the opportunities and dangers of incomplete bus rapid transit

One of Bus Rapid Transit's great virtues is that unlike rail, you don't have to build a complete, continuous piece of infrastructure if you really only need segments of one.  

Here in Portland, for example, the Barbur corridor — now being studied for BRT or rail — features a series of congested chokepoints with generally free-running traffic in between them.  Here, a BRT facility that got transit through the chokepoints reliably probably wouldn't need an exclusive lane in the free-flowing segments, because traffic in those segments would continue to be metered by the chokepoints and thus remain uncongested.  (Congested chokepoints meter traffic just as ramp meters do: they limit the rate at which cars can enter a road segment and thus reduce its chance of becoming congested.)

Unfortunately, Bus Rapid Transit can also be implemented in exactly the opposite way.  Severely congested chokepoints are generally expensive places to design transit priority for, especially if you're unwilling to simply take a lane for transit.  So we often see BRT projects that are missing where they are most needed.  The Boston Silver Line 4-5, like the Los Angeles Silver Line, can get stuck in traffic downtown.  New York's supposed BRT is so compromised that many refused to call it BRT anymore.  Even the world-class Auckland North Shore Busway disappears as it approaches the Harbour Bridge. 

Now we have the example of Seattle's RapidRide D, highlighted today by Mike Lindblom in the Seattle Times:

While the new RapidRide bus mostly lives up to its name in West Seattle, passengers on its sister route to Ballard are routinely stuck in traffic.

The service to Ballard, called the D Line, is d


elayed 10 to 15 minutes by late-afternoon car congestion leaving Belltown and winding through the crowded Uptown neighborhood, near Seattle Center.

That bottleneck is aggravated by traffic signals that haven't yet been re-timed by King County Metro Transit and the city of Seattle, to give the buses a longer or quicker green light. Metro acknowledges the D Line is just one minute faster than the local bus it replaced Sept. 29; the advantage is supposed to be six to eight minutes.

Transit managers hope to make gains by early 2013 after signal and road-lane changes are finished.

"We have a ways to go based on our early experience, but it is still too early to know whether the projection will be achieved," said Metro spokesman Jeff Switzer.

M674_0Just one minute faster than the bus it replaced?  Then the question arises: Why was it called Rapid Ride prior to the improvements that would make it Rapid?  There are some plausible if grim answers to this question.  Getting multiple big bureaucracies to move on the same timetable to the same deadline is hard.  The transit agency has to commit to a date months in advance, without being entirely sure whether its partners (typically in the City and the state Dept of Transportation) will be done with the improvements that are their responsibility.  So sometimes, the brand appears before the product does, causing this understandable blowback and also, more critically, tarnishing the brand.

RapidRide D raises a larger problem though.  Even when planned priority is completed further south there is still the problem of the congested Ballard Bridge.  Like Barbur's chokepoints in Portland, the Ballard Bridge is a familiar chokepoint that affects speed and reliability for all transit services forced to use it.  You can imagine the difficulty of demanding that RapidRide have an exclusive lane over the bridge, when that would leave only one for other cars.  (But what about a lane for buses + carpools + carshare cars + electric cars + etc. until you get a reasonable but uncongested lane volume?)

Sometimes, too, bridges can be metered, much the way the San Francisco Bay Bridge toll plaza meters traffic on that bridge.  At the approach point pictured above, a signal could have been placed at the bus merge point which meters traffic so that northbound congestion piles up south of the bridge rather than on it, and enters the bridge only at an uncongested rate.  That would have allowed buses uncongested operation without really slowing down cars much.  I'm not an engineer; there may be valid reasons why this wasn't possible, but it's the sort of solution that comes up when congested traffic is the reality anyway and the goal is to protect transit from it.

Transit agencies sometimes compromise BRT for their own reasons of budget.  Issues of boarding time associated with the lack of on-street ticket machines are coming up on RapidRide, as are concerns about reliability arising from the fact that two RapidRide lines are through-routed, transmitting delay from one to the other.  These are familiar struggles within transit agencies who are under pressure to spread a product over many corridors and can't afford to deliver every aspect of the product in all those places.  The result runs the risk of becoming symbolic transit; a bright red line appears on the map, but without the investment needed to make good on the promise that the red line implies.

I've received emails from Seattle friends on several sides of this issue, and sympathize with all of them.  I don't mean to criticize either the City or the State DOT or the transit agency, because what was done here is fairly typical historic American practice and the pressures involved are so routine.

But if there is a desire to aim higher than historic American practice, the question remains.  How much can we compromise BRT — tolerating its absence precisely in the congested chokepoint where it's most needed — and still call it BRT?  Might be better for transit agencies to refuse to implement BRT until the relevant traffic authorities have delivered the facilities it requires?  

31 Responses to the opportunities and dangers of incomplete bus rapid transit

  1. d.p. November 28, 2012 at 2:31 pm #

    Unfortunate coda:
    Seattle appears poised to one-up itself by building what it is trying to term a “rapid streetcar”:
    If built, it would operate in — you guessed it! — exclusive ROW where the traffic already flows freely, and in mixed traffic through every major chokepoint!

  2. Bruce Nourish November 28, 2012 at 2:43 pm #

    The whole RapidRide C/D rollout has been a disaster, and unpacking the causes has been the subject of a number of lengthy Seattle Transit Blog posts. My one-sentence summary is that the agency is under-resourced and demoralized from years of post-2007 austerity, and doesn’t have the money do anything which could seriously be called rapid transit, but the politicians the agency works for don’t want the political hit of canceling or seriously delaying a project promised to the voters in a local ballot measure.
    There’s also some gratuitous own-goals from Metro, like not publishing a schedule and claiming that every 15 minutes is “so frequent you don’t need” one.
    Believe it or not, the situation is actually going to get much worse for both the C and D in a few years. In 2016, the Alaskan Way Viaduct will come down, which is a good thing for the city, but will unavoidably increase travel times by several minutes for C Line riders. The state gave Metro some money to mitigate the increased travel times from West Seattle, but that will run out in 2014. The money was used to pad the schedules (to account for delays caused by construction) but also to add peak trips on West Seattle routes. Those trips are now full of riders, and nobody has any clue what to do about 2014.
    Between 2016 and 2018, viaduct demolition work means there will be no good arterial connection between Elliott Ave (the primary arterial to Ballard) and the south end of downtown. Drivers from Ballard will have to turn onto the single-lane West Mercer Place (where RapidRide runs) and go to the new tunnel portal at Aurora & Mercer (or slog through the downtown street grid). This is going to make Uptown a traffic disaster area, and the D Line will have to slog right through; again nobody has any clue what to do about it.
    The West Seattle-Ballard corridor is one of the few places in Seattle that just needs full grade separation, at least as far as Interbay and Delridge. Even if C/D had been done perfectly, the grid in Uptown and the potential post-2016 pathways are just not easy places for surface transit to move fast. Compare this to the future E Line on Aurora, which actually will have a pretty solid path into downtown, and could potentially have been a great Geary-like BRT project.
    As it is, RapidRide will mostly operate like SelectBusService, achieving little except further ingrain the idea that North American transit agencies are (with a few exceptions, like LA) hopelessly incapable of doing anything good with buses.

  3. Cascadian November 28, 2012 at 2:51 pm #

    This is why people distrust the whole concept of BRT, even though it can be done right in theory.
    Also, in the case of Seattle, the solution for much of this corridor could have been to replace the highway tunnel that is being built to replace the Viaduct with a transit tunnel that could have served stations along the same line as RapidRide D, and the northern part of RapidRide C. But the idea was a non-starter, because people were worried about congestion from the car traffic on the viaduct shifting to surface streets. Of course, projections have most cars doing that anyway with the tunnel, to avoid paying the toll.

  4. M1EK November 28, 2012 at 3:52 pm #

    If all US implementations end up this bad, and they do, then it is no longer honest to claim BRT is as good as rail. Even if your customers desperately want to be told it is.

  5. Geoffrey November 28, 2012 at 4:17 pm #

    This is not an issue of rail vs bus. A rail based solution without exclusive right of way would have suffered the same issues.

  6. Nathanael November 28, 2012 at 4:22 pm #

    Geoffrey, as d.p. mentions, Seattle seems set to make exactly the same mistake with rail.
    However, most places don’t. It seems to be possible to demand, and get, exclusive ROW for rail in most parts of the country.

  7. M1EK November 28, 2012 at 5:02 pm #

    In THEORY, rail can be done as badly as BRT always is. In PRACTICE, it isn’t. There’s that difference again! And it’s the dishonesty of pretending like shared running isn’t a killer, ported over from bad BRT, which is responsible for these (thankfully few) stupid shared lane rail proposals which are still mostly low volume proposals anyways. For instance, Seattle wasn’t dumb enough to share lanes on their LRT line, were they?
    When buses practically scream ‘compatible with cars’, it’s dishonest to pretend they’re NOT radically more likely to be forced to share lanes with cars.

  8. d.p. November 28, 2012 at 5:37 pm #

    M1EK, Seattle was dumb enough to share lanes on its streetcar starter lines (the South Lake Union and First Hill trolleys, the latter of which intends to fill a gap in the light rail subway left by the Bush-era FTA’s anti-urban cost-benefit bias).
    And Seattle indeed appears to be dumb enough to expand those same short-line streetcars to distances at which real rapid transit is necessary — — retaining mixed-traffic running where it hurts most.
    And Seattle may be dumb enough to think a bottlenecked “rapid streetcar” can preclude the need to build real in-city rapid transit in the future (they will be studied simultaneously and contrasted; political forces are sure to rally around the cheapest “option”).
    I’ve learned never to underestimate how dumb Seattle can be about transit. At least we’re on the vanguard of something.

  9. Geoffrey November 28, 2012 at 6:34 pm #

    ‘In THEORY, rail can be done as badly as BRT always is. In PRACTICE, it isn’t. There’s that difference again!’
    You are ignoring a whole host of examples of bus based systems with dedicated rights of way and rail systems with on street running.

  10. Bruce Nourish November 28, 2012 at 7:06 pm #

    “it’s the dishonesty of pretending like shared running isn’t a killer, ported over from bad BRT”
    WTF? You do know virtually every prewar street railway was primarily shared running, right? It’s not dishonestly you need to worry about, mate, it’s cluelessness.
    People on the internet should be banned from making statements such as “Rail lines always have (insert desirable feature) and buses don’t” until they’ve visited Los Angeles and San Francisco and travelled extensively on their transit networks.

  11. M1EK November 28, 2012 at 7:16 pm #

    Another dishonest tactic is to ignore clearly intended context – such as that we’re talking about decisions made the last decade or so, so prewar systems aren’t really that relevant.

  12. Bruce Nourish November 28, 2012 at 8:04 pm #

    No, that context is not clear at all, and you’ve still presented no examples of transit agencies saying “Hey, let’s not build an exclusive lane for this train, because … BRT!”, but you do seem to have the time to cast aspersions at people trying to apprise you of nuance which seems to be missing from your understanding.
    For example, Muni’s within-the-last-decade T Third is a major trunk route that runs in mixed traffic in a section south of the Dogpatch; presumably the reason is the same one which plagues bus operators: the neighborhood didn’t want to give up street parking, and the agency compromised to get the project built. At least T Third is grade separated in the city center.
    Can you show me some evidence they were inspired to do this by “bad BRT”, especially given that Muni recently finished design on two excellent, center-running, exclusive arterial BRT projects on Geary and Van Ness?

  13. Dexter Wong November 28, 2012 at 11:49 pm #

    Presenting an incomplete BRT plan hasn’t stopped politicians from promoting it as a panacea. In the last mayoral election in Honolulu, Ben Cayetano put out a BRT plan called “FAST” as being superior to the rail project that will be constructed once all objections are overcome. His BRT plan had many holes such as running on freeway shoulder lanes when many sections of freeway had no shoulders. His opponent Kirk Caldwell promoted rail and improved city services. When the election was over Caldwell won handily and BRT was no longer on the table.

  14. M1EK November 29, 2012 at 7:06 am #

    Bruce, this conversation started about recent projects – that’s the obvious context.
    Yes, some historical systems ran in-street. Yes, a few agencies still make dumb rail decisions – but to ignore the fact that an LRT system built in the last 30 years is 99% likely to run in its own lane even through ‘chokepoints’ while a “BRT” system is 99% likely NOT to run in its own lane through ‘chokepoints’ is dishonest.
    There is fundamentally something about BRT which makes it far more likely to be compromised into shared running. We all know what it is.

  15. M1EK November 29, 2012 at 7:07 am #

    @Dexter even worse than that – there was a previous “BRT” plan for Honolulu which was thankfully abandoned in favor of LRT a few years ago which would have run in shared lanes through an intolerable amount of its route. Note that’s 2 bad BRT plans in Honolulu that would have run in shared lanes; one LRT plan under construction now that doesn’t.

  16. Jeremy November 29, 2012 at 10:12 am #

    This is the problem I have with BRT. Although buses are great tools which can do great things, the reality is that they are never allowed to do those things.
    The proliferation of so-called BRT is a direct result of car-centric thinking. To most city planners and politicians, car throughput is most important, so “taking a lane away” is something they won’t do. And because transit agencies don’t control the streets, that take what they can get in the hope of making gains from stop consolidation and off-board fare payment.
    It’s also a fairly typical political issue in that BRT helps politicians say they’ve done something while not having to do very much at all. The way that it’s been framed is just wrong: it’s “like rail but inexpensive”, when in reality it is expensive to do it right.
    Because of the pro-car bias, transit only gets priority in uncongested places (see Silver Line Washington Street for a perfect example). NYC has found that bus lanes are often useless for this reason and worse than useless since they are always blocked by parked and turning cars. Most NYC bus drivers I’ve talked to will say directly that they don’t use bus lanes.

  17. Alan November 29, 2012 at 10:47 am #

    The NYPD is also a very car-centric agency. From how often bike and bus lanes are violated, it’s clear they have no interest in keeping them clear. I’m pretty sure the state legislature is holding up bus-lane enforcement cameras.

  18. November 29, 2012 at 12:12 pm #

    Bruce Nourish wrote: The West Seattle-Ballard corridor is one of the few places in Seattle that just needs full grade separation, at least as far as Interbay and Delridge.
    Yep. And as I recall, over a decade ago there was an agency that was looking at exactly that route. Elevated… from West Seattle to Ballard. Whooshing over cars. In fact I might’ve even interned for those guys, come to think of it. Set up and took down the meetings. Got into late-night AIM arguments with Dick Falkenbury.
    But that didn’t pan out, didn’t it.
    You’d almost think Seattle was incredibly effing mypopic about transit improvements, or something. I mean, what if, in some alternate reality, the lower deck on the Ship Canal bridge had been designed for heavy rail? What if the Feds had offered to mostly-fund a system like BART or MARTA in the 70’s? What if the CPSRTA had held a bunch of votes on a longer LRT system than what eventually got accepted as SoundTransit?
    Oh wait, all of that actually happened. Right. My bad.
    Good luck, I suppose.

  19. Joseph Singer November 29, 2012 at 4:25 pm #

    Just FYI: Checked with a speedometer app and LR the fastest it ever gets is 30 MPH. Buses in town that never use the freeway are mostly around 18 MPH with occasional bursts of up to 30 MPH.

  20. Matthew November 29, 2012 at 10:04 pm #

    Well, I took a moment at the public meeting to tell the head honchos over at MassDOT that the Silver Line is not BRT unless they put the bus lanes where it is politically hard to do so.
    Doubt they’ll listen.

  21. JN_Seattle November 30, 2012 at 2:50 am #

    I ride the bus described, and I like it. As with all transit rides, it could be better. It’s not a subway, nor a streetcar, but it was built faster, and costs taxpayers a lot less.
    The picture shown above that’s reprinted from the front page of the Seattle Times newspaper was I think meant to illustrate a new RapidRide bus “stuck in traffic.” But it’s not. The positions of the cars in the photo leads me to re-interpret the picture as showing heavy, mixed-mode traffic in motion, perhaps following an opening of the Ballard drawbridge to let a boat through on the Ship Canal that divides Seattle in half.
    For non-Seattle readers, notice the elevated bridge tender’s cabin, the raised gates, and the green lights at the draw span further down the bridge, about halfway across the bridge, but distorted to look further away by the camera perspective.
    I’m guessing that the RapidRide bus in the picture would move across the bridge at the same speed as the general traffic. Metro planned for nothing different. When the bridge lifts, everybody is stuck. Welcome to Seattle, and please turn off your engine to reduce emissions.
    Given the daily volume and mode mix on that 1917 bridge, a dedicated lane is not an option that will likely be painted in. Through 1939, the Ballard bridge had tracks of the Seattle Municipal Street Railway on it, part of a citywide network that was completely ripped out and gone by 1941. Now it’s just damn busy all day with cars, trucks, and a relatively few buses. Bikes and peds have to use a narrow sidewalk.
    I live near where the picture was shot, and often make the merge that this bus is shown making as a car driver. Under Washington State law, that bus has the right of way on entry. It is coming off an entrance ramp to the bridge that merges a bus station stop re-entry point with two streams of car/truck traffic. Note the “yield” triangle sign on the bus above the left tail-light in the original picture Seattle drivers will let that bus enter.
    The performance expectations for RapidRide, funding for which was voted in 2006 and talked about ever since, with publicity and hype climbing throughout 2012, were allowed to soar beyond what could be met. There are lots of ways in which this line can be and will be fine tuned into greater speed and reliability over the decade that will pass before a rail line is possibly built to serve some of the points in this corridor.
    Money matters. The Metro transit agency has to serve the entire county including all of Seattle and Bellevue, and the level of resources limits how much it can do in any one corridor.
    Finally, the term “BRT” has now morphed in practice to mean many different levels of service, many of which are not very close to “like a train on rubber tires,” which was a popular phrase ten years ago.
    King County Metro doesn’t use “BRT” or “bus rapid transit” in its latest web page description of the RapidRide service.
    Europeans have now introduced the term BHLS, meaning buses with a higher level of service, which I think fits the idea that there are all sorts of ways to improve bus service for more speed and reliability, including boarding through all doors, paying with smart cards, and bus stops further apart.
    Some of these techniques have been incorporated on parts of the Seattle RapidRide already, with more improvement coming in the future, phased in. I call this approach “incremental BRT” as described generally at

  22. PragmaticLRTsympathizer November 30, 2012 at 6:17 am #

    The thing is, some of the objections to street cars – blocking other traffic if there is a street car breakdown, incompatibility with cycling, etc – are precisely ones alleviated by a seperate ROW. So there is already incentive to give rail a seperate ROW quite apart from the benefits to transit. This is true despite the fact that many streetcars are mostly or completely in mixed traffic. Plus when you DO give BRT completely seperate ROWs, the economics don’t look nearly as favorable. You’re no longer getting the big capital savings over rail. To really get that big capital savings, but still get the benefits of seperate ROW, you have to do as Mr Walker suggests – run BRT in mixed traffic in non congested areas (so no incremental capital dollars, whereas rail would have required track anyway) but have seperate ROW only at chokepoints. To the extent that is not possible, it tends to weaken the entire case for BRT vs rail.

  23. Ntrain87 November 30, 2012 at 10:00 am #

    This radio program on the Grand Boulevard Initiative on El Camino Real in the SF Bay Area is very relevant, because the plan involves BRT in Santa Clara County. However, most of the planned exclusive lanes have been rejected by three cities along the route.
    How can we convince cities that efficient bus transit is in their best interests?

  24. Jeffrey Bridgman November 30, 2012 at 11:03 am #

    @Ntrain87: Unfortunately bus has such a bad name in America that I’m not entirely sure you could convince cities to provide a separate ROW unless you are going with rail… even then it will be a tough fight.

  25. Jack Whisner November 30, 2012 at 2:19 pm #

    During planning, I suggested the signal gate at the south end of the Ballard Bridge mentioned by Jarrett; I heard Grace Crunican, SDOT director, mention it in public. It was not used. With funding and will, it could be in the future.
    The key question transit lines is more basic. Is it better transit? Has service frequency, speed, reliability, and connectivity improved? Is more transit mobility provided? Are more riders attracted? Of course, complete grade separation is much better. Can it be achieved within budget?
    Is there a Yogi Berra paraphrase here: nobody rides the C or D lines any more; they are too crowded?

  26. d.p. November 30, 2012 at 7:17 pm #

    Has service frequency, speed, reliability, and connectivity improved?
    No, it hasn’t. Not even in the slightest.
    And in the mid-day and the late evening, service frequency has in fact dropped in comparison to the two services that were replaced.
    And then there’s the matter of no schedule and no real-time info.
    And what’s more, I know you know this, and it pains me to see you fall on your sword for Kevin Desmond and the other liars in Metro’s Department of Pathetic Spin.
    nobody rides the C or D lines any more; they are too crowded?
    Outside of rush hour, most avoid it like the plague. As your own quarter’s-end numbers will almost certainly show.

  27. JN_Seattle December 1, 2012 at 12:56 pm #

    I haven’t noticed low ridership outside of rush hour on RapidRide D, the time when I usually ride it.
    In any event, it seems King County Metro has an evolving case study in the service change brought by the advent of the RapidRide C and D lines in Seattle. Bring on the data!
    Will the data show that service is getting better or worse with the changes brought by RapidRide? Time will tell.
    The situation is fluid, since adjustments are still being made. Equipment is still being installed at some of the stops. Buses are being added. And patrons are learning that they can speed up dwell times by boarding through the rear doors legally with a transfer in hand, including a smart card that has already been tapped. My driver the other day on D line was training patrons as they boarded unnecessarily at the front door.
    Picking up on the blog post headline, I claim the opportunities overwhelm the dangers when we think of King County’s lite, arterial BRT as “incremental” rather than “incomplete.” Whatever will happen, it will happen pretty quickly compared to … building train lines.
    And BTW (here I go!), if you want to see “incomplete” and “dangers” in a transit system, consider Seattle’s signature system, Central Link light rail. Make that “near-permanently incomplete” as construction and funding woes continue for the next several decades. This is our incrementally forthcoming light rail subway line burrowing northward out of downtown at the price of BART, with operational speed and capacity to be limited by four-car trains and a sharp 90 degree turn under downtown Seattle.
    Subway builder Sound Transit is forecasting twice the daily rail ridership in 2030 than is forecast by the Metropolitan Planning Organization for 2040. Who to believe? It depends on whether you think Seattle folks in the 2020s and 2030s will show up to pack themselves like sardines into light rail cars.

  28. Erik Griswold December 1, 2012 at 7:28 pm #

    Why are C and D buses through routed Wast Seattle-Ballard and v.v.?
    Because Seattle Transit always routed the 15s and 18s through from Ballard to West Seattle starting in 1940!
    Why did Metro not reform the electric routes in 1980 when they re-did the electric overhead (used by ETBs of which the 15 and 18 were and had been slated to be again)?
    Why does Seattle have almost no cross-town buses??
    Because the General Manager of the Seattle Transit System (the private operator that preceded the public takeover by what is now King County Metro Transit) Lloyd Graber set these routes in Stone (and Webster!) and they must never be changed!

  29. Erik Griswold December 2, 2012 at 9:11 am #

    Here’s a good example of why a lot of transit supporters are wary of BRT:
    Look! It’s a BRT project!!
    Well, actually, it is a widening of US36 plus some queue jumping lanes. At least they were kind enough to add a bikeway.

  30. Carl December 6, 2012 at 4:33 am #

    So many BRT projects are promoted as “just like light rail but cheaper” or “like light rail but more flexible.” The problem is that cheaper comes from not providing dedicated right of way. If a BRT route is going to have dedicated right of way, then the implementation cost and time will be similar to light rail, with the main differences being that BRT has lower capacity due to station dwell times, and higher operator costs, but the flexibility to branch out at the (presumably uncongested) outer ends (vs. transfers to a bus) while having the reduced ride quality of rubber tires on roadsurface.
    No question that RapidRide in Seattle has been a poor implementation that was done primarily to gain federal money and political support by promising something, and like most BRT projects in America it really delivers nothing more than marginally improved bus service. It’s still a bus line, now with routes identified by letters instead of numbers (which sometimes confuses timetable lookups)