160 Replies to “Sunday Open Thread: A Day on King County Metro”

  1. With all the added buses going to reduce overcrowding on West Seattle (and Delridge) routes, I’d like to hear where else buses pass passengers by.

    1. 358! Between 8:30 and 9am, if a bus is a few minutes late, it’s guaranteed to be crowded and leaving a few people at the curb at 45th. I don’t think its really a problem at stops further north. Evening commutes sometimes fill up as early as pike. Seasoned commuters are used to it and have built it into their day, but there’s usually one or two outraged people. 358 is ripe for rapid ride.

      1. It seems like, ideally, there should be RR coaches, 60-footers, 40-footers, and 30-footers, standing by for all KCM routes.

        The WTA, in Bellingham, has operators report full buses, and the WTA uses these reports to determine when to add “shuttles”, which are unscheduled runs.

        (Though, they add the “shuttles” to the Trip Planner, which is really nice.)

    2. It occasionally happens on:
      – the 71/72/73, usually outbound from downtown in the morning;
      – the 41, usually outbound from downtown in the evening (especially if a bus gets delayed);
      – the 3/4 between Harborview and downtown in both directions;
      – the 8 on Denny in the evening, if a bus gets delayed.

      And it can happen on other routes under weird circumstances. I once had to pass up some passengers driving the 37… granted, it was a snowstorm.

      1. Oh, forgot the obvious one, because I wasn’t thinking of ST. 550s are regularly overloaded to the point of passing up passengers, because they need more frequency at peak than ST is allowed to put into the tunnel under its operating agreement with Metro.

      2. Also some weekday morning 44’s between Wallingford and the UD. It’s been a while, but some years I’d get passed up by a full bus at Wallingford Center occasionally, and was on full buses that passed stops closer to the freeway even more often.

      3. It is not ususual to be passed by a 71/72/73 local early on a weekday evening if you’re trying to go from South Lake Union to the U-district. Once, I even had this happen to me at 5:00 on a Sunday afternoon (although, in fairness, the Sunday headways have improved from 15 minutes to 10 minutes since this incident, so maybe that’s Sundays a problem anymore, but the evening headways hasn’t changed as far as I know)!

      1. Game nights are another whole kettle of fish. Almost any route serving the stadiums will pass people up on game nights. If we lived in a world of unlimited money, it would be great to have three or four 4-car Link trains come in a convoy out of the maintenance base, fill right up, and get people out of the area and into downtown. And pretty much any route leaving downtown in a southbound direction will have overloads on game night unless an extra trip is run.

      2. If anyone’s wondering why 80% of people drive to Mariner’s games, this is a big reason. If you have a high chance of being passed up on your trip home, and asked to wait for the next bus that isn’t coming for a full hour, you do not take transit to the game – if you have a car, you drive yourself; otherwise, you find a friend with a car to go to the game with.

    Martin, congratulations, you’ve made me do the most nerdiest transit geek thing I’ve ever done.
    Fast forward to ST2 completion, and try to visualize the vastly complex array of bus routes running around King County from the riders perspective. I taped a clear, sheet protector over my monitors screen, drew in the Link ‘Trunk Lines’ and added the stations. Now, sit back and watch all the bus movements, and try to visualize how over half of those will somehow end up on the trunk line.
    Appendix C of the ST voters guide calls for 165 million transit trips in 2030, of which 88.5 million will be on Link, with a transfer rate of 1.49 involving a bus boarding to start or end the trip.
    I’ll let you draw your own conclusions, as I did.
    One trunk line and 33 stations on rail will be great for many trips for many people, and many more that currently don’t use transit, but I can’t fathom about half of all those movements in the video being replaced by single Link lines.

    1. The ST2 election is over. We voted. Can we move on and help bring about the conditions that enable high-capacity transit to work?

      1. I couldn’t agree more Brent. HCT works when the seats fill up with butts, and those butts will come from the buses you see running all over the place.
        I’d like to help figure out the best ways to connect at least half of those dots with the new trains. If that’s not moving on, then I don’t know what is.
        Taken section by section, or route by route, I just can’t see half the dots disappearing from the map.
        Many are still grousing about lousy integration of Metro and Link in just the Rainier Valley and that’s the low lying fruit. Ballard, Fremont, W.Seattle, Vashon, Fed Way, the entire Green River Valley, Renton, Eastgate, Issaquah, Kirkland, Woodinville are a harder nut to crack.
        It’s a discussion worth starting before the trains start running.

      2. I couldn’t agree with you more that Metro’s service pattern in the Rainier Valley is quixotic. If I were in charge, I would remove all off-peak buses from the freeway on the segment between downtown and Rainier Beach/Tukwila.

        I think John Bailo had a great idea — just take the existing transfer point at Renton, and relocate it to Rainier Beach station. You end up with an all-day service pattern like this:

        – RapidRide from downtown Renton to Rainier Beach via the direct route
        – Frequent bus from Rainier Beach to Kent via 107/169 route
        – Frequent bus from downtown to Seward Park via Rainier Beach (106 north route)
        – Frequent bus from Rainier Beach to Fairwood (101/148 route)

        You’ve eliminated a whole bunch of redundant service; you have frequent and legible service from Seattle to Renton; and all of the remaining lines connect directly with Link, meaning that the total number of transfers is unchanged.

        Really, when you have a high-capacity train so readily available, there’s no good excuse to have a major transfer point that’s *not* at a train station…

        My real point is, even if Metro hasn’t done this yet, they can. There will come a time when Link is treated by everyone as the spine of the system. We’ll take it for granted that the best way to get anywhere is to take a frequent train/bus to a connection point, and then maybe a local service to our final destination. And ridership will bear that out. We’re still in the teething phase, so don’t take the poor integration as a sign that it will never change.

      3. Thanks Aleks for a thoughtful response.
        Assuming we did all that, how would you justify it to the current riders? Time Saved?
        One of my pet peeves is the reluctance to treat bus riders as customers that have other travel options besides a forced transfer. There has to be a significant benefit for them, such as time saved with a transfer, or more predictable service that is real, or creature comforts. Elasticity studies try to quantify those travel behaviors, but it’s the rider that ultimately votes with their feet.
        How far west of I-5 would your transfer shed extend, say at Northgate? West of Aurora? Could you justify everything funneling through Lake City to make a transfer at Northgate.
        Here’s the rub,
        Is Northgate being designed to accept all the transfers in an optimal manner, or will it look more like Husky Stadium with riders bitching about the lousy walk in the wind and the rain.
        The trains are beautiful. Stations impressive. Bus stops around the stations?
        Not so much.

      4. Metro’s record of restructuring to connect to Link in Rainier Valley is actually quite good. After next February, Metro will only have four Rainier Valley routes that could be considered to be providing redundant service: the 7, 7X, 36, and 106. I think the 106 will be involved in another restructure in the near future (and you may recall that it was restructured to create the first one-seat ride between Renton and RBS within a few months after Link opened for service). The 7 southbound finally has a stop alongside Mt Baker Station, where some buses wait for passengers from certain Link trips (albeit, hopefully not waiting long).

        The 106 could stand additional frequency, at least in the Skyway section. If the 106 were through-routed with the 169, I think that could become a highly popular route.

        I find it goofy for the 101 to run all the way downtown in the middle of the day. Maybe truncating it at peak is too politically difficult, but reverting to making a local equivalent of the 101 just going to RBS might not only save off-peak bus service for the apartments on Sunset, but also improve the waning frequency. The 101 is going into an off-peak and even worse counter-peak death spiral in its ridership with span of service getting degraded as well.

      5. I don’t think Bailo originated the Rainier Beach idea; it sounds too urbist for him. I first read about it in a comment by, I don’t remember, maybe Brent? I immediately thought it was a great idea and have been advocating for it ever since. But there are limits to it, which gets to my next point…

        “How far west of I-5 would your transfer shed extend, say at Northgate? West of Aurora? Could you justify everything funneling through Lake City to make a transfer at Northgate.”

        All transit lines have tradeoffs. Link is time-competitive with ST Express to Everett and Bellevue but not to Federal Way or Tacoma or Redmond. The same with replacing the 101. Link would be 7 minutes slower, I think Martin has calculated. That’s where human judgment comes in. Is the 7 minutes justified? My answer is yes off-peak, but no peak. I would cut the peak 101/102 too, but for the sake of peace with commuter-voters, it’s better to reorganize off-peak first, and make Metro double the frequency of the feeders, and leave the peak expresses for another year. After people have gotten used to the Link line and to more-frequent feeders, they may be more amenable to cutting the peak expresses or charging premium fares for them. The same goes for Federal Way, Tacoma, and Redmond.

        In north Seattle, Metro’s survey shows that almost all the Northgate P&R users come from east and west, not from Shoreline or Snoho. And in the Northgate station meetings, 99.9% of the respondents say they want more feeders to the station, and pedestrian paths and an I-5 ped bridge, rather than expanding the P&R or one-seat rides to downtown. So here Metro just has to do what the residents have asked.

        “Is Northgate being designed to accept all the transfers in an optimal manner, or will it look more like Husky Stadium with riders bitching about the lousy walk in the wind and the rain.”

        The station is being completely redesigned. To me it looks adequate, and comparable to the existing TC or TIB. You can look at the proposals and tell ST what you think of them.

      6. “redundant service: the 7, 7X, 36, and 106. I think the 106 will be involved in another restructure in the near future”

        The 7 is one of Metro’s highest-use routes, and serves a lot of intra-valley trips which Link and the 8 don’t. Likewise, the 36 goes across Beacon Hill, which Link doesn’t. The 7X is redundant, yes.

        Metro has been making the 106 non-redundant by turning it into a Skyway and Georgetown local. It now functions as two interlined routes from Rainier Beach, and is less attractive for Renton-downtown trips. It could be attached to the 169, I guess, although that would make a long route which could be unreliable.

        BTW, why does the 106 make that funny detour in SODO, from 4th to Industrial Way to Spokane to the busway. Why not just go straight north on 4th to Spokane or Lander, then turn to the busway? It doesn’t matter to me because I’m loving the additional route at the Costco stop, but I could see through riders being peeved at the detour.

      7. mic,

        The metric that I use is expected total trip time. That is, how long does it take to get from point A to point B, starting from the moment that you decide you want to make the trip?

        The on-vehicle time on the 101 from downtown to Renton is 40 minutes, and that’s assuming no unexpected traffic delays on I-5. The on-vehicle time on Link from downtown to Rainier Beach station is 27 minutes, and the 106 route is an extra 18 minutes, for a total of 45.

        From downtown to Renton, off-peak, you’ve saved 15 minutes of waiting, and added 12.5 minutes of waiting plus 5 minutes of travel time. That’s in the best case for the bus — any amount of delay on I-5 would make the bus trip longer. And the infrequency of the 101 increases the likelihood that problems with any one vehicle will delay your trip.

        However, consider someone going from downtown to suburban Kent on the 169. Currently, those riders have to connect from one bus with a 30-minute headway (the 101) to another bus with a 30-minute headway (the 169, say). Therefore, there’s actually 30 minutes of wait time.

        If the connection point were moved to Rainier Beach, then the first leg goes from a bus with a 30-minute headway to a train with a 10-minute headway. And the savings from running so much less redundant service would allow the remaining buses to be upgraded to 15-minute headways, which saves an extra 7.5 minutes.

        So, by moving the connection point, you’ve saved over 22 minutes of waiting, and you’ve only added 5 minutes of travel time at the most. That’s 17 minutes of savings for riders going anywhere not in walking distance of a 101 stop.

        And note that all of these calculations are for trips that go directly to downtown. If you’re going to Mount Baker, the Link trip gets shorter, while the 101 trip gets longer, since you have to backtrack. If you’re going to Capitol Hill or the U-District or Northgate, then once those segments of Link open, the Link trip saves you a connection.

        Every service change has winners and losers. Under the current arrangement, we spend a lot of money to provide a lot of redundant service which gives preferential treatment for riders to Renton TC and South Renton P&R but poorly serves everyone else. With the change, we greatly expand the number of destinations which have fast, frequent, and reliable service to downtown (and all Link destinations).

        When U-Link and especially North Link open, the same calculations apply. By consolidating nearby transfer points at Link stations, it becomes possible to provide faster and more frequent service to a much wider range of destinations. Here, the difference is even more dramatic, because Link will actually be faster than a freeway bus between the U-District and downtown. Therefore, a connected trip could save travel time as well as waiting time.

        It’s a bit early to speculate here, since so many things could change between now and then, but one improvement that I’d really like to see is the development of frequent gridded corridors connecting to each Link station. For example, a bus could combine the east portion of the 71 with the west portion of the 48, or the 44 could be extended east along Sand Point Way up to Magnuson Park, or the 48S could continue north along Montlake Blvd to replace the 68/243/372.

        I’m not saying that all of those changes should happen. The point is just that Link changes the game, and in all likelihood, it will be faster to transfer to a frequent train from Roosevelt to UW than it would be to stay on the bus. (And, if nothing else, passengers going directly from Roosevelt to UW will take the train, so the need for tons of bus capacity between Link stations goes down.) So there’s an inherent benefit in making our bus system have as little overlap with Link as possible; even if some trips are a bit harder than they are today, many other trips become much easier.

      8. Personally, I have no principled objection to providing peak express service that uses the freeway. But off-peak, there isn’t one southeast bus that meets Metro’s frequent service standards at all times. Only the 140 and 150 come close. By deleting the freeway segments of the 101 and 150 off-peak, and reinvesting the service hours in local service southwest from Rainier Beach, we could potentially achieve frequent service all the way through Renton and Kent.

        As far as the revised 106 being too long, if Rainier Beach were to become a major transfer point, I think Metro would probably split the 106 into two parts. One would provide local service through SODO/Georgetown/Beacon Hill. It would probably pick up part of the 107 — for example, terminating in Seward Park, or taking the long way to Renton. The other part would take the direct route from Rainier Beach to Renton, and would then head south to Kent. It’s that second route (let’s call it the 169) which would get the RapidRide treatment. It would essentially become the backbone of service for the southeast area.

      9. The 106 makes that annoying detour because of the Argo Bridge closure (Airway), and otherwise having to stay on its route.

        I can’t claim credit for the RBS connection idea, since it was a ridership assumption for Link.

        As far as Federal Way, I rode the peak-hour 577 to Federal Way Friday evening. It took 42 minutes from boarding at Jackson to alighting at FWTC. That probably beats the future Link travel time by not much more than a couple minutes. It is actually off-peak where the 577 could leave Link in the dust, but then that gets into the same math as the 194.

      10. When I refer to “redundant service” on the 7, 36, and 106, I’m talking about the riders who get on south of the station, and continue riding past the station all the way downtown, and vice versa.

        If the trolley wires allow it, there could be northbound runs of the 7 that start at Mt Baker, some reverse runs that only go to Mt Baker, northbound runs on the 36 that start at Beacon Hill Station (though the ridership on the 36 is already much lighter south of the station, as all the density is north of it), reverse runs that only go as far as Beacon Hill Station, northbound runs on the 106 that start at RBS, and some reverse runs that only go that far.

        With the 106, however, I suspect the Skyway section has higher ridership than the south Beacon Hill section now (in part, because of the Argo Bridge detour). I think it would make perfect sense to have some 106 runs that just go between RBS and Renton TC, to give Skyway decent frequency. Such an investment would help Metro with its social equity goal.

      11. One big irony about truncating off-peak service, but providing redundant peak service is that the peak period is precisely the period in which the redundant express service saves the least amount of time and is the most expensive to operate. Not only is the peak period the most likely time for traffic congestion to erase whatever time might be saved on paper, but the peak period is also when everything is running as frequently as it will ever get, making the train->bus transfers as painless as they are ever going to get. Similarly, from a standpoint of operations costs, extra service hours during the peak require the purchase of additional buses, which is far more expensive than running buses you already have for additional hours off-peak.

      12. asdf: I agree completely, and in general, I’m strongly in favor of focusing on the all-day network, with minor boosts to frequency when the extra capacity is needed.

        However, it’s important to remember that the principal problem which transit is designed to solve is that cities don’t have enough space for everyone to have a private automobile. If you live in the Seattle suburbs, then the only time you’re faced with this dilemma is during the peak commute to/from downtown. At all other times, and in all other places, there’s plenty of space for everyone’s car.

        Commuter express services may be expensive, but they’re a heck of a lot cheaper than building extra freeway lanes.

        The real solution is to do what virtually every other city does, and charge a premium rate for commuter express services. Ideally, we’d eliminate the peak premium on regular services at the same time.

        Charging a double fare for all peak commuter services (by which I mean, any bus which only runs during peak periods, and only in one direction at a time) would go a long way towards addressing the cost issue that you describe, and it would also provide a natural incentive for people to ride the services which are cheaper to provide.

      13. “One big irony about truncating off-peak service, but providing redundant peak service is that the peak period is precisely the period in which the redundant express service saves the least amount of time and is the most expensive to operate”

        Yes, but that’s looking at it from some abstract pavement-efficiency perspective rather than a human perspective. The buses aren’t running because there’s some imperative to rub the streets with wheels often enough that they don’t wear out, it’s to get people where they need to go. And in our society, and in every other industrial society, almost everybody goes to work at the same time. If people can’t get to work, or it takes two hours on 30 mph buses with long transfer times, the economy suffers in a major way. An interesting fact about many peak expresses that shadow a local route (like the 15X or 7X or 71X), is that the express takes about as long as the local does at night. In other words, it essentially makes up for the congestion that the local incurs at peak hours.

      14. Mike,

        asdf’s point is that, on a platform-hour basis, peak express service is much more expensive to provide than off-peak service. Even if the bus only runs in revenue service in one direction, it has to return to its starting point to make an additional run. And drivers generally hate short shifts and split shifts, so you often have to pay drivers for more time than they’re actually driving. Therefore, if you take an all-day bus and turn it into a peak-only bus, you may only save 20% of your money, despite an 80% reduction in platform hours.

        That’s bad enough for places like Sammamish, where the only transit market is the peak commute market. But it’s particularly vexing when peak expresses run at the same time as regular routes. If the peak-express riders all switched to the regular routes, it could save buckets of money, even if you had to run a few extra buses on the regular routes.

        Again, I think the solution here is simple: charge riders extra for the premium service they’re receiving. The riders who value it will pay; the riders who don’t will switch to the regular service.

      15. “If the peak-express riders all switched to the regular routes, it could save buckets of money, even if you had to run a few extra buses on the regular routes.”

        That’s not necessarily true. As long as the express trips are necessary to provide capacity, replacing them with an equal number of trips on a regular route doesn’t actually save anything – every additional trip during the regular route that operates peak-period, peak-direction only still requires an extra bus, an extra driver, and extra deadheading just for that one trip. In fact, if you try to do this, the extra trips on local route actually ends up costing more than the express route because you have to pay for additional fuel and driver time resulting from the extra stops.

        The I-90 commuter service is a good example of this. If every peak run of the 212, 216, and 218 were replaced with a peak run of the 554, we would be paying more, not less, than what we pay for the current service.

        Because of this, peak-only express routes, so long as they are mostly full and needed for their passenger capacity, do not bother me. Nor does it bother me when we have a peak-only route that saves riders huge amounts of time, but doesn’t have the demand to justify it during the off-peak hours (for instance, the CT 424 going from Monroe to Woodinville in 15 minutes, a trip that would take over 2 hours, off-peak, unless you drive).

        The peak-only express routes I do object do, however, are routes that:

        1) Are not necessary for capacity (plenty of empty seats on regular route buses to accommodate such passengers. If the regular route is a high-capacity train, this condition can sometimes be met, ever if the express route buses are consistently full.


        2) Only save riders a trivial amount of time over regular routes, or worse, don’t actually save riders any time, but merely eliminate a transfer or 5 minutes of walking, which gets made up for by a slower and/or more circuitous route.

      16. Aleks and Brent, you know I’ve been an opponent of simply cutting off today’s south end service to downtown at RBS without other changes; I think it’s passenger-unfriendly and not worth the large investment that would be needed to make RBS into a true transfer hub.

        But now you’re talking about bigger ideas that have got me thinking, and I’m revisiting the idea of the RBS transfer hub in connection with a major restructure of Renton and Kent service. The key is that we have to be able to improve enough people’s service to make condemning a bunch of land and spending tens of millions to rebuild the area around MLK and Henderson worth it. So here’s a proposal, organized by route. The idea is to match up corridors that will have a lot of through ridership and that will warrant similar service standards, and to prevent three-seat rides and make two-seat rides transferring at either Henderson or RTC as friendly as possible.

        Summary: Through-route 101S/169, 105/106S, 106N/107, 150S/180S, 164/180N, 166/168. Create new, more efficient peak-hour service concept for Renton and surrounding areas. Rationalize Kent peak-hour service.


        See 142, 143, 144, and 169 for replacement service.

        See 144 for replacement service.

        See 105 for replacement service between RTC and RBS and 107 for replacement service between RBS and downtown.

        See 142 for replacement service.

        Replacement service provided by 157 and 159 east of 104th SE, and by 164 and 168 connecting to Sounder west of 104th SE. No replacement service along the NE shore of Lake Meridian.

        Replacement service provided by 159.

        NEW ROUTES

        New peak-only route connecting downtown, S Renton P&R, and Renton Highlands. Would use current 101 routing through S Renton P&R. Would not serve Renton TC, but proceed into Renton Highlands via NE 3rd/4th. Once on NE 4th, would continue using current 111 routing. Would run substantially more frequently than current 111 to accommodate volume of passengers from current peak 101 and 111 service 142, 143, and 144 would be scheduled to create 10-minute peak service between downtown and S Renton P&R from 5:30-9:00 a.m. and 3:00-6:30 p.m., with possible extra trips in the peak of the peak if needed.

        New peak-only route connecting downtown, S Renton P&R, and Fairwood using current 102 routing, but extended along current deadhead route via Maple Valley Road so the AM start/PM end terminal is RTC. Schedule together with 142 and 143 as described above.

        New route covering local portion of current 180 in South Auburn, with service every 30 minutes until 7 p.m. and hourly until midnight. This is actually a revival of the old 151 that the 180 replaced (although with less service to match ridership). It could be extended into Algona/Pacific should funding become available.

        Peak-only route providing service every 20 minutes, in conjunction with revised 159, from Interurban Avenue to downtown. Would use current 150 routing from Southcenter to Tukwila P&R and current 161 routing from Tukwila P&R to downtown.


        Extend from RTC to RBS along Renton Ave. S. via Rainier & Henderson, using current 106 routing. Improve frequency to 20 minutes during the day weekdays and Saturdays; remain 30 minutes Sundays and nights until 10 pm; hourly after 10 pm.

        Extend to Downtown Seattle via South Beacon Hill and Georgetown, using current 106 routing through Airport Way and Spokane and current 124 routing into downtown. Alternate with 124 to create a 15-minute frequent-service corridor between 3rd/Pine and 13th/Bailey. Break current through-route with 148. Frequency; 30 minutes until 10 pm; hourly after 10 pm.

        Eliminate portion of route south of NE Sunset Blvd. Service to the area is provided by 105 and new 142. Add 2 trips each direction and revise route to start and end at St. Matthew Lutheran Church P&R, both to replace lost 111 service.

        Change routing to: 1) serve S Renton P&R rather than RTC, but then continue to Black Diamond via current routing; and 2) use DSTT. Add three additional short-turn trips to Wilderness Village. Schedule together with 142 and 144.

        On north end, truncate to RBS via MLK exit from I-5. On south end, extend to Auburn Station via current 180 routing. Create northbound short-turn trips to Kent to replace current early-morning 180 service.Keep corridor in mind for possible conversion to RapidRide as funds permit.

        Use current 158 routing between Lake Meridian P&R and SE 240th/116th SE. Service along current 157 routing between those points wil be provided by 159.

        Use current routing between Covington and Lake Meridian P&R. Use current 161 routing between Lake Meridian P&R and downtown. Service along inner portion of Kent-Kangley Road would be provided by 169 connecting with Sounder. No trips would be added.

        Extend from S Renton P&R to RBS via RTC (using Main Ave S rather than Rainier), S 2nd/3rd, current 101 routing from Rainier/Sunset to MLK/I-5, and MLK to RBS. Break current through-route with 166. Increase frequency to every 20 minutes weekdays from 6 am to 7 pm. Keep corridor in mind for possible conversion to RapidRide as funds permit.

        Eliminate route south of Kent. See 150 for replacement service between Kent and Auburn Station, and 151 for replacement service in South Auburn. Through-route remaining portion of route with route 164.


        Break current through-route with 107. Through-route every other trip with 155 at Fairwood Center during 155 span. (The eventual goal would be to rationalize these two routes into one route with similar combined frequency, but I’m only so ambitious tonight.)

        Through-route with 148 at Fairwood Center. (The eventual goal would be to rationalize these two routes into one route with similar combined frequency and span, but I’m only so ambitious tonight.)

        Through-route with new 180 (portion of current 180 between Kent and Burien). Break night through-route with 168.

        Break current through-route with 169. Through-route with 168.

        Through-route with 166. Break night through-route with 164.

      17. And.. I think I’ve figured out the 148/155 knot. This solution really wants the new 169 to be RapidRide in order to work well.

        Increase frequency to 20 minutes during the day and 30 minutes until 10 pm.


        Extend along current 155 routing from Southcenter to Valley Medical Center. Turn back in VMC parking lot.

      18. Dangit… forgot one other piece of the 148 revision: extend it west from Fairwood Center to serve the Cascade area currently served by the 155, and lay over somewhere in that area. The residents there would lose their direct east-west service but would gain vastly higher frequency.

      19. Would your revised 150 really need to get on the freeway at all? Or could it go Interurban-East Marginal-Boeing Access Road-MLK, or even Grady Way-Oaksdale-Monster Rd-68th-MLK?

      20. With regard to making RBS a major transfer point, I think there is a significant challenge due to it’s existing design and positioning in the MLK roadway. In order for a transfer point to be beneficial, it must minimize the “transfer penalty”. I know from my own experience using RBS, Othello and Columbia City stations that the current “transfer penalty” runs about 6-8 minutes (in addition to normal headway waits). That is the amount of time it takes to disembark the bus, walk to the intersection, WAIT for the cross walk at TWO sides of the street, travel into the station to wait for a train – one of which has likely come and gone while you’re stuck on the other side of the street.

        If we are going to expect people to efficiently transfer to Link then we need to design stations to accomodate buses out of the roadway. Here’s a url to a picture of the CTA’s Forest Park station as an example of a transit center with out of roadway bus to train transfers. http://www.transitchicago.com/assets/1/wayfinding/co_forestpark_pagehead.jpg

        Of course, we have this at TIB station. Pity it wasn’t applied at Mount Baker or other stations.

        Since RBS is already built and modifying it to create this kind of synergy would be expensive and perhaps disruptive of existing service, I re-float the idea of a Boeing Access Road intermodal station. Many more routes from the south could be truncated there versus RBS. Sounder could come into the picture. Sounder DMU service could be added.

      21. Morgan, I think it would be best to keep the 150 on the freeway for speed. The freeway is much faster than any of those surface routings, and multiple attempts to provide service in that area have met with near-zero ridership.

      22. @Charles,

        Adding an in-fill station would cost orders of magnitude more than installing bus layover and pick-up/drop-off areas.

        In the case of a Boeing Access Station, that would increase the travel time for all trips downtown from any of the routes we’re discussing, as well as add another minute for all through-trips on Link. Getting to such a station sends buses in the wrong direction, and would almost certainly involve a much more substantial time getting up to the platform. If the station is designed for Sounder transfers, that would be a multi-level mondo expensive station.

        There is space around RBS where bus pick-ups and drop-offs can happen just one crossing from the station, and not parallel to the station (leading to jay-running in front of and behind the bus). It may require buying some property, but there aren’t any pricey buildings down there now. Of course, in the case of through-routed stops, one way or the other will have two crossings, but I’ve never experienced a transfer taking close to 6 minutes to get to the platform.

      23. The biggest obstacle to making such extensive changes is convincing Metro that some of the tradeoffs are worth doing. The hardest nuts to crack are eliminating the one-seat rides on the 150 and 101. Metro has so far not shown any willingness to do this in spite of numerous suggestions by transit fans. There is a reasonable counterargument that these are basic service to south-central and southeast King County because Link is not close enough to them to adequately replace them. So, the only way to go forward is to convince Metro that it’s worth doing anyway. But how?

        This issue is similar to ST’s restructuring of the 512 and non-restructuring of the 594.

        I could tentatively support this restructure, although I won’t get into the weeds of specific local routes until it looks like Metro might be willing to budge. One question though will be, why restructure the 180? Metro seems to think a north-and-northwest route is worthwhile. There may be reasons for this, such as ridership on the blue-collar corridor to the airport. I do agree that straight north-south and east-west routes are ideal, but we also have to consider dominant trip patterns if any.

      24. Mike, the 180 came about because Metro needed to split the 150, which ran at the time from downtown all the way to downtown Auburn (and then Auburn Station), and was exceedingly unreliable as a result. The dominant trip pattern in the corridor is north-south. By breaking off the Seattle portion of the 150, you could extend it to Auburn again, which would be good for most riders in the Kent Valley, and which would match frequency with demand between Kent and Auburn. The 180/164 would make a very nice, complementary east-west corridor. Both routes have the potential to be expanded to frequent service at some point down the road.

        I agree that the political lift to eliminate one-seat rides is one of the hardest parts of this proposal, along with the substantial work needed to turn RBS into a workable, friendly transfer point. You could also go halfway with it, by truncating only the 101, which I think would be more palatable if you ran a bunch of peak service from S Renton P&R to downtown and through-routed the truncated 101 with the 169. In that case, none of my Kent/Southcenter proposals except the 157/158/159/161 consolidation would make any sense.

    2. “but I can’t fathom about half of all those movements in the video being replaced by single Link lines”

      But can you fathom a portion of that half using the closest, and fastest trunk line? Specifically, look at the denser areas of the animation including Downtown, I5 North, and I 90 – the areas that coincidentally have some of the most crowded buses. The problem many have with visualizing this is that transfers, by and large, are currently relatively painful. Improving the experience of the 2 seat ride is key here.

      1. The diagram also shows each bus with equal weight, even though some of them are much more crowded than others. Generally, the corridors that will have light rail and the most crowded ones. Also, the map shown is just King County Metro. A lot of the buses that will shift to Link are operating under the Sound Transit or Community Transit label.

    3. You could just as easily say, “I can’t fathom about half of all those movements in the video being replaced by a single freeway”. And yet, many people do go out of their way to use the freeway, because if your trip is long enough, the extra speed makes the diversion worth it.

      Imagine living in a city where every street was a narrow local access road with traffic circles. It would be pretty slow to get around!

      Conversely, imagine a city where there was an underground freeway every 4 blocks, with grade-separated interchanges. It would be really fast to get anywhere, but the costs would be mind-boggling.

      Thus, we compromise, and build a couple of spines (I-5, I-90, 520, 99, 15th Ave W), along with a comprehensive network of connecting roads to get people between those spines and where they want to be.

      Link is the new I-5.

      1. Exactly. If you look at how people use subways in the cities that have them, they function like freeways. People flock to them because they’re the fastest and most convenient way to get around. Subways also cost like freeways. Which means we shouldn’t just approve several freeways here and there and then become shocked, shocked, that a subway line will cost the same.

      2. The principal difference being that freeways presume the ability to access them from anywhere and to disembark them to reach anywhere.

        Subways must be built such that their passengers can reach their intended destination by walking or by excellent connecting transit. Fail to do that — and Link/Metro wholly fail to do that — and you end up having spent that money on something that simply doesn’t work.

      3. So getting taxed for more freeway lanes is somehow better than getting taxed for a rail line.(or a dedicated bus lane).

        The only difference seems to be that the transit users pay something back at the farebox, and the freeway users don’t.

      4. And that subways bypass traffic, and free people from having to spend $10,000 on a car plus gas and maintenance, plus they allow people to be part of the solution rather than part of the problem (by not adding to the congestion).

      5. With barely 1 or 2 exceptions, that simply is not true.

        You keep raising the spectre of the high-Soviet-era centrally-planned St. Petersburg system, because it’s literally the widest stop spacing you can find among functional metro systems anywhere in the world… and its gaps still pale in comparison to the way Link-style American light rails shaft their cities.

      6. Capitol Hill Station will be very walking-friendly, as will UW, U-District, and Roosevelt.

        U-District, Roosevelt, and especially Northgate will become major bus connection points.

        Metro has obviously done a pretty poor job at integrating the bus network with Link in the southeast. But there’s reason to expect that things will be better in the north, for two reasons:

        – Congestion on I-5 (and the poor design of the express lanes) means that Link will offer a real travel time advantage over even the most direct bus between the U-District and downtown.
        – The street grid in North Seattle is much more amenable to crosstown routes, like the 44 (which is one of six bus lines in the city that has frequent service all day every day).

        Therefore, I think that north end riders will welcome the reorganization of the bus network around Link.

        I do hope that Metro’s new focus on productivity will eventually lead to a reorganization of southeast bus service as well, but we’ll have to see.

      7. For what it’s worth, I do wish that Link had a few more stops, and I definitely think that we should have gone to Ballard (via 45th) well before Northgate. I will definitely do what I can to ensure that ST3 includes a Link extension to Ballard via 45th, and that it has an appropriate stop spacing (meaning, none of this 1.3 miles between stops business). But for the Link we’ve already got, I think the most productive thing we can do is to urge Metro to reorient its service to be perpendicular rather than parallel. The service hours we’re wasting on the freeway segments of the 101 and 150 are hours that could be buying us actual frequent corridors in that area.

      8. Sorry, I got diverted from the blog conversation, and now it’s too late to chime in, but tons of good thinking going on here. Maybe Bruce or Martin could host a discussion on just integrating a corridor at a time – say everything on I-90 when E.Link is done, or N.Link when Northgate opens. I really enjoyed the thinking outside the box on the whole Green River Valley service, as that’s where I spent a lot of hours working on the current network of routes during the 90’s on the S.King Sounding Board.
        My battleship brain is changing course – albeit very slowly, as the one seat ride syndrome is a killer to overcome, whether real or perceived. Human nature is fickle.

      9. mic, you guys and Metro did a great job with that ’90s restructure. The 164/166 were necessary and excellent additions to the network.

  3. With it very likely Mitt Romney is going to be President come 20 January – great for the economy, but a nightmare for Amtrak as we all know – is there some pro-Amtrak Cascades group being formed? I’m interested.

    I mean yeah I can take the Airporter from Seattle back to Skagit but still, business class seating ( http://www.flickr.com/photos/avgeekjoe/8121512840/in/photostream ) is well worth full, no-subsidy fare. I mean I like paying full fare and getting to get up & walk around a bit on those long train rides – especially last 4 July when I had to take Amtrak Cascades back from Tacoma and some great air show action. Sitting on a Greyhound? The schedules would not have worked, requiring a hotel stay or no trip when one can’t drive.

    So clearly I have a reason to, as you’d say, agitate for Amtrak Cascades to remain viable as either a private entity or as a state government-run organization. Thanks for your help.

      1. FiveThirtyEight, which Kyle linked to, is pretty much the gold standard of political prediction. Nate Silver is an experienced sports statistician who recently started applying his skills to politics. His methods are based on accumulating as much data as possible, figuring out about how reliable and/or biased each source is, and combining them together to create an accurate and stable prediction.

        As of this writing, Silver predicts that Obama has a 73.6% chance of winning the election. That includes data from the poll you linked to, and many, many others. That doesn’t mean that Obama will win — if we repeated these exact same circumstances 4 times, we’d expect Romney to win once — it just means that Obama will probably win.

        Note also that the poll you linked to is a national poll. The Electoral College math has surprisingly little to do with the national results. Everyone agrees that the chance of Romney winning the popular vote exceeds the chance of Romney winning the Electoral College. I happen to think that the Electoral College is an undemocratic anachronism, but for now, winning the popular vote doesn’t mean much.

      2. Let’s not forget Nate Silver correctly predicted every single state in 2008 with the exception of Indiana I believe, which was off by 1% within the margin or error.

      3. Even TPM, which has been the most Republican-leaning of the prediction sites this whole cycle, is still showing Obama with a significant lead in likely electoral votes. National popular vote percentages mean nothing when we still have the Electoral College.

    1. Even George W. Bush wasn’t able to completely gut Amtrak funding, in spite of having large Republican majorities in Congress during the earlier part of his term.

      As much as Republicans like to Bash Amtrak – and Big Bird – they are both still here and will likely remain, regardless of their rhetoric.

      And it is way premature to assume Romney is even going to win. The polls still indicate President Obama as the favorite.

      1. Obama slammed Romney’s naval ships comment saying we also have fewer horses and bayonets than in 1916.
        Amtrak is the naval ship of transportation. It’s a dinosaur. Times change. Modes of transportation change. It’s not a Republican or Democrat issue.

      2. Never mind that Amtrak exists chiefly because the private passenger railroads that preceded it couldn’t compete with the publicly subsidized national highway system. Why is there no push from the Republicans to privatize the Interstate Highway system? It’s not that they are for a different kind of Big Government, right?

      3. Europe and Asia called. They disagree that trains are the equivalent of horses in the military. They also mumbled some things about America being backward in several ways, such as creationism in textbooks, so they aren’t surprised.

      4. Mike Orr, speaking of backwards, how many decades away are western european countries away from electing a black man president? How many more decades are these non-backwards western european countries away from abolishing their monarchies?

      5. They abolished slavery before we did. :) And their monarchies evolved into constitutional monarchies, which means the king can’t interfere in the running of the country. As my British-Canadian friend described it, speaking of both Britian and Canada, “If Parliament decides to scrap the constitution and impose a dictatorship, the Queen can call in the army to defend the constiution. It’s quite a simple system and it works well.”

      6. It greatly depends on which Amtrak route you’re talking about. The routes that go all the way across the country which, half the time, show up within a couple hours of when they’re supposed to are, can be viewed as a dinosaur. On the other hand, shorter routes like the Cascades and Northeast Corridor attract a lot of people.

        Imagine how much more productive the Empire Builder would be if, for example, it just went back and forth between Seattle and Spokane, with trips that arrive and depart Spokane during daylight, rather than the middle of the night. (Ignoring for the sake of argument that the track sharing with freight trains makes this impossible for the foreseeable future without investments that would cost far more than they would be worth).

      7. “Mike Orr, speaking of backwards, how many decades away are western european countries away from electing a black man president?”

        How many years before we elect a female president? Germany and Britain have already done it.

        “How many more decades are these non-backwards western european countries away from abolishing their monarchies?”

        Name one major European country that is still ruled by a monarchy.

      8. Republicans only support reducing the size of government as a means of getting rid of something they personally don’t benefit from. As soon as the subject turns to something they want to use they’re all for BIG government.

    2. With it very likely Mitt Romney is going to be President

      Not sure how you get there from here, unless you buy in to the theory that the polling firms’ “likely voter” algorithm is broken.

      But even if he does win, despite his strong statements on the issue, Romney’ll have his work cut out for him actually defunding Amtrak without control of the Senate, because it’s looking like D’s will still be clinging to a slim majority. Even if they don’t hold it, the existing R controlled House actually collaborated with the D controlled senate and boosted Amtrak funding by a quarter billion for next year (granted, that was all in the form of capital spending earmarks, and the operations subsidy was cut). Republicans talk big about gutting this or that program, but when it comes down to voting, they usually find some excuse for why a project in their district is special and unique, and thus should be preserved. And there’s a super special unique project in every district.

      1. That’s the beauty of Military Spending. There’s always a military base in each state, and in many a representative’s district.

  4. I understand there are some Metro bus drivers who comment here. Are there any Link operators here? If so, I have an assignment for you. I want you to go out and buy a cam, like a GoPro or Contour. Then I want you to affix the cam to a helmet you would be wearing and record an entire Link trip, from your perspective as the train Operator, then post it to Youtube. Write a comment about it here when you are all done to let me know. Or maybe STB will even make a post about it. I want it to be done by November 15.

    1. As a bus operator I’m sure I would be suspended if I complied with your request. I can only imagine the rules governing rail operators to be more stringent. If you don’t like the rules, talk to this guy.

      1. I was just joking about demanding it, but I do think it would be a cool video to see an entire Link trip from the vantage point of the operator.

  5. This is great! If there was a little clock that showed the time elapsing on the side, it would be the greatest.


    In a previous thread, I posted a comment that was summarized as essentially turning RapidRide D into the 15 Express. I’d like to explain my reasoning.

    RapidRide was sold and intended as a frequent trunk line that would form a spine through our most heavily-traveled corridors. The failures of its implementation have turned it into a massive reduction of service and inconvenience to Ballard residents. The practical upshot to me, as a resident of Ballard’s multifamily core centered around 24th Avenue NW, is that I can not rely on my morning commute to be timely or consistent, nor can I rely on Metro to get me home past 11pm.

    The main implementation problem is that RapidRide doesn’t come nearly frequently enough. If we don’t have the platform hours to run the service, then we shouldn’t be running the service because it is a net harm to do so. Rather, let’s change our strategy. I think he implementation has enough redundancies that we can transform RapidRide from an all-day unscheduled and unreliable 15 local replacement into a core-hour high-frequency and reliable trunk line. In other words, an improved 15X which actually supports the connections people need to make to and from it, rather than leaving them stranded at varying times and thus unable to rely on their timed transfer.

    Delete the 15X, since RapidRide was supposed to replace the need for it in the first place. Instead, restore the 15 local with 20-minute service frequency during the day and 30 minute service frequency at other times. It will serve as RR D’s shadow, but it is overlapped with the 32, 40, and various Queen Anne buses for its entire route and thus does not need to be as frequent as its predecessor.

    Delete route 61 and extend route 48 to Ballard & Market, laying over in the old 75 location outside Canal Station. Route 61 is an abomination; a useless waste of cash that severely cuts service for 32nd Ave W residents, a place where we’re just starting to see pockets of increased multifamily development. Having ridden the North Beach loop, there’s no need to serve this neighborhood outside of peak, which is the decision that was made for Crown Hill when the 15 local was axed.

    Remove the Queen Anne dogleg from RapidRide D. Lower Queen Anne is more than adequately served by the 1, 2, 3, 4, 8, 13, restored 15 local, 29, and 32. RapidRide D would stop at 1st & Denny, which is a short, gradual slope from the Seattle Center. This is accessible because RapidRide provides no new net benefit to Lower Queen Anne residents.

    Remove RapidRide stops at W Armour, Wheeler, Newton/Armory, and Galer Sts and NW 60th, 70th, 75th, and 80th Sts. Basically, anything that’s not a “station”, and even one or two that are. The 15 local, 32, and 40 provide shadow service on these corridors. In order to make this work, we need to remove the RapidRide stop at Mary Ave NW and turn and have it stop at 15th & Holman southbound again to provide a transfer to morning 40 and 15 local commuters (85th is already a very popular stop for locals and for 48 transfers).

    Decrease RapidRide D headways to 7 minutes or less, which we pay for by cutting RapidRide hours to 6a–7p. Below a certain frequency which I have unscientifically estimated as 7 minutes, the schedule-free nature of RapidRide D makes it impossible to make connections. So stop trying! Let’s play the mode to its strengths and stop sending good money after bad on its weaknesses.

    1. “The main implementation problem is that RapidRide doesn’t come nearly frequently enough.” Sure it does. You’re just impatient.

      1. I know you’re just being even more contrarian than usual today, but the crux of my point is that without a schedule, transferring to or from a route with 15-minute frequency is broken.

      2. I don’t disagree, but that ship sailed when Metro agreed to ratify its new service guidelines, which specified that “very frequent” service means 15 minutes.

        Rather than improving frequency, I think it’s much more realistic to ask Metro to adopt the “every 10 minutes until” schedule format, where scheduled times are listed for all periods except peak.

        The most important think to remember is that, eventually, all of Metro’s core arterial service will be RapidRide. I’ll say it again: everything on this page will someday be RapidRide. In today’s budget environment, Metro simply can’t afford to run 10-minute service on all of those corridors. But 15-minute service, with extra buses during peak, is realistic. So let’s stick with that for now, and instead try to improve the usability of what we’re going to have.

      3. RapidRide also stops too much in Belltown. Every 4 blocks should be sufficient. Every two blocks is too much, especially with zillions of other buses serving those stops.

      4. We don’t need 4 dozen one-seat corridors running every 15 minutes, Aleks, any more than we need our current 8 dozen corridors running every 30 minutes.

        We need 2 dozen corridors running every 10 or less, reliably, so that getting from anywhere to anywhere, transfers and all, just starts TO WORK. 15-minute meanderers can’t do that, no matter how many of them you have.

      5. d.p.,

        We both know that “frequency is freedom” (to use Jarrett Walker’s words), and that 10 minute headways are leagues better than 15 minute ones.

        However, the official service guidelines state that “very frequent” service is defined as 15 minute headways during peak and off-peak, and 30 minutes at night. It also mentions that RapidRide services should have 15-minute frequency all day and night, but better at peak.

        RapidRide, despite the marketing, is not about providing a faster bus. Instead, it’s about legibility. This is Metro’s way of building a frequent service network bit by bit, and about teaching riders that a network of frequent gridded routes can be better than radial spaghetti.

        It may take ten years, but there will eventually come a time when everyone in the city is either within walking distance of Link, RapidRide, or a streetcar, or their main bus is a circulator which brings them to one of the above.

        When that happens, we’ve achieved something beautiful: every single one of Metro’s services has the same frequency and the same span, for the entire route. You don’t have to worry about whether this bus goes hourly after 6pm, or if that bus has a turnback, or if this pair of buses switches to a shuttle. Every bus is equally trustworthy.

        And even more importantly, once we’ve reached that point, *everyone benefits from RapidRide*.

        Today, if Metro tried to boost the frequency of RapidRide, people in neighborhoods without RapidRide would complain that Metro was biased against them. But if every neighborhood has a RapidRide, and Metro proposes to boost the frequency of the whole network to 10 or 12 minutes, then everyone benefits.

        Look — I wish that RapidRide had 10-minute headways all day. Really, I do. But, at least for the time being, we’re not going to get it, no matter how much we might like to. We’re not going to get it because the current political environment does not allow Metro to make radical system-wide changes, and the current fiscal environment does not allow Metro to have 10-minute RapidRide without instituting politically infeasible cuts elsewhere.

        The end of the current path is the same: we’ve rebuilt Metro’s network as a high-frequency grid. But so long as the county council has veto power over Metro’s plans, this is the only way we’re going to get there.

      6. And in the intervening decade(s), we’re stuck with a (partial) transfer-based network in which transfers are terrible, one-seat commuters (understandably) defend their one-seat rides with pitchforks, and a whole new generation learns that the automobile is the (obvious) best way to get around.

        I fail to see how that helps the political situation at all, how that ever allows us to move toward the functional future you envision.

        Basic, real-world-tested service principles and best practices do not alter with politics or time. Transit either works or it doesn’t. Endorsing years of service in the “fundamentally doesn’t work” column is no path to building a transit-enabled city.

      7. And in the intervening decade(s), we’re stuck with a (partial) transfer-based network in which transfers are terrible, one-seat commuters (understandably) defend their one-seat rides with pitchforks, and a whole new generation learns that the automobile is the (obvious) best way to get around.

        Central Ballard is one of the most transit-accessible places in the Pacific Northwest. Metro has 223 different routes, and only *six* have 15-minute or shorter for the whole route, 16 hours a day, 7 days a week. So for most of the Seattle metropolitan area, buses come every half-hour at best. They take a meandering route to downtown. Connecting anywhere else is basically not worth it, since you’re liable to wait at least 15 minutes, and more if the bus you’re connecting to happens to be delayed.

        RapidRide D may have reduced service to 15th and Leary, but between there and Crown Hill, it increased service to meet the 15 minute standard. Similarly, RapidRide C extended frequent service from Alaska Junction to Westwood Village. That means that a whole bunch of people now live within walking distance of a frequent bus.

        If your bus route gets upgraded from 30-minute to 15-minute frequencies, then your first connection effectively becomes free. Instead of waiting 15 minutes for one bus, you wait 7 minutes for the first bus and 8 minutes for the second. (Yes, it’s possible to time your trips to match the 30-minute bus’s schedule, but if you start counting from the moment that you decide to go somewhere, that’s only an illusory savings. You may be able to wait inside, but you still have to wait that much longer to get where you’re going.)

        So if we replace a plate of 30-minute radial spaghetti with a 15-minute grid, then we *have* achieved the main benefit of a connection-based network, which is that the number of destinations you can get to in any given amount of time has increased dramatically. Yes, it’s not as good as if the buses were more frequent, but on the whole, it’s far better than what we have now.

        Essentially, RapidRide gives up some frequency on the most well-served corridors, in exchange for establishing a minimum level of service for a greater number of people. I think that’s worth it.

        Basic, real-world-tested service principles and best practices do not alter with politics or time. Transit either works or it doesn’t. Endorsing years of service in the “fundamentally doesn’t work” column is no path to building a transit-enabled city.

        There are plenty of examples of real-world cities where “frequent service” means 15-minute headways. For example:

        Vancouver, BC
        Washington, DC
        Los Angeles
        Boston (to quote: Service operates every 10 minutes or better during weekday peak periods, every 15 minutes or better during weekday midday, and every 20 minutes or better during off-peak periods.)

        In fact, with the obvious exception of New York, I can’t find a single North American city which has explicitly created a “frequent transit network” map with headways that are consistently better than 15 minutes. Los Angeles had a 12-minute map at one point, but it seems to have been demoted to 15 minutes.

        Given Seattle’s size and density, and the ridiculous state of our current transit network (where half-hourly routes are the norm), I think that 15 minute frequency is a very appropriate goal.

        The goal of RapidRide is to define a single standard of frequent service, and to ensure that everyone who lives in an urban center/village is within walking distance of such a service (and everyone else is within one connection). If we defined RapidRide to have 10-minute headways all day, then some urban vilages would get left out. The political consequences of that are non-trivial. If every neighborhood has RapidRide, then everyone has a stake in RapidRide improvements. If only some neighborhoods have it, then it becomes a divisive issue, and the left-out neighborhoods will vigorously oppose improvements that they feel come at their own expense.

        I know I’m repeating myself here. I just really feel that the benefits of uniting the whole city around 15-minute service, 16 hours a day and 7 days a week, are worth much more than maintaining a higher level of service on a much smaller number of corridors.

      8. The problem with your list of examples is that in each of those cities, with the exception of Portland and post-recession L.A., a good chunk of the mapped frequent network beats that 15-minute “minimum standard” — in many cases by double the frequency. And that’s to say nothing of the subway networks in each city, running trains at frequencies between 4 and 12 minutes throughout the day, late into the evening, and all weekend.

        Seattle bests its own minimum standard… basically nowhere. Our first urban RapidRide lines were BRT, remember? They replace corridors that should be on the level of those other cities’ core-subway standards, not on the level of their lowest minimum-to-make-the-cutoff routes.

        At the core of those other cities’ networks are routes that aim to provide truly spontaneity-enabling service and painless transfers; the outer “frequent” routes are a downgrade by comparison. In Seattle, you advocate that we all suffer an insufficient standard on every trip, in every direction, at all times.

        Central Ballard is one of the most transit-accessible places in the Pacific Northwest.

        And yet, post-restructure, I make the rational determination to drive for nearly 1/2 of my trips — the trips for which I’m most likely to be able to choose a route that significantly bests the bus, and for which I’m most likely to strategize free parking. I am rarely disappointed by this decision.

        So how is it admirable that this is the best that we can do?

      9. In the 1980s, Portland completely replaced its bus network in one fell swoop. They ripped out a bunch of infrequent, circuitous routes to downtown, and created a frequent grid. (And yes, frequent was defined as every 15 minutes or less.)

        Meanwhile, look at what happened when Metro proposed a much more modest restructuring. There was vigorous opposition from virtually all affected parties. Outside of West Seattle, the changes that passed ended up being more modest still.

        Consider the 4. Metro proposed to kill it, in exchange for increasing frequency on the 3. But residents were so attached to their half-hourly bus that they fought Metro until it backed down.

        For a 15-minute bus, the expected wait time is 7.5 minutes. If the proposed 3 had instead been given 10-minute frequency, the expected wait time would be 5 minutes. Do you really think that 2.5 minutes would have convinced Joanna Cullen that killing the 4 was the right move?

        For better or worse, when it comes to transit, Seattle is highly conservative. Our institutions are almost purpose-built to discourage change. So I’m inclined to support any steps in the direction of a frequent grid, no matter how small. As much as I wish radical change were possible here, I just don’t think it is.

      10. Yes, I too read that post, and yes, I do think it’s impressive that a fundamentally small city with consistent low-medium density and no radial rail (at the time) could institute a dramatic grid-based restructure that achieved the bare-minimum standard of usable frequency across the board.

        But we’re a bigger city, with worse bottlenecks, and our core lines continue to replace the urban network that we haven’t built and that is barely likely to change our situation as proposed.

        Meanwhile, Jarrett Walker himself calls you out on your “expected wait time” calculations (read the last few paragraphs especially). A handful of Cullens of the world might have the capacity to withstand infinite delay for the sake of their one-seat rides, but many of the neighbors she whipped into a frenzy would have trouble arguing with such a dramatic improvement to their worst-case waits.

      11. From Chapter 12 of Human Transit:

        In the Direct Service scenario (figure 12-2a), a service runs directly from Neighborhood 1 to College. It runs every 30 minutes, so the average wait time is 15 minutes.

        So he can’t object too strongly to that type of calculation. ;-)

        Having said that, the numbers look exactly the same if you don’t divide by 2. Currently, if you want to catch the 4, your worst-case wait time is 30 minutes. If you take a more frequent route with a connection to another frequent route, your worst-case wait time is also 30 minutes. The difference is that the number of destinations you can get to with that 30 minutes of waiting has expanded dramatically.

        For me, it comes down to this: I think that the system we have now is, on the whole, better than what we had in August. I think that the continued deployment of RapidRide will continue to make things incrementally better. And I think that changing the region’s institutions and culture to permit more radical change is something that I have no power to do. Therefore, I think the best use of my time and energy is to accelerate and guide the change that’s already happening.

    2. I like the thrust of your idea, but I disagree strongly with two points.

      First, I don’t think we should be cutting RapidRide’s hours, and certainly not so severely. I like the idea of improving frequency, but if anything, I think we should match both frequency and span with Link. That means 7.5 minute service during peaks, 10 minute base, and 15 minutes early morning and late night. 7pm is just far too early. I’m not always home from work by then!

      It’s important to keep in mind that RapidRide really is Metro’s highest-quality product. All four of the RapidRide lines meet Metro’s most stringent standard of frequent service — every 15 minutes from the start of service until 9:30 pm, 7 days a week, for the whole route, with no branches or turnbacks. The only other routes that meet that criteria are Link, the 36, and the 44. I wish that Metro had decided to standardize on 10 minutes, but for now, they’ve chosen 15. (In fact, 15 minutes is the number specified in the official report of that committee in charge of establishing new service guidelines.) I think it’s better to pick one standard, and maximize the number of routes which meet it, then to pick a different standard for every route.

      Second, I don’t think that we can justify running a local shadow of RapidRide. The extra service hours required to do so are really expensive. The only real benefit to the local shadow is allowing you to have wider stop spacing on the rapid service. In this case, I don’t think that the time savings of 1-mile stop spacing on RapidRide would make up for the loss in frequency from paying for a shadow service.

      I think that the benefits of a fast and frequent all-day connection from Ballard to downtown far outweigh the loss of a one-seat ride from Ballard to LQA. So I think that removing the deviation stands on its own. To make up the service deficit in LQA, we can simply extend the 8 west, to provide the main Magnolia service off-peak.

      Everything I suggested above is completely in line with the official planning documents of both Metro and the City of Seattle. The only exception is removing the RapidRide deviation to LQA. If we focus on this one change, and come up with a coherent proposal that’s otherwise in line with things that Metro and Seattle have already agreed to do, I think it’s possible that we can achieve something here.

      1. What is with this new obsession with sending Magnolia service through LQA?

        Magnolia access to LQA? That’s what the D is for :P

      2. Extend the 8? One of the most unreliable lines we have that already meanders through difficult areas? And whose service has no duplication?

        Clearly they chose getting people on the bus and serving areas people live and want to go over a quick trip from Ballard to Downtown. I agree we need that but is this the answer? Is this the goal of Rapid Ride? It doesn’t seem like it. It seems like they tried to make a backbone of “frequent” service that hit many key areas which they will then build off of in the future. Basically it sounds like the complaints above are all a matter of a lack of resources… we know. If that weren’t the case we would have everything we ever dreamed of.

      3. I was on a mid-afternoon SB D changing to a mid-afternoon NB 24 today – the 24 did not have noticeably worse ridership. 18 going up the bridge, 7 continuing to Viewmont.

      4. The idea to send the 8 to Magnolia didn’t originate with me — it’s in Seattle’s Transit Master Plan, which was approved by the city council in April 2012.

        Before someone says it, yes, I know that the TMP is an analysis of corridors, not necessarily a proposal for new routes. However, if you read the document, it’s clear that the authors felt strongly that the corridors would make good bus routes. To quote:

        The TMP proposes service restructuring that moves Seattle transit toward a more grid-oriented design. This is best illustrated by the proposed FTN investments that link services between the Rainier Valley and the University District and between Beacon Hill, Capitol Hill, and the University District. Rather than traveling to downtown, routes would be modified to cross multiple FTN lines that offer convenient transfers to downtown (Link light rail, Madison BRT, and east-west priority bus routes). While some downtown-bound passengers would need to connect (transfer) to Link, others would have direct connections that did not previously exist (e.g., Rainier Valley to Central District and Beacon Hill to First Hill/Capitol Hill).

        The motivation for linking Magnolia to the 8 is simple: it’s the logical place for an east-west bus to go. The 8 will probably be a RapidRide route someday. Given how far west the 8 already terminates, extending it to Magnolia would be pretty darn cheap — much cheaper than the existing service. That means that Magnolia would get to have a RapidRide route, with a direct connection to Link and RapidRide C/D/E.

        The current Magnolia buses have to pull their weight based on Magnolia ridership alone. That’s why they all have 30 minute headways except at peak. Providing Magnolia service via the 8, or some other E-W route, means that Magnolia can get service that is much more frequent than it could justify on its own.

        Basically, either Magnolia gets hooked up to a frequent corridor like the 8, or its level of service will continue to drop. There’s no other option.

        Separately, I should note that the TMP doesn’t include Elliot Ave as a priority bus corridor at all. So realistically, the chance of changing the RRD route is pretty much zero. Of course, the “key improvements” listed includes transit signal priority at every intersection through LQA; if we can get that, then the cost of the deviation will drop dramatically, which is a win for everyone.

      5. The trouble with extending the 8 to Magnolia, even in a world where it has perfect reliability, is that it doesn’t suit the needs of Magnolia riders. Damn near all Magnolia riders are going downtown, in high volumes at commute time, and in low volumes off-peak. I think it would be better for real-life Magnolia riders to accept an LQA deviation on the current routes in exchange for more frequent peak service and much later night service than to change Magnolia’s core service to a RapidRide route that doesn’t go downtown.

      6. David: Of course they’re going downtown. When the only bus route in your neighborhood comes every 30 minutes, you’re not going to take it if you need to transfer. Thus, anyone in Magnolia whose destination isn’t downtown has already excluded themselves from the group of bus riders. :)

        Consider the following:

        – The chance of the RapidRide D route changing is pretty much zero.
        – Extending any existing route into LQA would be very expensive.
        – The 8 is the only route which comes from the east and terminates in LQA.
        – The officially-approved Seattle Transit Master Plan recommends extending the 8 to Magnolia.
        – Extending the 8 to Magnolia would allow for the deletion of the entire downtown segments of the 24 and 33. That’s a lot of extra service hours that could go towards actually running buses in Magnolia.

        To me, this is exactly parallel to the situation with RapidRide in West Seattle. During peak, there would be direct buses from Magnolia to downtown along the existing route. Off-peak, riders could take the 8 and transfer to RapidRide D for downtown, or to anywhere else along the 8’s route. Riders to downtown have a slightly longer trip; riders to anywhere else have a significantly shorter one.

        I don’t live in Magnolia, but I’ve advocated precisely the same reforms for my own neighborhood. I’d love to see the 49/9/60 replaced by a single “Corridor 3” route that ran from the U-District to Capitol Hill Station, then along 12th Ave, down to Beacon Hill and Othello. Even though I’d lose my direct route to downtown, I’d benefit by having a direct route to many destinations which are difficult to get to currently.

        If we keep making that tradeoff throughout our system, we’ll eventually end up in a place where you can get from anywhere to anywhere, at any time of day. That seems much better than assuming that ridership will always be as downtown-focused as it is today.

      7. Oh for… Magnolia doesn’t actually remotely need “more frequent peak service”. Buses head down the hill at 7:06,:11,:16,:25,:35,:45, and :55. They hit the bridge homeward bound at 5:00, :07, :16, :21, :27, :34, :39 :48, and :57. Yeah, they’re not all the same route, but if you miss yours there’s walking to another bus route (often relatively easy) or the de facto park and ride in the streets around the Christian Science Church, which is where all the downtown buses stop before leaving Magnolia. I’ve never considered my personal 30 minute peak schedule insufficient – even if I miss it, I’m not tied to 30 minutes, I can hoof it across Magnolia on foot and get on a different bus that’s only 20 minutes behind.

        Later service would be nice, but not worth what people are trying to put forward as an alleged good trade.

      8. Magnolia potentially needs more peak/off-peak -reliability- of existing scheduled service (I have vaguely overheard there are major issues with that late peak) but that’s also something we shouldn’t have to make tradeoffs for.

      9. Magnolia doesn’t actually remotely need “more frequent peak service”.

        Not once have I proposed changing the existing peak service in Magnolia in any way. I’ve only proposed changing the off-peak service.

        It’s totally fine if you disagree with what I’m saying; I just want to make sure you understand it. :)

      10. The critical transit corridor in Magnolia presently is up the bridge to the Christian Science church and along/below the first ridge. This is where people live, although it somewhat lacks businesses. Up the bridge is presently scheduled for 15 minutes all day. Along the ridge is staggered 30 minute service that, in a pinch, can be used as 15 minute service if you’re willing to hike. Many people use the 24 and 33 interchangably, despite the very significant vertical between them in places. They also connect to each other to provide additional frequent service at Government Way, which has 5 buses an hour peak and 4 off-peak until the evening.

        It’s actually not a bad set-up, or wouldn’t be bad if the buses were more timely. Service to Magnolia Village could be better, although 31 transfers are more possible outbound. These are largely good routes with -unproductive tails-. Running them on 3rd to synergize with the D and other buses has probably made them better (if slower) buses too.

        Do something about the tails and shipping buses out to the tails and the route productivity metrics will certainly improve. Terminate the 33 at the border of Discovery Park. Maybe loop the 24 back to the bridge instead of going to Viewmont. Look at the suggested Ballard restructure again and do it in a way that doesn’t screw the existing quasi-frequent corridor. Route productivity will probably improve anyway from the routing on 3rd, and night service has already been cut, so what’s the issue?

        Sticking our buses onto the 8 is just silly. Off-peak only is also silly (what happens to the 124 and the 27?) And anyway, it’d almost certainly be faster to get to the interesting parts of Capitol Hill the way I used to, by walking up from 2nd & Madison, than it would be on a through-routing with the 8. Post-Link the 8 will be even more obviously inferior. There’s no actual added utility involved.

    3. I feel very strongly that we cannot have an RR route without a full span of service.

      The RR route should be the easy one to use, and shouldn’t disappear at night and put riders into a haze of local buses they’re unfamiliar with.

      I agree 100% that RR C/D frequency is insufficient and is strangling the routes. The solution here is not to find service hours to run a local shadow, but to find service hours to increase RR D frequency to 7.5 minutes peak, 10 minutes off-peak, and 15 minutes at night.

      In perfect-reliability-land I very much like your idea to replace the 61 with a 48 extension. In real life, I’m not sure the 48 is reliable enough to fill that role, and I’m also not sure there is anything that can be done to make it reliable, given the sheer number of difficult intersections along core high-ridership parts of the route. The 61 definitely is useless as is, and I think the best way to fix it is the original restructure plan that would connect it with a modified 24.

      1. Bruce had a clever way to split the 48 that avoids the Montlake Mess. I believe that idea stands on its own merits, but particularly so with the extension to downtown Ballard.

        I don’t think the 48 extension would increase unreliability on route 48N nearly enough to offset the recuperation of platform hours currently wasted on the Fred Meyer Layover Express.

        I was a big fan of connecting the 24 to the 61 (to the point of sending a very angry email to Larry Phillips’s office that has resulted in me being on every mailing list his office sends out), and in fact if we split the 48N and 48S and reduce the frequency of the 48N in response, it might make sense to send the 24 to Ballard and the 29 to Magnolia.

        I’m also a huge fan of local shadows and 1-mile-spaced expresses (peak or even all-day), because I enjoyed that standard of service for years on the New York subway. Taking an express to fly between neighborhoods and then relying on locally gridded service is well-understood and common.

      2. (The 24 would serve the North Beach loop in my restructure. We would reorganize Magnolia service to eliminate double-backs up the bluff.)

      3. Kyle,

        Sadly, a year after tolling started, it’s clear that congestion on 520 (at least during peak) is pretty terrible. So I don’t think that idea will fly anymore.

        On a separate note, the reason that paired local/express service works in Manhattan is because of two reasons: sky-high demand (leading to mind-bogglingly low headways), and separate tracks. Dial down that demand (and increase the headways) even a tiny bit, and it generally turns out to be faster to take whatever vehicle comes first. In fact, it’s often the case that the chance of the express overtaking the local is zero, either by policy or technology (such as the lack of passing wires for trolleys).

        Therefore, you’re generally better off having a single line, with the 1/2 mile spacing and the shortest headways that you can afford.

      4. The patterns of transfers between expresses and local shadows only works when everything is really frequent – like “frequent” on the order of what you will never see outside of the New York Subway. Try to do that with 15-minute-headways buses and you’re probably best off taking whatever bus comes first, even if it’s a local and you’re headed to an express stop, or it’s an express that would require an additional 5-10 minutes of walking over the local.

        The primary real use of local shadows is to provide some basic functionality along the line for people who are unable to walk half a mile to the nearest express stop, yet are still in good enough physical shape to not quality for paratransit, all the while, still giving the bulk of the population a faster ride, in exchange for a longer walk.

      5. Prt of the problem is that people here think Manhattan’s headways are “mind-bogglingly low,” when other parts of the world would consider them the minimum standard of service.

      6. On the Lexington Ave line, a train comes every 2 minutes or less for a major portion of the day. Where in the world is that considered a minimum standard of service?

      7. So, I think we need to start with a minimum standard of frequency for core routes, of 15 minutes until 10pm, and 30 minutes night owl. I’d rather have 5 minutes daytime frequency, but you have to start somewhere. Then add runs (peak or off-peak) as necessary to address overcrowding. That seems to explain the 2 minutes on the Lexington Avenue line.

  7. Then I am sure you will agree, Link, with over 13 hours a week of scheduleless 15 minute+ frequency is also broken.

    1. Actually, yeah. The MTA in New York publishes schedules for all of its lines. It uses the “Every X minutes until…” format during periods where they need to use demand management; otherwise, they publish the exact times of trains, even when headways are shorter than 15 minutes.

      I think that Seattle is so excited by the idea of “so frequent you don’t need a schedule” that we forgot that schedules are actually useful. Once we get a handle on when demand management is useful and when it’s not, I’m hopeful that we’ll restore schedules for the latter periods.

  8. The lack of transit options south of downtown Bellevue is very clear in this video. Just lines passing by on 405, and a single line, the 240.

    Lots of suburban commuters living in Newport Hills, Renton Highlands, and Newcastle, with just a single line crawling though.

    1. But the other half is, that’s an overwhelmingly single-family area, and people don’t use the transit much even when it’s there.

    2. A lot of people bash Kemper Freeman, but you have to give him credit on one thing. He walks to work. He lives one block away from his office. He wraps circles around most everyone here in terms of green commuting.

      1. Yeah, why don’t more people have a luxury condo right across the street from the new shopping/office center they just built? People make such terrible choices.

    3. The two Newcastle local routes were just cut for total lack of ridership. They had more trips with 0 riders than with more than 0 riders!

      1. Route 240 is still there and I sometimes use it for hiking purposes, since some of its Newcastle stops are within a mile or so of various Cougar Mountain trailheads. It is possible to do a number of nice thru-hikes, where you take the 240 to a trailhead in Factoria or Issaquah, and end up at Issaquah Transit Center to take the 554 back.

        When I’ve ridden it (usually on a weekend), it wasn’t full, but it wasn’t empty either. At least half the people on the bus looked like school-age kids, just a bit too young to get a driver’s license.

        One often-overlooked benefit of public transportation is the ability of teenagers to get around town without their parents needing to shuttle them everywhere. This provides independence for the kids and greatly reduces burden on the parents.

      2. Yes, sorry — I was referring to the 219 and 925, which were oriented toward serving Newcastle, rather than just passing through it on the way between Renton Highlands and Factoria. The 240 does attract a number of peak Newcastle riders, although off-peak there are very few who aren’t just passing through.

      3. There is also the 114 which continues to pass through and attracts modest but consistent ridership to downtown on its few peak trips. See above for my Renton restructuring proposal which would have the side effect of adding trips to the 114 and shifting a few 111 riders to it.

  9. Anyone know what the number of toll-paying vehicles there are on hwy 520 each day? When I hear traffic reports, especially in the afternoons, it seems that 520 is nearly as busy as it was before the tolling began. Have people given up going the long way around via I-90? What about hwy 522 or I-405?

  10. Has leave from the back simply died as an initiative? It seems for the past two weeks, the percentage of people actually leaving form the back door is hovering around 10% (including myself) and nothing is actively correcting the behavior. I have not heard the messages on the buses and drivers never say a word to people leaving from the front, even if there is a line of people waiting to get on the bus (particularly infuriating on the few rapid ride trips I have taken and this has happened).

    Maybe it’s only my experience, but I think something needs to happen to change the ingrained behavior. People still tend to sit toward the front of bus by default, then want to leave from the front quickly, while disrupting the flow.

    I know we always say things will take time, but unless there is some active initiative to change behavior, I do not think we will see a change in riders at large.

    1. I’ve started leaving from the front when there’s nobody getting on and the driver has positioned the back door in a way that opens into a tree.

      Like at Leary at 15th on the inbound 40.


    2. In an ideal world, the best system is the one where, like on subways, people enter in the closest door and leave at the closest door.

      This is ideal for two reasons. First, in general, it’s the most efficient. If doors are evenly spaced, using all doors for entry and exit is the easiest way to ensure that people are well distributed across the vehicle, and to avoid the endless calls of “move to the back”.

      But also, it’s the most freeing. The more rules and procedures that people have to follow when riding the bus, the more they perceive the bus as something noxious and restricting. Every time a driver tells riders to “move to the back”, in an angry voice, some of those riders are probably thinking, “If I were driving, no one would be able to tell me what to do”.

      Now, having said that, there’s two obvious catches. The first is that most of our buses don’t allow all-door boarding. And the second is that, at certain times of day and on certain routes (or route segments), the patterns of loading and unloading mean that a unidirectional passenger flow (enter at front, exit at rear) will be more efficient.

      So I agree that rear door exiting should be encouraged, but I don’t think that front door exiting needs to be strongly discouraged. Let people do what’s natural.

      1. No it really does. Maybe we are just use to the willy nilly ways we get on and off the bus, but other locales keep this rule pretty well enforced because it is ALL about efficiency. The quickest way to load and unload a bus is to do it simultaneously. People get off in back while people pay in front, so few buses have the ability to pay at all doors so that point makes little sense to me. Its amazing how much time is wasted as people walk to the front. Sure if you are right at the front and it is crowded, just zip off. But people here make an effort to get off in the front, and it needs to be enforced, for the most part, because load/unload time is one of the biggest issues with out bus system.

    3. Anecdotally, I haven’t had a single operator refuse to open the back door for me during this pick. I also haven’t heard the exit-at-the-rear message at all since the first week.

      The ORCA push seems to be over, except for the cost to advertise Saars on some buses. As far as I can tell, the free ORCA offer seems to be perpetual, though the transit agencies don’t want to create that impression. But Saars remains nowhere near any major transit hub, unless you count a mile from RBS as a major transfer hub. Why not offer free ORCA at the Metro Customer Service office on King Street once their vending machine is (has already been?) installed?

      What are people’s experiences with whether change fumbling remains a menace to bus movement?

      1. I always assumed Saars was paying some discount rate for the ORCA cards as a marketing tactic to bring more customers in the doors. Perhaps the agreement with Saars is exclusive?

      2. Change fumbling remains a problem in the tunnel. Tunnel ops are still not perfectly smooth, especially in the morning (!). ORCA users are not the reason why.

      3. The amusing thing is that if you go into Saar’s, there’s no signage whatsoever about ORCA, at least there wasn’t when I went into the Rainier Beach one a few months back. Not just nothing indicating that you can get a free card, but not even any signs indicating that you could add funds to an existing card. The agencies really don’t seem to give a damn about promoting adoption.

    4. The practical question is, are riders choosing the back door when a significant number of people want to enter, when more people are entering than exiting, and when they aren’t sitting in the first few rows? Those are the cases where using the back door has the most benefit.

  11. An engineering flow analyst would conclude that the node called “Downtown Seattle” is a huge bottleneck and the system needs to be leveled across the rest of the network.

  12. Transit update from Pete Donohue: Subways could be out for days All under-water tubes linking Manhattan and Brooklyn are flooded – raising the possibility that subway service could be out at least in part for days. The 4, 5, 2, 3, A , C, F and R trains use the five tunnels under the East River. There also is flooding on the No. 1 line in lower Manhattan.


    Good thing people still have cars and surface transit!

  13. Getting off at the back can be hard if someone has any kind of leg, knee or foot problems, because of the clearance. The driver can’t always get the entire articulated bus next to the curb, especially if it’s on a hill. Next to a driveway can have the same problem. These people might not look physically challenged and may not be under normal circumstances, but may want to avoid problems, so they try to exit at the front if possible.

    I’d love to use the back exit but have had problems when I’ve tried to use it. Once when the Rapid Ride C stopped in front of the driveway of The Bridge pub at 35th and Avalon, I almost got off the back, but it looked impossible for me so I had to rush to the front.

    Now I try to stay near the front if possible. I wait for the next bus if possible. I want to comply and I know that it’s a problem and ticks some of you off, but I don’t have a car. Otherwise I’d drive so as not to clog up the buses.

  14. I hung out at Bay A in Westlake Station today for about an hour during peak.

    One 41 was crushloaded, and the operator kept it sitting there for about a minute, asking passengers to please push to the back three times, but there was no more room. The loading assistant informed the operator, and let him know there were more 41s coming right behind. The bus then had to take about 20 more seconds to transition to diesel mode to exit the tunnel.

    I saw no buses open their doors twice, nor runners let on once the doors were closed. I also saw no runners complaining. One who might have was told by a loading assistant that another bus was right behind. Slick!

    Every rear-door line was done well before every front-door line. However, most of the boarders at the front were using ORCA or flashing transfers. Sometimes, one of the loading assistants would wander to the front door to make passengers away of the rear-door option. In that respect, the loaders were useful, but are still not fully utilized.

    One of the loading assistants helped tie down a wheelchair. However, the wheelchair-bound passenger was let on last, after the rest of the passengers at the front had boarded, and taken up all the front seats, causing the assistant to ask several passengers to move.

    I detected no platooning errors. All buses, including 41s, were evenly spaced so as to not have more than one of a route number in any platoon. Only two platoons came between each train during the period I observed. Excellent! There were a couple pairs of 72/73 buses that arrived in the same platoon.

    At 4:57, I jumped on the train to do the time trial. Much to my surprise, travel time traversing the tunnel had actually gotten worse. Westlake to Stadium Station took just under eleven minutes. The return trip took just under sixteen minutes. Something went awry between my observation period and the return trip.

    So, let me offer a few suggestions, some new, and some a reiteration:

    1. Give first priority to passengers with mobility aids boarding at the front. The boarding assistant could direct the rest of the passengers in line to the rear door, and tell those paying with cash to be patient.
    2. Treat routes 71-74 as one route for purposes of distributing them among the platoons.
    3. Provide lightweight portable sandwich-board signs to the boarding assistant teams. The signs could say something like “ORCA and transfers: Please board at the rear door when you see this sign.”
    4. Redistribute the routes so as to better utilize Bay B. Consider also giving the 41 its own bay, taking up just one bus length.

    Let me get more specific with the math on point 4.
    Outbound northbound trips originating 4:30-5:30 p.m. at ID Station:
    41: 14 trips
    71: 2
    72: 2
    73: 2
    74: 3
    76: 4
    77: 4
    255: 7
    316: 4

    There are 35 Bay A trips during the PM peak-of-peak hour. Assuming two platoons between each train, the platooning algorithm only allows 32 trips per bay per hour.

    If Bay A were shortened to one bus length, and all but the 41 were moved to Bay B, Bay A would have 14 trips during peak-of-peak hour, while Bay B would have 28. That is tight, but unlike the current situation, doable.

    Alternatively, if the 71-76 were moved to Bay B, Bay B would have 20 peak-of-peak trips, while Bay A would have 22 trips. This provides plenty of cushion to keep northbound outbound buses from piling up at the tunnel entrance. This shift does not require waiting for the next pick. It just requires signage, voice messages, and a little retraining. The platform hours saved might be needed to keep Metro from busting its overload contingency budget.

    1. The second redistribution you advocate might well work. But rather than moving the 71-76, I’d move the 71-74 and 77. The 76 and 316 need to be at the same bay because substantial numbers of riders use them interchangeably to get to I-5/65th.

      There are two problems with loading wheelchairs first. The first (especially now that we have low-floor buses) is that it’s very difficult to keep people from boarding immediately after the wheelchair user boards and getting in the way as you’re trying to secure the wheelchair. The second is that it contradicts the behavior we’ve trained everyone to get used to, and will result in more passengers doing the wrong thing on the surface.

      1. Regarding the wheelchair boarding, there are two personnel at the front of the bus. The boarding assistant is securing the mobility aid. The operator is sitting and waiting, but she/he could be blocking the entrance until the assistant gives the all-clear.

        The fact that passengers are boarding at the rear at all confuses riders. The sandwich-board sign might make it more clear that that is the exception. Is your concern about the safety of letting mobility devices board first the rest of the time, and then having passengers board while the operator is still securing? In that case, standard procedure could be to close the door until securement is complete. That may seem rude, but the operator could still say it is for the safety of the passenger being secured. How is this handled elsewhere?

      2. This is typically handled the way Metro does it: allow everyone else to board, and board the wheelchair passenger last. No, it’s not perfect, but it’s easier than dealing with boarding passengers or closing the door (which will generate a lot of customer complaints). People get antsy when they’re trying to board and are a lot calmer and easier to deal with once they’re actually on the bus.

        When I drove, I would typically remind people as they entered not to sit in the securement area, or, if they did, just ask them to move once I got up to raise the seat and prepare the belts.

      3. Okay, how about a voice message recording, played from the bus:

        “Please board at the rear door unless you are paying with cash, to help make boarding faster. Thank you for your assistance in keeping Metro moving.”

        And for more general use:

        “Please leave the front seats open for riders who need to sit at the front. Thank you for your courtesy to your fellow passengers.”

        If that doesn’t work: “Hey, you see that 8-month pregnant lady? She would like to sit down. Don’t be a jerk.”

        The bay distribution you suggest seems fine to me, since the split is as even as possible. FWIW, there are only 15 inbound northbound buses 4:30-5:30 pm. Of course, distributing them is less of an option. They arrive when they arrive, and we can’t keep the passengers waiting. So, I’d actually be happy with just moving the 71-74, leaving a 26/16 split, which means all but six platoons (out of sixteen) would need two Bay A buses.

        For long-term pondering: Consider the effect of reducing Link headway to 6 minutes with the opening of U-Link. If only two platoons fit between each train now, then only one platoon will fit between each train after 2016. Even removing the 71-74 and pretending bus capacity won’t be increased, we’d still have 48 northbound tunnel buses and 41 southbound buses in the DSTT at peak-of-peak hour, when even a perfect platooning algorithm (and lucky spacing of inbound buses) would only allow 40.

        Or, we’ll just have to run four-car trains at 7.5-minute headway.

  15. I made up a map to try to understand David L’s proposal. I’m not going to try to take a picture of it, but here’s a list of routes sorted by origin. Apologies for any mistakes in the segments; it’s hard to keep track of them all in my head. CAPITALS indicate Link station (L), Sounder station (S), or transit center. RR means RapidRide.

    107: (L,S) DOWNTOWN – Georgetown – (L) RAINIER BEACH – South Beacon – RENTON.
    150 RR: (L) RAINIER BEACH – Southcenter – (S) KENT – (S) AUBURN.
    169 RR: (L) RAINIER BEACH – MLK Way – RENTON – S Renton P&R – East Hill – (S) KENT.
    105: (L) RAINIER BEACH – Renton Ave – RENTON – Renton Highlands.
    180/164: BURIEN – (L) SEATAC – (S) KENT – East Hill – Green River CC.
    166/168: BURIEN – (L) KENT-DES MOINES – (S) KENT – East Hill – Covington – Timberlane – Maple Valley.
    151: (S) AUBURN – SE Auburn.

    Sounder: (L,S) Downtown – (S)Tukwila – (S) Kent – (S) Auburn – (S) Tacoma.
    157: (L,S) DOWNTOWN express – (S) KENT – Lake Meridian.
    159: (L,S) DOWNTOWN express – (S) KENT – Covington – Timberlane.
    144: (L,S) DOWNTOWN express – S Renton P&R – Renton Highlands.
    142: (L,S) DOWNTOWN express – S Renton P&R – Lake Kathleen.
    143: (L,S) DOWNTOWN express – S Renton P&R – Maple Valley – Black Diamond – Enumclaw.
    160: (L,S) DOWNTOWN express – Tukwila P&R.
    Unchanged: 114, 161.

    My only comment is, what about exchanging the 180 and 166? I’ve always thought that SeaTac – Kent – Covington is east-west grid-correct, and then Des Moines – Kent – Green River CC would both be southern.

    1. Nice summary! I’m increasingly wishing for a lesson in mapmaking from Oran. The only mistakes I see are in the peak:

      1) the 143 would terminate at Black Diamond, as it does today;
      2) the 144 would go to Fairwood after S Renton P&R, not Renton Highlands, and then double back via Highway 169 to terminate at RTC (to take advantage of a deadhead route; and
      3) the 160 would go to Southcenter after Tukwila P&R. But your mistake seems better to me than my original idea. Tukwila P&R is the only place where there will be more demand than can be satisfied by my revised 159, and Tukwila P&R is very close to South Base, so 160 trippers starting from there should be cheap to run. There is not much peak-direction demand between Southcenter and downtown.

    2. Oops — forgot to comment on the 180/166. I chose the 180/164 combination based on current and potential ridership, which is higher than on the 166/168 combination. The 180/164 might justify running some artics. But the plan would work just fine with 180/168 and 166/164 combinations.

    3. New one-seat rides
      Renton – MLK Way – Rainier Beach
      Renton Highlands – Renton – Rainier Beach
      Auburn – Southcenter
      Burien – SeaTac – Green River CC
      Burien – Highline CC – Maple Valley

      Lost or downgraded one-seat rides
      Renton – Downtown (now slower)
      S Renton P&R – Downtown (now peak only)
      Kent – Downtown (now peak only)
      SE Auburn – Kent – SeaTac – Burien
      Green River CC – Maple Valley (was evenings only)
      Burien – Highline CC – north Benson Road – Renton

      1. Don’t forget the biggie of new one-seat rides: Kent East Hill – Link stations (both Airport and RBS).

        Something that would enhance this restructure and restore the Kent-Downtown ride would be the 594/577/578 restructure we’ve seen proposed by several people here. Add a Federal Way stop to the 594 and reroute the 578 via Kent to be a true off-peak Sounder shadow.

        Highline CC to Benson Road is indeed a bit of a loss, but I think Auburn-Downtown and East Hill-Link are more important.

  16. Busmageddon on 3rd Avenue right now. I can see most of the length of 3rd between Seneca and Olive out my office window. There is a disabled D60 at Pine, and at least 25 buses lined up behind it trying to get around.

    Operational issues aren’t limited to the tunnel…

    1. At least I only has one disabled bus issue today – 19 broke down in the Village, and so the dedicated bus riders of Viewmont tromped en masse to a happily trailing and very empty 31 once the driver declared things hopeless, doing the extra loop around the Village before getting to 28th & Blaine just in time to catch the 24 had the 24 been on time rather than 8 minutes late. After some waiting around, we provided a giant lump of SRO.

      ‘Course, the 19 returned the favor in the evening, as the 24 6 minutes ahead of it was at least 12 minutes late instead of the usual 5, so plenty of SRO to go around…

    2. A couple of days after the end of the RFA, during the evening rush on 2nd avenue a tow truck clipped mirrors with a CT bus and promptly pulled over and parked in the bus lane at 2nd + Seneca so he could talk to the driver and call his boss. This backed up buses on 2nd as far as at least Stewart. After about ten minutes the guy finally moved his truck after a lynch mob of wet and angry 212 riders began to form.

      1. I see that the six upstairs ORCA loading assistants have been reassigned to other duties. Perhaps it would be wise to assign a roving team to go to bottlenecks, wherever they may occur, and expedite the buses trying to get around the bottleneck. If a bus blocks a zone, forcing other buses to pull up one by one, it makes sense to try to double the boarding speed.

  17. I was at a bus rider meeting tonight in Magnolia, and apparently some night 24 service has been promised for February, as well as a study in 2013 of Magnolia to look at rationalizing stop placement.

    Also, apparently 24 reliability has overall gotten -even worse- since the restructure. Which is pretty impressive.

  18. The 301, 306, 308, 312, and 522 are headed back to Pike and Union (couplet), with a 4th Ave stop moving a block, effective next Monday. The Renton Reporter appears to have scooped the Metro website.

  19. I rode the 60 today during peak. Nobody got on or off at the VA. I wish we could eliminate the VA knot from the 60, as Metro originally proposed, and donate the platform hours to more frequency on the 50, which would provide frequent front-door service to that stop without taking ten minutes to serve it.

    I hope Metro is tracking 50 vs. 60 usage of the VA front-door stop. If most of the usage is coming from the 50, the above suggestion is a serious one.

    1. Actually, the hours that could be saved on the 60 during mid-day, it turns out, would be almost enough to get the 50 down to 20-minute headway mid-day.

Comments are closed.