Virginia at 8th Avenue, turning from two-way into one-way. Time for a contraflow transit lane?
Virginia at 8th Avenue, turning from two-way into one-way. Time for a contraflow transit lane?

Amid the general hand-wringing about growth in Seattle lately – be it from Danny Westneat, Crosscut, or innumerable KUOW radio hours – there has been no shortage of discussion about the relative lack of transit service in South Lake Union. A combination of fewer transit options, abundant parking, and an affluent workforce have yielded a drive-alone rate in SLU (46%) that is more than double that of the traditional downtown core (22%). Though no one would argue that transit has kept up with growth, our agencies are working hard to catch up, with many potential projects to address the problem:

Aside from an ST3-funded subway – a line at least 15 years away if all goes well – the good projects above still generally tinker around the margins while continuing to treat SLU as a peripheral neighborhood. But SLU deserves transit service befitting what it has become, which is the northern half of Downtown. That means a lot of peak bus service, at least until 2023.

But if you look at the current peak network operated by Metro, Sound Transit, and Community Transit, you could be forgiven for thinking that the respective agencies still view SLU primarily as layover space for buses. Aside from Route 309, the closest any I-5 buses get to SLU is the view they get from I-5 while slogging towards Stewart Street. From the south, it’s much the same, with all routes petering out in Belltown or Denny Triangle and either deadheading back to base or laying over. From the eastside, the 554’s routing is particularly disappointing, with the last stop on 4th/Lenora in Belltown, from which it then deadheads into SLU to layover. Despite all the growth, the peak network still acts as though Downtown ends at Stewart. And of course, Mercer Street has no transit at all.

A perfect storm is brewing, with massive growth in north Downtown and SLU, Convention Place likely closing a couple years early, ever fewer buses in the tunnel, too few Link vehicles to mitigate lost tunnel capacity, and progressively degraded surface transit pathways. We need more transit, and we need more surface right-of-way (ROW), especially in booming Denny Triangle and SLU. Fortunately, these two neighborhoods have two wide arterials that are not choked with traffic, have a direct connection to the I-5 express lanes, and could have a relatively uncongested pathway into Downtown: Fairview and Virginia.

A proposal:

  • Shift most non-SR 520 peak service away from Stewart/Olive/Howell to Fairview/Virginia, drawn from the following routes:
    • Metro Routes 74, 76, 77, 111, 114, 157, 158, 159, 177, 178, 190, 192, 301, 304, 312, 316, 355
    • Community Transit routes 402, 405, 410, 413, 415, 416, 417, 421, 422, and 425
    • Sound Transit Routes 510, 511, 512, 513, 554, 577, 578, 590, 592, 594, and 595
  • Add two-way bus lanes on Fairview between Mercer and Denny
  • Add two-way bus lanes on Virginia, including a contraflow bus lane between 2nd-8th
  • Add a bus-only turn lane from northbound Fairview to the Mercer on-ramp
  • Add a bus-only turn lane from the Mercer off-ramp to southbound Fairview
  • Make the Mercer off-ramp from the express lanes HOV/Transit only
  • Remove the bus-only lane on Howell Street

Here’s how it could work.

SLU Transit Service Draft-01

Between Mercer and Denny, Fairview is 84′ of ROW, with 56′ available for travel lanes. It is your standard late-century 5-lane arterial, with two travel lanes in each direction plus a turning lane. Converting the outer lanes of Fairview to BAT lanes would be relatively easy, as they are currently used off-peak for parking.  At Mercer, Fairview also has the rarity of enjoying double turn lanes to/from I-5, in both directions (see here and here). Could we take one of them for transit?

Current Fairview Fairview with Bus Lanes

South of Denny Way, as Fairview forks into Boren and Virginia, things get a bit more complicated. From Boren to 8th Avenue, Virginia is a two-way, 66′ minor arterial, with 42′ available for travel lanes. Currently it has two eastbound lanes, one westbound lane, and one westbound parking lane. West of 8th Avenue, Virginia is a one-way, 5-lane eastbound arterial, with 2 parking lanes and 3 travel lanes.

Eastbound on Virginia, a curbside bus lane could shift to the interior lane at 8th Avenue, with no stops between 8th and Denny, allowing buses to proceed straight to Fairview while cars queue in the right lane for I-5 via Boren and Howell (see second image below). Westbound on Virginia, this proposal would simply extend the westbound lane from 8th Avenue to 2nd Avenue as a transit-only contraflow lane (see fourth image)





Points in favor: 

  • Cheap dedicated ROW: though I am not qualified to calculate capital cost, all that is likely needed for this proposal is removing street parking and restriping/resignaling the street. Shifting Stewart’s inbound trolley wire to Virginia could be done as funding allows.
  • Puts most of SLU and Denny Triangle within 1/4 mile of peak express service (see map above)
  • Serves Amazon HQ’s new front door at 7th/Virginia
  • Could be served by any peak route except for SR 520 routes. The Fairview portion could also be used by First Hill routes 63 (proposed), 64 (proposed), and 309.
  • Encourages concurrent improvements to Route 70 and telegraphs the eventual build out of the Roosevelt to Downtown HCT project, providing robust 7-day service for the thousands of new residential units being built on Fairview.
  • Simplifies the network. Buses could use the same two-way pathway regardless of whether their trip uses the express lanes. No more Stewart/Howell/Olive express lane dance.
  • Easily extends the 577/578/590s into SLU while allowing them to keep their Eastlake layover.
  • Neither Fairview nor Virginia are slated for protected bike lanes under the Bicycle Master Plan, making these perfect transit arterials.
  • The pathway is only 1/10th of a mile longer than the current Stewart pathway. Done right, there could be zero time penalty for commuters, and the split pattern of many Metro and CT routes could continue, with half serving SLU first and half (such as the 303, 355, etc) serving Pioneer Square first.
  • Provides an alternative pathway during the highly disruptive Stewart Street construction planned for the Denny Substation project.
Points against:
  • SR 520 routes would need to continue using Stewart, as the 520-to-Mercer weave (5 lanes in 1 mile) isn’t allowed for buses for safety reasons.
  • Many peak routes would either need new layover space or would need to directly enter service from I-5, potentially adding service hours.
  • South Downtown and Pioneer Square commuters from the north may object to an earlier exit from I-5, even if actual travel time is roughly equivalent.
  • Those who currently work in NE Denny Triangle (REI, Met Park, Hill7 etc) would see their walking distances increase by a block or two.
  • Parking garage entrances on Virginia conflict with bus lanes.
  • The parking on Virginia between 8th-9th is reserved for use by the Seattle Police Department, and even if approved by SPD, they would need other parking provided nearby.
The point of this proposal is to think hard about surface transit pathways in greater Downtown, and to recognize that whatever our success in building out our urban and regional rail network, that surface bus transit will remain vitally important and should not be neglected. This proposal attempts to carve out new capacity from surprisingly underutilized streets, and whatever its merits, I hope that our agency planners think creatively about better serving our booming job centers.

38 Replies to “Fairview & Virginia: New Transit Pathways for SLU?”

  1. This is outstanding. I really like this a lot. The advantages will grow and disadvantages will shrink over time as well. For example, this doesn’t serve the 520 buses, but it isn’t clear whether there will be a lot of 520 buses once U-Link opens. Maybe those buses just go to Husky Stadium. Regardless, as you mention, this would be a key part of the Roosevelt to Downtown HCT project. That alone makes this a worthy project.

    Looking at the possible pitfalls, I only see a few:

    * Parking garage entrances on Virginia conflict with bus lanes.

    That could be a problem, but I wonder how common that is, and how much of a problem that is. I assume there are similar situations like that in various parts of town. Are there possible remedies (other than allowing those people to share the lane)?

    * At Mercer, Fairview also has the rarity of enjoying double turn lanes to/from I-5, in both directions. Could we take one of them for transit?

    Maybe. Hard to say. Taking parking lanes is easy. Taking a lane in an area that was part of a huge project (to somehow fix the Mercer mess) might be a little harder. It is definitely worth pursuing. What would it mean if the city says no? Would buses have to merge and then slog through traffic — if so for how long?

    1. Potential parking lot/garage conflicts:

      North side of Virginia at 8th & 3rd as well as the South side at 4th, 5th (possibly not used by the Westin since they have an entrance on Westlake), and a service entrance on 7th. There are similar conflicts on RapidRide C/D on Seneca St between 3rd & 4th that aren’t terribly problematic, even during the rush hour. I’m relatively confident these can be dealt with.

      One thing I’d add: Any parking garage exits that conflict with bus lanes should be metered with priority given to buses whenever present.

  2. How would this work where Fairview meets Denny? Fairview just south of Denny has 6 (!) lanes: 2 lanes NB, 2 lanes SB, and 2 NB turn lanes onto WB Denny. Would 5 lanes (2x BAT, 2x general, 1x turn) be feasible given the traffic in that intersection?

    The intersection is a choke point and has a lot of signal phases. It is also a short block.

  3. Another reason I like this is that it would enable another great bus route — from Boren to Fairview. Similar improvements might have to be done with Boren. Such a bus route would be quite popular right now (connecting First Hill directly with South Lake Union and Eastlake). When the Madison BRT gets finished, that route would be even more popular. I believe the closest we have to that is the 309, which doesn’t travel very often (5 times a day) or go north of Mercer.

    1. As noted in the post, if the ULink restructure is approved, Routes 63 and 64 will serve Fairview->Boren in peak too, providing new service to SLU and First Hill from Lake City, Wedgwood, Ravenna, Green Lake, and Northgate. But yes, it would also set up a nice all-day route in the near future.

      1. Good point. I could see either one of those routes be extended to all day. From what I can tell, neither one uses the express lanes (or has to use the express lanes). But I could also just see a bus route from Eastlake and Fairview to Boren and Yesler. First Hill is an all day destination, and with increasing residences in South Lake Union, it makes sense to connect it that way.

    2. It would be great if there was frequent bus service from MtBaker “TC” to SLU via Rainier and Boren.

  4. I wonder how much South Lake Union bus service has simply been held back until these last few years’ massive construction is done with?

    There’d be a good argument that for the time being, buses are faster where they are now. And the neighborhood is really just getting started.

    For the time being, however, there are some measures that can make current transit faster without much capital expense.

    First, streetcar lanes can reserved through rush hours, and signal pre-empt restored- if it hasn’t been already. Which would also benefit the Route 40.

    Second- and also cheapest, easiest and most important- two streetcar adjustments:

    1. Give streetcars a signaled and enforced path from north terminal to Lake Union Park at pm rush. Traffic across tracks invariably blocks inbound cars up to ten minutes.

    2. At Whole Foods northbound, transit only lane starting two blocks south of Denny. With traffic turning right onto Denny signaled to turn until streetcar was ready to leave.

    Depending on when DSTT becomes rail-only and Convention Place Station closes, might be worth it to let I-5 buses use the CPS yard at rush hour- as was done in 2005 for the last conversion.

    If Tunnel bus fleet is reduced, I still think it would be good to put ST Express Everett and Lynnwood routes in there- like northern 550’s ’til total conversion.

    Also, I wonder if we can work an agreement with the new residents and employers in South Lake Union to adjust parking lot leave times so rush hour isn’t “stuck” hour.

    If we could promise rapidly-improved service in return for clearer streets, younger and more transit-friendly clientele might go for Happier Hour at local establishments.

    We might Work a Deal! without having to put a badger on our heads, destroy the dignity of the news media, and trash the Republican Party. Precedent seems to be working.

    Mark Dublin

  5. I like this. As another example, I commute via the 193, and I think it would make sense for it to have routing like the 309, that serves SLU and First Hill. Right now, it kind of takes a go-down-every-road approach to bring door-to-door service to most hospitals, but because of traffic in first hill, it takes almost a half hour to go from Seattle U to I-5. If people can accept compromise and walk from Boren, this could be a big improvement.

  6. •The parking on Virginia between 8th-9th is reserved for use by the Seattle Police Department, and even if approved by SPD, they would need other parking provided nearby.

    Great reason to remove street parking elsewhere!

    1. There’s little reason to alter westbound Virginia east of 8th. I drive through there frequently to drop my kids off at school and the traffic is quite light, as it is not a convenient way to get to downtown. Adding a bus lane here might cause more problems than it’s worth — besides the SPD and the school dropoff zone there’s the fact that when you get to 8th all the cars must turn right, thus conflicting with a right-side bus lane. Getting rid of a few parking spots near 8th to create a right turn lane separate from a bus-only straight lane could be useful, though.

      Though I do wonder if this plan could just as well be made to work by having inbound buses go one block further to Stewart where there is already a bus lane?

  7. So far you have not called out where the 522 will go. Will it continue as is, or parallel the 312?

  8. With Amazon slated to have almost 70,000 people in or around SLU, is there any hope of them shedding their “don’t know, don’t care” stance and have them lobby the Council for these minor improvements that would greatly help their (and other) employees?

    1. RapidR, very often young people affect indifference as a cover for deep insecurity caused by such things as being trapped all night in an underground parking garage with no strobe lights or ear-splitting music.

      So I think future number you’re quoting is transit’s best hope. As yet, South Lake Union is still more of a construction zone than a city neighborhood.

      If it’s true that South Lake Union’s future population will be the subset of people whose demands are never either futile or met with pepper spray, this wasteful subterranean land use will soon meet a swift end.

      To paraphrase President Lincoln

      ” I believe this neighborhood cannot endure, permanently half traffic jam and half food court.

      I do not expect Amazon’s lease dissolved — or rents to fall into affordability.

      But neither do I expect the battery-powered bowling ball-sized unicycle I saw at Kakao Friday to require parking beyond the owner’s sports bag.”

      Same for the beautiful machined steel scooter a young woman in a beautiful red office suit brought aboard the streetcar same day.

      Remember- the Jetsons only had the original Monorail and the Space Needle for an example. When it was still orange.


  9. While I am generally supportive of modifying many of the peak-only routes to serve SLU, the time sink of traveling from one end of downtown to the other end on a bus is already pretty bad, and I’m very weary of anything that would make it worse. Already, a bus takes nearly half an hour to get between 4th/Jackson and the Olive Way entrance ramp to I-5, and this slog effects not just people headed to downtown, but also everyone who must pass through downtown to get to where they ultimately want to go.

    For peak-only routes, the concerns about going through SLU are mitigated by alternative routes being available that use I-5 exits at 5th/Columbia, Pike St., or the DSTT. But if you reroute the off-peak 512 to SLU, anyone coming from Lynnwood has no competing bypass options, short of Uber or a private car. So, I think sending the 512 to SLU would be a mistake.

    Also, if any non-trivial number of buses are going to take the Mercer St. exit ramp, transit priority down the exit ramp is a must.

    Another item badly needed here is transit signal priority along Eastlake. This is something that should have happened years ago for the 71/72/73, but even when those routes go away, route 70 will still get plenty of use out of it.

    1. I agree with asdf2, this must be contingent on improving the mobility of buses downtown, particularly on the named streets (SW/NE). In the afternoon It is not unusual for these routes to take as much time from Westlake to the freeway ramp as from the freeway ramp to Mountlake Terrace. This is when the problem intersections go from “vehicles per stoplight cycle” to “light cycles per vehicle”. As someone else pointed out, the delays are also unacceptable for much of the morning due to the lack of transit priority at the downtown freeway exits, and the fact that non-aggressive bus drivers always lose out in the weave ahead of Stewart and Denny.

      I also have some issues with the shaded squares (1/4-mile drainages) as a methodology for stop placement. My commute involves a 1/2 mile walk to Belltown and I have always felt that it is the stops near the *corners* of the route which should get the most attention, because they serve a vastly larger area than the linear stops: to first order, this is a “2-D” versus “1-D” thing. So e.g. there should be a stop as close as reasonably practicable to 2nd and Virginia, because the marginal reduction in walk times is much greater at that corner than it would be in the middle of a long straightaway.

    2. As a Snohomish County rider, I must agree with asdf2 too. 4th Ave has become a nightmare as of late for afternoon commuters. Adding an entirely new neighborhood to serve would add to the travel time and inconvenience majority of riders who board between the city/county admin area and Westlake.

      Additionally, the 512 has become crowded in of itself. I dreaded ST combining the 511 & 512 because I knew overloads would occur. Sure enough, they’ve have been occurring and quite increasingly too – making for longer rides. Having the 512 operate through SLU would only add to the frustration of riders who are being passed by full and late buses.

      The positive aspect of this plan, however, is that it does bring riders much closer to the Seattle Center. Currently, riders who venture for a weekend day at the Needle have to transfer to another bus.

      1. I haven’t been on Fourth Avenue in the afternoon peak lately – could you elaborate on what the problem is? Is the bus lane being violated, are the intersections being blocked, are right-turners clogging the lane up, or is something else going on?

      2. I haven’t taken the 511 during the afternoon peak, but I have been on crush-loaded #512 trips on weekends. Typically, it’s a bunch of little things adding up. Outbound, there are a few problematic intersections with long light cycles (e.g. crossing 5th Ave. from Olive Way), but most of the delay is actually dwell time a bus stops to load the passengers. This gets exacerbated when the bus is crowded and the driver has to repeatedly ask people to “move to the back” and when special events attract large number of riders who don’t take transit often enough to justify getting an Orca card.

        Once, when coming home from a Mariners game, I had the good fortune to get a bus with a broken farebox, so the driver was letting everybody on board for free and opening both the front door and back door to speed up the boarding process. The result was about 15 minutes between 4th/Jackson and I-5, rather than the 30+ minutes that would be typical with all the seats and standing area full.

        Inbound, the biggest problem is stoplights and traffic, rather than fare payment. The light at Stewart and Denny can sometimes take several cycles to get through, and the lights along Stewart between Denny and 5th are timed for cars not buses. This means that when you drive (or bike) down Stewart, you make all the lights, but when you stop every other block to drop off passengers, you miss all the lights – even with a dwell time of just 15 seconds or so per stop. 5th Ave. itself is also a big problem during the PM peak, as it lacks bus lanes and is often very congested.

        Meanwhile, if you throw SLU into the mix, you add in yet more stoplights to cross Denny and Mercer. At some point, enough is enough.

      3. Hey William C., I know this is a little late . . .
        4th Ave is becoming more gridlocked with mere traffic. That’s it. Each time I see the line of buses stretching for several blocks along 4h Ave – barely moving – I sometimes wish there’s some blocking incident that’s causing the gridlock. But there’s nothing. It’s simply traffic and it’s getting worse. There have been times where I had to choose the E-line over the 301 because it was quicker to exit downtown (only to see other 301 riders doing the same). Once, buses were idling for so long in traffic that they became standing-room only as more riders piled on at the same stop.

    3. I definitely dreamed of the WSTT today as it took a full 20 minutes to crawl from Seneca to Virginia on a packed E line.

  10. This could be a game changer, however, the bottleneck is the Mercer exit. It’s a MESS. It can take three light cycles to move from the southbound exit to left turn lanes going onto Fairview. I think to fix this would be a massive investment, especially in light of LINK, where many of these busses could meet the LINK at some northern station, taking the busses completely off the freeway. The better improvement for this is to create a really robust SLUT (dedicated ROW, signal prioritization, etc).

  11. What about using the new transit lanes on Westlake? That would bring frequent service into a single corridor, probably better than two corridors separated by a couple blocks. This might also be politically easier too, since it’s far less invasive.

    The only challenge would be how do you get from I-5 to Westlake and avoid the current Mercer mess? What about take busses up the old Fairview -> Valley route? Valley will be pretty unused now that traffic is on Mercer. If you build transit turn lanes / TSP on the Fairview -> Valley left-hand turn and the Valley -> Westlake turn then it probably won’t be too bad.

  12. This is a very good contribution – and not the least is that it points to an important organizational issue: “…whatever [the proposal’s] merits, I hope that our agency planners think creatively about better serving our booming job centers.”
    I doubt Zach meant to cast doubt on individual planners’ creativity, and I surely don’t, but it does beg some questions that might well be put to the agencies:
    1. What do SDOT and Metro (and perhaps ST) think is the best way to improve service to SLU over the next 8 years? More specifically, what measurable improvements in service (jobs, or sq ft of existing and committed office space, in the walkshed of frequent service) and performance (real commute time from representative neighborhoods to SLU) are achievable at reasonable cost, and what would that cost be – net of additional revenue?
    2. Should SDOT’s priorities for Prop 1-funded service and for related infrastructure projects be adjusted to achieve these improvements?
    Such questions will be familiar to those who have done technical planning work for large infrastructure-based businesses. My own experience happened to be in telecom, but I bet there are also some folks who work umpty-ump hours/week on such questions in those big new buildings in SLU. With all of the emphasis on public input and outreach, are we ALSO giving SDOT and Metro planners the opportunity – and the mandate! – to do this sort of optimization?

  13. The biggest issue in SLU is the daily several hour long afternoon backup of cars trying to get on I-5 SB at Mercer. This creates near gridlock everyday in SLU. I-5 NB and SR520 is not the problem.

    I love that we completely rebuilt Mercer and spent ZERO money on transit infrastructure for it.

    Unless you figure out a way to fix the Mercer I-5 SB issue, everything else will just be stuck in traffic.

    Here’s what should be done (but won’t): Expand Express Bus service to/from I-5 South and I-90 to SLU on mercer.

    Charge a vairable “congestion mgmt” toll on all SOV’s getting off and on i-5 at Mercer during commuting hours. The toll point could be on city ROW on the Eastside of the Mercer/Fairview intersection.

    Congestion based tolls at all I-5 on/off ramps and a large expansion of express bus service could solve I-5 and it’s associated Seattle gridlock today if our politicians had the desire to do so. (They don’t and won’t).

  14. the cluster of an intersection at Boren/Fairview/Denny/Virginia is DESPERATELY in need of a complete redesign … could be incorporated into this excellent idea of Zach’s

  15. For those who need to transfer from the D line to the C line (and vice versa): instead of having a one-seat ride between Ballard and West Seattle as they do now– where would such transfer point be?

  16. It’s a complete embarrassment.

    A significant growth area right in the backyard of downtown and after decades and billions spent people still drive to work there!

    All it tells me is there’s absolutely no reason to build anything in or near downtown because the dynamics of are exactly the same as if you had built it in Issaquah. Except if Amazon had headquartered in Issaquah they’d probably be a lot closer to a place where their employees (apparently) want to live and work.

    1. “A significant growth area right in the backyard of downtown and after decades and billions spent people still drive to work there!”

      Compared to downtown Seattle, yes. But compared to almost any other part of the city, no. A 46% drive-alone rate is still low, in the context of the whole Metro area. If Amazon were located in Issaquah, the drive-alone rate would be close to 100%. Microsoft, between the 542, 545, and zillions of privately-funded Connector routes, illustrates the best reasonable case for minimizing the drive-alone rate to a suburban campus. And the drive-alone rate there is still around 67%.

      It’s also very presumptuous to assume that all the Amazon employees live or want to live in Issaquah. Amazon is primarily interested in the younger demographic (most of their employees don’t last more than a year or so) and most of the younger generation wants to live in the city. In fact, many of the people that drive to SLU actually live close enough to work so that the best chance to getting them out of their cars is not transit at all, but feet and bikes. Move Amazon to Issaquah and the 2-mile slog through the neighborhood becomes in addition to a 15-mile slog on the freeway.

      And, even for those that do live in a suburb, if you live in a different suburb from where the job is, the commute becomes truly miserable – the road congestion is often worse than commutes to/from downtown Seattle, and the transit options get much worse, as commuting patterns are too dispersed to run separate 5-minute frequency buses to every employment center.

      For example, once, on an attempt to hike Tiger Mountain after work, I discovered that getting to Issaquah is actually quicker during afternoon rush hour from downtown Seattle than it is from Redmond. On this trip, I took the Connector from Microsoft (non-stop) and others took the 218 from downtown Seattle (almost non-stop). With 405 being the parking lot it always is, the Connector trip took about twice as long as the route 218 trip.

  17. It takes too damn long to get to and from downtown on a bus. I do not understand why we let single-occupancy vehicles get in the way of buses.

    I propose we require every elected official who works in Downtown Seattle to commute by bus. The problem would be solved, one way or another, before a month was out.

  18. Great proposal – this would be a huge improvement to SLU transit service with minimal cost to other service, parking, etc. I am concerned about what would happen with the onramps, but this would be a huge step in the right direction.

  19. I will throw support behind the general principle of bus-and-streetcar-only lanes. This is the only way to get reliable surface transit service — and you have *gobs* of road space to do it in.

    Whether on these specific streets, or on other streets, it *should be done*.

Comments are closed.