Fun fact: the first escalator on the London Underground was installed in 1911 at Earl’s Court station. They hired a man with a peg leg to ride up and down the escalator all day to reassure people of its safety.

32 Replies to “Sunday Open Thread: Tube Escalators”


    “Roads, bridges, airports, water and transit systems are in bad shape, according to civil engineers. The report card from the American Society of Civil Engineers gives U.S. infrastructure a D-plus.”

    First and hopefully last time I’m in a quad-trillionth of agreement with present tenant of the White House. Ever since dirty underwear skid-marks started appearing on our flag this last Inauguration Day, I’ve always held my nose at the idea that our country has to be Made Great Again. In the Inaugurant’s definition.

    So let’s just say it’ll clear up any misunderstanding when we start building PCC streetcars again. With as much up to date technology as does not render the car weaker, more overcomplicated, and breakdown prone than when St. Louis Car Company closed its doors.

    If memory serves, Otis build elevators of same quality. When I was five, I thought there’d still be a job for me at age seventy-five, sharp uniform and visor cap, polished hardwood folding seat, brass wheel and all. Every morning when I got off the PCC streetcar I drove ’til I got my new job along with my pension.

    Before my union obsolesced along with the caliber of machinery it created. Meantime, classically appropriate British understatement at frame 5:02 on the video: “Danger. Deep Void.” Put that on the sandwich sign across the bridge from the airport LINK station. And everybody’s office door who’s responsible for fixing it.

    Mark Dublin

  2. Miway (ie Mississauga Transit) in Mississauga, Ontario is in the midst of a 5-year plan to “gridify” if you will their network. You can read about it here:

    “The MiWay Five Transit Service Plan (2016-2020) is moving Mississauga’s transit system from a design that radiates from the city centre to a grid network that will allow for more frequent buses along main corridors.”

    Miway is already largely grid-based as you can see by looking at it’s weekday service map (, but it will become even more grid-based.

    When I look at Kc Metro’s long-range plan, the system is still not grid-based enough for my liking, even by 2025 and 2040. An example of what a grid-based system would have if a route that runs from South Bellevue LINK/P&R (or even further south) straight to Kirkland TC (or even further north) without serving Belleve TC, as that only would add time for through passengers. This is essentially what Miway did with their Route 26. Route 26 used to go around Square One shopping centre to serve the City Centre Transit Terminal every day. This now occurs only on the weekends. From Monday-Friday, it now runs straight down Burnhamthorpe Road, as you can see from the route map:

    Compare this to the Seattle area where there is no route running straight down Northgate Way/110th Street bypassing Northgate TC. Or no route serving downtown Bellevue that runs in a straight line and bypasses Bellevue TC (like a fictional line from South Bellevue to Kirkland TC as I described).

    I think transit agencies in the Puget Sound region should be much more aggressive in redesigning their networks to conform more to a grid, without having mid-route diversions which just add travel time for through passengers.

    1. A grid is one important model but it can’t be elevated to an absolute. Seattle has two factors that work against a complete grid solution like Chicago: geography and land use. Water, cliff, and highway barriers prevent buses from going very far in a straight line. The difficulty in the Capitol Hill/Madison restructure is that you can’t go more than a mile or two in any direction without hitting a barrier, and straight routes contradict the density and travel patterns which are more in a stick-shift shape.

      The second problem is land use: strict grid routes go through large low-density single-family areas with low ridership and few destinations, and they bypass nearby destinations most people are going to. The biggest problem here is the low-density, residential-only part: it means people from outside the neighborhood have little reason to go there. In more gridded cities like San Francisco and Vancouver, the average density is higher and more mixed-use, so that the combination of residents, people going to locations further along the grid street, and people going to major destinations, all add up to healthy ridership on all grid routes. In contrast, a route on 15th Ave NE from UW Station to Mountlake Terrace looks good on paper, but it bypasses the destinations many people are going to: Greenlake, Roosevelt, Northgate, the Crest Cinema. Instead it gives you a one-seat ride through no-man’s-land, useful only to people living directly near it. And because that area is mostly large-lot single-family, that’s not very many people, and they mostly have cars in their garages and drive so they’re not on the bus. That makes an all-15th route a weak route. Metro has surely studied it and been interested in it just as I have, but it just never has been strong enough.

      But the LRP now has a compromise: a line straight north to 175th, then northwest to Shoreline and Richmond Beach. That makes the route stronger by putting more destinations at the north end: the library at 175th & 5th, the Link station at 185th & 5th, central Shoreline on 185th, the commercial destinations and E transfer on Aurora, and Richmond Beach. It’s an example of an L-shaped route, another strong model, where two routes that cover much but not all of their grid are interlined.

      The 49 and 45 are L-shaped routes, and possibly the 11. I always thought the 49 should be replaced with a straight north-south route on Broadway, but instead Metro has been gradually positioning the 49 as the main route on Capitol Hill, with the most frequent service and night owl. There may be something Metro knows about travel patterns there that I don’t.

      Likewise there are people who want to turn the 45 and 62 into a NW 85th-NE 65th route. That has some niceities but but it ignores the fact that more people on both sides want to go to the U-District, not to the other side. The U-District is where a 50,000 person university is, hundreds of small businesses, dense housing, and the most transfers to everywhere. People get a little prissy at having to transfer to get to the U-District in exchange for a one-seat ride to nowhere, and it’s worse if they have to transfer twice: once to the U-District, and again to Bellevue or wherever they’re going. With all this, a NW 85th-NEW 6th route is pretty unacceptable before Roosvelt Station opens. After it opens, maybe it would be OK, but you really have to evaluate the tradeoffs and dominant travel patterns.

      One idea I’ve been thinking about is to treat the whole area from 65th to the Ship Canal, I-5 to 15th, like downtown. More and more routes are being positioned so that if they go anywhere downtown, they go all he way through from Pioneer Square to SLU, because so many people are going to all parts of it. What if we did the same thing in the U-District; e.g., extending the 48 to 6th. The 45 already goes to UW Station so it would stay the same. If Roosevelt RR goes to 65th, that would do it for the 70. All this would make a lot of people happy, such as those on STB on 23rd who wish they could get further into north Seattle without transferring.

      In the case of South Bellevue to Kirkland, I assume you mean a Bellevue Way route. The problem is that a lot of people are going to the transit center to transfer. Perhaps the real problem is that the transit center is in the wrong place; it should be on Bellevue Way. Then there would be no conflict between a most-gridded, most-urban bus route and the transit center. Likewise with Northgate. Not going to Northgate TC would inconvenience a lot of people, and with the Link station coming it will be impossible to bypass it. But again, maybe the transit center is in the wrong place. The region should not have conflated transit centers (where people transfer) with P&Rs (where people park and take the trunk line, not the other routes). Transit centers should be right on main drags like Northgate Way and Bellevue Way. P&R should be somewhere else outside the pedestrian centers. If it’s necessary to add a station for a P&R, that’s better than putting the P&R in the neighborhood center. The split between Bellevue TC and South Bellevue P&R is pretty ideal in that regard.

      As for bypassing transit centers, I lived in Bellevue before the transit center, when routes crossed in downtown Bellevue wherever they happened to. I didn’t like having to memorize which route stopped on which street, or having two routes to the same area at different stops. The great thing about the transit center is you can just go there, walk across the platform, and get your next bus.

      TIB station has also come up repeatedly, as in why can’t the F go straight on Southcenter Blvd rather than detouring into the TC. The first question is Link connections: there must be a good transfer to Link. But walking from the street to the Link entrance is less than thirty seconds so it looks feasable. Another factor is that the F has a lot of through riders not going to the station, because it connects Burien to Southcenter. That’s not as much the case for your Northgate or Bellevue routes, where there is no Burien at one end and Southcenter at the other to draw a lot of people. Instead a lot of people just want to get to the transit center so they can transfer to another route.

      On paper a like on 15th Ave NE from UW Station to Mountlake Terrace looks like an ideal grid route bu

      1. >>In the case of South Bellevue to Kirkland, I assume you mean a Bellevue Way route. The problem is that a lot of people are going to the transit center to transfer.”

        But many of those people could transfer elsewhere. For example, if you want to transfer to the 271 you could do so at NE8th. There should also be a route from the future East Main LINK station straight down Main Street to the Main Street business district in the SW corner of downtown Bellevue. People transferring to that route could transfer at Main Street. Some of the routes coming from the east – places like Redmond or Issaquah, could be extended a couple blocks further west to Bellevue Way. That way anyone using those routes could transfer at Bellevue Way.

        Look at the Miway Rte. 26 map I posted. From Monday-Friday if you want to go to the City Centre Transit Terminal (where more than 20 routes converge), you are now forced to transfer. But the tradeoff is that it has cut approximately 10 minutes each way for people travelling through. If youre travelling 5 days a week in both directions thats 100 minutes per week, which is huge. Plus if you want to transfer to another route, most likely you can do it somewhere else. If you want to transfer to Rte. 19, you can do so at the corner of Burnhamthorpe and Hurontario. If you want to transfer to Rte. 28, you can do so at the corner of Burnhamthorpe and Confederation Parkway, etc. If the east-west routes in downtown Bellevue were extended to Bellevue Way, you could transfer to all of them at Bellevue Way, so no need to go to the Bellevue TC.

      2. I do like the idea of a full Bellevue Way route as an addition, and to replace fragmented Bellevue Way service,. My concern is just taking away from another route. If you move one route away from the transit center, what would be the full effect?

      3. There’s a big difference between nothing and lock step grid. The idea that there is no simple east west route from the sound to Lake Washington somewhere in north Seattle (north of 45th) is atrocious. Please, 125th/ 130th from Sand Point to 3rd Ave NW.

      4. Take a look at the proposed service map:

        There is a local Bellevue Wy route proposed, the 3989, which goes from the Kirkland TC to Factoria. However, it does divert slightly to connect to Bellevue TC. Why? Because it’s a local route and is designed to feed riders to various key destinations (Kirkland TC, Bellevue TC, S Bellevue station, etc.). It’s not the best one seat ride from Kirkland TC to anything south of Bellevue TC, so the diversion to the TC doesn’t hurt anyone.

        Basically, when looking at a secondary corridor, the route shouldn’t be design with an end-to-end rider in mind.

      5. There are three other factors that keep a pure grid system from operating in Seattle. First, some streets are too narrow to work for bus routes. That’s why we end up with buses on the few wider and hence congested streets that are pinch points in our system. For example, there isn’t a good street to continue Route 12 south of Union Street because 19th Avenue is too narrow.

        Another is that much of Seattle’s community destinations are in north-south oriented districts. It seems cruel to make everyone who is going east-wet have to walk or wait or a us to go 8 blocks off of the crosstown route. It’s easier to walk or drive than to use transit in these situations. Routes 3, 27 and 50 east of 23rd are routes which are essentially grid routes, but get weak ridership in the CD and SE Seattle because they only skirt past only a few commercial/ community destinations and everyone else must transfer.

        A final general problem that never gets discussed with grid systems is what happens at the end points. If neighborhoods hit a body of water or a park, there isn’t a good end point. The middle may be well-used but the end segments end up terribly unproductive. There are plenty of routes that end near Lake Washington or Puget Sound that seem to run almost empty near the end points.

        I think a better theoretical model for Seattle would first be routes which are ribs, but run along a commercial spine for a decent distance to provide connectivity to community destinations. The letter that comes closest is a “Z” but with the middle segment being perpendicular rather than angular. To make this work, we would need to have better bus lanes and signal priority for buses on these spines. Then, tying two end points of routes on a parallel street can be considered to create a two-directional demand that would provide access to more destinations. Whether it’s Rainier Avenue, California Avenue, Roosevelt Way/12th Avenue or Aurora Avenue, a Z-shaped route with the middle section on the commercial street is probably going to be more useful and productive than a true crosstown grid route is.

      6. >>It’s not the best one seat ride from Kirkland TC to anything south of Bellevue TC, so the diversion to the TC doesn’t hurt anyone.<<

        Huh? If it's not the best one-seat ride from Kirkland TC to anything south of Bellevue TC, then what is the best one-seat ride? It appears to be the only one-seat ride for anyone going from Kirkland TC to South Bellevue TC for example.

      7. “It appears to be the only one-seat ride for anyone going from Kirkland TC to South Bellevue TC for example.”

        The question is how many people would make that trip. South Bellevue Station is not a transit center; it’s a station and a P&R and nothing else. The only thing in walking distance is a nature slough, blueberry farm, four or five houses, and the trail to Mercer Island. The few people going to those can take the detour to the transit center. A more common scenario would be going from the Kirkland TC to Old Bellevue on Main Street or the apartments south of it, or possibly Bellevue High School (but if the bus comes back on Main Street that;s just as close to the school). That’s still only a few people.

        For many years before RapidRide B there was a bus from Kirkland to upper Bellevue Way to the TC interlined with a route on NE 8th Street to Northup Way and Redmond. That part of Bellevue Way got so little ridership it was reorganized into a coverage route. Lower Bellevue Way is hard to tell because it has been subsumed in the 550 for twenty years and in the meantime more apartments have been built there, so it’s hard to say how much ridership it would get on its own. But we do know that with the 550, a large chunk of people would get on at NE 4th & BW, a few maybe at Main Street, then practically nothing until the P&R. (Even though I lived in both upper and lower Bellevue Way before the 550 and used those stops, not many others do.) So if Metro judges it all deserves a coverage route, maybe Metro is right. And coverage routes don’t have to be straight like RapidRide lines. It’s more coverage to go to the largest transit center in the Eastside, and incidentally past the entire parade of large businesses in downtown Bellevue that may be people’s destinations.

      8. @Mike – S Bellevue will be a major transfer point for buses coming from the south and east, even if it’s not formally a transit center. So from a grid standpoint, it’s still a major node.

        @ChrisC – sorry, I should have said “it’s not the best route.” Yes, it is the only one-seat ride, but it will be likely be faster to do a 2-seat ride using a RR or ST Exp bus to get to Bellevue TC and then transfer to another bus (or Link) to get to the final destination.

        If you prioritize a one-seat ride, then sure you can take that route. But from Metro’s point of view, it’s not a key part of the overall grid, and connecting it to the TC (and Bellevue’s core) is more important than staying linear.

      9. An odd-looking route can sometimes be explained as the interlining of two routes. In this case, Kirkland to Bellevue TC via Lake Street and north Bellevue Way, and Bellevue TC to South Bellevue via south Bellevue Way. In this case, the anchor end of the two routes is the transit center, which they both share, so they’re interlined. A route that bypassed the transit center would arguably get fewer riders and be less useful, at least that seems to be Metro’s reasoning, and it’s plausable for Bellevue and the Bellevue TC station.

    1. I’m happy for the improved service, but won’t be down there to partake of it while it is free.

      Having Pierce Transit join the ORCA LIFT sub-pod would be really cool.

  3. Now that all the trolley buses have off-wire capability can we petition Metro to remove the overhead switches at Plum St. and Rose St. on Rainier Avenue? Keeping those switches requires every trolley to slow down to 10mph on every trip. I know it would cost several thousand dollars to clean up the wire and it would only save a few seconds on every trip, but it sure would be nice to eliminate those 2 slow spots.. The Plum St. switch is there for buses to deadhead to the Route 4 southern terminal and it’s completely unnecessary now. The Rose St. switch used to be for a regular turnback on Route 7 but it’s only function now is as an emergency turnback when there’s a blocking incident further south on Rainier Avenue which, with off-wire capability, is now completely unnecessary.

    1. I’m all for fixing slow spots, so long as it doesn’t cause trollies to switch to off-wire mode.

      But taking out trolley wire currently not used should be done only after we know for sure buses will never run under them again.

      1. The Plum St. switch and crossover is just south of I-90. The bus is doing 30mph and has to slow down to 10mph and then accelerate back to 30. The switch is only there to allow early morning Route 4 buses to get to their terminal. The Rose St. switch used to be used be used for the Rose Street buses, but since Metro built the Henderson loop the Rose St. turnback hasn’t been used.

        There’s also a turnback at Graham St. but that doesn’t cause as much disruption because the Graham St. stops are almost always used by passengers and the bus is already slowing down to serve those passengers.

        I’m thinking that a line crew could clean up the switches and cross overs in less than a day.

    2. Does anybody know how flexible our auxiliary power systems are? Is it possible to switch power on and off while moving? I suppose we can put “wiring pans” – those little fiberglass housed that used to be in DSTT staging- at different places around the system.

      But right now, while poles come down with the flick of a switch, driver has to get out and put the poles back up. And I don’t think anyone would dare drop poles with the coach in motion. If a pole on its way down snags in the overhead, momentum of the coach can drag down a 600 volt Brooklyn Bridge.

      Somebody presently driving trolleys: are new buses more likely to “throw poles” than last generation? And how well is the Instruction Department training drivers? I’ve seen drivers solve the slowing-down problem by just not doing it.

      Reason I’m mentioning training is that trolley driving requires very different use of power and brake pedals than with diesel equipment. Also, learning to work with momentum and road gradient. A driver should come out of training able to roll under a switch without passengers noticing the bus slowing down.

      And get enough extra pay to encourage drivers to stay on trolleys long enough to learn them. Usually takes at least a year. Meaning that far too many brand new operators drivers get work they hate as their first experience with transit. And last time they’ll ever touch a trolleybus. Replaced by their exact duplicates.

      However, some developments in Sweden could help. Pantographs. Like LINK. Semi-trucks that’ll lower and raise contacts at highway speed. Hybrid package powers coach when poles are down. Two wires- like trolleybus. With separate pantograph mechanism under each contact. Limited side-to-side traverse.

      Would’ve given anything to have this propulsion system for the DSTT all these years. Metro really did seriously study wiring the Route 7 into the Tunnel, with entrance to the I-90 ramp at Rainier and Dearborn. Regular trolleywire. Nixed because it cost $12 million.

      Know voltage difference probably makes this impossible, but would be great to see the embattled Mercer Islanders’ reaction if they got offered this kind of service on I-90 through construction. Quiet and environmental. Flexible.

      But bet me somebody wouldn’t file suit about the overhead ruining the view of something scenically concrete.


  4. I wonder when STB will cover the whole streetcar fiasco and what is the STB Editorial Board’s view on it.

    Personally, I’m soured on advocating for streetcar expansion. I’d rather see electric trolleys and hear a cogent case from Martin & Frank on the podcast why further streetcars.

    Very respectfully;


    1. Joe, the amount of time the First Hill Streetcar’s been down is a little short to be a fiasco yet. A car had a malfunction. Any chance that similar things have happened shortly after the opening of many freeways?

      The bad lane arrangement on Broadway also fixable without any capital expense at all. Just have trains and cars switch priority, and some signals re-programmed. For trolley buses to do same work as well or better, they’ll need exactly the same thing.

      It’s true mechanically that a trolleybus can steer around a blockage in its lane. In any kind of traffic- half the story. Bus driver has to wait to pass until lane alongside is clear. Also, fact that buses can change lanes to pass means that nobody official sees any problem making them do it. All the time.

      Given minimum railroad conditions- meaning priorities above- streetcars can carry larger loads than trolleybuses, and a lot more comfortably. Even in reserved lanes, the side-to-side variance makes the ride less comfortable for a bus. Especially with a standing load.

      Also, electric or diesel, buses have to run on street pavement, which I think takes more maintenance than track. Does anyone know cost figures?


      1. Solution to streetcar lane blockage should be an immediate tow. People will learn pretty quick.

      1. MAX cars surely won’t start from a standstill if their brakes are set at full power. Static friction is powerful! But when static friction is broken things are dicier!

        Public accounts of the Seattle incident state that the car lost power while in motion, moving downhill, probably on wet tracks considering recent weather. The parking brake set (as designed), but this locked the wheels and the car slid for a couple blocks before coming to a stop. If anything, the failsafe brake was too powerful

        … or, from the little physics I remember, the vehicle was too light, or the contact area too small (this is where track brakes come in I guess). I’d guess the cars’ failsafe mechanism is pretty good in most conditions — wet descents are tough to handle with no electrical power. From a distance, if I was investigating this incident, I’d first look into why the car lost power, and second, whether an additional emergency braking system could be added that would be effective in the (hopefully extremely rare) case where the wheels are locked and there’s no electrical power. It’s easy to imagine that a fully manual, fully mechanical track brake could be built, that’s easy enough to control that it could be used without destroying the track (I’ve read SF’s cable cars have such a brake, for example). I’ve also read that, in automobiles, the simple cable attachment for the parking brake is at least as likely to fail as the powered braking systems used for the main brake, in part because it tends not to be tested and maintained as well — any additional braking system for streetcars would have to receive routine testing and maintenance to be reliable itself!

    2. It was a bad vehicle selection. It’s just one of many issues about the FHSC.

      The basic problem: I don’t think it’s wise to have a non-transit-operator department (City of Seattle) choose vehicles and design routes as the primary sponsor. While City support is important, it needs to be done in a supporting role and not in a main role — because they just don’t fully understand the implications of their decisions on day-to-day operations (speeds) and safety.

      The same goes for hiring project managers and department directors who are light on experience of good transit design; these people bend to the narrow interests of others more easily and think that they know good transit design — but they haven’t really done many projects to know better how tradeoffs should be made. They often also don’t know when it’s time to provide some community education about why incorporating other things can be a big problem for transit operations and safety rather than to add them to projects at a detriment to transit operations (speeds) and safety.

  5. In same frame of mind as my first comment- that civil engineering report confirmed worst of everything I’ve been observing last few decades- I wonder why anybody -like the Democratic Party-hasn’t publicly insisted that along with everything else downgraded, transit is a life-and-death National Defense expenditure.

    And hospitals. And fire-fighting And water quality. And schools at all levels. How dare anybody, let alone everybody, get away with talking about these things competing with, and having to be sacrificed to, the defense of our country? How long can any unit in the field keep fighting without medics? And logistics?

    I think a good case can be made that the real superiority of our armed forces, certainly starting with the First World War, is the unmatched economic strength of the country behind it. What was the production rate of the “Liberty Ships”, the mass-produced freighters for World War II?

    But more than that, our real fighting spirit: our people’s up-to-now reasonable expectation that we are strong enough to fight a major war without the devastation and suffering that generally falls upon the average country in the most just war possible.

    Because given condition of every one of the systems that really powered our troops to victory- the enemy might not have to do very much but wait until something collapses, catches fire, or explodes and take credit. Leaving us angrily demanding that the world give us credit for our own body count. No matter whose overpriced jet fighter just shot down the other guys’.


  6. BART’s Warm Springs/ South Fremont Station opens on March 24, and regular service begins March 25. This is several months later than anticipated, due to delays because of an antiquated communications system.

    Travel time to the next station (Fremont) is 6 minutes according to the schedule. Only San Francisco trains will run there; not Richmond trains according to the schedule so that’s only one train every 15 minutes.

    Station opening video is here:

    With the East San Jose extension opening in just one more year, I believe that Santa Clara VTA isn’t going to run buses to Warm Springs BART when it opens — and will wait to restructure routes next year.

    Guesses on first full month average weekday station boardings? My guess is 5,000 — with 1,000 coming from Fremont station (which has about 9,000 average weekday boardings now). For comparison, Angle Lake has 2,800 and the number of parking spaces is about 1,000. This station will have about 2,000 spaces available. Keep in mind that Fremont sends many more workers into Silicon Valley than it does to Oakland or San Francisco.

    1. “Only San Francisco trains will run there; not Richmond trains according to the schedule so that’s only one train every 15 minutes.”

      Yes it does. It has to because the Fremont-San Francisco line doesn’t run Sundays. Go to, click on “Schedules by Line”, choose Warm Springs – Richmond or Warm Springs – Daly City, and choose a date in April. Both lines are extended to Warm Springs.

      1. You are right about Richmond trains serving the station at off-peak hours when the San Francisco train isn’t running. I had missed that. Still, the service during peak times is posted to be only every 15 minutes.

        It will be interesting to see how much station switching from Fremont Station will happen. The net extra time savings is about 8-10 minutes from driving each way but there is a higher fare of 45 cents each way to San Francisco and frequencies are half of what they are at Fremont station (noting that Fremont station riders can take Richmond trains and transfer at Bayfair to go to San Francisco).

        There isn’t much within walking distance of this new station yet either. Big future TODs are in different stages of development. Its market when it opens is almost entirely park-ride and drop-off/pick-up for now.

Comments are closed.