Gateway to SODO
Dave Honan/Flickr

Back during my first round of gripes about rider-unfriendly choices for the Connect2020 construction delays, I suggested that Sound Transit might have run trains more frequently outside the downtown transit tunnel. At the time, ST said that this would likely result in significant train bunching. After further discussions, they appear to have backed off this objection in favor of other ones, which readers can judge for themselves.

But first, let’s show how the right plan would minimize train bunching. While there is (infamously) no fixed schedule, the current operational model for Central Link trains is something like this:

0 Train cleared to depart Northbound from Sodo
2 Serves Stadium NB (NB platform)
4 Serves Chinatown NB (NB platform)
6 Arrives at Pioneer Square
8 (est) Leaves Pioneer Square
10 Serves Chinatown SB (NB platform)
12 Serves Stadium SB (NB platform)
14 Arrives Sodo SB, next train cleared to go north from Sodo.

Imagine a second pair of trains, A and B, interleaving as follows. At the start of the cycle, A is loitering just south of Stadium on the SB track, and B is on its way up from Angle Lake.

0 Train cleared to depart Northbound from Sodo
2 Serves Stadium NB (NB platform)
4 Serves Chinatown NB (NB platform)
5A departs for Sodo
6 Arrives at Pioneer Square
7A serves Sodo SB, B serves Sodo NB
8 (est) Leaves Pioneer Square
9B arrives at the loitering spot
10 Serves Chinatown SB (NB platform)
12 Serves Stadium SB (NB platform)
14 Arrives Sodo SB, next train cleared to go north from Sodo.

Note that in both directions, this offers headways of about 7 minutes at Sodo and below. In priniciple, one could park trains at the unused track at Stadium and thus extend service to that station; however, ST is using this opportunity to do some repair work at Stadium as well.

I believe a similar pattern would work at Westlake.

One downside is that the turning trains will have to drop everyone at Sodo. While tere is lots of bus service there, ST points out that emptying the train can “introduce rider safety and crowding issues” for those awaiting the next one. Longtime riders recall that Sodo is a traditional dropoff point for trains experiencing a maintenance issue and destined for base. There is risk in making this a routine part of operations, rather than something for urgent situations.

Furthermore, there would be some tough decisions on the fleet. There are probably not enough rail cars to uniformly run 4-car Pioneer Square trains and even two-car Sodo or Westlake trains in the peak. As crowding is not as severe between UW and Westlake, it would make sense for some or all of the Pioneer Square trains in this segment to have 3 cars.

Although not as significant a problem as the lack of real-time information, more frequent trains would have solved a few problems. Westlake is a huge destination in its own right. Moreover, both Westlake and Sodo provide nearly infinite bus transfers at peak to travel further downtown. It’s easy to call for smaller safety margins from the sidelines. And yet, much like the Connect2020 bike restrictions, this seems like an overabundance of caution, one that has the disastrous effect of halving frequency on important segments of the line.

19 Replies to “How Connect2020 trains might have been more frequent”

  1. Good gawd. It hasn’t been the end of the world and it is almost over. Give it up.

    It will be all over in a couple of weeks.

    1. You are correct. An agency that cares very little about rider experience is, indeed, not the the end of the world. Perhaps you should spend your time on a blog focused on world-ending emergencies.

      1. that’s sort of hilarious. I think the point is that they’re damned if they do, and damned if they don’t. If they had attempted more frequent service outside the tunnel, and the trains had bunched and people had been left on platforms as a result, they would’ve come in for harsh criticism about their rider unfriendly construction mitigation program.

      2. Which is to say, this is a temporary situation that was bound to cause discomfort of various sorts, they managed it cleverly with the center platform in PSS, it’s nearly over, and the payoff will be a one-seat connection to/from the east side. More good than bad here.

      3. They are coming under harsh criticism about their rider-unfriendly construction mitigation program. Would train bunching have been worse? Maybe, maybe not. The greater frequency would have been a benefit. Right now it’s making me reluctant to take Link at UW Station because I never know if there will be a 15-20 minute wait. That makes it not as competitive with the 47, 70, and 48 as it usually is.

    2. If drawing attention to this issue keeps ST from making the same mistakes while building ST3 (e.g. not planning for likely expansions) then we frankly are not talking about this enough!

      1. Wasn’t Link running 25 minute headways this week with the single tracking near beacon hill? Glad I work from home and don’t need to commute.

    3. I think people see Connect2020 differently, depending on whether it affects them or not. People who seldom take Link don’t understand why riders are complaining about a disruption that is temporary. They liken it to a road being closed for a few months. It’s annoying, but only temporary. Regular Link riders aren’t arguing the disruption was unnecessary, but think it should have been handled differently.

      Sam aka Captain Obvious

      1. We’re complaining that ST doesn’t take simple steps to make the mitigation better, and it doesn’t design the system with expandability and resiliance (redundancy) in mind. That would be like WSDOT not future-proofing a highway interchange, or closing it unnecessarily.

  2. “ST said that this would likely result in significant train bunching. After further discussions, they appear to have backed off this objection”

    Every single thing ST says should be taken with a grain of salt. They have a penchant for making up excuses, maybe we should figure out why this is? Are they lazy or incompetent?

  3. I think this is actually a fairly decent plan for non-peak hours, and really bad one for peak hours.

    In my experience riding the Link into downtown from South Seattle, probably 80+% of passengers boarding in South Seattle or further south have a destination of ID/Chinatown or northward. What’s going to happen during rush hour is the packed 2 car trains will entirely empty out onto the Sodo platform, and wait for the next (hopefully 3 car?) train to cram into. Not a great rider experience either; in fact I think I prefer the current one!

    Honestly if they could have just solved the real time information problem it would have improved things 3x for me.

  4. A few issues I see with this otherwise clever proposal.

    1. Having 3-car trains meet 4-car trains would asymmetrically lengthen the transfer time at Pioneer Square due to crowding/queuing in the doubled-up car, especially during peak. Even 30 seconds can snowball and throw things off when things are so tightly choreographed.

    2. Simplicity and elegance are rider experience principles too, not just frequency. There would be a non-trivial hit to rider experience from having 4 types of trains at single platforms at SODO, Stadium, and Westlake. It’s valid to design your system for the low-information rider, the occasional rider, the new rider, or the rider who doesn’t speak English. Minimizing the possibilities for stranding or involuntary 3-seat rides is a worthwhile thing, as is not having to even further saturate riders with annoying messages about which train is continuing, which one is terminating and turning around, etc.

    3. I’m not sure there would be enough operators to do this. Right now they have operators at both ends of the train at some stations, which keeps the transfer time low as they don’t to walk the length of the platform.

    4. Your instinct about fleet issues is probably right. If my math is right, they’re running basically traditional off-peak capacity all the time right now, just in a 4.5 x 4-car configuration instead of the usual 6 x 3-car. If usual peak is (6 x 3-car + 4 x 2-car) then you only have two 4-car sets to play with, and they’re already keeping one set in reserve at the Stadium pocket track (and it seems like they’ve used it quite a bit as other issues have come up?). So in your proposal trains would be nearly as crowded, with the added layer of potentially a majority of folks changing trains either once or twice(!).

    5. The SODO dump off would be a pretty awful rider experience. During peak, SODO buses are at their fullest and there are no buses available to add capacity. SODO transfers can work in weekend closures because there are two 60-footers that *board empty* at SODO specifically for that purpose, running more often than the trains. It’s arguably a worse customer experience if riders (with luggage? bikes?) tried to board a 590 or similar bus but were turned away only to return to the SODO platform to get on the same train they would have gotten on if they’d waited 7 more minutes?

    6. The current frequency sucks since we’re accustomed to 6 minute peak/10 minute otherwise. But our current apocalypse (4-car trains every 14 minutes, mostly grade separated) is roughly equivalent to having trains slightly more than twice as good as any given Portland MAX line (2-car trains every 15 minutes, mostly at-grade). Perspective.

    1. You’re ignoring the fact that the trunk route in Portland between Beaverton and Gateway has 2 lines (blue and red) and Rose Quarter to Gateway has 3 (also green line). For my stop, that means I don’t need to look at a schedule if I want to go downtown (5min headways or better). Added to that is the fact that the blue line headways improve to better than 15 during rush hours, especially in the afternoon.

  5. Even though Connect 2020 is almost over, I think operational alternatives like this are needed like this. It’s called “contingency planning”.

    One of my biggest concerns is that service disruptions in the DSTT will happen. With Link normally at 6 minutes in each direction with three-car trains, one solution works. When it goes to four-car trains with Northgate next year (!), or 4 minutes in 2023 and then 3 minutes in 2030 (West Seattle / Tacoma), problems will evolve quickly — in just a matter of minutes.

    These plans need to address how to both operate trains and how to get riders in and out of stations. Questions like “if full trains stop at this station every 3/4/6 minutes and have to let everyone off, what will the train then do? Can the station vacate all of those riders on elevators, escalators and stairs before the next trains arrive? What will the exiting riders do to continue their trip?” need to be quantitatively answered. Then, these plans need to be in a reference manual for staff to use, and any major problems need to be fixed before we start flooding DSTT station platforms with likely 300 to 500 percent more riders as soon as in the next five years.

    Connect 2020 is a one-time event but periodic disruptions are inevitable. ST needs to be prepared — and key staff should have no valid excuse (that is get fired or at least denied a raise) to give the Board if there isn’t one.

    1. Once the second downtown tunnel is built, for contingency’s sake it would be nice if trains had the ability to take either tunnel from either direction. I don’t think that’s in the plans though is it?

      1. It would be pretty foolish to configure tracks in a way to not allow train switching! Of course, nowhere in the SODO configurations In the early planning have I seen any suggested switching tracks. Given ST’s track record of “we didn’t think about that when we made a decision”, I don’t hold much hope.

        If the four SODO tracks were instead designed with tracks heading in the same direction (northbound or southbound) next to each other rather than the currently proposed approach (West Seattle tracks paired and Beacon Hill tracks paired), it would be a piece of cake. ST has suggested that it’s too expensive to have cross-platforms between lines at SODO but that’s being myopic (or lazy engineering BS) because building switching tracks in the current configurations is going to be more expensive than simply moving one Southbound track to Beacon Hill to connect in a different place.

        Even without switching, trains could be stopped at Westlake and riders could go to the other tunnel platforms and trains. It’s one more reason that Westlake internal circulation deserves close quantitative scrutiny of its escalators, elevators and stairs when the new line/ platform opens.

Comments are closed.