worked but now it is time for Ottowa to move on.
    The underground link was probably not built for the same reason light rail was not built, and BRT was chosen instead: ridership & lower cost.
    It’s the same in Brisbane. There is no underground tunnel from the South East Busway to the Queen Street Mall underground bus tunnel & station. Why? Cost avoidance. An underground from Roma St station to Queen St and a new King George Square cost $333 million AUD to do.
    BRT is precisely cheaper because you use roads in parts, you run on city streets and you delay (but not avoid) having to dig tunnels in the CBD and put off LRT and all that until later.
    Even Brisbane’s Queen Street tunnel and King George Square station combined cannot hope to accommodate all the city’s buses. A substantial number of buses still spill on to Adelaide and Elizabeth Street.
    Cultural Centre could not cope with the sheer volume of buses if many services did not exit the busway onto the Captain Cook Bridge, use the freeway and Elizabeth St. Other services use skip stop or zone express operation to avoid creating congestion.
    Many busways are sold on the idea that “they are just like LRT and can be converted to LRT later”.
    For Ottawa, I think that time for conversion from a successful BRT to a successful LRT has now come.

  38. Matt Fisher July 7, 2010 at 7:29 am #

    David in Ottawa,
    You’re right. John Bonsall and others are portraying Ottawa as something it’s not. While I believe that the Transitway is responsible for Ottawa’s high transit ridership, and I am currently satisfied with existing OC Transpo service – most people I know in Ottawa ride the bus every day, including many students I went to my now closed high school with, and now at Carleton, where I’m a student – the policy is another factor.
    On January 28, 2010, I wrote a letter to the editor of the Ottawa Citizen (our main daily newspaper) about the BRT issue. I made reference to the Curitiba comparison. You should read it. 🙂