Atomic Taco (flickr)
Atomic Taco (flickr)

The Downtown Seattle Transit Tunnel (DSTT) and Link’s Stadium Station will be closed this Saturday, August 9th from approximately 5-11am. The closure will allow Metro and Sound Transit to test the joint U-Link and bus operations that will be in place from 2016-2019.

Link trains will terminate at SODO, and Shuttle Route 97 will operate from SODO to Westlake, serving all tunnel stops. Buses will operate their full routes, albeit on the surface. See Metro’s Alerts page for more details of each route’s surface routing.

U-Link will boost peak frequency from 8 trains per hour to 10 per hour, and those tunnel slots either have to be accommodated via improved operating efficiency or by surfacing buses. Metro obviously has every incentive to keep its buses in the tunnel, as the tunnel saves riders time and saves Metro money. Sound Transit, of course, has every incentive for U-Link to be as fast and reliable as true rapid transit should be. These incentives being at odds, tests like these try to find the operating level where Metro maximizes its bus throughput without damaging U-Link’s reliability.

Long term, of course, the tunnel is slated to be rail-only, but not until 2019.  In the event that joint ops at 2016 service levels bring the tunnel to a standstill, some additional buses will need to be removed. For a prioritization framework about which routes are worth keeping in the tunnel, I wrote on the subject in 2012 when the RFA was ending, and I think the analysis still holds up fairly well.

47 Replies to “Transit Tunnel Closed Saturday Morning”

  1. One train every 6 minutes.

    Seems like at that level, it would be possible to remove all the southern routes from the tunnel and terminate them at Royal Brougham. You have a nice easy cross platform transfer there, and the busway adds capacity.

  2. I actually think that some of the deleted routes from the eastside would also loosen the traffic going through the tunnel…maybe those routes lost will allow for more LINK traffic?

  3. While unrealistic, I hope they have a few thousand volunteers ready to mimic rush hour. With a few in wheelchairs, a few with cash, and a few that try to chat up the drivers on the best route to x. Also have some runners to challenge the busses….heck I would do this. :)

    1. More unrealistic thing is idea that planned test can be at least a trillion years from accurate without for-real simulation of passengers. Wheelchairs, questions, fare collection and all. At least mandatory departure of fareboxes would be apparent beyond all powers of stupidity.

      Mark Dublin

  4. I hope their testing methods account for real life situations that hold up bus traffic on a daily basis:
    Loading wheelchair users
    Customers holding up the line to ask the driver where the bus goes
    Customers fumbling for exact change
    People running for the bus after the doors close, and the driver reopening them
    etc., etc., etc.
    If they don’t then the results will over-estimate the number of buses that can co-exist with the trains in the tunnel, and come 2016 things will bog down at least as much as they do today, if not more.

    1. Two of those can be solved easily. Rear-door boarding will greatly speed up wheelchair loading (so that everyone else can get on the bus), and for heavens sake, close the doors and go — no more runners.

      1. True, but you know that there will be drivers that will still reopen the doors, and there will still be people that want to enter by the front door, even if the rear door is an option, and will wait behind the wheelchair user just to do so. So you have to include these in your test methodology, or reality will sink in real fast in January 2016.

      2. Exactly. Let’s hope that the test criteria is gathered through the simplest method possible — a few Metro folks standing about on the platform during the evening rush one or two evenings a week. Old habits die hard, it’s true.

    2. I hope their testing methods account for real life situations that hold up bus traffic on a daily basis:

      Or maybe just leave the tunnel open and operate as normal? You can’t get any more real life than actual real life.

      1. +1

        Run Link at 6-minute intervals for a week, and run the bus routes Metro hopes to see in the tunnel after 2016 for a week. Glenn, you are a genious!

      2. Nope.

        I got to watch TriMet do this very thing on the Transit Mall, as all bus routes were changed to 3rd & 4th, then install MAX, then put everything back.

        Sure, they started with the computer models. This helped them get the timing of the traffic lights down. They used experience from the transit mall to gauge what appropriate wait times would be.

        If I remember right, the process used was something like this (and this could be mis-remembering what was going on):

        1. Their first tests were midnight to 5 in the morning, so if something blew up there would be fewer witnesses. Remember, everything is on the surface here, so closing off access to stations is impossible. There are hotel entrances, business entrances, and parking structure entrances that come off the transit mall. When TriMet starts a new service, this seems to be the hours of the morning they do it.

        2. The second set of operations happened on Sunday morning, as that is the lowest ridership time during the weekday.

        3. Opening day was a Saturday, but more service was operating than normal due to the expected crowds. In reality, my perception was that “Delays due to the opening day crowds” is something like a fudge factor TriMet uses any time opening day operations doesn’t go as planned, and they have to make adjustments.

        It seems to work OK. The average bus speed in the transit mall is much better now than it was before the reconstruction took place (mostly because MAX displaced a bunch of bus routes so the overcrowding on the mall isn’t anywhere near as bad, but also the shift to stops every 4 blocks from stops every 2 blocks allows for faster throughput), though in my opinion it is quite a bit more difficult to find bus stops on the transit mall than it used to be. Based on the timetables, it seems like MAX runs the length of the transit mall at about the same speed Link runs from one end of the downtown Seattle tunnel to the other. It isn’t great, but it is better than what it replaced.

        Now, if we could only get TriMet to recognize the value of timed transfers like they did during the Cowan administration in the 1980s….

      1. Actually, my (ot) post ended with a call for Metro to buy a new model low-floor trolleybus fleet designed specifically for downtown hillclimbs and to run on 1st Ave instead of streetcar. I’ve argued this point before many times here and elsewhere but Seattle Transit officionados and wannabees censor contrary viewpoint in their minds.

        The 2nd/4th Ave couplet could run limited stop diesel bus service on restricted curb lanes.
        Combined with more frequent east/west trolleybus service, 1st and 3rd Aves could run nearly all trolleybus between Mercer and Jackson. The design is called The Seattle Circulator Plan and it has been submitted to City Hall. Anyone could consider its intricate design, but the Seattle way is PR manipulation and mind control.

      2. Metro already runs “limited stop diesel bus service on restricted curb lanes” on Second and Fourth Avenues, at least at the peak hours when the reservations matter most. First Avenue is completely unworkable for buses; that’s why Metro removed them fifteen years ago. They didn’t do it because they hate on the Pike Place Market and Pioneer Square. They did it because their buses were getting lost in a sea of cars.

        And bus reservations won’t really help.

        Because buses have right-hand doors priority lanes for them must be right-hand running and therefore subject to all the slings and arrows of BAT lanes, horribly at Stewart, Pike, Spring, Marion and Yesler and to a lesser degree at University, Madison, and Columbia.

        The streetcar reservation will be center-running with no blocking traffic.

      3. Anandakos, several bus lines run on 1st Ave, and as I understand it, the Left-Lane Center Station arrangement proposed will run streetcar lines in traffic. Several related safety concerns should be addressed. 1), Patrons are likely to dart between curb bus stops and median streetcar stations. 2), Rail on 1st Ave will pose a skidding hazard for traffic, especially downhill, also braking in both directions. 3), Traffic following the streetcar (in the Left Lane) will regularly pull into the curb lane to pass on its right.

        Keeping all transit stops on the curb lane is presumably safer, thus my preference for the 4th/5th Aves Couplet Streetcar line. 1st Ave median streetcar stations will be complicated to locate and manage traffic. The 1st Ave streetcar route as proposed is likely to result in poor performance and a high accident rate.

      4. The First Avenue, Center City Streetcar will run in right of way shared with buses, but not regular traffic. From the City Council Resolution 31526:

        WHEREAS, SDOT analyzed streetcar performance and impacts of a streetcar operating in exclusive transit lanes with transit signal priority or in mixed traffic; and
        WHEREAS, SDOT recommends that the streetcar operate in exclusive transit lanes (that may be shared with buses) with transit signal priority throughout the Center City Connector alignment since the analysis shows that the streetcar has faster travel times, lower operating costs, and higher fare revenue and ridership with exclusive lanes;

  5. All this cost and disruption could be avoided if Metro just moved its buses to the surface in 2016 instead of trying to deny the inevitable. ALL buses will be leaving the tunnel eventually. Metro just needs to get on with it and stop trying to hold onto the past.

    1. But that’s a huge waste of tunnel capacity with a train coming every 6 minutes and only going to the SE corner of the UW. As nice as it would be to have Link be reliable for every Link user, it would drastically impact all other transit users in the region as the surface would become more congested with both buses and cars. As it stands, downtown is a mess on the surface for all users and there isn’t anything that can be done to improve it. Advocating to put more transit up there is crazy until we have a viable alternative. ULink is not that. Northgate and East Link are.

      If anything, Metro should be allowed to keep its buses in the tunnel until at least 2021 when Northgate is open. Then, we can finally alleviate some of the pressure on the surface streets by truncating routes at Northgate and not having them come into downtown.

      1. Maybe forcing all the busses out of the tunnel will in turn force the city to make some real adjustments to the street usage including exclusive bus lanes.

      2. I agree, Mike. If this round included the U-District, not just Husky Stadium, then I would agree with lazurus. But this round of changes only go as far as Pacific and 513. There are plenty of people there (school, hospital and the occasional sporting event) but from a bus truncation standpoint, it is terrible. There is no way that a 71, 72, 73, or 74 should be rerouted to serve that area, and then turnaround. It would take too long. You are better off with the buses doing what they do now — head to the nearest freeway on-ramp (it is closer).

        This is an awkward time, and I’m sure everyone and Sound Transit wished that we had light rail just a bit further, but we won’t for several years. So, for those years, we make do as we always have, and muddle along. As luck would have it, the next round of changes will not only include the U-District (which means that the 7X buses can be truncated) but Northgate as well. This means that the 41 (the third most popular bus route) can be truncated along with all of the UW buses. That will be a happy day, for all concerned.

      3. It will be interested to see what percentage of the 71/72/73 riders continue to ride the bus downtown after 2016 vs. divert to the train. My guess is that most riders headed from campus and points north of campus (especially along the 71 and 72 corridors) will switch over to the train, in spite of the deviation through campus, away from the station.

        While I think the enough riders will remain on the 71/72/73 to justify the continued existence (mostly residents of the U-district, west of campus), there are still some good opportunities to truncate some peak-only routes between downtown and northeast Seattle, while providing at least a peak-hour connection to the train that is fast and without deviations. For instance, I would strongly favor truncating the 74 (regular stops east of 25th, then along 25th through the U-village, then express down Montlake to the station). Besides allowing for drastic frequency improvements, 74 could operate via Sand Point Way in the reverse direction, service commuters to Children’s Hospital. Similarly, sending the #76 south along 25th Ave. to the Link station, in exchange for more trips each day, could also be a possibility.

      4. I think the 7X buses serve a lot of purposes, including:

        1) North end neighborhood to the UW.
        2) North end neighborhood to downtown.
        3) UW to downtown.

        Obviously the first group of riders isn’t changing. I don’t think the second group is likely to transfer to the train, which leaves only the third group. There is a substantial distance between the train station and the bus stops. I have no doubt that a lot of riders right now travel west to take a bus, simply because the alternatives are so much slower. Now the reverse might happen. But I think it really depends on where the rider is. From Red Square, for example, the bus stop on The Ave is much closer then the train station. But there are plenty of counter examples as well (such as the hospital, or different parts of campus). All in all, I would expect less use of the 7X buses, but not a huge decrease.

        As for truncating the 71, 72 or 73, I go back to what I said about traffic. If the buses can navigate to that area easily, then I’m all for it. But I don’t think they can. As a result, you would simply aggravate the bulk of the riders (those headed to downtown). I just don’t see it happening at all (I think these buses remain the same until North Link is built).

        I agree, the 74 seems like one of the few candidates for truncation. If it doesn’t go downtown, then it is the same thing as the 30, which runs all day. If it operates along Sand Point Way, then it becomes very similar to the 75. I could see these being combined, but that means a loss of coverage along 55th. I think I would keep that section, but reroute the 74/30 along 25th to Montlake (thus skipping the section by Ravenna).

        This begs the question, though — should the bus go through campus or on Montlake to the train station? For the next few years (while U-Link exists and North Link is being built) I would argue for running it along Montlake. That would have the added bonus of providing a fast (and hopefully frequent) connection between UW Hospital and Children’s Hospital. Once North Link is built, I would reroute it through campus, since it would be very easy for it to serve the Brooklyn station that way. It could also do both, like the 25, but I think that would make it very slow.

        Generally speaking, though, this is very small potatoes, from a reroute perspective. The 74 runs 8 times a day, so this truncation is minor. The reroute of the 74/30 to avoid Ravenna and the area north of campus would be a significant change. It might also prove to be unpopular. My guess is that Metro won’t touch it, because they don’t want to stir up a controversy, especially when it would make sense to change the routes again in a few years.

        In any event, none of this will effect the tunnel very much, since the 75 doesn’t go there, and the 74 only makes a handful of runs through the tunnel.

      5. Mike,

        During the DSTT retrofit ALL tunnel buses operated on the surface and NONE of the dire consequences you mention occurred. In fact, you had to look real hard at the data to see the effect (and that was before the restructuring that has already occured, and before the first round of cuts we expect this fall).

        Basically there is plenty of room on the surface for more buses, and we have the data that proves it.

      6. When considering how residents of Northeast Seattle are going to get downtown, you can’t forget the impact of Metro’s planned restructuring for 2015 – the 71/72/73 will be consolidated into one route and everyone in Northeast Seattle is going to be forced to transfer to reach downtown anyway, whenever their peak expresses aren’t running – it’s just a question of whether they transfer to a bus or the train. A quirk of the current bus routing gives the bus transfer less walking. But the train transfer would be much more reliable, and would save considerable time most of the day, even with the extra walk.

        I’m personally of the opinion that buses should stay on Montlake, especially during off-peak hours when traffic isn’t really that bad, and focus on connecting people to the station, while running as efficiently as possible. Especially since 20-year-old college students are more than capable of hoofing it up the hill and those that drive to campus and parking in the stadium lot have to do so anyway. That said, given how Metro considers the UW very important, with northeast Seattle, a mere afterthought, I’m not holding my breath for Metro to change anything whatsoever.

        Fortunately, there are lots of good biking options to the station, including the Burke-Gilman Trail, so nobody has to rely on Metro that doesn’t want to.

    1. This is a reasonable suggestion, since the east side doesn’t gain anything directly from the next two rounds of light rail. This is different than the biggest user of the tunnel (7X) and third biggest bus user (the 41). But the 550, while substantial, only represents a minority of the bus service, and well under a quarter, if my math is correct. East side service in general through the tunnel is large, but not dominant. It also splits from buses coming from the north. Those buses, unlike the 550, would benefit from light rail to the U-District. So, basically, like some of the buses coming from the south, it is simply a matter of trading off bus service with rail service.

      My guess is that at some point, it just doesn’t make sense for the train to interact with the buses. It is easier to manage one type of vehicle, with one type of payment system, one type of boarding, etc. I think that point is when North Link is complete.

  6. Why does there need to be so much planning for simply removing buses from the tunnel? Oh wait, you mean they’re not doing the smart thing, and are going to continue with this festival of unreliable joint operations?

    1. Careful, Kyle. Some few of us have enough seniority to have helped plan joint operations into the DSTT and a good enough memory to remember original design- and therefore know that present festival of unreliable operations stems from a 20 years long World’s Fair of laziness, cheap-outs, and lack of imagination at highest levels.

      Which can be canceled and the elephant crap cleaned up in very short order if anybody at said levels really feels like it.

      There’s also nothing smart about putting buses on the street when there are at least a couple of billion dollars worth of underground right of way to carry them. Let’s have somebody in accounting tell us cost of every lost minute of bus operations.

      Also, Zach “These incentives being at odds” assumes something incorrect. As one day of operations incorporating the communications, coordination, and training as originally intended would prove.

      Mark Dublin

  7. Is there capacity to run the 218 up the hill from 2nd to 5th in the afternoon, or along 5th for that matter? That stretch seems to constantly back up.

    Once ULInk opens they’ll be able to run longer trains. Do they really need 6 minute headways? Politically, I suspect that keeping at least the suburban tunnel buses in place at least until it cannot become a campaign issue is essential to the passage of ST3.

    With regard to moving all or none of the Eastside buses upstairs, the choices are generally pretty clear. The 554 and 550 serve mostly disparate destinations. To the extent they overlap, outside the peaks, it’s rarely worthwhile walking further to catch the 554, the added frequency and better reliability generally counterweigh any advantage you could achieve. In the peak, only the 212, 550, 214 and 218/219 run often enough to be worth trying for. Once Metro dropped the Eastgate stops from outbound 218s, the choice between the buses can be made more or less based on destination. The only real exception is Ranier Ave, for which whatever you choose will probably be just fine.

  8. I’ve never understood why there needs to be so much separation between the buses and trains anyway. I’ve been in subways around the world and there’s trains travelling much faster then Link in and out of a station every 2-3 minutes. By that, I mean that a train leaves and another one comes into the station 3 minutes later.

    Surely it can’t be that hard to implement here when its been done all over the world.

    1. Or why, once it gets into the platform the train takes over two minutes to clear it again. sub-two minute headways are fairly easily achievable if you have level boarding (or are willing to screw over wheelchair users) and off board payment, but you have to hold dwell times under a minute.

    2. Have seen trains of exact same length and equipment caliber as LINK follow equivalent buses into standard streetcar stops on “Tverbanna” , Swedish for light rail, in Stockholm. While negotiating street running past parked and parking cars. Fact that First World countries can easily handle this safely is proof that we need to get ourselves out of World Number Three.


      1. The only thing that i can think of that would cause over the top safety conscious regulation is the fear of lawsuits. Seriously, pretty much anything that happens all over the rest of the world without any issue whatsoever is quashed here or altered beyond any form or recognition due to the sue happy american public and lawyers looking to make a buck.

    3. If I understand it correctly, the reason that buses and trains can’t be in the tunnel at the same time is because of the FRA

  9. Yes, there’ll be lots of problems between 2019 and 2023. But when fully built out, Link will replace the 41, 71/2/3/4, 212/4/5/8/9, 510/1/2, and 550. Those’re 16 all-day round trips an hour, even ignoring the peak expresses and any possible SR520 or SR522 truncations.

    Another bus-only street would be nice to have, but not necessary on a permanent basis.

    1. So we build a temporary one. If nothing else, I believe that locals have shown enough resilience and willingness to adapt to handle an upheaval every few years. Eventually we’ll reach the ideal state, but each new segment brings some healthy excitement to the game.

  10. I think William addressed your concerns quite well below (he may have meant to reply to your comments). Looking at the previous post by Zach ( I think I would say that the light rail is going to replace plenty of routes. As I see it, the routes are like so:

    7X and 41 — These represent the lion’s share of the bus traffic, and they are replaced (quite nicely) by light rail.
    255 — This could be chopped at Montlake or the U-District.
    101, 106, 102, 150 — Arguably should have been kicked out of the tunnel years ago (since Link has been servicing the tail end of the route for years).
    301 — Can serve Northgate instead.
    550, 218 — The only substantial bus traffic that will miss the tunnel. But, all things considered, I’m afraid that is just the price we pay so that the trains will run a lot more smoothly. I don’t know what the east side buses will do once they are kicked out of the tunnel. My guess is they will turn around at the south end of downtown (SoDo or thereabouts) which is just fine by me. It might even mean more frequent service (if the bus can turn around quickly enough). That does mean a transfer for someone trying to get to the north end of downtown, but that seems like a small price to pay. Besides, the light rail system is supposed to provide exactly that type of service (quick rail from the south to the north). Let’s hope it can provide it really well once it has the tunnel all to itself.

    1. 7X goes away in the first round of cuts though. I assume the tunnel portion gets replaced by something else?

    2. The 7X doesn’t go in the tunnel; I assume Ross’s talking about the 71/2/3/4.

      (And yes, I did mean my below comment to be a response to Scott.)

    3. Yeah, sorry for the confusion Glenn, William is right — I did mean the 71/72/73/74. I should probably write it as 7*, but that is confusing, too. I think I’ll just avoid abbreviations.

    4. Ross,

      255 is Kirkland’s trunk; the town has service to the U on the 540. The 255 should continue to downtown. It will be there most of the time sooner than Link would be with the Montlake Mess®, accessing the track level, and Linking downtown.

  11. Pingback: A 6/10 Compromise?

Comments are closed.