The downtown tunnel will close early at 11 pm September 15-17 for planned maintenance. This is only in the evenings and won’t affect daytime service. The last Link trains through the tunnel each direction will be:

  • 10:30 pm to Lynnwood: 2 Line westbound/northbound at Judkins Park.
  • 10:38 pm to Federal Way: 1 Line southbound at Capitol Hill.
  • 10:44 pm to Redmond: 2 Line westbound/northbound at Capitol Hill.

Or looking at the entire train runs:

  • 2 Line westbound/northbound: Downtown Redmond at 9:04, Judkins Park at 9:40, Capitol Hill at 9:53, and on to Lynnwood.
  • 1 Line northbound: Federal Way at 9:45, SODO at 10:30, Capitol Hill at 10:44, and on to Lynnwood.
  • 1 Line southbound: Lynnwood at 10:10, Capitol Hill at 10:38, SODO at 10:51, and on to Federal Way.
  • 2 Line southbound/eastbound: Lynnwood at 10:16, Capitol Hill at 10:44, Judkins Park at 10:57, and on to Redmond.

1 Line shuttle buses will run every 10-15 minutes between Capitol Hill and SODO stations. 2 Line shuttle buses will run every 10-15 minutes between Capitol Hill and Judkins Park stations. An earlier announcement said the shuttle buses would start earlier at 9pm, but Monday’s announcement doesn’t mention this,. so they may or may not start that early.

Check Sound Transit’s service alerts for the latest updates. A multilingual translation like is available at 800-823-9230 for passengers with limited English proficiency.

The reason for this is to extend Link’s after-hours maintenance period this month to make the system more resilient. I’m eager for Link to get more reliable someday so I’m glad for this maintenance.

31 Replies to “Link Evening Closure Tuesday-Thursday”

  1. The first 2 Line shuttle that I see in Google Maps leaves JPS at 10PM while the first 1 Line shuttle leaves SODO at 10:45PM. But I suppose shuttle schedules are to be taken with a grain of salt.

    I wish Metro and ST would run more runs of the 8 and especially the 48 late at night for thru riders. It’s significantly faster than the shuttle. I know that they provide some of this routing information, but you do have to dig and if you aren’t that familiar with the bus routes in the area I can imagine it being hard to navigate.

    1. I wish Sound Transit would run express shuttles between convenient train stations on each end without stopping at intermediate stops. So, from Capitol Hill, run south on Broadway-Boren to Judkins Park and then Mount Baker.

      1. The 48 basically does this already, it just skips Capitol Hill and serves UW. For a through rider, that’s almost as good since 23rd is generally faster to get through than Broadway/Boren. The 48 making all local stops is only 7 minutes slower than the Link from Mount Baker. I think it’d be easy enough to just run extra special trips on the 48 rather than have a local and express shuttle

  2. I keep asking this same question to different people, and I’ve never gotten satisfactory answers. Why does Link require maintenance so much, so often? The local train I used to take to go to school in Kobe, Japan, never stopped running due to maintenance for at least six years, and quite possibly way longer than that. Every time I mention this, I hear “Well Japan is different.” So how different? Do they use better materials? Are their train cars somehow easier on the tracks? Are their pantographs more sturdy?

    2. The question should be … Compared to other light rail systems in the US, where does Link rank in terms of shutting down segments of track for maintenance? Do they shut down more than most other systems, or less? If no one here knows the answer to that question, claiming Link shuts down for maintenance too often is a baseless claim.

    3. My uninformed suspicion is that the maintenance requirements are met by Metro staff even though it’s an ST operation. It doesn’t appear that ST can direct Metro to look at managing maintenance in different ways. Plus, ST hires attorneys and managers rather than people who are experienced in the field — so they don’t have the field knowledge needed to explore approaching maintenance needs differently.

      1. This is not true, ST employs people who are maintenance and engineering experts. Metro does end up dictating the amount of time and ability to actually do the work, though, so blame them for wider windows and worse work.

        STs only options other than Metro ate to shop around for a new private vendor (like keolis or hertzog) or bring it in house if they feel that Metro doesn’t do a satisfactory job. Whatever the case may be in the future, ST doesn’t seem to be able to do much about it right now.

        1. “Metro does end up dictating the amount of time and ability to actually do the work, though, so blame them for wider windows and worse work.”

          That’s the core problem as I see it. ST should not be relying on Metro. ST should have their own crews doing it.

          While ST has some technical resources, these experts are clearly operating with one hand tied behind their back. And the layers of management above them don’t appear to have worked extensively in places that do it differently.

          The fact that the only choices you see are either Metro or the private sector tells me that you don’t see the underlying structural problem. Almost every other light rail system does their major maintenance with their internal staff — but not ST. It’s a structural problem that needs to change or the problem will not dissipate.

        2. I literally said bring it in house after mentioning new private vendors, which I think would be the preferred option (but I did not specify in my comment).

          In any event, you’d be surprised how many systems actually use contractors for maintenance and operations. It’s more than you’d expect (not that it’s necessarily the best practice).

        4. It’s a gigantic leap from “Link has a lot of breakdowns” to “(we don’t know what exact components are failing or why or what ST could do that it isn’t doing because we aren’t train maintenance experts or know the state of all components)” to “the primary problem is the ST/Metro relationship”. We don’t know that. It’s highly unlikely that that relationship is the cause of so many problems or the inability to fix them, or that an internal staff following the same maintenance regimen would do any better, or that Metro can only be half as effective as an internal crew (“one hand tied behind its back”). That sounds like too much speculation.

    4. We don’t know why Link has so many more unplanned outages than peer networks. But increasing maintenance time makes it more likely it will improve reliability, while not increasing maintenance time means it won’t improve.

      ST threw a lot of maintenance into Link last year and early this year, and outages went down from almost every day to a few times a month. It managed to work flawlessly through the first couple World Cup matches, then power failures and signal problems stared returning. The outages are still less than they were earlier this year, so that’s something. But clearly ST needs to put more maintenance into it to bend the curve down more decisively.

    5. All rail systems require extensive inspection and maintenance programs. Trains are heavy machines that move very fast. The infrastructure wears out and has to be rehabbed or replaced on a regular basis. Generally, it’s better to schedule shorter, routine inspections on a regular basis than to wait for things to break and cause a major service interruption. My impression is that ST has allowed its infrastructure to wear out and is now trying to catch up. That’s why we have so many late night closures and unexpected, peak hour service interruptions.

    6. The issue isn’t the amount of planned maintenance, it’s the amount of unplanned disruptions. It’s not normal to have one almost every day or a few times a week. Those BART, DC, and New York examples are decades-old or century-old systems that had no maintenance budget due to 20th-century thinking and opposition to transit-tax increases. Link has “state of good repair” built into the budget so that’s supposed to include maintenance and end-of-life fleet replacement. Yet Link keeps having several kinds of power outages, “signal issues”, and “mechanical issues”. That’s on top of something unauthorized on the tracks, or “police activity” closing a station, or planned maintenance.

      All that adds up, and passengers may not make a precise distinction in their minds between how much of each occurs — the totality is so overwhelming. In any case, when they’re on a train or platform, they don’t know why service is interrupted, just that it is.

      ST probably has a normal amount of planned maintenance. The issue isn’t really prescheduled closures, where ST sends a bunch of announcements and has several station agents at every entrance and shuttle stop. That’s more than any other agency does, and there will always be people who don’t know about it or don’t understand the situation no matter what ST does. The issue is unscheduled disruptions that happen anytime, and happen a lot, and aren’t because Link has had no maintenance budget for decades like the NY or DC subways.

  3. Most subway networks don’t have power outages all the time or endless “signal issues” or “mechanical issues” that Link has. If they did we’d know about it, because enough of us have lived in other cities and we’d experience the outages and hear the complaints.

    1. Gonna be honest NYMTA, BART, WMATA, MART, MBTA and the CTA all have near constant signal and maintenance issues just like link. A cursory search of reddit subs for these systems reveals a plethora of similar ailments about signal issues or switch problems or even derailments in WMATA and the MBTA’s cases. This is not unique to link and it’s silly to see it as such, because you then search for a unique silver bullet that would somehow solve the issue.

      1. Following up on this. In the last month, I found threads for the following.

        BART:
        Service suspended on the green line due to substation fire (aeging infra).

        MBTA:
        Red line delays due to signaling issues. (Many instances across entire system).

        WMATA:
        RedLine track problem with single tracking and delays.
        green line single tracking with delays

        MARTA:
        Track Issues apparently, though headways are so bad nobody can tell (ouch)

        CTA: literally the entire blue line is full of slow zones. Imagine if the 1 line could only go 35mph the entire way to federal way. Plus Bustitution and signal issues.

        These are issues from threads in the last few weeks. The reality of a rail rapid transit system is that it’s down for maintenance or other issues at times, but for link I’ve hardly had it down more than any other system I’ve experienced as of late.

    2. I think a bit more important issue is how many times there must be newly scheduled disruptions for several hours. Simply put, it’s in Metro’s best interest to overestimate disruptions.

      It makes Metro look good for getting the work done early.

      Metro gets paid more hours to run replacement bus service.

      Parallel Metro routes can pick up more riders (and collect more taps) when Link isn’t running.

      My skepticism is rooted in the Columbia City tile replacement project a few years ago. Link service was substantially reduced for four weeks on a project that appeared to take much less. Crews that I observed just worked normal daytime shifts.

      Meanwhile, I see WSDOT routinely do most major projects starting after 10 at night with things cleared by the morning commute.

  4. Is it really Metro that maintains Link trains and the trackway? I thought Metro just operated the trains.

  5. So they originally designed it to be not resilient and now they’re having to fix it with resiliency… typical government project. Unbelievable. And will this really achieve a goal or is it just the standard maintenance being parroted as an improvement? Seems like they are not promising any level of service improvements like when they claim the escalators they are replacing will be more reliable than the meme-inducing broken ones they originally installed.

    1. I can’t imagine ST intentionally installed faulty signal systems or short-lived ventilation fans or unreliable power systems. Something unknown is happening that doesn’t happen at other agencies with such new infrastructure, because ST keeps fixing things and the same things keep breaking again.

      The escalators are a lot better than they were in the 2010s. Don’t you remember when UW station had a different escalator breaking every day, or one in the morning and another in the afternoon. And similar although not as bad issues with escalators and elevators at Capitol Hill, Westlake, and the other tunnel stations. At least now when an escalator is closed, it’s not nearly as often, it reopens much sooner, and you’re more likely to see somebody currently working on it or new treads piled up ready to install.

  6. Have these constant maintenance disruptions made the system more resilient?

    Or is it more a reflection that ST/Metro don’t have good management or experience on how to maintain a system? Or don’t have a culture to prioritize keeping the system open with single tracking and use the already scheduled nightly closures to get the work done?

    For a relatively new modern system, Link has more closures than you’d expect.

  7. Meantime, the Seattle Times prints a headline that reads:

    “Seattle light rail train strikes empty school bus”

    Like the Link train got off its route and hit the bus which was in a proper place. Maybe a more accurate headline would be “Negligent school bus driver backs empty bus onto Link tracks”

    1. The management / contracting out of school bus operations is something we rarely cover here.

      The political cycle of firing the school bus operator every time the school board majority changes makes it a de facto temp job.

      1. First Student has run the school bus program for Seattle School District for as long as I can remember. It did change names and possibly ownership a while back. UK-based transportation company, FirstGroup (who owns Greyhound) is the corporate entity behind First Student. I guess part of the contract was given to Zum a few years ago. Not sure if they are still a player. Laidlaw is the other name you see on school buses but they are also owned by FirstGroup.

        Hiring and retaining school bus drivers is difficult. Seattle School District should be running this in house the way Bellevue and Lake Washington do. Then drivers become eligible for Washington State pension which is a huge plus in retaining drivers.

        1. “FirstGroup (who owns Greyhound)”

          I thought FlixBus bought Greyhound. Does FirstGroup own FlixBus?

  8. Late-night closures are inconvenient, but extending the maintenance window seems worthwhile if it genuinely improves reliability. The 10–15 minute shuttle frequency should at least make the disruption a little easier to manage.

    1. Having 10-15 minute shuttles serving every station is better than having Link end at 9 or 10pm and having half-hourly ST Express skip most of the Seattle stations. That may be where we’re heading toward, depending on what ST decides to do to extend the regular maintenance window. These additional special closures are a step toward a permanent Link reduction strategy, which ST hasn’t decided on yet.

      ST had a public survey asking which strategies were preferable to get more maintenance time. Some of the concepts were Link ending earlier, having more weekend closures, having more single-trackings, or truncating Link in the late evenings. It’s unclear whether these would be every day, a couple days a month, or irregularly.

      Single-tracking forces a frequency reduction to 15-40 minutes depending on where it is and how ST mitigates it). Truncating Link might be OK if it runs in the highest-ridership Seattle segment, if Link serves the area the ST Express night owls bypass. But ST needs to do maintenance in Seattle too, so would it reverse-truncate? Where? We’d need to know those details in order to evaluate the tradeoffs and impacts and determine whether we’d be OK with some tactic.

Comments are closed.