Yes, when it’s busy and the escalators are so long most don’t walk.

55 Replies to “Sunday Open Thread: Stand on Both Sides of the Escalator”

  1. I really hope there’s a STB story on Round II of the ST3 hearings. TVW link: . Flickr photos the blog can use in your write-ups, so I put them in your Flickr pool:

    I have to say the last half hour of the hearings with David Habewhiz (SP?!?!), Maggie Firmia and the Eastside Transportation Association really make me want to revoke my support of the hearings. I would think having a Lyft driver complain about tabs and a transit rider praise Sound Transit and a civil libertarian talk about ORCAleak would have been more productive. Don’t get me started about Timmy Eyman…

    That sour grapes given, I’m happy something close to the truth dribbled out in the rest of the hearings. The incompetence of e-mail management and the Mass Transit Now campaign manager leaving TCC senior staff to twist in the heat really speak as loudly as an EA-18G Growler at OLF Coupeville…

    I await the blog post.

  2. Two quick questions:

    – Why are there no ST express buses at night? There is 0 way to get to the eastside from UW after ~10pm

    – Why didn’t ULink take the I-5 express lanes like ST3 is taking the I-90 express lanes. Seems like an awfully huge waste to dig a brand new tunnel when we could have just used some existing ROW no?

    1. Think about it, Andrew. Capitol Hill. UW. Brooklyn. Roosevelt. We’re not talking about the bullet-train from Canada.

      LINK is supposed to be the fastest ride possible to serve a line of neighborhoods full of passengers. Very different from same map-space both sides of I-90 between the shores of Lake Washington.

      Express buses, on the other hand…excellent question.

      Mark Dublin

      1. I’m not sure what you mean. All those neighborhoods lie on the I-5 corridor, no?

        At least this opens the possibility for an I-5 express rail.

      2. North of Roosevelt, sure, but by that time you’re getting close to the end of the express lanes. South of there, I-5 runs along the edge of each of those neighborhoods. Have you ever tried walking from there to the center of Capitol Hill? Or even the UW? It’s a significant way.

        As you say, I-5 is more suitable for long-distance express trains.

      3. Well, I-5 tends to skirt busy neighborhoods (for good reason, no one wants the center of their neighborhood to be a highway, and no one makes a neighborhood center at a highway). Walking from I-5 to UW is already pretty bad from the 70 bus stop there. I-5 next to Cap Hill is several stories down. And so on and so forth.

        Highway rights of way are only good if you expect your ridership at that station to mostly be bus transfers or P&R, because no one wants to walk to a highway stop.

      4. >> I’m not sure what you mean. All those neighborhoods lie on the I-5 corridor, no?


        I understand your confusion. ST has emphasized “stations” as essentially serving a “neighborhood”, or even a city. West Seattle, Ballard, Tacoma, Everett. These are all huge areas, and a single subway stop can’t possibly serve all of the people from those areas.

        To judge a stop, you really have to look at the specific cross streets. With that in mind, assuming you could run alongside I-5 and save a few bucks (and it is no means clear that you could) just where exactly would you put the stations? I just don’t see anything that is even close to being as effective as what ST did. Stations close to the freeway usually perform very poorly. That is because it is difficult to get to them. The freeway itself takes up a significant amount of the available space, which means that you end up with very few people that can actually walk to the station. This means that you are largely dependent on bus connections and park and rides. Park and rides don’t scale, make no sense in the middle of the city, and almost always perform worse than urban stops. Bus integration then becomes the only way you can make a freeway based stop worth the money. In the case of the area you mentioned, the choices for that are just not very good, because there aren’t many cross streets (and those that exist are often congested).

        Running next to the freeway has other disadvantages. One is the difference in speed versus driving. Capitol Hill Station is a great example. It takes a while to get there, making a subway trip faster even at noon. But with freeway stations, the time savings for transit only exist during rush hour. So not only does a freeway station have far fewer places nearby, but rarely is it faster to drive to the station. This means that for a typical end to end trip (to a popular destination) it is almost always much faster to drive. This is not a good system to build, as you rarely get your money’s worth out of it. At best you are spending a huge amount of money on what is then essentially a commuter rail line. No commuter rail line carries more people than the transit system serving the central core.

        If you make cheap compromises, you often fail miserably to attract riders, because those compromises make the system difficult for too many riders. Even the station at Mount Baker struggles because they didn’t think of the little things ( On paper, that should be one of the most popular stations in our system (the 7 is really popular, and this is where the two major south end transit lines cross). But the station isn’t very popular, because they didn’t get the details right.

        That is a long way of saying that if we had built the stations next to the freeway, they would suck. CHS and UW are the second and third most popular stations (trailing only the Westlake station downtown). If they were closer to the freeway, those numbers would be much lower, as they simply wouldn’t work for as many trips.

      5. RossB okay sure forget UW/Capitol Hill Station then. What about UDistrict Station? It’s like a thousand ft from the freeway. Roosevelt Station is around 500 ft. Northgate Station is pretty much on the freeway. Seems like a real waste not to have used that existing ROW.

      6. You have to look at it one by one:

        U-District — It has to come out of the ground somewhere, and work its way back to the freeway, under your scenario. You might as well go pretty much where you are headed, then emerge from the ground somewhere between the U-District and Roosevelt (close to the freeway).

        Roosevelt — That was the original plan, if I remember right. They were going to use the freeway right of way there, and put the station next to the park and ride (under the freeway). But folks from Roosevelt basically begged to move the station closer to them. It ended up costing a few hundred million, and the old commentator d.p., thought that was a waste. Mainly because you don’t gain that much. You have better service for the Roosevelt neighborhood, but worse for folks close to Green Lake. I don’t feel that strongly about it, although I think that is a reasonable argument to make, and you aren’t the first to make it (or at least that variation of it).

        Northgate — Since the tunnel went to Roosevelt, it made sense to keep digging, and emerge on the side of Maple Leaf. It does that. At that point, it is more or less using the freeway right of way.

        So basically, the only variation that could have used the freeway more would be to start before Roosevelt, which was the original plan. But like I said, the folks from the neighborhood pushed hard to have the station where it is now, and since ridership might actually be higher, that is what they did.

      7. The original plan was for Link to come out Convention Place directly to the express lanes like the peak buses do with a freeway stop at 45th, or maybe to next to the freeway near Eastlake. That raises the specter of a U-District station where the 512 stops are and a Capitol Hill station way down at the bottom of a steep hill that nobody would go to. Urbanists pushed really hard to get it rerouted to Broadway and University Way and succeeded. Having a train one or two blocks from your destination is excellent, and having it five flat blocks is pretty good. Brooklyn Ave puts it right in the middle between the freeway (six blocks west) and all the UW classrooms and north campus dorms (six blocks east). If you’ve ever used the 45th freeway stops going from say the U bookstore or a classroom, you’d much rather have the station right at University Way or Brooklyn rather than walking over to the freeway at the edge of the neighborhood and then have to stand next to the freeway. I think of London’s Leicester Square station, which is just around the corner from bookshops and Soho and a lot of other destinations: this is what a station at 43rd & Brooklyn or Broadway & John is like. A station at the edge of the neighborhood is much less useful, gets less ridership, and is less of a compelling alternative to driving.

        Roosevelt Station was originally going to be around 65th & I-5. That would have been in the middle between two smaller neighborhoods, Roosevelt and Greenlake, a 5-block walk from either of them. Neither of them is as bustling or has as many jobs as the U-District, so there’s less justification for moving it east to Roosevelt Way, especially since that puts it further from Greenlake. But Roosevelt has a more precise neighborhood identity and political clout, so it managed to do it anyway. And from my years of experience going to Roosevelt, it punches above its weight in terms of unique things to go to. (stereo shops, eclectic shops, the first Whole Foods in Seattle, the Friendly Foam Shop, the Monkey Pub, although the last two are a little south). So Roosevelt is a good place for an underground station, it just wasn’t as urgent as University Way and Broadway were.

      8. I guess these all make sense. I don’t really dig the Roosvelt argument, in that it was done for no reason but political pressure, but whatever, what’s done is done.

    2. All three ST Express buses from downtown Seattle to the eastside have runs leaving after 11:30 pm on weekdays, and after 11:20 pm on weekends.

      Some of the service drops to hourly on weekends, but it does exist.

      Metro service to the eastside shuts down earlier.

      1. That does leave a gap between UW and Bellevue, though. Metro should run a few later short-turn 271’s; they could probably do it with just two buses.

      2. It’s too bad ST doesn’t have a horizontal conveyance device to whisk people between UW and ID Station all the way until midnight.

      3. What about at the wee hours. If you’re trying to cross the lake between ~1am and ~4am, you’re SOL.

      4. The last 255 to Kirkland leaves Montlake Station at 12:25 am on weekdays and an hour earlier on weekends.

        There used to be a suburban night owl, Route 280, that ran a big clockwise loop from downtown via 520, Bellevue, Renton, Tukwila, and back downtown but that was cut a few years ago in the recession. There were two trips at 2 am and 3 am.

      5. The night owls are now largely funded by Seattle. Talk to Bellevue and Kirkland about helping fund cross-lake night owl service, and it could happen.

      6. The night owls are now 100% funded by Seattle. Metro was going to delete all of them but the city council put in money to keep the Seattle ones.

        I forgot that the 280 gave a kind of UW-Eastside service.

      7. Actually. not all. Metro may be paying for the suburban part of the E and 120 since Seattle is paying most of the cost. And the A and 180 night owls are funded by Metro. So I guess Metro didn’t cut all night owl funding, just Seattle and the Eastside.

    3. Actually, you’d still have to buy the ROW along I5; you’d save far less money than you think. You would capture the Convention stop, but meh, not enough.

      Transit, to be useful, has to stop at a place you can easily get to and has to stop at a somewhere or can easily develop as a somewhere. There’s much, much more somewhere at the current CapHill, Udistrict, Roosevelt stops than there would be at their I-5 equivalents.

      And I say that, even though I’m a booster to the Seattle 130th stop, which would be an I-5 stop. The idea there is that the stop should be built precisely at the 125/130th E-W arterial connecting with a very frequent bus route, developing a somewhere. STs non-ability to build a station easily serviceable by Metro is a comment fest for another day.

      1. Huh. That’s a good point. We’d be giving the state more money to mess around with suburban freeways.

    4. The express lanes are still in high demand, especially by buses from Snohomish County and Shoreline. Closing it before light rail is expanded far enough north would be politically impossible. Having the stations further from I-5 helps increase the amount of developable space around the entrances and makes things more pleasant for those who live/work in the new TOD. Freeway noise and air pollution drop off fairly sharply even after a few hundred feet of distance.

      1. > Having the stations further from I-5 helps increase the amount of developable space around the entrances and makes things more pleasant for those who live/work in the new TOD. Freeway noise and air pollution drop off fairly sharply even after a few hundred feet of distance.

        I see. That being the case, why ARE the stations further north on the freeway then? Why were they not routed to the respective city centers? Lynnwood station is at lynwood transit center if I’m remembering correctly, which is good, but the rest are in random areas that don’t have anything around them but suburbs. Plus, as you say, freeway noise/pollution.

    5. Andrew, for one thing the ship channel bridge is a major bottleneck and I doubt it would ever be politically viable to take away any car lanes there. As for I-90, simple answer is there is geographically really no other way to cross! And if anything, the Link extensions are going to follow I-5 a bit *too much*, skipping most of the walk sheds, work places, and businesses along the corridor as well as effectively cutting the walk sheds and bike access areas in half. Freeway ramps are decidedly hostile to pedestrians and bicyclists–they’re for cars!. Finally, forcing bus transfers everywhere is a great way to cut your ridership in half–it’s NOT the same as driving to the freeway exit then driving on local roads. Instead imagine if the freeway offramp only opened once every 15 minutes!!!

      Northgate station is a good example–the heart of the walkable part is further north and east, so it will most likely remain a park-and-ride station and less of an urban neighborhood station (though the pedestrian bridge across the expressway mitigates some of this). That station should be located at the northeast parking lot area of the mall or an elevated station above 5th Ave. And there would be zero delay for those opting for the park-and-ride option–zero riders lost there.

      1. Gotcha!

        In my view, perhaps this has some benefits. Maybe Seattle can take this as an opportunity to build more freeway lids around dense MFH in places like Northgate, UDistrict, and Roosevelt stations.

  3. I just think it would be nice if ST did dynamic escalator direction changes at UW based on sporting events and morning rush.

    1. For sure. But in Pittsburgh, I think it was, I saw something better. Escalators could be quickly switched to change direction. Under game-day loads, best to have station agents on the switch, organizing loads of passengers to take turns in both directions.

      Considering the depth of some of our stations, though, if confined space made two-way escalators impossible, I’m surprised the authorities didn’t insist on controlled bi-directional operations from the beginning.


    2. I’ve been harping on the lack of switching escalators to match peak ingress/egress direction for awhile.

      But in the case of UW, where riders are both going downtown and coming to school, I’m not sure which is really the peak direction.

      1. I think this only makes sense for game days. I like Mark’s idea, have someone at the switch. The rule of thumb is that if you have an event that requires traffic cops, also have some escalator monitors. Before the game, they are going out of the platform. After the game, towards it. The monitors can then decide when it is time to just let them go back to their regular pattern.

      2. They don’t need to staff up. They already staff up for Husky football games. All that is needed is some training on which buttons to push.

  4. It’s nonsensical to me that we built two subway stations without stairs. I’m happy to use the stairs if the escalator’s full of people standing. I’m not happy to spend the first block of my 6-block walk on the only mode of transportation that’s slower than walking.

    1) In station in that video, there was still a place where people were allowed to walk.

    2) Walkers are less theoretically efficient simply because we tend to leave larger gaps between us. If I walk close behind the walker in front of me, I’d like to think I’m not contributing to the problem, or at least mitigating it.

    1. I think the bigger problem is that there simply aren’t enough walkers. You basically have three lanes: One for walking on the escalator, one for riding, and one for the stairs. But not that many people walk on the escalator. With this change, no one does. That lane has been given to those that like to stand. Since there are way more of those people, there is less crowding. In addition, many of those that like to walk on the escalators now walk on the stairs (which aren’t that crowded).

      I could be wrong, but I think if everyone walked on the escalators, it would empty out just as fast as if everyone stood. Everyone takes that first step onto the escalator the same way. You basically take the step after the person in front, whether that person continues to walk, or just stands there. In terms of emptying the platform, it really doesn’t matter, as long as everyone uses up initial step.

      In other words, the problems isn’t the walkers, it is that too few people walk. A story for our times, I guess.

      1. Ross, I remember reading a year or two back that the most efficient way to run a crowded escalator is for everybody just to stand still and ride. Dynamics of flow, or whatever technical name is.

        Part of the calculation is that riders standing, that is, letting the stair move them, still take up less space than climbing ones, letting the conveyor belt (good way to think of an excalator) carry more load per unit of time.

        Brent, I wouldn’t forbid passengers to switch escalator direction without supervision all the time. Though our extremely long escalators might have more coordination problems than ones like at Westlake, about one story.

        But for heavy crowds under pressure, too much chance competition for direction could cause some serious trouble. Which will also slow things down even worse re: police and medics trying to get down there to settle affairs.

        Could be unfair stereotype, but soccer in the United States doesn’t seem to have as high a casualty count as in some other places. Where police have been known to control post-game loading so everybody on each train is on or with the same team.

        But: do any of the English soccer clubs have women’s divisions? Because while the Swedish coach dismissed being called a coward by pointing our that her team won, Hope Solo might get a different reaction from a lassie from Liverpool, Manchester, or Finland. And her respective team-mates.

        Just our luck, will probably happen same time as a simultaneous collision at Othello and a shooting at Rainier Beach. At Friday evening Holiday Season rush hour with a game at Safeco Field as well.

        Remember, a whole tunnel-full of trapped passengers can’t run up the tube to from Westlake to Capitol Hill like it was an escalator. Because while it would definitely make more sense for mobs to use both tunnels for panicking in opposite directions…

        Brent, we’d better have some uniformed fingers on those switches. If only because it provides more of them to keep crossed.


    2. just wait until these stations are old and the escalators are too. terrible design. these stations are mobbed and all passengers cant fit in an elevator.

      1. Plus they’ve already experienced breakdowns with elevators/escalators at multiple stations. As new as these facilities are, I find these issues rather troubling.

    3. All the non-at-grade stations have stairs. Some of those stairs just happen to be hidden behind signs that say “Emergency Exit Only”.

      One such stairwell has been opened up for public use, at the north end of SeaTac Airport Station, last time I looked.

  5. Following up on the discussion about improving evening service from Beacon Hill Station to Georgetown, I did a deeper look at the current schedules for the 60 and 107 and how they could be changed to improve evening service along 15th Ave. S. Currently, Metro schedules both routes almost simultaneously along 15th S. which doesn’t make a lot of sense or provide good options for riders who may want to transfer at BHS. In my first look, I saw that the evening schedule for the 60 is pretty tight and that moving the 60 runs more than 2 or 3 minutes would necessitate adding another bus to the schedule. I also assumed that the 107 was working on a pulse schedule at the Renton TC and it would be detrimental to many other riders if the 107 schedule was changed significantly to benefit 15th Ave S. riders. But another poster suggested that it might be possible to tweak the 60/107 schedules by a few minutes and provide better coverage on 15th Ave. S. Because Link is mostly at 10 minute headways in the evening, it isn’t really beneficial to run a bus every 15 minutes–an evening schedule at 10 and 20 minute headways would match Link headways more closely. There are several times when small adjustments in the schedules would actually create much better service for Link-to-Metro transfers in the evening.

    1) The 107 trip leaving BHS at 759p should be moved ahead by 3 (or 13) minutes and the 60 scheduled for 804p should leave 2 minutes later.
    2) The 107 at 831p should leave 5 minutes earlier (this creates a slightly longer layover at Renton TC for the continuation to route 148) and the 60 at 834p should leave 2 minutes later
    3) the 107 at 855p should leave 1 minute later and the 60 at 903p should leave 3 minutes later
    4) the 107 at 931p should leave 5 minutes earlier and the 60 at 933p should leave 3 minutes later
    5) the 60 leaving at 1003p should leave 3 minutes later and the 60 leaving at 1005p should leave 9 minutes earlier
    6) the 60 leaving at 1033p should leave 3 minutes later and the 107 leaving at 1035p should leave 9 minutes earlier
    Later buses also could also be adjusted between -2 to +4 minutes to create a schedule that more closely matches Metro service to Link arrivals.

    There may be some contract requirements that I don’t know about that prevent some of these changes, but these tweaks are designed to allow 5 minutes between the arrival of southbound Link trains at BHS and the departure of a southbound 60 or 107 bus and eliminate duplicative scheduling southbound on 15th Ave. S.

    1. I’ve noticed the same, but since I live in South Park, I have less standing to grieve the missed opportunity for 15-minute-headway than you do. Would you be willing to write a post about this?

      If Metro hears from enough riders along 15th Ave S who want the frequency, I think it is just a matter of time before Metro does it.

  6. Interesting thing is that same engineers built the light-rail subways in both Pittsburgh and Seattle. Though the former railroad rights-of-way paved for joint ops didn’t do any underground running.

    But another thing both systems have in common: an old Downtown with space underground barely large enough for platforms or stations. Necessitating corridors and stairways really pushing standard “fits.”

    When budget finally allows, secret plans for station agents with huge hats and big crooked “staves” for accurate directions. Meantime, though, good thing we’re building another Tunnel.

    Also, considering other unknowns including when (name your favorite “car sewer”) is going to freeze over, might be a good idea to at least check out right of way that could be grooved-railed to handle joint ops for a couple of decades ’til we get pillars and undercuts for full-speed LINK.


  7. I’m envious of London having four escalators next to each other! In the DSTT there is only one escalator on each end of each platform. Some of those are so narrow that a person can’t pass someone standing anyway.

    With the magic number of 80 passengers a minute as shown in the video, we now know some data about how fast it will take to clear train riders leaving a platform — although this is for London. Seattle people seem to move slower and have more rolling things so I would put that capacity number about 60 – 65.

    Can we please push ST to study escalator deficiencies after ST2 and then ST3?

    1. I’m super disappointed with UW station’s design. It was designed 100% with sportsball at the forefront of the architects’ minds, an event that happens 5 times a year, and not commuters, the way more important demographic who use the station on a daily basis. A much better design would have featured a significantly wider platform, and have replaced one of the escalators with elevators. Ideally, there’d be something like 8 elevators IMO. That way, at least one elevator could be made always to be available on the surface and one on the platform, and for commuters the typical time to get out of the station would be no more than 60 seconds

      1. UW is a truly horrifically designed station… station location, station layout, station design… epic fail.

      2. It is terrible, but it wasn’t because they prioritized football (or any other event). Basically, the original (much better) station placements were killed for a number of reasons, mostly do to the inability of Sound Transit to cooperate with the UW.

      3. I was under the impression that the awful transfer environment for UW Station was due to UW intransigence about giving up part of their parking lot, no? And ST couldn’t use eminent domain on them because UW is a state agency.

      4. Epic fail indeed. Not to mention 10 years late and, combined with the CHS, some $2.3 million over the adopted budget.

      5. Pat. I recall ST paying an additional $10 million to the UW for the mitigation issue of the loss of parking spaces at Husky Stadium.

      6. It was the UW that wouldn’t allow any other station location. From the UW’s perspective, it’s the absolute corner of campus and can’t be put to better use, so that’s where it dumped the station. Plus it’s next to the stadium for game days, and a block away from a large medical center which generates lots of passengers all day and evening.

        The parking has to do with the lack of bus layover bays at the station. That’s where we want to put the Eastside buses, but UW wouldn’t give up or reconfigure its parking.

    2. Keep in mind that the London escalators move 50% faster than the escalators on Link. US safety codes limit escalator speeds to no more than 100 feet per minute or 0.5 m/s. London’s (and many other world cities’ transit) escalators move at 0.75 m/s.

  8. I was on a northbound Link train at midnight recently, and operations were single-tracked in the vicinity of Mount Baker station (and necessarily at Beacon Hill also), so we had to wait for a train to pass before switching to the southbound track and entering MBS on the west platform.

    Makes me wonder, since Link has to close every night for maintenance, I wonder if they could simply have alternating single-tracked (with timed train passings at turnback points) hourly (or half hourly) service, with maintenance being done on the opposite track. I could see one car trains being sufficient.

    1. I don’t recall the source, but I’ve heard the maintenance window is mostly about systems rather than physical infrastructure. That is only possible after the last train has exited the system.

  9. Metro has an update on the 520 restructure. It summarizes the public feedback from last spring. It doesn’t have a specific proposal but it says that will come in March. It also says, “Metro and Sound Transit are working with the Seattle Department of Transportation and University of Washington on plans for improvements to make the bus-rail transfer as convenient and seamless as possible.”

    It restates the three concepts: (A) no change, (B) reroute all 520-downtown buses to UW Station and make them more frequent, (C) reroute daytime buses but keep some evening ones to downtown.

    The feedback summary is, quote:

    “First, people are open to transferring to light rail for a more convenient and more reliable commute. They also want more frequent, reliable service and less overcrowding during peak periods. They also liked the idea of direct service to South Lake Union. Secondly, people want a seamless transfer between buses and light rail. People had concerns about giving up their “one-seat ride” and wanted more information about what it will look like for customers to regularly transfer between the train and bus at the Montlake Triangle.”

    Additional concerns:

    “Traffic from the Montlake Boulevard exit off of SR-520 and at the Montlake Triangle. Ensuring Link light rail has room for additional riders at UW. Transit capacity during Husky games and other big events. Impact of Montlake Bridge openings on bus service. Walking distance between the light rail platform and bus stops, especially for seniors and people with disabilities. Exposure to rain and elements. Extra time needed to transfer to and from Link. Transfers between bus and Link for cash-paying customers.”

  10. Well, it sounds like all agencies involved have the same right concerns as when they built UW Station to the point it’s at now.

    Question is whether they’re now concerned enough to extend their plaza across Pacific to the Hospital, and do whatever’s necessary to lanes and signals to keep those buses moving. And those passengers transferring.

    Because my own guess is that same level of active control might be needed to get people between buses and trains at rush hour. Let alone game days. It’s also time for some stats, or guesses, about likely performance for buses that stay on 520 all the way Downtown.

    My guess is that whatever transfer trouble passengers face at UW Station, they’ll still get Downtown faster than if they’d stayed on the bus using I-5 and Stewart. After DSTT won’t carry the 255 anymore.


  11. I think the STB Management – and Sound Transit management – needs to quite frankly wake up.

    This car tabs thing is almost a daily subject for right wing talk radio. Where’s OUR talk radio?

    The car side of the aisle is about to have their Seattle International Auto Show with beautiful models on four wheels and two heels. Where’s OUR transit expo this year? Oh and unlike this year, I will NOT be wearing a Sound Transit cap through the car show.

    There are a lot of car games on X-box and other such platforms. Where are the Sound Transit games?

    There you go.

Comments are closed.