A few months ago, with lots of help from STB regulars, I created the Frequent Network Plan to show how we could improve Metro’s Seattle/North King County bus network without adding any service hours.  That’s great, wrote guest writer and commenter extraordinaire Mike Orr.  But, he asked, how much money would it cost to get the bus network we really want?  Or, since it’s mid-December, what would Santa have to bring Seattle’s passengers to make it the best Christmas ever?

At first, I was reluctant to look into the question because I figured the results would be ridiculously unrealistic, especially when we are still trying to fight off a network-killing 17% cut.  But I started playing with maps and steadily got more interested.  I drew up an “ideal” network closely related to the FNP, but with the goal of making the best possible bus network regardless of resources, rather than using a fixed level of resources more efficiently.  Pictured is a small bit of that network.

SLU/Queen Anne
A bit of the “ideal network,” in the 33% funding scenario.
Red = 6 min. Orange = 7-8 min. Yellow = 10 min. Green = 12 min.

Then I put together a preliminary estimate of the service hours needed.  The answer surprised me: only about a 33% increase in service hours from today’s level.  That could actually come to pass, if there were a solution to the 17% cut, a few good years of economic growth, and maybe one more funding vote premised on meaningful improvements.  It’s realistic enough that the City of Bellevue considered a 30% increase as the best-case scenario in their 2030 Transit Service Vision Report.  A 33% increase is an attainable goal for medium-term political advocacy and makes for a credible network vision, not a fevered hallucination.

As I did with the FNP, I’ve created maps of this “ideal” network:

  • Color-coded by route (the labels reflect the +33% scenario).
  • Color-coded by frequency for both the +33% and the +15% funding scenarios.

Of course, a 33% increase, while imaginable, would be an uphill struggle.  So I created one more scenario, intended to show the lowest funding level at which the “ideal” network is meaningfully superior to the FNP network even though it aims for broader coverage and thus sacrifices a bit of efficiency.  I found that to be a 15% increase in hours from the current level.  The +15% scenario uses the same network (with two extremely minor changes), just with not-quite-ideal frequency levels on thinner routes.

Much more explanation follows below the jump.

Both scenarios allow for established core routes to gain frequency; allow for improvement of services in neighborhoods that have potential for substantial growth (often radical improvement, in the +33% scenario); and offer enough coverage with frequent routes that a trip between any two points in the service area should be possible in a reasonable time except possibly at PM peak.  In the +33%  scenario, no all-day route would run less often than every 15 minutes, and no core route would run less often than every 10 minutes.  (In the +15% scenario, a few routes would have 20-minute frequency.)

In addition to the maps, I’ve published three reference documents:

As always, a few explanatory notes are needed for this proposal.

  • As with the Frequent Network Plan, this is a weekday and Saturday daytime network.  Both scenarios would allow for significant improvement of night frequencies, and each would look far better than the FNP night map, but night service levels would be lower than pictured here.  Sunday service levels would also be lower unless Metro’s work rules were revised to allow partial shifts on Sunday.
  • Also as with the FNP, this work is entirely unofficial, done without any collaboration with Metro other than careful reading of their public documents.   It’s not intended to set this vision in amber, but to demonstrate an operationally realistic example of how a good network could be put together at these levels of funding.  Other suggestions are very welcome.
  • Neighborhoods where there is lots of room for growth would see improvements in service ranging from important to radical.  Major transit improvements of this nature could help take some controversy out of growth, particularly in historically car-centric areas where residents have objected that new arrivals can’t possibly be expected to live car-free and therefore growth means Certain Parking Doom.  Here is a partial list of the improvements:
    • Crosstown service throughout South Seattle would be revolutionized.  The current two east-west routes (50 and 60), with 5 buses per hour, would become four evenly spaced routes (50, 52, 51, and 59) with 16 buses per hour in the +33% scenario.  In addition, a fifth crosstown route within West Seattle would connect SSCC, north Delridge, the Junction, and Alki.  Virtually every point in West and South Seattle would have access to at least one east-west route.  North-south service on California Ave SW, which also functions significantly like crosstown service, would run every 7-10 minutes all day.
    • Georgetown would gain new service to the Alaska Junction, High Point, and Othello Station, all running every 15 minutes in the +33% scenario.  Its existing service to downtown, Beacon Hill, South Park, and Tukwila would also become much more frequent.
    • South Park would gain new frequent service to Rainier Beach and Alaska Junction, both every 15 minutes in the +33% scenario.  Its existing service would also double in frequency.
    • Boulevard Park would gain a new connection to Rainier Beach, and would see its current service double in frequency.
    • Seward Park would see service frequency double, and a new connection to Georgetown would be added.  (Routes 50 and 52 are through-routed, which may not be immediately obvious from the map.)
    • Alki would go from 2 buses per hour, serving an indirect Admiral-Alaska Junction route, to 8 buses per hour in the +33% scenario, with a much faster connection to Link and new service to multifamily housing on Beach Drive.
    • Admiral District would gain an extremely fast connection to Link on revised route 50, as well as 6-8 buses per hour to the Alaska Junction and Seacrest Dock.
    • Making up for our failure to build a Summit light rail station, Summit would gain 10 buses per hour to Capitol Hill Station, as well as a frequent and reliable connection to downtown on expanded route 47.
    • Service in South Lake Union would be radically improved.  Route 8 would avoid much more of Denny Way, improving reliability.  New frequent service to Belltown and the waterfront would be added on route 34, and to First Hill on route 7.  Frequency would double on the now-underserved route 40 corridor to downtown and Fremont.
    • Madison Park would go from 2-4 buses per hour to 11-12, evenly divided between the 8 (service to Capitol Hill Station, SLU, Uptown, and Magnolia) and the 12 (“Madison BRT” one-seat ride to First Hill and downtown).  I’m bullish on the urban part of Madison Park as a place that could absorb quite a lot of growth without losing its attractiveness, but its current transit service is a disgrace.
    • Magnolia Village would gain new crosstown service on the 8, much more frequent service to downtown and Ballard on the 24, and connections to all of Magnolia between the 24 and 33.
    • W Manor Pl, the densest part of Magnolia, would have 7-8 buses per hour traveling directly through it on two different routes (24 and 31), collectively serving downtown, Ballard, Fremont, the U-District, and Magnolia Village.  Currently, residents have to walk uphill or downhill to access service on half-hourly routes.
    • Roosevelt and Fremont would both become transfer hubs, each gaining lots of new service to new destinations, and seeing existing service become more frequent.
    • Central Wallingford would gain a new connection to Fremont on new route 63 and faster, much more frequent service to downtown on streamlined route 16.
  • Now that ST has announced likely station sites for Lynnwood Link, I’ve added a “likely version” of Lynnwood Link to the map, with stations at 130th, 145th, 185th, and Mountlake Terrace.  More optimistically than some commenters, I think the 130th station is an inevitability as long as ST continues to accommodate it in plans.  It’s just too logical not to build, and the marginal cost is very low.  I’ve restructured Shoreline and far north Seattle service with these Link stations in mind.
  • Because this is now a 2023 vision, I’ve also added the relevant portion of East Link to the map, which includes a very useful transfer point for north-south riders in the Rainier Valley and the Central District.  I’ve also changed a couple of routes near South Lake Union to take advantage of the reconnected street grid there.
  • The map does not make the Queen Anne service pattern totally clear, and I should explain.  Routes 1 and 4 are through-routed, with a pause at Queen Anne and Crockett in both directions.  For example, outbound route 4 buses would arrive at Queen Anne and Crockett southbound; lay over for a short time (likely 5-8 minutes); and continue as inbound route 1 buses.  The same would occur in reverse, laying over northbound.  This routing would have a number of positive effects.  First, it would allow for what amounts to a crosstown route within Queen Anne.  Second, it would give everyone in Queen Anne a one-seat ride to the Queen Anne Ave N commercial area, and to a Queen Anne Link stop should one be built.  Third, it would continue providing service to the current route 2N wire.  (Route 13 would remain as is except for the Fremont extension and the frequency improvement.)
  • I haven’t prepared a comprehensive FAQ for this plan like the one I prepared for the FNP, because this plan is less about showing exactly how improvements would work and more about establishing a baseline amount of resources we ought to devote to the system.  So I’ll provide a few basic numbers here instead.  Where the current all-day network in this area uses roughly 324 all-day buses and the FNP uses 337 (through absorption of hours from some redundant peak-only routes), the +33% scenario here uses 441 all-day buses and the +15% scenario uses 387.  (As with the FNP, these numbers assume elimination of peak service wholly redundant with the improved networks.)  These plans also offer more generous recovery time than either the FNP or the current network, so they should improve reliability significantly.

110 Replies to “What Should Santa Bring Seattle’s Bus Riders?”

  1. Home rule for taxes on the County level, and for Counties (all of them) to be able to fund transportation by taxes either with a County vote or by ballot vote. Leave the choice up to the County Councils. Cut Olympia out of the mix.

    1. With county home-rule, it will be interesting to see a highway drop from four lanes each way to two lanes plus an HOV lane at the county line.

      1. Back where I used to live, there was an expressway which really did drop from three to two lanes each way just past the county line. Admittedly, traffic volumes also dropped there.

      2. In the early 80s I had friends in north Spokane and just outside Yakima, and I remember the local streets becoming less maintained at Spokane’s border, and paved roads turning into dirt roads at Yakima’s border.

      3. I’ve seen a lot of this on the Kitsap/Mason border: roads will change from asphalt with full lining and shoulders to narrow chipseal with barely a line in the middle.

  2. How many non-redundant peak-only busses are there left? Could this plan be met by absorbing the remaining peak-only routes? Arguably, with a network this good the need for many one-seat rides is far diminished.

  3. How many here remember Exec Gary Locke’s decision to cancel a capital project and plow the money into increased service?
    That was 20 years ago, which by cancelling the CNG Conversion Program netted about 600,000 additional service hours (IIRC), or a 25% increase. Ridership really shot up from that, and the networks improvement showed it. I was on the S.King Sounding Board at the time, and it was challenging to spend the new hours wisely, but we got a lot of new service like the 164, 153 and splitting the 150 from south of Auburn.
    I only bring this up because the Puget Sound is spending like crazy on capital, and starving it’s base riders out of existence. ST was born by Metro. The two are joined at the hip, and voters should have the ability to determine how each operate and maybe change their minds when things don’t work out so well.
    It doesn’t have to be either or.

    1. Do you have particular Metro capital projects in mind, or are you suggesting funding Metro service by delaying ST capital projects?

      1. Metro has no capital budget left. It got cannibalized over the last 5 years of dithering with audits, blue ribbon panels, and endless navel gazing as they approached the abyss..

  4. This would actually suck for Wallingford. Running the 16 down Stone skips the center of the neighborhood and since you’ve deleted the 26 there’s no direct connection to Downtown during the day (the 63 looks like it just stops in Fremont).

    1. Agreed, this takes away transit service from significant portions of Wallingford. I also noticed that parts of Broadview lose service as well thanks to what appears to be the elimination of the 28X. I hope there aren’t other cuts included here – the entire point of a map that is predicated on more funding is that everyone keeps what they’ve got and we then add new things as needed, whether it’s a new route or new frequency to existing routes.

      I also don’t understand the objection to a single-seat ride for buses. Once Link is built through Northeast Seattle it makes sense to tie bus routes in that region to Link stops, but that’s because Link provides far superior frequent and fast service than a bus ever could. Single-seat rides in places like Northwest Seattle should be preserved until the subway is built out there. Under Option D for Ballard, for example, the current single seat rides to downtown could stop at the Ballard subway station, but until that is built, they ought to continue on to downtown.

    2. Wallingford’s street network makes it one of the most challenging neighborhoods in the city. There is no good solution.

      This solution is designed to provide much faster service than today for those on the west end of the neighborhood. Those on the northeast end will get their fastest trip downtown, counterintuitively, by using the 63 and transferring to Link at Roosevelt. Link is so much faster than the current 16 routing through the middle of Wallingford as to be the best option. From the southeast corner of the neighborhood — an area where the street network makes fast service absolutely impossible — the best option is to use the 31 or 44 to transfer to Link.

      I’d anticipate retaining the 26X, possibly shortened on the north end, for peak service from the southeast.

      1. One more thought: using the 63 to get downtown via Fremont is also a reasonable solution, because of the extreme convenience of the transfer. In the 33% scenario, there are 14 buses per hour — an average of one about every four minutes — going from the Fremont transfer point to downtown. I’ve tied the 63 in knots in Fremont, using a whole extra bus just to reach the layover point, so that the transfers in both directions are as seamless as possible. (I wanted to use this solution for the 28 in the FNP but couldn’t find the hours.)

        Of course, the same also applies to transfers in Fremont from the 31, which covers the territory that’s furthest from the north-south routes on the map.

      2. Not being able to get from Downtown to the center of Wallingford without transferring seems like poor service to me.

      3. Not being able to go anywhere without zigging and zagging three dozen times on a bus that’s always five minutes early or twelve minutes late — that’s today’s 16 situation — is far, far worse service.

        Wallingford is an east/west-oriented neighborhood with a large lake directly betwixt it and downtown. You have to go around the lake, be it on one vehicle or two. If the final result yields better absolute trip times in every circumstance — which David’s frequent network most certainly does — than you have improved Wallingford’s transfer situation.

        You need to stop presuming transfers are inherently horrible, just because most of the ones in today’s Seattle are.

        (Anyway, the real solution is the east/west subway, giving every person north of the Ship Canal a painless, reliable tap-in to all of the north-south spines to downtown… and to everywhere else.)

      4. Personally, I’d rather ride my revised 16 to Stone Way/45th and walk three blocks to the “center” of Wallingford (which is, in any case, moving west) than ride the current 16. Speed is important! If it’s not, there’s always the 63.

        If we had full-time bus lanes along 45th, then I could have kept the 16 on something sort of similar to its current route without turning it into a pit of slowness. But we won’t anytime in the foreseeable future.

        Fully agreed with d.p. that the only way to really untangle Wallingford is to build the east-west subway. Until then, the goal is to provide the fastest possible service, and unfortunately in Wallingford that means missing the center by a couple of blocks.

      5. My experience taking the current 16 to the center of Wallingford is that, with typical traffic on 45th through there, I can actually get to the center of Wallingford faster if I get off the 16 at Stone Way and walk 3 blocks, then if I stay on the bus. If my ultimate destination is the U-district and I have a 44 to catch, sometimes, walking a few blocks can even allow me to catch up to a stuck-in-traffic #44 I would have missed if I just stayed on the #16 bus.

      6. Personally, as someone who lives in the neighborhood and is a frequent rider of the current 16, I’d rather have the 16 go on 45th between Stone and Meridian and take the speed hit in order to serve the center of the neighborhood and the walkshed in the eastern part of the neighborhood. The reliability of the bus has improved considerably since they got rid of the Seattle Center routing (even with the construction). Your reroute reminds me of the old 6 which Metro folded into the 16 because the new routing hit more people.

        Clearly a subway is the best answer. Until we get it a direct connection to Downtown from the center of the neighborhood (as we have today) is a requirement of any bus network.

      7. I don’t live in Wallingford, but I live close enough that when I go there I walk because it’s faster than taking any bus and biking or driving is more of a hassle than it’s worth. I ride every part of today’s 16 often enough, so…

        – Stone Way isn’t that far from Wallingford Ave, and its speed advantage for those going through Wallingford is great.
        – The walkshed on Stone Way north of 45th kinda sucks. In the long term, we’ve gotta blow up Green Lake Way between 46th and 50th to restore the street grid… anyway…
        – The point about Wallingford being oriented east-west is extremely apt.
        – Central Wallingford doesn’t have any more residential density than the rest of Wallingford, nothing warranting tearing up your network to connect specifically to downtown Seattle.
        – It’s notable as a commercial center, but not regionally notable… mostly to people either (a) along today’s 44 (b) along David’s 63 (pretty cool) (c) close enough to walk.
        – For going downtown, the extra walking to David’s 16 is fine for a lot of people; for the rest, this works to the extent that funding is stable enough for a frequent 63 and that the 44 is reliable enough to avoid bunching and crowding.

      8. Of course, I love the idea of running the 16 on Stone Way, which probably isn’t surprising, given my prior history on the subject. ;) But I’m curious why you decided to take the fast bus (i.e. the 16’s route) and cut it off early. It seems like you could do a bit of shuffling around and end up with services that are more useful to almost everyone:

        – 16: Starting from downtown, take Aurora to 38th St/Bridge Way. Continue east on 40th to Wallingford Ave. Turn left and head to 45th. Turn right on 45th and continue on your 63 route, terminating around Roosevelt Station.
        – 63: Starting at Fremont Station, head east on 34th/35th to Stone Way. Continue north along Stone Way to East Green Lake Way. At Green Lake Center, cross over to Woodlawn, and continue north on your 63 route to Northgate TC.

        In terms of service hours, this shouldn’t be much more expensive than what you currently have. It might even be cheaper, if you run both routes at 15-minute frequency. In exchange, what you get is one bus that runs express until it gets to Wallingford and then goes local, and another bus that runs express until it gets to North Green Lake (or whatever you call the neighborhood around 85th/Wallingford) and then goes local. Both of them have plenty of rail connections.

  5. Very nice. Looking at the north end, the one thing that stands out to me is that there is no direct connection to the 145th LINK station from 145th & Bothell Way. Riders from Lake Forest Park, Kenmore, and Bothell would presumably take the long, meandering 88 through North City to catch LINK at 185th.

    Riders on Bothell Way between LFP and 145th who want to catch LINK, however, would be left with the following options, none of them all that attractive:

    1) Take the 69 to 145th and walk 3 blocks west to catch the 78 at 30th.
    2) Transfer to the 78 at 130th and backtrack to the 145th station.
    3) Transfer to the 75 at 125th to catch LINK at the 130th station.
    4) Transfer to the 15 to get to Northgate Station.
    5) Transfer to the 78 or 71 to get to Roosevelt Station.
    6) Stay on the 69 all the way to Husky Stadium and catch LINK there.

    The only one of these routes that has better than 15 minute headways is the 15, and it has to slog through a bunch of traffic and signals to get to Northgate. If ST 522 still exists in this scenario, I think most riders will just use that to get downtown instead of using any of the above options to catch LINK.

    1. I expect there would be peak-only service from Bothell Way to 145th, replacing the north part of the 306/308/312.

      All-day service would remain on the 522 (not pictured here because it’s a ST route), probably truncated to 130th.

      Also, the 75 has 12-minute headways in the better funding scenario.

    2. I really like this. The east-west routes are great. I’d love to see a more direct route which runs east-west from Lake City to Ballard along the relatively uncongested 125th/130th – Greenwood – Holman – Ballard route, instead of getting bogged down in the Northgate transfer mess. That would likely be sufficient for east-west demand the next half-century, unless there is far more growth than I am expecting up there. No need to build a subway. Money would be far better to put a subway Bothell to Downtown via Lake City/Roosevelt.

      1. I suspect more people are travelling from Ballard or Lake City to Northgate than travelling between the two. Personally, I love the fact the 75 terminates in Broadview and would love to see it further extended to incorporate the entirety of the only segment of the 28 up there that actually had some reason to exist (the bit along 8th Ave NW), but that would probably require an extra bus to serve an extremely un-dense and rich-snob area.

        In a perfect world where this blog existed when Sound Transit was doing its planning, ST was actually planning ahead, and the various suburban subareas put more emphasis on getting it right than getting it now, the route we’re building now would ultimately head up Lake City/Bothell Way and the Ballard route would follow 99 north of 105th or so all the way to Everett. Even then, though, you might want to have another crosstown route or connection somewhere further north than 45th.

      2. Morgan–the scenario you envisioned with the current route travelling up through Lake City and a Ballard route turning up 99 is basically what was envisioned 40+ years ago…even predating STB! Unfortunately for many reasons we got a line that was considered unacceptable by the planners of that time due to its inaccessibility from several of what now are called urban hubs/centers.

      3. We should revisit it. I don’t see any tracks next to the TOD wasteland along I-5 yet. It’s all still just paper and a some millions on planning do-overs. A drop in the bucket compared to doing it right.

        Do it right.

        Shoot a prong from Roosevelt through Northgate, and up to 99, and take advantage of all that TOD potential and urban retail along Aurora. It would revolutionize the Paigow and Ho strip into something not just the underbelly of our city spends time at (not that there is anything wrong with Paigow and Hos). Then on to Lynnwood via 99.

        Then shoot a prong up Lake City from Roosevelt station all the way to UW-Bothell. That prong can go cheap – mostly at-grade, except a few crossings, because of the bluff. Elevate at Lake City and again at LFP town center and maybe Kenmore/Bothell.

        Now you have great coverage of both NE and NW.

        Do your obligatory Ballard to downtown line, and maybe a cross-town at 45th, and the north-end is golden. You DO NOT need a crosstown subway, with two transfer points from East to West at 45th and and Roosevelt.

        Probably much cheaper and much more sense than the Northgate crosstown “we are nascent Manhatten” silliness.

      4. The people you’d have to convince would be the Snohomish County politicians who don’t understand their own constituents, let alone the importance of walkshed and the impact of freeways on it; they’ve already spent tons of money with laser focus on Mountlake Terrace and Lynnwood freeway stations. But I still say Aurora makes more sense as an extension of the Ballard line; Northgate is an awkward spot to serve on the way to 99 or Lake City, but it’s easier to take the Lake City route, and even beyond that I don’t like this notion of building a “spine” that’s parallel to another route; Ballard would actually be a semi-express for Snohomish County commuters compared to Northgate. But Snohomish County and other suburban politicians don’t care about what makes sense or even actually understand transit at all; all they know is they want their rail and they want it now, because heaven forbid their tax dollars “subsidize” anyone else just because the only return they get is buses (imagine that in the most dismissive voice you can think of). Like I always say, don’t build long-term infrastructure to meet short-term needs.

  6. Is this a 33% increase in County-wide bus service hours to fund network/frequency improvements in Seattle & Shoreline only? Is it funded countywide? Or is this a proposal for a Seattle-only funding measure to buy more service hours from Metro?

    1. This is a 33% increase from the current all-day network in the Seattle/North King County service area. If we had a 33% increase in countywide funding, then similar improvements could happen in SKC and on the Eastside. Seattle could also go it alone on a Metro funding proposal to make something like this happen independently of the rest of the network.

  7. Excellent map!

    I’ll probably have more questions later, but the first thing that leaps out at me from skimming the route descriptions is: Is the 4 through-routed with the 3, the 14, or both? Also, have you considered whether the horribly long 14-1-4-3 through-route will really be reliable?

    1. The 4 is through-routed with the 14. There is a “pulse” (short layover) in Queen Anne to ensure some level of reliability; it’s not like the bus will go 3-1-4-14 without any stop (which would be horrible). There are also long layovers on the south ends of the 14 and the 3.

      1. Yes, I noted the Queen Anne pulse, but it won’t really be able to be too long if you expect people to ride it through from one side of Queen Anne to the other. The through-route still seems pretty long to me, but I’ve never studied layovers, so I’ll defer to your judgment.

      2. I expect the “pulse” to be 5-8 minutes depending on time of day. There are very few riders who will ride all the way through; most E-W riders in Queen Anne will be riding from Queen Anne Ave one direction or the other. I’m OK with making the few through-riders wait through the pulse.

        Another solution would be to add two buses and have the “pulse” be long enough for each bus to overlap with the next. The challenge with that is a lack of curb space along Queen Anne Ave.

        It is a long through-route, but I think it should generally work OK – of all the routes involved, only the 3 has meaningful schedule problems, and those will be significantly ameliorated by the move to Yesler.

  8. I like very much the improvements in east/west service and connecting neighborhoods that were previously isolated without a major detour.

  9. Creating frequent service by interlining different routes, and keeping headway between them, seems like an unrealistic expectation. From your experience during your Metro days, David, are there ways to actually achieve it?

    1. It depends on the nature of the routes.

      It’s very easy where you are interlining one end of two short routes that share a terminal (as in the 2/10 case).

      It gets harder when the interlined portion is in the middle, or when the routes are longer. It gets harder still when one route has a reliability problem. Good scheduling really, really helps, but there will always be a delay somewhere. The point to consider is that the interline may suck, but a single route would probably suck too.

      In the case of the 1/3. 14/34, and 34/35 interlines, I have little expectation that they will actually provide even 6-minute service at all times, but there is no reason not to schedule them that way, and a single line would probably not do any better. The 131/132 interline should do pretty well because all the routes involved should be pretty reliable. The 28/40 interline is the toughest one, particularly because the 124 has some reliability challenges, but breaking all those routes apart would be stupendously expensive in hours — hard to justify even for an “ideal” network.

      All of these would be better than interlines we have today just because all of them would have much shorter headways, period. The worst case was the old 5Northgate/5Shoreline/54/55 interline — the interlined routes were half-hourly and one of the two had a major reliability problem.

  10. Santa should bring a 1% SIIT (State Investment Income Tax) to make all the transit girls and boys dreams come true on Christmas night…

  11. I’m a big fan of the southern leg of your 78 route along 65th. Is this possible? I emailed Metro earlier this year regarding the potential of running a bus over Phinney Ridge on 65th. This is the response I received:

    “Thank you for the suggestion and there was a mini-bus route that used that street back in the early 1990s. However, there was very little use of the route. Also, our safety department has become much more conservative with routing buses and is unlikely to approve operation on NW 65th Street at the intersection of Phinney Av N. Further complicating matters is our funding situation in which we have no additional resources and may have to cut service up to 17% in 2014.”

    Would a reconfiguration of the 65th/Phinney intersection to make this route palatable to Metro? I would love to see this route. Thanks.

    1. I think you’d need more than just a rebuild of the intersection — you’d also need a rebuild of 65th itself immediately east of Phinney. I note that in my route-by-route notes.

      I think a full route from Ballard to Roosevelt would also attract much higher ridership than the infrequent circulator Metro referred to in their email.

      1. Yeah, that’s a really cool route–having lived near the north end of the proposed 78 for years the idea of a direct bus to Roosevelt and Greenlake/crosstown to Ballard really opens up areas of the city that were basically inaccessible by bus from NE Seattle.

        If I still lived in the area, I’d miss having the bus to U Village and the UW, though, as that was easily a more likely destination. Roosevelt Station is probably as far, time-wise, as U Village was from the upper end of 35th NE; at least there you can transfer to Link but there is definitely a time penalty for an area of town pretty heavily connected to UW. U Village is now quite inconvenient as you have to transfer from a 15 minute bus to a 12, meaning there’s always the chance of a 10+minute gap. Just a small observation based on many years of traveling by bus and car from that area…understood that with the complete lack of cross-town possibilities from that part of the city it will always remain somewhat of a transit desert. There’s really no good way to access Link (boo no Lake City line!) so Roosevelt is probably the best possible alternative.

    2. I would love to see this route too, but the street really does need to be improved. The existing right-of-way just west of Phinney Ave is too narrow. As it is, it’s a two-lane street with no parking and a sidewalk on only one side of the street because there’s no room on the other side. To make matters worse, 65th jogs a little bit from one side of Phinney to the other, which is likely difficult for any long vehicle to navigate safely.

      If SDOT wanted to improve the situation they would probably need to purchase the Home Espresso Repair business and the house behind it and demolish them to both widen the right-of-way and allow the street to curve more gradually from one side of Phinney to the other.

      That would probably be the bare minimum amount of work needed to make the road usable for transit. I think it would be a worthwhile investment in the long run. We currently don’t have any east-west streets in the northwest quadrant of the city between Market and 85th St (with the possible exception of 80th Street) that are suitable for transit. That’s too big of a gap. A grid system is designed to make people transfer, and that’s fine. Where it fails when you have to transfer multiple times because you’re trying to go straight west, but the only way to do that is to go north, west, and then south again because there’s a two-mile gap between east-west routes. Might as well walk halfway if that’s the case.

  12. Good job once again. I have one comment:

    Some of the new routes such as the 51/52 pair and the 78 maybe restricted from being able to use a full sized bus (even a 30-foot) due to the hills and the turning radii. Could a paratransit van handle the passenger volumes along 35th NE?

    1. As noted in the detailed notes, for the 78 to work you’d need a capital improvement project at 65th and Phinney, to accommodate normal buses. I think a 40-foot bus would be sufficient for expected ridership on the 78.

      The 51 and 52 would need 35-footers, which would need minor improvements to the Seacrest Dock parking lot to turn around, but would otherwise work fine.

  13. David: This is awesome! Amazing map! Thank you for all the work! I have three questions, each about the only bus routes I’ve used very extensively:

    1) Have you considered terrain in your routing? I’m still a Magnolia resident and Manor Pl alumni: I’m just curious how much of a consideration some of those hills (Manor/McGraw) that the 24&31 are taking were. I know buses get up Queen Anne Ave everyday, but I honestly don’t know, do articulated diesel buses make it? Trolley wire in Magnolia is a stretch of my imagination. Does the frequency make up for lack of articulated buses?

    2) The very unofficial, but obvious Magnolia Transit hub is at 28th Ave W and W Blaine St. It looks like you’re attempting to move it to the Magnolia Village, but then having the 31 miss it. Do people really need transit from UW into Discovery Park instead of the major connection point? Maybe they do? I thought the FNP 31 routing was the best idea for the 31 I’ve seen–I’d just route it on one block south on W Hayes to the 28th and Blaine stops. I have no idea how the McGraw hill works for buses, but routing the 24 and 8 down Condon Way to maintain the 28th and Blaine transfer point might at least be less circuitous? Are the neighbors a factor in your change to Magnolia Blvd (fewer old, annoying neighbors there)?

    3) Neighborhoods like Magnolia are–to put it nicely–set in their ways. I’m all for routing the 24 to Sunset Hill, but it’s always been a tough sell as the 24 is full during peak commutes. Are you counting on the 17/18X to make alleviate that? Do you have any thoughts on how to slowly, incrementally implement plans like this one? Are you counting on frequency to be the carrot for the major overhaul?

    So now that I have no more questions on minutia, how do we make this happen tomorrow?

    1. Sorry, in 2) I meant routing the 24 down 28th Ave W and 8 down Condon Way to maintain the 28th and Blaine transfer point might at least be less circuitous?*

    2. 1) Manor Place is not a place that’s ever had bus service before, but it is not too steep or curvy for any type of bus. I’m sure Metro Safety would get very anxious about it for some reason or other, but as a former operator I would feel totally confident taking a 60-foot coach up or down it several times a day. McGraw is really not that steep at all.

      Metro itself proposed service on Emerson, which is steeper and narrower than either McGraw or Manor Place, for the revised 33 in the 17% cut scenario.

      2) I want to abolish the 28/Blaine hub. It makes no sense. It’s remote from anyplace in Magnolia that people actually want to go. Serving it pisses off residents along Condon Wy (as you noted) and slows down riders because of the 15-mph slow order there. Using Magnolia Blvd and Clise Pl to reach a Magnolia Village transfer point would be far faster and friendlier to riders and residents alike; the “only” issue is the need to get SDOT approval.

      As for the 31, I replace its east/west connection to Magnolia Village with the 8. I would have UW riders either use the 8 (or downtown service) to reach Link, or use the 24 or 33 to reach the 31. This is one of those trips where the one-seat ride has to disappear, but the transfer ride is a whole lot better than today’s.

      3) I think Magnolia residents would like this proposal, because it increases service so much. There shouldn’t be any capacity issue because of the dramatically improved frequency (improved even at peak hours). In any case, the 24 will pick up very few peak-hour commuters north of Fishermen’s Terminal.

      The FNP proposal for Magnolia is a tougher sell, but it still has major advantages.

      1. As a former operator, what are your thoughts about turns from 15th Ave to Dravus St towards Magnolia and vice versa, both northerly and southerly approaches? Those ramps/turning radii look cramped at least to turn south in 15th from Dravus.

        I only ask because another Magnolia route through Dravus to downtown might make sense someday: it has provides grocery access to east Magnolia.

      2. The only problematic turn at that intersection is SB 15th onto WB Dravus. The signal programming includes a dedicated right-turn phase, which means that the bus must turn out of the turn lane without splitting the lanes. (If it could split the lanes, there would be no problem.) I have the 31 making that turn, but noted in the notes that you would probably need to restripe Dravus, to move the centerline south a bit, to make that turn fully safe.

        In an earlier draft of the plan I had the 33 going into Magnolia using Dravus, and the 31 covering Thorndyke and ending in Magnolia Village. I changed the plan around to ensure more coverage of the residentially dense areas of Magnolia.

  14. Another slightly-less-tiny question: In both this and the original FNP, how much Renton-Seattle service is put into your bucket of hours to be redistributed? Obviously, the old 106 and 107 would give way to your revised versions, but what about the 101 and any other routes I’m forgetting?

    I’m all for replacing the 101 by a Link shuttle, but if that’s what we’re doing, we should come out and say so.

    1. The 106 and 107 were included in the bucket of hours because each one is primary service for a part of Seattle. The 101 was not included in the bucket of hours or in the plan. It would be part of my South King County plan, if I ever get around to finishing that one.

  15. Another great round of changes. I have some thoughts about Magnolia, held over from the last go round.

    I am a bit concerned about the big hole in western Magnolia. That is a sparsely populated area, but so are a lot of places in Seattle — and few, if any, are so far away from service. Just from a political standpoint, it might hurt to see a big hole there. Fortunately, I don’t think it would be difficult to solve this problem.

    Make one of the buses do a big lollipop loop in western Magnolia. The loop consists of 34th and Emerson, Emerson and Viewmont, Viewmont and McGraw, 34th and McGraw. I think you would lose very little frequency this way. but gain a lot of coverage. Keep in mind, there are no stop lights in Magnolia, so adding on this extra distance would cost very little time. With the new plan, I think the easiest way to accomplish this would be to modify the 33. In that case, it would probably be best to do the loop clockwise (take a right on McGraw from 34th, take another right on Viewmont, then a right on Emerson, then take a left on 34th). You lose direct northbound service from Magnolia Village, but I think that is a small price to pay.

    Speaking of Magnolia, I’m not sure the 8 should serve Magnolia Village. If time allows, it would be nice to have it go down Thorndyke, then take a left on 22nd, then a right on Bertona or Ruffner. This keeps most of the existing stops on 22nd. There are way more people in that area of Magnolia then by the village.

    1. I had the 31 end in western Magnolia in the FNP. I decided it was more important to serve the two densest places in northern Magnolia: 34th/Government and Manor Pl. There are essentially zero riders west of 34th except during peak, when the 19 would be running.

      Still, a loop like you suggest for the 33 isn’t a bad idea. That part of it is a pure coverage route anyway, so it wouldn’t have a negative effect on much of anything. The only question is where you lay over — you need a layover in Magnolia, because the 33/132 combination is quite long.

      1. I think there are lots of places where a bus could layover there. Emerson is very wide and has a greenbelt on both sides a bit north of 36th. A layover at 34th would probably be ideal, as it would allow folks who are switching directions (e. g. someone who came from western Magnolia who wants to get to the village) to stay warm and make the connection more often. The loop from there should be very consistent (again, no stop lights and fairly minimal traffic).

    2. Neglected to address the last part of your comment. I am trying to allow for growth in Magnolia Village, and to establish it as a transfer hub. It’s the one place in Magnolia that could be urban in our lifetimes. As such, it makes a very logical westernmost point on a mid-city crosstown route. Sending the 8 up to the apartment areas and the 31 to Magnolia Village would create spaghetti in place of what is sort of a grid.

      1. Why does the 33 stay on Thorndyke/20th/Gilman rather than improve its walkshed, create better spacing with the 24, and serve existing stops on 22nd? I get that the transfer with the 31 would be virtually nonexistent on the north end unless the 31 deviates to 22nd too, but…

        While I’m in the area, what do you think of the 24 using 24th and Thurman to get from Manor Pl to Emerson Pl rather than the Emerson-Gilman-Emerson Z pattern? Yeah, there wouldn’t be a single place to transfer between all three routes, but…

      2. 22nd is a Pit of Slowness(tm) and narrow enough that it wouldn’t be approved today by Metro Safety. As a passenger, I tend to prefer speed where the walk to the faster route is reasonable.

        That block of 24th is not wide enough for buses and wouldn’t be unless you banned parking on both sides. I’ll let you come to the community meeting and explain that one to the residents…

      3. I would argue that eastern Magnolia is fairly urban, but I know what you mean. It is possible that Magnolia Village might catch up and pass it (although I wouldn’t put money on it). Avoiding a bit of slowness sounds like a good idea.

      4. Ross, east Magnolia is quite dense in parts, but it’s not urban, and won’t be without significant rebuilding. There’s essentially no commercial activity and no place to put it in. The one area that would allow commercial development (a few lots along Gilman) is remote from the residential density.

        By contrast, Magnolia Village is tailor-made to become a small urban center. All it needs is some housing on top of the existing commercial uses, which is already happening in spots. (This being Magnolia, it also likely needs a couple of expensive parking garages under said housing, but that’s fine.)

      5. 32nd north of McGraw has some land use action signs these days. Although it’s going to be a long time before the SFR around there rolls to MF. But if you bring transit, maybe it has a shot.

  16. This is great, and it even agrees with what I would do in places. Something like this could motivate people to push for more funding, especially since Olympia is constantly pushing Metro to be “more efficient”. This would seem to achieve maximum efficiency with the minimum pissing off of constituents (except for perhaps the Will Douglases of the world).

    Some comments and questions:
    *How much more would it cost to have an all-Meridian 86, an all-15th 87, extending the 78 to follow your 86 route to Aurora Village, and rerouting and extending the 67 to 1st between 130th and 145th, swinging back to 5th to 155th, and terminating near 155th and Aurora/Westminster? Alternately, what about having an all-Meridian 86, an all-15th 87, keeping the 67 as-is, extending the 78 to Aurora Village, and rerouting the 78 to switch from 145th to 155th at 5th or, if that doesn’t serve the Link station well enough, 1st? I get you’re trying to serve the Link stations and the 78 is already one hell of a milk run, but the Shoreline grid just looks like a mess.
    *I notice the 65 is much shorter than the 75. This is my completely uninformed idea of ridership distribution, but would it help make the 78 less milk-run-y if the 65 stayed on 35th and inherited the route of the 69 to Kenmore, while the 69 did everything I have the 78 doing above?
    *Again I get you’re trying to serve Link stations, and I read your comment on ridership from Ballard to Roosevelt, but W Green Lake Way serves absolutely nothing, the 50th alternative is too far out of the way and too redundant, and the 78 mostly duplicates other routes between Aurora and Roosevelt/75th. What would the consequences be of routing it Linden-Winona-Wallingford-80th-Roosevelt-75th?
    *It’s almost like the 28/40 render the SLUT almost entirely pointless… hmm.
    *The 47 is still a ridiculous stub of a route! I guess I can kind of accept it when it’s only running every 15 minutes in a plan where 15 minutes becomes only “almost” frequent, but what do you think of this idea, which I’m not asking you to institute, just throwing out there: truncate the 35 at Capitol Hill Station and extend the 47 along Bellevue Pl, Boyleston, Prospect, and the 35 routing to the U-District. With Link I can’t imagine there’s much call for the 49’s current U-District-Capitol-Hill connection function. I’m assuming the Aloha Extention gets built; if not, the 35 will have to run for longer, but perhaps it could be through-routed with the 10 along Boston.
    *Why does the 59 serve Arbor Heights instead of the 21? Is it just so both routes serve as much of Westwood Village as they can?

    1. Shoreline is the weakest part of this proposal, no question about it. I’ve been thinking about ways to make the north-south service there less spaghetti-like while still serving Link stops. My current thought is to straighten the 87 and eliminate the 86 north of 145th, “replacing” it with another east-west route parallel to the 78 that can also take over the Richmond Highlands segment. Folks in Meridian Park would lose north-south service but would have east-west service every half mile that would take them to Link. Expect a revised Shoreline proposal from me in time.

      The idea of the 65/78 division on 35th is that passengers south of 75th are better served by a direct route to the UW Campus while almost everyone north of 85th who needs to get to points south is better served with a quick Link transfer using 75th.

      I thought about having the 78 go around the north end of Green Lake. Traffic did the concept in. The routing around the south side is quite a lot faster. The important connection here is the one between Phinney and Roosevelt, and that connection is best served by the southern routing. Both the 78 and the 16 are using West Green Lake Way not because it has a lot of ridership on its own, but because it’s the fastest way to travel between ridership nodes.

      The 47 only exists because ST stupidly didn’t build a Summit Link station. The point is purely to provide a connection between Summit (a very dense, high-ridership area) and downtown that wouldn’t exist in the absence of the 43. Fortunately it’s cheap to run. I’d rather keep the 35 as the main north/south route in the Broadway/12 Ave S corridor rather than bisect it at CHS.

      The 21/40 combo is already too long, and serving both Arbor Heights and Westwood Village with it would require an awkward deviation. (Of course, the 59 itself makes an awkward deviation to serve Westwood Village and is rather long in combination with the 6, so maybe that’s not a compelling argument.)

      1. I get the sense that too much emphasis on serving Link stations, especially in North Seattle, ends up conflicting with trying to create a clear grid. Part of that, of course, is the lack of possible east-west routes forcing the use of north-south routes for part of the journey, and part of it is poor station spacing (where’s my Mapleleaf station dammit?!?), but I still get the sense you’re the least Walkerite-fundamentalist of STB’s regular contributors.

        Was a Summit Link station ever in the plans? Would it have been taken seriously by people who don’t live there if it was? If, as I suspect, the answer to both questions is no, it’s not so obvious that to exclude it was “stupid”. Remember, ST was originally going to deviate pretty far to the south to serve First Hill, and unless you want to deviate just as far to the north, creating an even worse winding effect unless you change the profile of Cap Hill station (and even then the winding would still be noticeable, especially if you don’t use the Convention Place space), you probably can’t put a Summit station further north than Pine and Bellevue. It is pretty close to Cap Hill station if you ignore topography.

        What would the impact be of the east-west segment of the 51 serving Arbor Heights and the north-south segment of the 51 inheriting the rest of the 59?

      2. A Summit Link station was definitely not compatible with the First Hill plans. But once those were scotched, it should have become part of the plans. It would be at Bellevue and Pine, which is at the south end of a huge pocket of density that is topographically very separate from Broadway/John. Westlake and CHS are just too far apart, and Bellevue would be the logical place to put a station in the middle.

        The 59/51 swap you describe would result in an extremely long 6/59 combination. (I actually had it that way at first, but the route was too long and was guaranteed to be unreliable.) I like the 6/59 through-route because it connects Boulevard Park and much of South Seattle so naturally and quickly.

      3. Another issue I forgot to mention with a 6/51 combination: 35-foot buses that are needed for the 51 are likely inadequate for the 6. I think the 6 would need mostly 40-footers with a 60-footer here and there when the schoolkids are commuting.

      4. Bellevue and Pine station would have been amazing. That entire pocket is now squarely between Link stations, on a hill.

        Perhaps the only consolation is that the hill is monotonic: this way, you’re not completely and utterly screwed if you can’t walk downhill xor uphill.

        Maybe Santa will hire some mole rats, plus a few beavers, to deliver us this, via driverless sleigh.

      5. Regarding the 47, I’m curious how much it would cost to build northbound wire on Bellevue Ave. Summit Ave is a ridiculous street to run buses on — it has *traffic circles*! Shifting service to Bellevue Ave in both directions would simplify the route, and would probably lead to operational savings, too. But it’s certainly not a high-priority capital project.

      6. Summit Station was never in the original discussions that I heard. Basically, when ST revived University Link from its coma and changed the Ship Canal routing and later deleted First Hill Station, it did not review any of the other stations or potiental stations; it just stuck with what it had decided before.

  17. I’m not sure how 3rd Ave can handle all those buses. I would think that 2nd Ave would have to be usurped for additional local service. I would wonder how you address the Downtown routing for Metro Expresses and south/east local lines that aren’t through-routed, Community Transit services would be curtailed to Lynnwood in the 2020s so all the 400s would go away, but you still have the Central Connector and consolidating of SoundTransit Expresses (all to 4th and 5th?). It is physically impossible for 1/3, 4/14, 5,13, 15, 16, 21/40, 24/131, 28/124, 33/132, 54, 58, and 120.

    Also, what happens to routes from SE and SW King County to Seattle? Are they truncated?

    All-in-all, this looks really good though.

    1. The increase in volume of buses on 3rd is actually not that dramatic. You’re increasing a lot of frequencies, but you’re also removing two of the most major routes entirely (7 and 36). If we have a transit-priority downtown street, let’s use it to the fullest.

      South King County is not considered in this proposal except for those routes that are primary service for communities in Seattle (106/107/120/124/125/128/131/132).

      1. Separately, what are the service hours for this now and what would we need under the 15% and 33% levels?

    2. I think you’re being a little over optimistic. Even if we pushed off-board payment and changed some policies to fare penalty for onboard payment, I don’t think that would increase boarding efficiency dramatically to compensate. Plus, each platoon of buses in the peak core area literally takes up the full blocks. The two-block grouped stops aren’t good enough and buses can’t really jump each other when ready to depart. My point is, while you might not being adding all that many more buses to the soup, the soup is already fucked. What more do you want for TSP? It’ll come at further cost to the east-west buses that also flood the peak core.

      1. What David is proposing here is a system solid enough for all-purpose transit usage to become ubiquitous among Seattleites.

        One nice thing about ubiquitous-transit cities is that the public trends toward near-total smartcard adoption. You’d be amazed how much faster things get when virtually no one is paying with cash.

        (Frequency also gives drivers a reason not to indulge dumb, long-winded questions. “I’m sorry. If it turns out you needed my route, there’s another one along in 7 minutes. In the meantime, there’s a system map over there.”)

      2. As someone who has a view of 3rd from University to Virginia from his office window, I disagree that 3rd isn’t working. Signal timing could still use some refinement (although it’s gotten dramatically better since the 3rd “mall” opened) but the only time I ever see jam-ups on 3rd is during PM peak or when there’s an obstruction. What’s more, just one stop — Pike northbound — is the source of 90% of the jam-ups. Get the 7 and 36 out of that stop, and rebalance the remaining routes a bit, and there will be no problem.

        Belltown is worse, but SDOT is planning improvements to that part of the corridor to make it work as well as downtown.

  18. Looking at this some more, I’m a bit concerned that you are over-serving Northgate, and under-serving 130th. The main reason we need a station at 130th is to serve Lake City, which is bigger than Northgate, and Bitterlake (which is comparable). But Lake City to the train happens only every 12 minutes. I would like to see that doubled. I would like to see a bus that goes through that little 15/69 loop in Lake City, but then just connects to the western part of the 75 (in Bitterlake and Broadview). This would be a relatively short, fast bus.

    In general I think serving Northgate should happen less (since it is crowded mess in there) and buses should serve 130th more (because it isn’t). The biggest advantage to the Northgate transit center (it has express lane entrances and exits) goes away once the train station is built.

    Another possibility is to alter the 75 and 15. Starting from Sand Point, go to Lake City, then head south on Lake City Way to Northgate, on the line now being served by the 15. Turn around at the Northgate Transit Center. Now, the 15 doesn’t go to Northgate anymore. From Ballard, it follows the same route, but takes a left on Greenwood (or Aurora) then takes a left on 130th until it gets to Lake City. I think this would be much faster. For example, if you are at 105th and Greenwood, you could get to Lake City or just the train much faster. You lose a little service along Northgate Way, but that part of Northgate Way is nothing special.

    Along the same lines, you could move the 73 over to serve 130th, assuming there is a good place to turn around (which might be the hardest part of this idea). That still leaves plenty of direct service to Northgate from various spots (Lake City, 15th, Roosevelt, Northgate Way, Maple Leaf). I think you can get from 15th and Northgate Way to the 130th station faster than you can get to the Northgate station.

    These ideas are in priority order (I like the first one the best).

    1. I understand the concern with the focus on Northgate, because Northgate is the ultimate pit of slowness. The reason for it is that Northgate is not just a Link station, but is also an “urban village” (in designation, not form) that is going to be the locus of some of the most dramatic growth in Seattle in the period we’re looking at here. Greenwood, Crown Hill, and south Lake City are all also set up for major growth. For some reason (that I don’t quite understand) 105th and Aurora is also a stop with extremely major ridership. I think a crosstown route using Northgate Way is extremely important, even if it’s doomed to be a bit slow, and I think it warrants the 10-minute frequency the 15 will provide. The crosstown route on 130th is very important too because of its speed for Link connections, but it serves fewer urban centers and (on the east side) would be boosted, I expect, by ST service operating between Lake City Way and 130th Station without stopping.

      Similarly, the 73 hits Northgate because it’s a key destination for Maple Leaf residents on its own. 68 ridership is very good in that segment, and if I’m going to take service off Roosevelt Way entirely in Maple Leaf — an unnatural thing to do — then I should ensure everyone can easily get to their nearby commercial center. People transferring to Link from the 73 will almost certainly do better using Roosevelt.

      The most valid place where I might want to rethink Northgate connections is with the north-south service immediately north of Northgate. As already mentioned in another comment, I’m not happy with this proposal for the 86 and 87, and I’m thinking through better ways.

      1. I agree, Northgate will grow. But I don’t think Northgate will ever be as big as Lake City. There are just too many natural obstacles. The freeway cuts right through it. Even with a bridge (to the college) you have lots of land being used up by all those lanes. In the other direction there is a creek and a hillside to contend with. Meanwhile, Lake City is already much bigger, and continues to grow at a very good clip.

        Much of the growth around Northgate is happening on Northgate Way, close to 8th avenue. Thus, you are absolutely right — the corridor between Northgate and Lake City is a good one. I would definitely keep a line there.

        I could easily see Bitterlake grow as well. There is so much underused land close to Aurora it boggles the mind. The area just needs a bit of momentum. You can see the tide turning now in the southern part of Aurora (two new buildings between 40th and 45th being built as we speak) and I think it is only a matter of time before the same thing happens to the north.

        One of the remaining really nice things about the Northgate Transit Center is that it is a great place for buses to turn around. So I would definitely have a line that goes from Lake City to Northgate. Doubling up and serving the section of Northgate Way between 15th and 5th (as you do with the 73) also makes a lot of sense.

        But the section of the 15 between Aurora and the Northgate station is not only a pit of slowness, but has very few riders. Based on what you said, going up Aurora makes sense (so you can pick up those riders on 105th and Aurora). After 105th the ridership gets a bit dead (so to speak) but there are a few apartment buildings squeezed in next to the cemetery. If Aurora grows the way I expect it to grow, then the section north of there could be reasonably dense. I think the time savings from going that way could be huge.

        One key question: Can you get from 105th and Aurora to the Northgate Transit Center faster than you can get to 130th and the freeway? It is hard to say. It probably depends on the time of day. f it is about the same, then I think it makes sense for the 15 to use Aurora. If you are coming from Lake City, then the Aurora 15 is much faster to Ballard, and becomes the fastest way to Ballard. Coming from the north, via a train, it would be little bit faster (one less stop). Northgate loses out to Lake City, but like I said, it will never be as big. South of Northgate, you are probably better off taking a different bus (regardless of where the 15 goes). Mainly, though, I want to speed up the 15, because I think it could be the best way to get to Ballard from the north end.

      2. Oh, speaking of the 86, I don’t know if the bridge over the freeway changes anything. A bus could easily just end inside the college (close to the bridge) and just turning around. That would be nice for the folks in the school and not bad at all for everyone else (I could walk across that bridge faster than a bus could go all the way around). I don’t know if that would save much time for the bus.

      3. Because of the college and the general proximity to the freeway and Green Lake south of there, Meridian north of Northgate is a conceptually different route than south of there. That said, if the bridge allows you to terminate the 86 *at* the college, there’s no reason not to go *through* the college and tie it off at Roosevelt Station.

  19. I notice that even under the “ideal” network, all the West Seattle buses still have to make a one-block detour to serve their stop at the West Seattle junction. Is there any reason for this? As far as I’ve been able to tell, there is absolutely no reason why the stops can’t be moved half a block and the 54 rerouted to turn left directly from Alaska to California.

    1. The block of Alaska between California and 44th is ideally suited for major bus stops. It would take a lot of reconfiguration to create an Alaska Junction that made sense without using that block, especially because buses would be stopping immediately after turning, which tends to block traffic at major intersections.

      What I’d rather see than elimination of the jog is aggressive transit priority at the Edmunds/California/Erskine Way light. It’s that light that makes the jog time-consuming.

    2. It’s not so much the turn from Alaska to California but the reverse. The way the light is set up, there is a signal for right turn only for the right lane. The bus would either have to take up both lanes to make the turn or it would have to waste the turn only light to wait for an opening to swing out to turn right.

      No turns on red on any corner, btw, as it’s one of the 2 “Walk all ways” crosswalks in Seattle.

      1. There are more than 2 :) Alaska/California, the one on Beacon Hill, the one at the Market, and two more further south on 1st Ave downtown.

  20. I made a summary of David’s detailed page which makes it easier to grasp the whole if you’re familiar with the existing route network.

    1. Thank you for pointing out the error with the 124. I’ll figure out what the night service level should be once I’m at home with my spreadsheets. It’s either 15 or 20 minutes.

  21. This looks amazing! Again, major props for your hard work.

    Of course, I’m happy to see that many of my comments about the initial map have been addressed. I have no idea if they’re because of my suggestions, but it doesn’t much matter either way. :)

    I have a few minor comments/questions:

    * I love that the 15 now goes to Northgate and Lake City, though it kills me to see the transit center deviation. But I guess it’s not so bad, because through-ridership will probably be very low.

    * Some thoughts about Shoreline:

    – Straighten out and extend the 5; skip Shoreline CC, and extend it north to 200th and east to Aurora TC.
    – Keep the 86 on Meridian all the way to 200th, terminating at Aurora TC.
    – Keep the 87 on Roosevelt, all the way to the 236th Link station.
    – Extend the 67 north on 5th to the 185th Link station.
    – Reroute the 88. Heading east, stay on 185th until 10th Ave NE, then turn left, and continue on NE Perkins Way to Bothell Way.

    Virtually every corridor that’s served by your map continues to be served by this proposal. Shoreline CC has the 78. Meridian has the 87, including the part that has a “gap” on your map. 5th Ave has the 67. 145th has the 78. 185th has the 88. 15th NE has the 87. The only “lost” service are the short E-W segments on 155th and 175th, but in both cases, all of those riders are within the walkshed of either the 86, 67, or 87.

    * Every time I look at the 3, I wish that it took Cherry instead of Jefferson. That would involve new wire, and so it’s probably not worth it, but you would remove a handful of turns from the route.

    * How come the 7 detours on Jackson, instead of just staying straight on Boren? Is that just to connect with the 34?

    * I’m still curious why the 78 and 65 play musical chairs. Why not just send the 78 south on Montlake, terminating at UW Station, and then have the 65 act as the 65th crosstown, with the 40th Ave terminus loop?

    * I know I brought this up last time, but I think that the argument for removing the 7 from downtown is identical to the argument for removing the 70 from downtown. I believe that the demand to get downtown from Eastlake is comparable to the demand to get downtown from Rainier. If we think it’s okay to require a connection from Rainier, then I think it’s okay to require a connection from Eastlake. In that case, through-routing the 7 and the 70 would save some service hours and enable some new trips.

    * Why have the 78th cross over on 130th? That seems like a pretty small street for a frequent bus. Why not just cross over at 125th?

    1. Thanks, Aleks. I appreciate the thoughtful feedback as always. A few answers…

      it kills me to see the transit center deviation. But I guess it’s not so bad, because through-ridership will probably be very low

      Indeed. I’ve shortened the deviation from how it is on the current 40 (and previous 75), but I think it’s pretty essential.

      Some thoughts about Shoreline:

      I think I will try to straighten the 86 and 87, although serving Link stations has to be the first priority here because virtually all riders from these areas are heading either to Aurora or to somewhere served by Link.

      I’m reluctant to extend the 5 or 67 further because those routes are already so long and both prone to delays. Also, the 5 has to serve SCC — it’s the largest destination up there by a mile.

      The 88 uses the routing it does east of 185th station because the two most major destinations to be served crosstown in that area are North City and the Shoreline library. Using Perkins would replace those with a low-density residential area.

      * Every time I look at the 3, I wish that it took Cherry instead of Jefferson.

      The problem with that is James. The impassability of James in the PM peak is why Yesler is proposed (not just by me, also by SDOT). Once you’re using Yesler anyway, you might as well use Jefferson rather than James/Cherry — it passes by the Providence front door, closer to Garfield, and by juvie (sadly, a major destination).

      * How come the 7 detours on Jackson, instead of just staying straight on Boren?

      The zones at 12th/Jackson in all directions will be a major transfer point. Transfers will be available with frequent service in every direction on the FHSC, 7, 14, 34, and 35. It’s a tiny deviation that’s worthwhile for a place where I think the majority of riders may be transferring.

      * I’m still curious why the 78 and 65 play musical chairs.

      As explained in other comments, I continue to think that the most useful single connection for Meadowbrook riders (of which a majority, I think, are Hale students) is to Link at Roosevelt. It’s faster to UW than direct service all the way along 35th and it’s much faster to any other destination, particularly downtown. There is also the need for a Roosevelt-Lake City connection.

      The early drafts of the FNP had a straight 65, but I abandoned it after doing the math on travel time. I like grids, but I’ll abandon them for the sake of speed.

      * I know I brought this up last time, but I think that the argument for removing the 7 from downtown is identical to the argument for removing the 70 from downtown.

      The difference is that the 7 is nearly shadowed by Link, and connections are to Link. There is no place on the 7 that is more than a few minutes’ ride or walk to a Link stop, and the trip via Link is almost always faster than the one-seat ride on the current 7. Connections to a non-downtown 70 would either be to a slow bus/streetcar (every bus in the SLU/Denny Triangle area is slow) or to Link via a very long and congested backtrack; both would add time to current trips.

      Also, I’m unwilling to through-route the 7 with anything else during the day. It’s already got too many reliability issues.

      * Why have the 78th cross over on 130th?

      That one could go either way, but my feeling is that more riders who aren’t transferring want to get to Fred Meyer than to 125th. Transfers are easy either way.

      1. Thanks for the responses. :)

        I think I will try to straighten the 86 and 87, although serving Link stations has to be the first priority here because virtually all riders from these areas are heading either to Aurora or to somewhere served by Link.

        I see; the musical chairs is to ensure that every route touches a Link station. I understand the rationale, but in a world where all of these routes are super frequent, I don’t think it’s as important as you claim. An extended 67 would cross Link several times (of course). The 87 would run between two Link stations (Northgate and 236th). The 86 has two options: it could turn right at 185th and terminate at the Link station, or it could continue north to 244th, and then meander up to 236th. Both of these options still accomplish the legibility and speed objectives of having relatively straight and sensible routes. Consider how much more ridership the 5-Shoreline has than the 5-Northgate did, despite the fact that the latter one detoured to a transit center.

        I’m reluctant to extend the 5 or 67 further because those routes are already so long and both prone to delays.

        The 5 would be about as long as the 58, and the 67 would be shorter. Admittedly, the 58 is also a long route, but I think that shows that it can be done. But also, unless traffic is much worse on the northern segments than I realize, I don’t think the extra length will significantly impact reliability on the route.

        Also, the 5 has to serve SCC — it’s the largest destination up there by a mile.

        It strikes me that the best way to get to SCC under your proposal would be to take Link to 145th, and then the 78 the rest of the way. I would rather see a super-frequent 78, than having the SCC frequency split between three different routes.

        I’m making an assumption, of course, which is that most people going to SCC are already transferring from another route, rather than living within walking distance of the 5’s unique stops. Certainly, this change would make it harder to get to SCC from Greenwood or Phinney Ridge. I don’t have the ridership data to quantify whether that’s significant.

        For what it’s worth, there’s another possibility. Instead of switching to Dayton, the 5 can stay on Greenwood. From there, it can take Arden Way to the college, and then continue onto N Greenwood Drive, to loop around back to Dayton. It’s a bit circuitous, but it makes it possible to serve the college and still continue north, and the deviation is smaller than it is today.

        Also, I’m unwilling to through-route the 7 with anything else during the day. It’s already got too many reliability issues.

        SDOT has suggested splitting the 7 in half. The section south of Mount Baker would be spliced onto the 48S. That seems like it would help a lot with the reliability issues, at least on the northern segment.

      2. *Even if Jackson Park Golf Course gets redeveloped I’d want a route on 5th north of Northgate to deviate to 1st between 145th and 130th; there actually is something on the freeway side of the street, specifically Lakeside School, and a more direct interface with the Link stations.
        *An all-Meridian 86 could still serve the segment Aleks wants to serve with the 5 between Shoreline CC and Aurora Village. It wouldn’t be able to connect to Link except at Northgate, but it would connect to not only the 58 but Swift as well.

        “SDOT has suggested splitting the 7 in half. The section south of Mount Baker would be spliced onto the 48S. That seems like it would help a lot with the reliability issues, at least on the northern segment.”

        I have a feeling people in the Rainier Valley would scream bloody murder if that happened; yes, there’s a Link entrance on 23rd, but I have a feeling a one-seat ride to Little Saigon is fairly important for at least some people south of Mount Baker. The TMP is pretty good at identifying corridors for investment, but I came away from it thinking SDOT should leave routing to the professionals.

  22. Another thought.

    Metro’s 17% cut proposes to take the 28 out of Fremont. I think the main reason for doing this is to avoid the need for articulated buses on the revised 28X. This way, the buses can be used on another route that needs them more, like the 40.

    I would like to see the 40 have 10-minute frequency. I think it needs it. I think you could save some of those service hours by cutting back the 16 to 12-minute or 15-minute frequency (see my post above on changes to the 16/63). And if you do that, then I also think it makes sense to have the 28 take Aurora and use 39th, at which point a 40-foot bus will probably suffice.

    Of course, if we end up building a Fremont Link station, then we’ll want to send the 28 to meet that station. But in the meantime, we might as well have it take the faster route, and avoid the capacity mismatch of serving Fremont.

    (This is another reason that I think the 5 and 40 are the best routes to serve Fremont. Both routes have strong demand north of Fremont as well, and so 10-minute articulated service on those routes will be well-used.)

    1. I don’t think the 40 by itself will need 10-minute frequency anymore once the 15 is clearly the faster option from Ballard to downtown, which it will be with no Uptown deviation. The real ridership crush is along the Westlake-Fremont corridor, and that’s why I want 8 buses/hour there.

      It also seems a bit perverse to me to a) cut off 8th Ave NW from Fremont, which is clearly the transfer hub for the entire north-central area even without Link; and b) give the relatively few riders along 8th Ave NW a special fast bus to downtown that the great bulk of riders from the rest of the area don’t have. If anyone from north of Fremont (other than Aurora riders) should have a special fast bus to downtown, it’s 5 and 16 riders, given their numbers and the lack of alternatives. That works out more naturally with the 16 than the 5, and I still feel a bit bad that my plan makes Greenwood riders’ trip downtown so slow.

      1. That’s fair. I guess there are a lot of folks for whom it’s a tossup between the D and the 40. Personally, even with the proposed 15, I think that folks who want to go to central Ballard will take the 40, because it goes right there. But I might be wrong.

  23. Very impressive work. I’m short on time to contribute to this discussion right now. But I had to express my gratitude for all the effort that went into this.

    Thanks David!

  24. I made a summary of the details page to make it easier to see the whole network at once. It also lists which ST2 Link stations each line meets. Speculative stations on the Ballard, 45th, and West Seattle Link lines are in [brackets]. I did not consider other Link lines, or streetcars beyond the FHS and SLUS.

    I tried to post this yesterday but somehow my Firefox and STB stopped getting along, so I’m trying in Chromium.

  25. David, I have nothing intelligent to say beyond that this is phenomenal work. As someone who is traveling more and more between the Rainier Valley and various points in West Seattle, I think the network you propose here would be transformative, and would get me out of the car for a significant number of trips.

    Your proposal doesn’t address my particular hobby horse about the 50 diverting through Sodo, but I have to do a bit more research before condemning what you’ve done here.

  26. I see several interesting innovations here. I don’t have much to say about specific routes because nothing is really must-have or must-avoid; the main issue is getting some kind of frequent network in place, and I’m seeing some encouraging commonalities between this plan, the FNP, the TMP, Metro’s ideas, Aleks’s ever-astute suggestions, etc. The commonality is probably what we’ll get if we can convince the city/county to commit to this level of service, and the details of route termini and coverage will take care of themselves. As I told David originally, the main reason we need this FNP is not to say exactly how to get from Arbor Heights to Rainier Beach, but to show the level of service we want, how it would function as a network, and how much it would cost. Then you can start trying to make it happen.

    New interesting innovations I see here:
    1/4: Similar to my suggestion of a Queen Anne shuttle loop from Uptown Station combining the 1/2/3/4.
    2/10: Interesting to put the 2 on Pine to partially make up for the 11. I can see this working well. I can also see complaints from those wondering where’s the bus to Virginia Mason.
    6: This gets the “south part of the 8” to Broadway.
    8: Like. Better to overserve Madison Park than underserve it. A full crosstown route from Magnolia to Madison Park would help overcome the isolation caused by geographic barriers.
    12: Would rather call it the 11. Is 19th really losing bus service? And nobody noticed?
    28/40: Interesting making a common corridor with these routes.
    47: I worry about 1st & Yesler killing reliability.
    51/52: Puts Seacrest on a regular route. Good connections between West Seattle and southeast Seattle. Perhaps saturating the market? But overservice is better than the perennial underservice we’ve had, and it would encourage people to take transit.

    The main flaw is the spaghetti on Meridian and 15th in Shoreline, but that’s been discussed above. It’s actually hilarious. Three routes on different parts of Meridian sounds like a satire of Metro’s bad old days.

    1. I think 19th is losing service in all worlds. The only reason it still has any service at all is that Metro wants to continue providing service along Madison up to 19th, and the 11 would need a frequency upgrade to do the trick — something that Metro can’t currently afford.

      Regarding Virginia Mason, I had an interesting thought, which also solves another issue. What if the 7 continued north only to Pike/Pine, then followed the 10’s wire to downtown, live-looping on 2nd Ave? You reconnect the 7 to “downtown” (3rd and Pike), and you maintain a one-seat ride to Virginia Mason (albeit at Boren instead of Terry). The only coverage you lose are a couple blocks in the vicinity of the I-5 ramps, which hardly seems like a travesty.

      I guess you also lose a one-seat ride from Rainier Valley to SLU, but that doesn’t seem like a travesty, either. Another possibility would be to continue north to Stewart/Virginia and use the 70’s wire to downtown. That would give you a stop at Boren/Fairview that’s within spitting distance of the Amazon buildings, but it would probably cost more service hours, compared to using Pike/Pine. My instinct is that the shorter routing would be better, but I’m prepared to be wrong. :)

      1. Aleks, that 7-to-Pine idea is a really interesting idea. You couldn’t do the 2/10 live loop because the trips are too long; you’d have to use the current convoluted 49 routing to a layover on Virginia Street, which would require 1-2 more buses than my proposed 7 routing. You’d lose the SLU (Westlake)-First Hill connection which I’m trying to add (and which I think would have very high ridership once people internalized it), but you might gain a lot from the downtown connection. I’ll think it over.

        19th should lose service. One poster complained about it in the FNP comments, but I didn’t find the complaint convincing. The area has N/S service less than 1/4 mile away in both directions and nearly zero ridership, particularly ridership that can’t be served by the E/W service on the 6/8 (in ideal network world) or 8/43 (in today’s world).

        The 8 turns on Westlake because the Westlake/Denny intersection is a major destination and there is nothing in the area of Fairview/Harrison. I debated that one, but I think the coverage is worth the minute or two extra delay at PM peak.

        While the 5 will serve Fremont, I’m trying to promote the 28/40 as the “principal” Fremont corridor; it’s faster, it serves the heart of SLU, and it serves more of Fremont. Tons of riders live between the 5 and 358, and I’m trying to promote commonality between those two routes. (You’ll also notice I moved the 28/40 to Virginia and Stewart, where the 17 ran before the SLUSC was built, to maximize straight running in the Westlake corridor.)

      2. Aside from the common corridor, here’s my two arguments for the 5 routing:

        – I think it would be slightly faster to downtown. This is my biggest argument. Given the importance of the 5, and that we’re already penalizing it by going through Fremont and taking Dexter, I would like to see it made as fast as possible everywhere else.

        – As you say, Westlake/Denny is an important intersection. I think you’d get a lot of ridership between Phinney/Greenwood and SLU.

        For what it’s worth, I’m still wondering if it would make sense to have the 5 use Westlake and the 40/28 use Dexter. Especially with your proposed 15, I think the 5 would have the most downtown riders of the three services. So the 5 should get the fastest route of the 3. And once all the construction is done, Westlake should be noticeably faster than Dexter.

      3. I was wondering about reversing the 5 and 28/40 as well, from the other perspective: Dexter is denser and is always going to be denser, so it makes sense for it to get the more frequent service (besides not slowing down the long-haul commuters on the 5).

    2. Mike, your big-picture thinking is exactly right. There are two reasons I go into such a high level of detail in these proposals, and neither one is because I’m attached to these particular details. I’d be thrilled with any network built using the principles I used to build this one. The reasons I go into so much detail:

      1) I want to show convincingly that what I’m selling (i.e. a FNP with no more hours, or a revolutionary network with 33% additional hours) is actually operationally possible. Too often, transit activists propose things that work in the big picture but just fall apart when you look into the details.

      2) Occasionally, I do have specific ideas which I haven’t seen in Metro documents or in other people’s plans (for example, building a terminal/turnaround at Seacrest Dock, or turning around a U-District trolley route at the Campus Parkway turnback) which I want to get into the conversation.

      Now on to a couple of your specific items…

      12: Would rather call it the 11.

      One principle I’ve been sticking to is to prioritize the highest-ridership segment. First Hill has more ridership — much more — than Madison Park. Thus I keep the route number for First Hill.

      47: I worry about 1st & Yesler killing reliability.

      Indeed. It’s the 47, so I don’t care too much. And it’s not too bad except late on Fri/Sat night or at game time. I might consider a reroute during those times, to turn around using Madison and Marion.

      51/52… Perhaps saturating the market?

      Probably. One thing to keep in mind is that these routes would be using small 35′ buses. On California, I think there is potential for wild growth in ridership.

      Further south, the 51 will do fine. I think the 52 would have very low ridership for its first few years, though. It relies on 1) High Point continuing to develop; 2) people internalizing the connection between High Point and Georgetown/Othello; and 3) people in Seward Park finally using Link like they used their old service once they have frequent connections.

      1. By the way, I want to echo Mike’s big-picture comments. I’m talking about details because it’s fun for me (and presumably for you too :P), but the overall message — “frequency is freedom!” — is much more important than any of the little things. :)

  27. Some more thoughts :)

    Why does the 8 turn on Westlake, rather than Fairview? It seems like the more of Denny you can bypass, the better…

    Why does the 5 turn west to 7th Ave (nee Aurora), rather than east to Westlake? If we’re creating a common corridor, I’d rather unify the 5 with the other services going to Fremont (i.e. the 28 and 40). I would also expect the Westlake routing to be a faster way to get downtown (once 7th Ave becomes part of the grid again), but I could be wrong.

    1. Another advantage of Westlake is that it is the bottom of the both Denny hills, preserving the 41.4% (ok, it’s not exactly 90 degrees, so a little less) shorter / one-seat path from Capitol Hill to Belltown / Whole Foods – Amazon / Fremont+ buses.

      (Sorry, I keep whining about not being able to go downhill even though my leg is now better, but it did really change the way I had to get around. I imagine that walking downhill is tough on the elderly as well, extrapolating from someone mentioning that walking down stairs was harder than walking upstairs.)

Comments are closed.