2 Express Diesel
Photo by Atomic Taco

While much of the debate about the possible Queen Anne-Downtown-First Hill-Madrona restructure I presented last week centered on the long tail of the 4S, some commenters raised good questions about current ridership patterns on Queen Anne, and whether abolishing local service on the northern tail of the 2 in favor of more service on the 13 left too big of a gap in all-day service at the top of Queen Anne.

If you’re not already familiar with the details of Queen Anne service, I recommend opening this Google map and this Metro map in separate browser windows, and examining on each the alignments of the 2 and 13. The Metro map gives a better sense of context, but the Google map will allow you to zoom in and see the street grid in more detail.

Data and analysis after the jump.

Ridership Patterns on Route 2

Ridership Chart for Route 2
Ridership Chart for Route 2

As always, an explanation of the meaning of these charts is available in my post about Route 36. Here’s what I see in the data:

  • 1st & Mercer stands out as an activity center. Other routes that serve this stop (or the southbound stop on Queen Anne & Mercer) include the 1, 8, 15, 18 and 30, making this an important transfer point.
  • Density drives ridership, which will come as no surprise to long-time readers. You can guess the density of the walkshed very well just by looking at the boarding activity at a particular stop.
  • People like riding buses up big hills, another shocker. Note the asymmetry between southbound and northbound boardings, in the section between Galer and Mercer.
  • Ridership is strong across the board until Galer St. After that, something interesting happens. Ridership to Downtown essentially disappears in all time periods. Northbound ridership almost disappears, except in the PM peak and evening, where on average a handful of people use the bus, some almost to the end, on every trip.
  • 6th & McGraw is unusually strong for northbound deboarding. This stop is only three blocks from the 13 and four blocks from the terminal of the 3, making me think that some riders are using the 2 (plus a short walk downhill) as a substitute for those routes.
Towards the end, I’ll examine and discuss even more comparative data for the tails of the 2N and 13.

Ridership Patterns on Route 13

Ridership Chart for Route 13
Ridership Chart for Route 13

Between Downtown and Galer St, this chart is basically the same as the 2N. Between Galer and Crockett, the 13 exhibits more ridership and boarding activity, although still an essentially residential pattern of loading up inbound and unloading outbound. The service is little-used past McGraw.

Ridership Patterns on Route 2X

Ridership Chart for Route 2X
Ridership Chart for Route 2X

Operating only the peak, the 2X behaves as you’d expect from an commuter variant of the 2N, with only the 1st & Mercer stop showing any significant churn of riders. Interestingly, the AM trips to Downtown are stronger than the PM trips to Queen Anne, all the the way to the northern terminus. This goes against the general observation thet the PM peak is usually stronger than the AM peak.

Comparing the Tails

It’s evident from the chart that the tail of the 13 (i.e. north of Galer) is utilized rather more than the tail of the 2N, but the raw data allow me to quantify how much. This table summarizes the numbers:

Tail Ons+Offs and Load table

Some caveats about these results:

  • If you look at the schedule for the 2 and 13, you’ll note that they have similar frequency and span of service (if anything, the 2N has more service than the 13) so this should be an apples-to-apples comparison.
  • I have defined the tail of the 2N to begin at the second stop on Galer after leaving Queen Ann Ave, as the stops adjacent to Queen Anne Ave are extremely close to the 13.
  • I have omitted values for times when the 2X operates, as ridership loss there from deleting the 2N tail would presumably be insignificant. I have also eliminated early-morning service, as ridership on either tail is insignificant.
  • This analysis slightly overstates the importance of the 2N tail, as some stops near the end and on Galer are still very close to stops on the 13.

The Value Judgement

Ultimately, whether you think it’s right to delete the 2N in favor of (more than) doubling the frequency on the 13 comes down to whether you’re willing to make a smaller number of people walk further than most to the nearest bus (and make small fraction of that group switch to paratransit) in order to simplify and streamline the bus network and make it more convenient for a rather larger number of riders. My generally-utilitarian philosophy, the existence of other distantly-served little pockets of the city, and the compelling need to reduce costs and drive ridership suggest to me that it’s the right thing to do.

74 Replies to “Why Delete Route 2N?”

  1. I think this is another one where you might want to throw up a map of the two routes before there are howls of protest. :)

    That said, I don’t have anything personally invested in these routes (and have only ridden them to Seattle Center area) but wonder: if they are so close together on Queen Anne hill, why is it? Are there some very steep gradients they are separated by? It’s not obvious to me with a look at the route maps because I don’t know that area of town.

    Merging frequency on these routes sounds like it would be awesome: I see so many 2 buses downtown when I’m waiting for other buses that I’m imagining the resulting frequency would be close to every 5 mins (peak already seems to be sub-10 if you count express and local together).

    1. The problem is that Metro’s individual route maps aren’t to scale (and scale is the crux of this issue), Metro’s systemwide map is too visually busy to excerpt usefully, and I’m not going to ask Oran to make a map for every one of my posts, as they take hours to make and he presumably has other things he needs to be doing with his time. This post also went out on somewhat-short notice, and things might have been different if the piece had sat in the queue for a while.

      In the piece I provide links to a Google Map and the Metro system map, and recommend people familiarize themselves with the context. I do think this is good advice :-)

      1. Yes, but people are laaazzy! (Me included!)

        I understand why system maps aren’t to scale but it’s really inconvenient to not have real world route overlays. Actually it can’t be too hard (or at least it’s a solved problem) – onebusaway overlays routes onto google maps in the iPhone app at the least. All we need is multiple routes overlain at once. Hrm.

      2. Onebusaway’s current interface won’t let you show two routes at once but if you zoom in without searching for a route you can see the stops only a block or two apart. Interesting.

      3. Matt, how steep is the hill between 6th and 10th (i.e. the 2 and the 1), practically speaking? I’ve done a bunch of elevation data point-checks, and it seems like it’s no worse than the gap between 6th and QA, but it’s highly possible things are different on the ground.

      4. It’s quite steep. The street to watch is Highland, that turns into 8th. That’s the “Queen Anne Boulevard” that starts at the view park on the south and loops around the hill at roughly a constant elevation. It has such nice views because it’s a steep drop on the downhill side, sometimes looking like this.

    2. “If they are so close together on Queen Anne hill, why is it?”

      Because streetcars did it a hundred years ago, when Queen Anne and Seattle were much different places.

      “Are there some very steep gradients they are separated by?”

      Click the “Terrain” link in the Google map to see the gradients. My recollection is that 6th W to QA between McGraw and Galer is relatively flat and easy to walk, but some others have disagreed. The steepest hills are between Queen Anne Ave and Taylor Ave.

      1. The top of Queen Anne Hill is basically flat. It was made that way by the developer. There’s small sag at 3rd between 6th West and Queen Anne, but it’s nothing to complain about.

        I used to live at 6th West and West Galer when there was no 13, only the 1, 2, 3, and 4. The 2 did a land office business in those days. Grant, most of it was on the counterbalance, but folks did ride the “tail”.

        That was in the days of the tensioned overhead; you could hear the wires singing before you saw the lights of the bus. It was nice.

  2. how would removing bus service from whole swaths of a neighborhood that is at the top of a large hill help anyone? remember Queen Anne hill isn’t exactly a flat area

    1. That’s my question too. But Bruce is asserting that for most of their routes on the hill, they are very close together. Are they not actually due to hills? Where exactly? I don’t know the area well…

      1. On Galer west of Queen Anne Ave, there is a moderate depression centered around 3rd Ave W. Heading north, it quickly flattens out.

    2. Yes, I used to live up here, and it’s essentially flat in a large area from 4th Ave N on the east to 8th Ave W on the west, and from McGraw on the N to Galer on the S (plus various small pockets north of McGraw around the cemetery and near the tail of the 3.) If you can get to the top of the hill it’s actually a fairly easy walk for a healthy individual. The challenge is that Queen Anne Ave is the commercial core so it makes sense to keep service there, but it’s unfortunately not ideally centered on the flat area. The 1 covers some of the west area, but there is a fairly steep, if short, hill between 8th and 10th Ave W.

  3. I have to recuse myself on this one. This week I failed basic knowledge of the 2 by riding from Queen Anne to Virginia, then getting off and walking up First Hill (my destination). Onebusaway’s map showed it ending at 3rd and Pine and I forgot it is supposed to go to First Hill… I realized my mistake when I encountered the same bus again at Terry and Spring.

  4. Bruce, you mention “Density drives ridership” but didn’t make the obvious connection: the 13 travels on Queen Anne Ave N, a much denser corridor than MT 2’s 6th Ave W. So a more frequent 13 might not only be able to serve 2N riders willing and able to walk a few blocks, but also capture a higher share of trips in the denser location.

    1. Yes, I failed to mention that. Queen Anne Ave has a couple of supermarkets, lots of restaurants, and quite a bit of new residential mixed-use midrise.

  5. I’ve been on rt.2 and it’s a very empty ride north of Seattle Center. Keeping 2X is a good idea for the commute. Having rt. 13 run every 15 M-F (and maybe mid-day on Saturday)is better for riders going to Queen Anne and SPU.

  6. Excellent analysis as usual Bruce, though not the “disemboweling” I was anticipating.

    Something that jumps out at me is that on the 2 SB, the only activity along 6th south of Crockett appears to be alightings, not boardings. Are people really using the 2 to travel a few blocks on top of the hill, or is there something else at work here? But if true this assuages my main concern, which is boarding/alighting in the vicinity of 6th & Galer. People along the 2 north of McGraw are 3-4 blocks from 3rd Ave W, which under this proposal would have 7.5 minute service all day and evening. They’re also 3-4 blocks from the 1 along mostly flat terrain. Those folks have it good.

    But someone at say, 7th & Garfield would have a choice: about a half-mile walk to the 13, which would be a 10-minute walk to a 15-minute service – barely an improvement over 30-minute service a block away – or a 200-foot climb/descent via a series of staircases to the 1, which would be difficult for the disabled/elderly and probably impossible for anyone requiring a mobility device.

    (And before someone mentions Access, read the very apropos post on Human Transit.)

    It does seem that the service on 6th Ave W would be missed by few, and so I think this plan is workable. There’s going to be some folks who raise holy hell about it though. It’s also going to be a challenge that these routes are among the most cost-effective that Metro operates, and so the cost-cutting argument will ring hollow for some folks who will point to much less productive service elsewhere.

    1. With Bruce’s numbers, you can’t possibly argue that the 2N/4N/4S/etc. are cost-effective because of their long tail. The segments closer to downtown are what makes these routes so good, and none of the plans floating around will change that. (Honestly, I think it’s a shame that Metro’s performance data isn’t broken down by corridor. I’d love to see the data for the 2/13 shared segment… I have no doubt it would blow the 2’s numbers out of the water.)

      Anyway, I actually think that proposals like these would help our cause. The first thing anyone says when you try to cut a route like the 42 is, “You’re just picking on poor people”. But if you’re cutting routes in Queen Anne at the same time, now it’s *their* argument that rings hollow.

  7. On top of Queen Anne, where I live, the distance between 6th Ave. W. and Queen Anne Ave. N. is about 3/8 of a mile. It is 6 blocks, and those blocks are about 110 yards long, including the width of the streets between the blocks.

    I see the #2 bus on 6th W. in the afternoon all the time, and there are never many people on it in the afternoon, and often NObody on it in either direction. There are usually a few people on it along W. Galer in both directions.

  8. I used to live on top of QA and I’ve used bus service up there many times this summer. Most of the ridership on 13 is to and from SPU, which would be impacted if Metro moves or adjusts route 17, so more service from SPU to downtown may be needed. My suggestion would be to delete the 3N tail and move that service to SPU. SPU would have the 13 and 3N for connections downtown. If the 3N neighborhood wants peak hour service, the 2X could begin/end on that loop.

    The idea that deleting the 2N to provide better service on the 13–if people are willing to walk a little further–overlooks one fact: the option of walking to the other bus already exists. If I’m living along the 2N route and I realize that “Oh poop, I just missed the 2, I guess I’ll just walk 3/8 mile and catch the 13.” I already have that option. Is it an improvement to make that option mandatory? No, I think in the long run, people will choose to not walk and drive somewhere. Also, it’s not a given that all of the 2X ridership is generated by commuters only. How much of that ridership rode the 2N one way and is using the 2X for the other way?

    1. “Oh poop, I just missed the 2, I guess I’ll just walk 3/8 mile and catch the 13.”

      If they’re willing and able to sometimes walk 3/8 of of a mile to a service that runs every 30 minutes, they’re most likely willing to walk the same distance to a service that runs every 15 minutes, and many more people are in the unique walkshed of the 13 vs the 2. Only in one small corner of Queen Anne (around 7th & Galer) does the walk to the 13 (or the 1) suck, and the lost ridership there will be more than made up by increased ridership at lower cost elsewhere.

    2. And what happens if, while walking that 3/8 of a mile, you miss the bus you were trying to catch?

      Your idea overlooks another fact, which is that service on both of these routes currently sucks. The 2 and the 13 each have 30-minute service for most of the day. The 2 has 20-minute service for a small part of AM peak, and the 13 has 15-20 minute service for both peaks. But that’s it.

      If you want to have the choice of either bus, then you have to walk to QA/Galer. For many riders, that’s a much further walk than 3/8 of a mile.

      During peak, this restructure means that you can walk 3/8 of a mile (at the absolute worst) and have a bus which comes every 10 minutes. That means, without needing to check a schedule, you know that you’ll be waiting about 5 minutes on average.

    3. You’re missing the riders who don’t walk to the 13 because it’s half-hourly and they don’t know if they’ll just miss it. Consolidating two routes into one gives a predictable, evenly-spaced headway at all times. When two routes overlap, they may be evenly spaced part of the time but not at other times, and you have to study both schedules to find out. Plus, one route is subject to different delays than the other, which also throws the spacing off-kilter.

  9. As an expert 2(N), 2X, and 13(N) rider, I have some comments.

    1. “AM trips to Downtown are stronger than the PM trips to Queen Anne” This is due to the wonderful 3rd & Pine, late buses, and probably OneBusAway. On your way to downtown, you get a choice of buses, and unless you’re at QA & Galer you have to stick to your choice. Bus is late? Too bad, you’re stuck waiting. But on the way home you get your choice – just check OneBus. 2X isn’t coming for 15 minutes? Take the 13 and walk a little. Or the 4 and walk a little more. Or the 3 and take a stroll across the top of the hill. Or take the 1, 15, or 18 to the Met, do some shopping, and take the first 2 or 13 that arrives.

    2. [Matt L] “Those folks have it good” Yes, we have a lot of choices. But remember it usually takes at least a half hour to travel the 2 miles home from downtown. So it’s not all roses.

    3. QA Ave on the top of the hill is not the most reliable street. There are several stop signs that are often backed up for half a block or more, and people are pulling in and out of angled street spaces for shopping. 6th, on the other hand, is a wide and fast street. It could actually use a serious road diet (the speed bumps help, but speed bumps are annoying).

    4. The numbers on top of the hill don’t look that great. But remember you’re looking at numbers spread among frequent service and 5 seperate bus lines. In agregate, there are a lot of bus riders on the hill. That said, for purposes of bus planning, I admit it wouldn’t be a terrible idea to reduce the number of bus lines to bump these numbers up.

    5. Kill the 2, or the 13 (and modify the 2), or the 4. Do not do more than one of these. The proposed cuts killed both the 4 and the 13, leaving people on the east side with a very very long walk to a bus, and a very very long walk back with a very steep hill in the way. That’s just mean.

    6. If you kill the 2, you leave some people with a very long walk to the bus. See Norman’s comments. But ignore the bit about nobody being on it – QA is the very beginning and end of the line – of course the bus looks less full than its midpoint downtown.

    My recommendation: I lean toward the idea of modifying the 2 to get down the SPU, and killing the 13 instead. Then you get the benefit of a faster, more consistant 6th Ave W. QA Ave would still be served by the 3 and 4, and you don’t increase walking distance to almost anybody.

    1. Oh, but of course combined with a gondola to either the Seattle Center or SLU. That would serve the QA core, and probably anyone that can walk there since it would cut their travel time at least in half.

      1. Or if Puyallup is not done with the Skyride yet, bring back the Bubbleator and rebuild it as a funicular down Galer. Doesn’t MOHAI still have that?

    2. “QA Ave on the top of the hill is not the most reliable street.”

      That can be fixed with bus bulbs. I will write about this more in future.

      “The proposed cuts killed both the 4 and the 13”

      Huh? The proposed cuts killed the 4 and 2, extended the 3 and 2X to SPU and boosted frequency on both the 3 and 13.

      “Then you get the benefit of a faster, more consistant 6th Ave W. QA Ave would still be served by the 3 and 4, and you don’t increase walking distance to almost anybody.”

      Sorry, that makes no sense. You’re putting resources where they’re least used, making people on the dense main drag of Queen Anne either backtrack to Crockett or walk to Galer, keeping the pointless turnback loop of the 4, while providing front-door service to the less dense part of West Queen Anne where people aren’t using it.

      1. Sorry, I mixed up the 2X with the 2 on the cuts.

        “You’re putting resources where they’re least used” Not really. I agree in theory we’d have high ridership on the QA strip. But look at your charts. You have about 7 people on the 13 before QA and Galer (where the bus lines meet), compared to 6 people on the 2. On the way back you have 10 people on the 13 vs. 8 on the 2. Sure, the 13 performs slightly better, but SPU makes up all of this difference and more. Connect the 2 to SPU and it outperforms the 13.

        That said, we’ve been adding a little density on QA Ave (yawn, 4-5 stories – I want skyscrapers), so maybe the 13 will catch up in terms of ridership. But this still leaves more people with a further walk. Keeping the 2 on one side and the 3/4 on the other serves more people.

      2. Random idea for the 4. Instead of a turnback loop, how about having it turn off Boston at Nob Hill, turn north on Blaine, hop back over to QA Ave, and run north to the 3’s route. This should only add a minute or two to the 3 route, and still serve the 4’s area. It would also cover QA Ave better and we’d only need to add 6 blocks of wire.

      3. Keeping the 2 on one side and the 3/4 on the other serves more people.

        QA Ave is by far the densest commercial corridor on Queen Anne Hill. It’s the only part of the hill that could be considered a destination. If any road on the entire hill should get a frequent bus route, it’s this one.

        I have no problem with serving dense residential areas, especially when they show strong ridership or when they’re along (or the natural extension of) an existing corridor (e.g. the 49 along 10th Ave, the 10 north of Mercer, the 11 east of 23rd). But the 2 has none of these properties, and in all likelihood, it never will.

      4. I agree with all of that. But if Bruce’s numbers are right, it looks like the 2 would actually have higher ridership than the 13 if you include SPU. I’m curious to know why that is – there are certainly apartments and condos spread throughout the hill, and there are 30′ property line streetcar suburbs along the 2. Would that be enough to beat QA Ave’s dense strip? I wonder if it would make sense to make a second QA strip – raise the zoned ceiling on 6th ave and add highrise condos there. The street is wide enough for angled parking and retail, and after a few floors the condos and apartments would have great views. Connected with the existing retail on McGraw and Galer it would make a very nice and walkable dense circle.

      5. I wonder if it would make sense to make a second QA strip – raise the zoned ceiling on 6th ave and add highrise condos there.

        It’s definitely possible. I’m always in favor of more mixed-use corridors.

        That said, one of the things I hate most about Seattle zoning is that we seem to have an almost pathological aversion to rectangular, rather than linear, density zones. Almost every neighborhood in Seattle has one or two dense streets, surrounded by blocks and blocks of low-rise apartments and/or SFH. Downtown, Belltown, and Capitol Hill are the only places I can think of that have *any* mixed-use areas wider than one road.

        So, in an ideal world, I would much rather see Queen Anne Ave expand directly eastward/westward.

        Either way, though, the fact remains that Queen Anne Ave already has density (and is getting more all the time), while plans for 6th Ave are in the distant future. If Queen Anne comes up with an official neighborhood plan to turn 6th into a major corridor, then we can talk.

        (And finally, if and when 6th becomes a major corridor, it should have 15-minute service too! Half-hourly service on the major thoroughfare in a central city neighborhood is just ridiculous.)

      6. I’ll buy that, and I think widening our dense neighborhood strips should be the way to go throughout the city, and I could certainly be convinced that the 13 is the way to go. But I just can’t figure out why the 2 outperforms the non-SPU section of the 13. Maybe our retail is in the wrong spot!

      7. I suspect the reason the 13 doesn’t appear as strong along QA Ave is due to the extra coverage afforded by the 3/4. People headed to/from downtown can just hop on the first bus to come by since travel times are comparable between them. Going to/from SPU you only have the option of the 13 or 17 so that segment looks stronger. If the 4N gets axed or revised as has been proposed, or if the 13 doubles its service hours, I suspect you’d see a big uptick in usage (conversely, sending the tail of the 3N to SPU would probably make the 13 look worse there, but overall you should see more ridership in the corridor due to more frequent service.)

        In an ideal (flat) world you’d have north/south corridors on 10 block spacing (e.g., the 1, 13, and 26/28), and some east/west corridors similarly spaced (Mercer, Galer, McGraw?), and the area would be nicely served in a grid like North Seattle. Sadly the reality of topography makes that an impossibility, especially in the E/W orientation, so instead we’re given this tangled web of seemingly illogical route ends. Even if there were one bus-friendly E/W through street connecting the 3/4 with the 1 somewhere around Blaine St it would make routing a lot more logical.

      8. “it looks like the 2 would actually have higher ridership than the 13 if you include SPU”

        If the 13 were frequent, it would outperform the other routes. It’s the straightest and shortest way up the hill, and Queen Anne Ave attracts both residents and non-residents. The ONLY people for whom the current routing is optimal are those who live within two blocks on either side of the 2, 3, and 4. Those who live between 3rd W and 3rd N can just as easily walk to the 13 as to those other routes.

        So the 13 gets: (1) residents in the densest housing along QA Ave, (2) visitors to businesses along QA Ave, (3) those living between 3rd W and 3rd N, (3) SPU riders, and (4) a few people from Fremont. It would also attract NEW RIDERS; i.e., those who currently avoid going to Queen Anne because of its infrequent, scattered bus service.

        Strengthening the 2, 3, or 4 at the expense of the 13 would benefit only those people who live within two blocks of those routes. It would not attract new riders (or rather, it would lose more riders than it would gain), and it would make trips from SPU and Fremont longer.

      9. I suspect the reason the 13 doesn’t appear as strong along QA Ave is due to the extra coverage afforded by the 3/4. People headed to/from downtown can just hop on the first bus to come by since travel times are comparable between them.

        As long as they:
        a) have OneBusAway, or
        b) are willing to dash across the street when they see a bus coming.

        On QA Ave the shared stops between the 4 and 13 have opposite destinations.

    3. I think the 3/4 tails are much less crucial than the 2 tail, so I’m somewhat puzzled why Metro would want to cut the 2. The tail from QA & Galer to the end of the line only takes about 4-5 minutes to run and according to Bruce’s charts, it looks like each midday trip serves about 9 riders per trip. Is that ridership really an inefficient use of resources?

      1. I’m writing about the tail of the 2N in the previous post. The 2N appears to serve 9 riders per midday trip.

      2. The problem is that Queen Anne currently receives a ton of service hours, and yet there’s no all-day frequent route on Queen Anne Ave.

        If Metro had infinite money and the roads were infinitely wide, then sure, run a bus with 2-minute frequency on every arterial in Queen Anne. But in the budget-constrained world we live in, I’d much rather see a frequent route on the major commercial corridor, even if it means cutting service on the side streets.

    4. “I lean toward the idea of modifying the 2 to get down the SPU, and killing the 13 instead.”

      Queen Anne is crying out for a frequent route along its central commercial corridor, which is also its apartment/density corridor. (Funny how that happens.) Axing the 13 in favor of the 2 would bypass this urban center in favor of a residential area. That would be counterproductive to the goal of concentrating density and transit at urban centers, and providing good links between the urban centers.

      1. And I agree. I was just following the numbers with the 13 comment, and realized traffic is slow through the commercial corridor itself. I think the problem is that we don’t have enough density in our downtown strip – the apartment/condos are only half a block wide and limited to 40′ tall, and outside of that is SF5000 zoning.

  10. I see a poitential problem with eliminating the 2N and 4S–the cost of dismantling the overhead. There is no point in leaving the wires in place if they’re not being used, but the cost of dismantling the overhead is high.

    1. The 2N wires would probably stay, as they are used for the snow route of the 1 and 13, or if some other reason exists to avoid the south side of Queen Anne Ave. The cost of dismantling the 3N, 4N wires is not great, and part of the 4S wire could be used for electrifying the 48S as described in a previous post.

      Taking down wire would cost money, but it doesn’t have to be done urgently. It could be done in stages, potentially over several years.

      1. Actually I would suggest keeping the 3N wire as a turn around when buses get too far behind and in snow/ice conditions do not allow buses to go down 3rd Ave W. I agree that the 4N wire could go towards wiring the 48S, if it is broken from the north end. Although it would be cool if the entire route were under wire and buses could use their off-wire capability to go from the base to the end of the 44 and 48 using 15th Ave W.

  11. I used to drive the 2 alot and always thought the scheduling of the 2 & 2X was nuts.
    Sometimes the trolley leaves 8:00, followed 2 minutes later by the 2X, then another express at 8:10. The 2 does all the work, and the expresses blast into town after Denny.
    Sometimes the express leaves first at 6:57, then the trolley at 7:00, doing nothing until Denny.
    I used to see how soon I could catch up to the express – ‘nothing else to do’
    There really isn’t much difference between the times on either, that a little stop diet wouldn’t resolve.
    Trivia Dept: If you look left into the cul-de-sac at the 2 layover, you can see where the old streetcars used to turn around.

    1. I used to have a 2X driver that would announce facts as he drove. Entertaining when I didn’t have a book or iPod.

      All of those stops between Denny and downtown really kill the 2 (for us from QA, I’m sure the Belltownians love it). I learned long ago not to take a 2 or 13 if a 2X was coming. The 2 would do fine until Denny, then would be left in the dust. I factor in about 18 minutes on the way to work on the 2X, or 28 minutes on the 2. So the 2X has to be 10 minutes away in order to take the 2. Of course then I learned I can take the 2 down the hill and catch the 15 or 18, which used to effectively be an express on 1st. I wonder if they stop as much as the 2 does now that they travel on 3rd.

      1. They are faster, if only because they miss the awkward jog on Broad Street. That light would be a great place for a smart signal to advance or extend the protected left on Broad if a bus was present. Of course, sometimes traffic on Denny will be gridlocked, and then you would have been better on the 2, so there’s really no 100% win. Really, Lower Queen Anne should have fully grade-separated transit of some sort, but for the moment I’m just trying to maximize the assets we have now.

  12. Interesting discussion – a few points from another ex-QA resident who still uses the 13 frequently:

    1) Most 13 riders are NOT going to SPU. Bruce’s chart indicates more offs along QA Ave than the 2 stops closest to SPU (Dravus and Bertona). SPU students don’t have U-PASSes like UW students do and far fewer of them are commuters.

    2) The 3N/4N/13 have poor reliability in the PM commute. The 3N/4N suffer from severe traffic along 5th Ave N between Denny and Mercer, primarily caused by box-blocking and heavy traffic volume. The 13 sometimes gets stuck at 1st & Broad and also slows along Queen Anne Ave at the top of the hill. Even so, I actually prefer the 13 to the 3N/4N at rush hour, despite the 13 being timetabled to be 2-3 minutes slower to QA Ave / Boston, because the 13 is marginally more reliable.

    3) The 13 tail along 3rd Ave W runs in a valley between two reasonably steep hills – riders who live at the top of those hills probably get off at an earlier stop or ride the 2N (if living west) or 3N (if living east). Not surprising that the tail has poor ridership.

    4) The 13’s reliability has worsened since SDOT added a bike lane along part of QA Ave – the traffic lanes are so narrow that buses can barely fit between parked cars and oncoming traffic. Illegally/poorly parked cars don’t help.

    5) As Matt the Engineer said, although QA has lots of service, it is slooooooow.

    6) Delete the 4N before anything else – the tail doesn’t make much sense (if anything, it should be served via the 13 route) and it is time-consuming.

    1. Regarding point one: If this were done in conjunction with discontinuing the 17 (as a part of Rapid Ride D implementation) there would be more people taking the 13 to SPU.

  13. Bruce,

    Please note that the 2 current runs 90 minutes later at night than the 13. Back when I lived in the general vicinity of SPU, it was a 2-hour discrepancy.

    Any proposal to transfer the service hours from one to the other must retain the 2’s service span. Otherwise, the northern reaches of the hill are left with no service whatsoever after the last 17 (11:15) and last 13 (11:45).

    Also, I’ve long had the impression that prior preferential treatment of the 2 over the 13 (higher rush-hour frequency, more expresses, longer service span) was owed to its faster travel time. This also helps your observed service bump at 6th & McGraw: in addition to being a small but popular business area, that intersection has easy access to part of the 13’s walkshed and reaches is 5-10 minutes faster when traffic on Queen Anne Ave is heavy.

    The fault lies with the all-way stops at Boston and McGraw, which have not handled the increased auto flow to and from the Aurora Bridge or the increased arterial flow over the hill very well.

    I feel strongly that a consolidation proposal, while wise, should be tied to some sort of bus-priority mechanism for cutting the line at these intersections.

    1. I’ve never actually taken any Queen Anne bus past Boston St, but I have noticed how bad that all-way stop is. I’m assuming a traffic light plus bus queue-jump would make sense at those two intersections, but I doubt Metro would or could put up money for that in the current circumstances.

      The restructure schedule I’m looking at has the last bus departing SPU at 1:34 on weekdays, slightly later than the last departing 2 under the current schedule. It looks pretty comparable in other respects.

      1. I don’t think you’ll get a stoplight on the hill. Stop signs keep traffic slow, and we like that. Our downtown strip is very pedestrian friendly, and I’d guess that more people walk to QA Ave retail than drive. We just put in the stop sign next to Safeway, I believe to slow down traffic further. Change these to stoplights and cars start speeding down the street (where speeding = 20mph or more).

        This is one reason I was pushing for 6th, but you’ve more or less convinced me that even with the reduced area of walkshed and slower buses, the 13’s the way to go.

      2. The restructure schedule I’m looking at has the last bus departing SPU at 1:34 on weekdays…

        So then the post-11:45 outbound 2s would just become outbound 13s? Good.

        I’d guess that more people walk to QA Ave retail than drive.

        My experience living on one of the hill’s slopes did not reflect that. I think most people from parts of the hill more than 3 blocks from the commercial strip drive to it, then circle around endlessly looking for parking (sometimes they pass through the QA/Boston all-way half a dozen times).

        The strip’s pedestrian-friendliness does still get enjoyed, as once those people have parked, they tend to remain on foot for as much time as they spend in the various businesses. They might actually walk twice as far, while remaining on the Ave, than they drove to reach it. Funny how having street activity changes the sense of distance.

        I don’t think you’ll get a stoplight on the hill.

        The bus could still get a queue jump in the form of a block-long rush hour no-parking zone/bus lane and a “yield to buses” mandate where it merges back in at the 4-way. If the political will could be found.

      3. actually … they replaced the all-way stop signs with a traffic light there a few years back … was an unmitigated disaster … completely messed up ALL traffic on top of the hill … so they made the traffic light only flash red (all way stop) … which actually works

      4. “I think most people from parts of the hill more than 3 blocks from the commercial strip drive to it, then circle around endlessly looking for parking”

        That reminds me of when I was in high school visiting a friend on upper Queen Anne in the 80s. His mom was taking him clothes shopping for school and I went along. I had just discovered the wonderful Seattle bus routes with 30-minute frequency until 1:30am and whisper-quiet trolleybuses (as they were then before the continuous fans). So different from the hourly buses in Bellevue whose last inbound trip was 9:30pm. So I naturally assumed Queen Anne people took the bus downtown to buy clothes. Instead his mom drive to Northgate. When I asked why, because downtown is so much closer, she said, “Because Northgate has parking.”

        (I don’t remember whether she actually said, “Northgate has parking” or “Northgate has free parking,” but either way it’s almost the same.)

      5. …so they made the traffic light only flash red (all way stop) … which actually works

        Define “works.”

        Not saying it was better on a cycle, but this ain’t workin’.

        Taking the 13 end-to-almost-end at rush hour was the nightmare that kept on giving.

      6. If you want speed through QA, put the route on 6th. If you want speed to SPU, loop around the hill (i.e. the 17). If you want speed right through the retail strip, add timed lights, remove parking, and put up pedestrian barriers to keep people off the highway, um I mean street. If you want to serve the retail strip without killing it, you’re stuck with a slow bus.

        I do think removing parking for a bus lane has some merit, though I’m not sure how much time it will really shave.

      7. Another option: Close QA Ave to cars. Add a parking structure with 1st floor retail somewhere close by (maybe tear down a few homes on 1st near Galer?). Route cars to 6th. Make QA Ave a pedestrian and bus only street. They do this with streetcars in Zagreb and Istanbul.

        Of course with a counterbalance we could actually bring a streetcar up there, but I won’t press my luck.

      8. In a city that just voted for the Deeply Boring Tunnel, pedestrianizing Queen Anne Avenue would definitely go over well. Especially if what DP says is true, that many people on Queen Anne drive five blocks to Queen Anne Avenue and then circle for parking. I can just see the campaign ads: “What’s worse than a Nickerson Street road diet? Car-hater Mayor McSchwinn has a plan to ban cars on Queen Anne Avenue.”

      9. I learned to use the 17 when I lived over there, even though it was an eight-minute walk (vs two minutes to the 13). But its bridge-addled reliability and its evening frequency are a killer.

        Also, it’s always bothered me terribly that the 17 usually passes by five minutes after the 13 heads up the hill; the most logical transfer point to the top of the hill from points north simply becomes moot.

        I do think removing parking for a bus lane has some merit, though I’m not sure how much time it will really shave.

        It can cause a minute or two delay any time of day or on a busy evening, but it’s during the evening rush hour that it’s totally horrible. So I think my above suggestion — just on the last block before the all-way stops, just during the afternoon rush, and with some system in place to make cars yield to the bus when it reaches the intersection — could be minimally invasive but effective.

        Close QA Ave to cars.

        I’ve had many, many debates with Seattlites who think pedestrianization schemes are the bees knees. Their success or failure is absolutely dependent on spacial math: pedestrian density / width of the right-of-way.

        Which basically means:
        Tiny medieval or colonial streets (plus the occasional extremely well-used square) = pedestrianization will work, bring vibrancy
        Wide boulevards = pedestrianization makes the area seem abandoned, undesirable, kills retail for a generation

        Queen Anne Ave might seem small by Seattle standards, but really it’s 4.5 lanes wide (including the angled parking), plus already ample sidewalks. It’s a grand boulevard compared to İstiklal Avenue! And without a fraction of the pedestrian demand.

        The best American point-counterpoint is (as are many of my examples) in my hometown.

        Here’s Boston’s Downtown Crossing. Pedestrianized in the ’70s as part of a slate of misguided urban renewal plans, it gets desolate after 5pm, shuts down entirely by 7, and is sketchy and avoided after that.

        It’s not about access: it sits atop one of the busiest subway stations in the western hemisphere. Cars, if let through, would amount to the tiniest fraction of traffic through the area. What they would do is contribute a sense of multi-modal business and a sense that the area is tied into, rather than isolated from, the rest of the busy, desirable city that surrounds it. They’d also provide a constant stream of eyes of the ground, contributing to the perception of safety that allows “good” usage of the area to snowball and squelch the “bad” usage. (Tremont Street, the parkside auto/pedestrian through-street just a block to the west, is busy and perfectly safe at 1:00 in the morning.)

        For the counterpoint, let me transport you a mile to the west, where Newbury Street continues to be the busiest and most desirable (read: expensive) commercial strip in New England. The traffic is calmed to the point where you can jaywalk in front of it with impunity, but it’s there, and it’s part of what makes the street feel like an exciting urban nexus.

        Of course, it doesn’t hurt that the public transit speeds below the street, impervious to the traffic slowdown.

        FWIW, those who want to go to malls, like the mom of Mike Orr’s friend, will always go to malls. Trying to make the urban experience mimic malls has long been a desperate and ineffective strategy to lure them back. (Look up Chicago’s State Street experiment for one of the most disastrous examples.)

        And well-meaning progressive who pedestrianize “because I love the way it works in Europe” — without the requisite transit and pedestrian density that underpin centuries of European success — failure at math.

        Seattle has but a single street with the density of usage and the paucity of space to justify pedestrianization. And it’s already pedestrian!

      10. Man, that’s a lot of typos.

        There’s something about this comment-box font that just won’t let me edit the way I could post-posting.

        Anything about that rendered terminally unclear by the myriad mistakes? I’m happy to clarify. 99% of pedestrianization schemes in the U.S. have been unmitigated disasters. And it’s not because we love our cars. It’s a population-density/transit/street-width thing that’s just never going to be fixed in our lifetimes.

      11. It was all clear. I again agree with you (despite it killing my suggestion). Cars are important there, and there’s not enough density to remove them. Which kind of leaves us without options for the 13 (except your suggestion, which I still don’t see shaving more than a minute or two).

      12. Matt, am I right in thinking you live not too far from McGraw?

        I urge you to run a test someday soon: two days in a row, take a single bus all the way from work to north of McGraw (even if it means backtracking home a bit).

        The first day, take the 2 to 6th & Smith. The second day, take the 13 to 3rd & Smith, allowing you to experience the slog through all three all-way intersections: the major junction at Boston, the turn at McGraw where all the traffic from Aurora arrives to muddy up the journey, and the five-way confusion turning onto 3rd.

        Do this test at the height of the evening rush. Calculate the difference between the two entire journeys. And also note the total time you spend on those five blocks around and between the three all-way intersections.

        Trust me, it’s more than a minute or two. If that becomes the trunk, some sort of fix is definitely in order.

  14. Another thought would be to combine the tail of the 2N with the tail of the 1. This requires no new wire and keeps service along 6th including to the 6th & McGraw business district (which I suspect is responsible for the large number of deboardings there). This was proposed for several of the scenarios for the Rapid Trolley Network plan.

    It’s funny how some of the same ideas for restructuring transit keep coming up over and over again. It would be nice to finally see a few of them implemented.

    The ETB network should be the core of Metro’s frequent service network. We need something like the Rapid Trolley Network (if the service hours can be found) or at the very least the restructures Bruce has proposed in order to make that so.

    1. Alternatively, the tails of the 3/4 be chopped off and instead rerouted down McGraw to 10th to Gilman to 15th to Dravus. Although this would require some new wiring, it would also provide better access to Rapid Ride D, Magnolia (with Magnolia routes presumably rerouted via dravus upon Rapid Ride D’s implementation), the commercial area of 6th and McGraw, and the steep western section of Queen Anne Hill.

      1. The steepest part is south of Howe. You’d still need some form of the 1; on a map, it’s actually closer to the 2 than the 13 is between Galer and McGraw. I’d prefer McGraw-10th-9th-Barrett-11th-Dravus, connecting to RapidRide and potentially extending into Magnolia, possibly allowing a route there to be cut. Then the 1 could be rerouted up Gilman.

Comments are closed.