Yesterday, I had a chance to discuss with Metro staff the details of the revised Fall restructure proposal that was released last week. From that discussion,  a few clarifications to my original post about this restructure arose:

  • RapidRide D’s schedule did not change at all, but the headways listed in the public documents were changed to more precisely conform to Metro’s time period definitions. In particular, the D Line will operate every 15 minutes until at least 10:30 PM, seven days a week; and every 10 minutes in the peak direction from 6-9 AM and 3-6 PM (slightly less in the reverse-peak direction). In the common segment of Routes 15 and 18 from Downtown through Queen Anne and Interbay to Leary, riders are losing Monday-Saturday midday frequency (10 to 15 minutes), but riders on 15th Ave NW are gaining full-time frequent service.
  • Adding back a handful of trips to the 15X and 55 was indeed done primarily to save money on RapidRide coaches that would be used only for one or two trips a day. Metro considered operating standard coaches on RapidRide routes, but decided that would dilute the RapidRide brand more than restoring those trips. In the case of the 15X, it also served to restore coverage to a pocket of Blue Ridge. At some point in the future when Metro has more money, these choices could be revisited.
  • I was sloppy in my discussion of the stop-level data for Route 37. Of the inbound riders on that bus, 20% are at stops shared with Route 56 in the Alki area; 48% are between Alki and the West Seattle Bridge, who have access to Water Taxi shuttles; 32% are south of Hinds St. The last number includes a cluster of riders around Beach Dr & Carroll St, who, depending on how far you believe people will walk to useful transit service, may or may not be considered cut off from the from the bus network if the 37 were deleted. I remain unconvinced either that this route is viable in terms of Metro’s performance criteria, or that the small number of city residents who will unquestionably lose service (west of Me-Kwa-Mooks park) provide enough “geographic value” to offset performance figures that will surely be terrible.
  • The increased midday headway on the 11 was indeed a typo, and never planned in this restructure.
  • Also, within the next couple of weeks, I will elaborate on what I meant when I opined that the extended Route 1 was a “mess”, as well as a modification to that extension which would, I believe, be more useful to the riders served.

Finally, this isn’t quite in keeping with the topic, but I can’t find anywhere else to put it: Metro’s application for a TIGGER grant to help pay the costs of electrifying Route 48 was, unfortunately, not funded. This project was already in jeopardy due to the failure of Prop 1, which would have provided the local matching funds.

Thanks to those Metro staff for taking the time to meet and discuss these issues, and answer my many questions.

41 Replies to “Clarifications on Fall Restructure”

  1. Did you ever get clarification on Route 31? It would be a real shame if they abandoned the Dravus St service.

      1. Ugh. You can walk from Fishermen’s Terminal to the stops on the east side of 15th faster than buses can cross that overpass a lot of the time.

        Anyone who works a Terminal sort of job can easily manage the 0.4 mile walk over to Gilman as well.

        The Dravus routing A) avoided the worst of bridge back-up traffic; B) helped to create grid-structured service at Interbay’s busiest transit stop (15th/Dravus); and C) was more direct between the 31’s major demand points, period!

      2. What’s more, the Dravus routing would have made RapidRide-to-Nickerson transfers (relatively) easy. Using the Emerson overpass in both directions, such transfers will be forbidding.

      3. Yeah. That routing down Dravus made a lot of sense. The routing around Ballard Bridge always makes the 31 unreliable, especially in the afternoon. Traffic from 15th to Nickerson usually keeps moving even when the bridge is up.
        And you’re right d.p., transfers at the Emerson Street overpass are a pain. You have to walk under 15th though an overgrown, poorly lit area. And coming from Nickerson the closest stop with a crosswalk is Nickerson and 11th, and it adds like 10 minutes onto transfers to a SB 15th Ave route. Transferring at Dravus would have been much simpler.

        Bruce, any idea why they chickened out?

      4. Poor pedestrian access conditions on Gilman and 20th, I’m told. I’m not familiar with the area, so I can’t elaborate.

      5. There are no sidewalks on the East side of 20th and on certain parts of Gilman.

        That said, the New 31 was supposed to run south to Thorndyke, while the New 24 ran north to Gilman. Does that mean the New 24 is now off Dravus and on Emerson too?

        If so, oy!

        Just run the things on Dravus to 21st (if it’s wide enough) or 22nd (which already has buses on it). Either of those would serve the neighborhood just as well or better than 20th anyway!

      6. Sure there’s poor pedestrian access on those roads in Magnolia, but with little traffic and wide shoulders it’s much easier to navigate than the pedestrian black hole that is the Emerson St/15th Ave W overpass.

  2. I’d like to mention here again: North Delridge, a very transit-dependent, high-ridership corridor, will have a net loss of service under the revised changes. No new service hours on the 120, no new route 40, no diverted route 128, though all three of these improvements were in the original proposal. About the only thing kept from the original proposal is the removal of the 125 to downtown on weekends. This net loss of service in a zero-sum restructure is unacceptable for one of the highest ridership corridors in the county.

    1. The failure of governments to speak with each other and to properly outreach to the neighborhoods is also frustrating… Jack Latteman indicated that Metro can’t run the route 40 up Genessee because the funding for a traffic signal at Avalon on top was in the failed Prop 1. On the other hand, SDOT has indicated that it is funded – although even here, one person was told funding is in place for this year, and another was told in 2013.

  3. Did the Metro people explain why they dropped so many changes that so many of us transit nerds were so excited about? Like, why is the 355 still going to make that ridiculous detour through the U-District, instead of being streamlined like they had initially said? Was there really that much negative feedback to override their data-based planning?

    1. I suspect it’s been pushed back to the E Line restructure, along with the Wallingford stuff.

    2. Because in spite of all the traffic, getting from Greenwood to the U-district is still much faster on the 355 than on the glacially slow and unreliable 48. The 48X was a bit better, but that’s going away, leaving the 355 as the only decent transit option for that trip.

      As to riders going downtown, doesn’t the 358 today take the route you would prefer the 355 to take (except for the Linden Deviation)?

      1. I’ve ridden the 48X a few times and one thing I have to say about it is that it’s definately crowded. Enough so that travel time from one end of the U-district to the other is actually slower than the regular 48, negating some of the express’s time advantage.

        The thought of all those people being forced onto the 48 local seemed awful – it made me glad I no longer live somewhere where it would matter.

        We do, however, have a new express alternative for some of the 48 riders, and that’s the 542’s tail between the U-district and Green Lake P&R. I’ve seen a few people use the 542 for this purpose. I suspect more would do this if it were better publicized.

      2. Yes, the 358 does run along Aurora where they had proposed to run the 355. But it stops a bazillion times, whereas the 355 wouldn’t have. In addition, there’s the “intangible” of riding a bus that is not filled with crack heads, and, for folks living on the west side of Greenwood, that is an awfully long walk to get to the 358 on Aurora.

        I’m not saying the 48 is great service–it’s not. I’m just saying, is it worth the detour given the ridership.

        Also, if they have to keep that route going to the U-District, couldn’t they at least come up with a routing that stays closer to the freeway? I know they’re detouring to get to the express lanes entrance, but the time spent getting over there has got to outweigh whatever they save by taking the express lanes instead of the main line.

        I agree about people taking that bus more to/from the south end of downtown than up to the shopping district. I don’t think it needs to run the whole length of 3rd either.

    3. I’m a regular 355 rider on that detour. I did not comment on the proposal, as I have other options.

      As for the 355 detour being “ridiculous”, from looking anecdotally at ons/offs, it appears to me that the majority of the 355 riders have destinations in south downtown, where using the express lanes actually does buy you something in terms of speed. Quite frankly, I wonder why they don’t end the 355 in south downtown and not run it north on 3rd.

  4. 15-minute headways weekday daytimes are disappointing on RapidRide. The standard should really be 10-minute headways. Also, adding a few express bus routes creates complexity for riders. I don’t think it would matter if some trips were operated by different coaches. With the various bus wraps and occasionally an ST coach operating an MT route and v.v. it is not the end of the world if occasionally a different coach operates RR, and it will probably happen anyway.

    1. It should. Central Link set the standard of 10-minute minimums, so the concept finally has its nose in the door. And people who move to Rainier Valley or will move to future station areas are swayed partly by the 10-minute frequency. Someday when we have more money we can upgrade all the RapidRide corridors to 10-minute minimums.

    2. Or at least 12 or 13 minutes. That is the true and absolute threshold for “so frequent you don’t need a schedule.”

      Subway lines in Boston and Washington are able to drop to the 12-13 range — though never before 7:45 at night — without riders consulting a schedule. And people do start getting edgy around minutes 10 and 11, though real-time arrival info helps to take that edge off.

      It’s actually sort of funny to obsess over choosing between two round numbers for a schedule-less line. When you become accustomed to actually not checking your watch as part of the transit-riding process, the roundness of the number shouldn’t matter.* Meanwhile, on the agency’s end, a 60-minute hour divides just as easily by 12 as it does by 10 or 15.

      *(In the daytime, when demand warrants, high-volume subway systems routinely push towards other non-round numbers — 9 minutes, 8 minutes, 7 minutes. At that point, every minute closer between trains benefits the customer experience; under 10 minutes, no one complains about not being able to set their watch by it.)

    3. We absolutely can afford it. We just need to consolidate nearby service.

      Today, there are 6 buses an hour, all day, on 15th NW. There are 8 buses an hour, all day, on 23rd (43/48). There are 9 buses an hour, all day, on Broadway. There are 6 buses an hour on 15th E and 19th E (10/12) combined, and 6-8 buses an hour on Pine, all day. There are 6 buses an hour from downtown Seattle to Alaska Junction via the Viaduct.

      If we decided it was a priority to have 10 minute service, we could have it tomorrow. Some route pairs (Capitol Hill to Montlake) would require a transfer, and some corridors (19th Ave) would lose service entirely. But it’s almost unquestionable that most people would be better off with such a change.

      1. I question that most Capitol Hill residents would be better off with those changes. They would require getting rid of the 43, which is one of the highest ridership buses in the system. Your plan doesn’t just get rid of a direct Capitol Hill-Montlake bus, it also gets rid of a direct Capitol Hill-U District bus for anyone living near Group Health. As a consolation prize you say we will only have to wait for 8 minutes for a transfer on 23rd Ave E. Thanks.

      2. Yep, Morgan is exactly right. You’d have to wait 4 minutes for a bus to take you from 15th to 23rd, and another 4 minutes for a bus to take you to the U-District. Incidentally, that’s about how long you’d have to wait to catch a single bus today. But in return, any other trip that requires a transfer — such as Capitol Hill to Queen Anne or Ballard or West Seattle — now involves a significantly shorter wait time.

      3. Metro needs to set the goal now, even if it can’t fully fund it in the foreseeable future. “10 minutes until 10pm, 7 days a week, between all neighborhood centers”. If it creates a performance metric for it, it’ll at least measure all future proposals against it. If it does nothing, we’ll have service dropping to 30-minutes at 7pm and all day Sundays forever.

      4. I can see some value in exchanging the Madison and John-Olive-Pine segments between the 43 and 12. Then reducing the 12’s frequency to 30-minutes and applying the hours to the 43 or 10. On the other hand, I see a lot of people travelling between Montlake and John, who would not be happy with Madison, and who would be really unhappy with the lack of a direct Link transfer on Capitol Hill or downtown. If only the DSTT had a Madison station!!! So I’m undecided on changing the 43.

      5. What are you talking about, Mike?

        Aleks is talking about killing 19th Ave service and the 43 as we know it, and making the straight portions of the 48 and the 8 actually frequent and actually work for the first time ever. Madison improvements, though good, are entirely unrelated to his example.

        And before you tangent more, yes, there is a huge difference between forcing a (painless) transfer for right-angle service — eliminating the uncoordinated 23rd overlap with half the buses going one way and half going another — and forcing an (inherently less convenient) transfer by enacting your silly single-corridor service-splitting methodology.

      6. What are you talking about, Mike?

        Aleks is talking about killing 19th Ave service and the 43 as we know it, and making the straight portions of the 48 and the 8 actually frequent and actually work for the first time ever. Madison improvements, though good, are entirely unrelated to his example.

        And before you tangent more, yes, there is a huge difference between forcing a (painless) transfer for right-angle service — eliminating the uncoordinated 23rd overlap with half the buses going one way and half going another — and forcing an (inherently less convenient) transfer by enacting your silly single-corridor service-splitting methodology.

        Aleks is talking about “no-brainer” transit: You walk to the stop and the thing comes. If you need to change directions, you get off where the lines intersect and the other thing comes.

        He’s talking about making trips easier and making total trip-times quicker, and pointing out that, with aggressive restructuring, we probably already have the hours to do so.

        While there are many kernels of good in the 2012 restructure, and while I want to give Metro encouragement for moving in the right direction, the plan is still saddled by a lot of complicated routes with questionable transfer utility. It will still take rocket science combined with clairvoyance to predict the fastest choice from point A to point B.

        That is what Seattle seems to be best at: taking simple problems and making them more complicated.

  5. The transfer from a southbound 48 to an eastbound 8 would be relatively painless. The transfer from an eastbound 8 to a northbound 48 would be awful. Wait for the light to cross 23rd. Then wait for the light to cross John. Then wait for the 48 on a god awful street. I would much rather wait 8 minutes for a bus along Thomas than for 4 minutes for a bus along 23rd.

    Two other points. If you run 8 buses an hour the average wait is only 3.75 minutes in fantasy land. In reality the buses bunch together so that the average wait is longer. With a schedule and 4 buses and hour you can probably get an average wait time of 6 minutes pretty easily. With a transfer you would in reality probably be 8 or 9 minutes. Add in the penalty to cross 23rd and John and it may be five minutes longer to get from Group Health to UW. More people transferring means more people getting on and off so the buses would be a little slower as well.

    Also this plan means a longer walk for a lot of people (near 19th and John or Olive Way) for a direct bus downtown.

    1. Wait for the light to cross 23rd. Then wait for the light to cross John…

      Or… you could just move the stops to facilitate the most common transfer patterns.

      With a schedule and 4 buses and hour you can probably get an average wait time of 6 minutes pretty easily.

      Yeah. Because that’s totally the way things are on Seattle’s semi-frequent (15-minute routes) now. [rolls eyes]

      Bus bunching on surface streets is a fact of life. But it is well-proven that the degree of deviation in timeliness from one vehicle to the next (one bus 8 minutes late, the next 2 minutes early) are exponentially reduced by adding frequency, as there’s less of a “cumulative delay” effect on any single vehicle.

      It may be five minutes longer to get from Group Health to UW.

      Option 2: Walk (or bus, if you have to) five blocks, take the train one stop. 10 minutes total.

      But I still disagree with you; Option 1 (8->48) would run more smoothly than the 43’s status quo.

      Ever lived anywhere with “no-brainer transit,” Jeff? People really can’t appreciate how easy it makes life if they’ve never had it before.

      1. I have lived in Rio and DC where my commutes involved “no-brainer” transfers. They were both much better than the 8 to 48 transfer would be.
        Those experiences are a large part of the reason that I think getting rid of the 43 would be idiotic, as painfully slow as it can be.

      2. Then you’re only envisioning the way things currently are on those two buses.

        Seattle isn’t special. No-brainer route maps and frequencies can work as well here as they do anywhere else.

  6. Where in the world does “no-brainer” transit involve 30 people getting off one bus, waiting on a god-awful street, and then getting on another bus?

    1. Yes, it is truly Hell on earth, isn’t it?

      30 people off one vehicle and onto another vehicle is perfectly commonplace in “no-brainer” transit. In some places, believe it or not, that even includes bus-to-bus transfers.

      All of your complaints about this transfer could be taken care of with high frequency, better stop placement, better lighting/shelters in recognition of its role as a transfer point, better walk-light timing, and maybe some flashing lights in the crosswalk.

      Switching to a transfer-based system will also require a major push for ORCA ubiquity and against cash payment.

      (p.s. When are there really 30 through-passengers on the 43, outside of rush hour? Most 43 trips throughout the day have about 12-15 passengers on them. And at rush hour, the transfer would be near-instantaneous.)

      1. I would like to know where in the US such a bus system exists. If such systems are as common as you say then would you be kind enough to provide an example.

        To keep the same frequency on 23rd and John you would have to double the frequency on the 8 and 48. There is hardly anyone on the 8 when I ride from Columbia City to Capitol Hill. I am not on the 48 south of John very much but it has never seemed like it needed extra capacity. When I ride the 43 during rush hour 30 people is a light load.

      2. Jeff,

        In Boston, the four subway lines, as well as a number of key cross-town bus lines, run every 10 minutes or less all day — often much less. (The Green Line runs every 2-3 minutes for most of the day.) Here’s a train schedule.

        In New York, every subway line has frequencies well in the single digits for most of the day, as do a large number of bus routes, especially ones which cross the subway, rather than running parallel. For example, the M15 Select (New York’s version of BRT) runs every 4-7 minutes for most of the day, only dipping below 10 after 9:30 pm. Here are schedules.

        In San Francisco, every rail line has frequencies of 10 minutes or below until 6pm. The key rail and bus lines have mid-day frequencies well below that. Here’s a full chart. Note that the real numbers are often better than the chart reflects. For example, Geary has both local and limited service until 9pm, which combine to provide 16 buses an hour until 9pm.

        In DC, the Metro lines run every 6 minutes on common corridors for almost all day. I’m not familiar with the city so I don’t know which are the important bus routes, but there are definitely a few trunks with very frequent service: here’s an example.

        However, if you really want to see the difference, just look at the maps:



        San Francisco



        Notice how the other cities’ maps show a nice collection of straight routes with simple intersections, while Seattle’s looks like spaghetti.

        We wouldn’t be the first city to abandon a downtown-centric radial network for a grid-based one. Portland did it in 1982. Tallahassee did it in 2011.

        Oh, and for what it’s worth, a truly grid-based system would probably also involve splitting the 8 into an east-west and a north-south segment. So while the east-west part of the 8 could justify 5 or 6-minute frequency for most of the day, the north-south part could probably make do with less. (But it would also connect with the 3, which should run every 5-6 minutes as well, meaning that it wouldn’t be any harder for riders on the southern portion of the 8 to get downtown.

  7. Most of those systems are rail based or have much greater frequency than 8 buses per hour. With 6 minute headways and the efficient boarding and unboarding of the DC Metro transfers work great. With 20 buses an hour and payment after you enter the bus (like in Rio) a transfer based system worked fine. With 7-8 minute headways and the painfully slow boarding/unboarding of King County Metro I am very skeptical that this system would be as good the current system. If you can find the money to improve frequency and ways to speed up the boarding that would be wonderful.

Comments are closed.