North Seattle Restructure after Lynnwood Link

17 Replies to “North Seattle Restructure after Lynnwood Link”

1. The theory behind a grid system has one major design weakness in North Seattle: most commercial districts run north-south. To offer residents opportunities to make local trips without transferring, some sort of zig-zagging will remain useful.

   I’ve wondered if there could even be utility in having either C-shaped or O-shaped routes ending at Link stations as a design alternative to a zig-zag approach.

   1. Much of the zig-zagging in the north end is due to the Northgate Transit Center. The author manages to avoid a lot of that by killing off the 347/348 and 41. The only new bus headed to Northgate is a bus on 5th, which is much more of a straight shot than the 41 it replaces. Meanwhile, you still have service on 15th/Pinehurst Way (like the old 347/348) except instead of going to Northgate, you go to the U-District. You are still connecting those north-south neighborhoods, you are just connecting them in different combinations. For example, Pinehurst is still connected to Lake City, but it loses its direct connection to Northgate. In exchange, though, it gets connected to Maple Leaf and Roosevelt, which would probably be a faster connection to Link.

      As far as C-Shaped routes, I would expect it. Actually, I think there will be a bunch of L shaped routes. Swift is going to head down SR 99 (as it always has) but then make a turn and head to Link. It wouldn’t surprise me if Metro adds a similar run north of 145th.

      The drawback to that approach, though, is that it gets very expensive. You redundantly add service on SR 99. You probably still want a bus that just goes straight from downtown to the county border (the E) so that means every bus that cuts over is somewhat redundant. But that is the trade-off, and one worth considering. Do we have a bunch of ‘L’ buses, or add enough east-west and north-south service to make the transfer painless? Personally, I would choose the latter.

   2. Most commercial districts run north south because that’s how Seattle transit is aligned. Reorg with more east west routes would generate more of a X or balled up commercial district.

      Interesting how the 45 would through route to Brooklyn rather that Roosevelt, which it almost does by a mere 2 blocks now. I’d let 44 serve Brooklyn and let the unnamed Rapid-Ride esque Roosevelt service handle the in-between.

2. I like this idea. I think one of the more interesting things about it is that you avoid a lot of curves by sending fewer buses to Northgate. This seems counter-intuitive, since Link adds a stop there. But one of the current advantages of the Northgate Transit Center is that it sits right next to the freeway. The 41 runs often enough to make transfers easy, and thus once you get to the Northgate TC, you can get downtown very easily.

   But when Link gets to Northgate, it also gets to Roosevelt. Someone who is riding the 347/348, would probably be better off if the bus kept going on Roosevelt all the way to the U-District. That way the rider gets a direct connection to the UW, and getting to downtown is just as fast (via the Roosevelt Station). That isn’t exactly what you designed, but that is basically the idea, and I like it.

   I also like the idea of sending the 75 east to Aurora. There is a disconnect between the Lake City to Northgate section and the Lake City to Sand Point section of the 75. This will only increase if you get rid of the 41. Lots of people will want to go between the two big neighborhoods in north Seattle, and not that many people are headed to Sand Point. It makes sense to separate them, which could allow you to really beef up service on the more important connection.

   The only problem with the whole thing is service on 5th. It sounds great on paper, but crossing 5th close to the freeway is very difficult. You have cars coming off the freeway, and they aren’t required to stop: https://goo.gl/maps/GwgQhupPYxq. Thus a bus could be stuck there for a very long time. I suppose the city could change that intersection, but if not, then I’m afraid that just wouldn’t work. The alternative is to go around the east side of Jackson Park, but then you lose one of the advantages of the proposed change (just staying on 5th the whole time).

   I would be temped to just keep the current 347/348, but not send them to Northgate. Send them both to the UW (via Roosevelt) instead. Keep the 41 (from Lake City to Northgate) but get rid of the 73. You could also get rid of the 67 at that point. Along 5th you would have the D and 41. Along Northgate Way you have the D. Along Roosevelt you have the 347/348 north of 65th, and a bunch of buses south of it. If you do keep the 67, you could ramp it down a bit (to every half hour, hopefully offset from the 347/348) as much of the route would be redundant.

3. Northgate is not just the transit center, it’s the entire urban center which is the largest jobs/retail/education/service center in far north Seattle. That’s why all the buses go to it because a lot of people are going to it, the way a lot of buses go to the U-District because it’s the other urban center in north Seattle. I have long liked some aspects of an all-15th 73 (or 377), but the fact that it goes through low-density residential areas and passes close-but-not-close-enough to Gtreenlake, Northgate, and 155th & 5th is a weakness because it goes where people are coming from but nobody is going to, and to get to a place that even those residents are going to you have to transfer or take another route. This is another problem with Seattle’s grid, that several grid lines don’t have any anchor destinations or just have one at one end, so they don’t get the full two-way traffic that grids in Vancouver or San Francisco get, where one person is going one way from his house while another person is going the other way to a destination. North Seattle has the largest rectangular area in the city where a grid can work, but the distribution of non-residential destinations is still clustered in islands, which makes trip patterns still more node-based than in cities with mixed-use and commercial destinations everywhere, and thus a lot of people are going to these urban-village islands, and aren’t served by too rigid grid routes that basically serves nobody but are grid-correct. I still tentatively support an all-15th 73 (377) because it’s the closest thing to a major straight strategic corridor northeast Seattle has, but it’s still a tradeoff because there’s not much on it directly.

   Anthony’s 45 map has a feature not in the article: it includes the 17’s tail.It would be nice if it went all the way down the hill to Golden Gardens, which has no other transit. As it is I don’t think it would help pedestrians get to the park easier because they’re supposed to use the stairs rather than the switchback road, but I wish we could do something to make the park accessible.

   Why do you connect the 75 to the 45? The 75/31/32 is well-used and a proper grid corridor. Those coming from the western 45 to U-Village would do better to transfer to the 44 which has your U-Village extension that SDOT/Metro are also planning. There is one advantage to keeping the 75 in north Seattle though: it avoids the Fremont Bridge congestion.

   1. The road down the hill to Golden Gardens has way too many sharp turns for a bus. I think a Seaview route would be better, like the Route 68 I proposed in my Ballard restructure. Or maybe a bus that runs on Seaview between Golden Gardens and the Locks, then runs to Downtown Ballard. The problem with such a route is that there isn’t much activity along Seaview itself.

      Right now the 75 runs through the UW campus via Stevens Way, and I think keeping that connection makes sense. In that case, it makes more sense to route the 75 to Brooklyn Station rather than Husky Stadium Station. Sending the 31/32 to Brooklyn Station would result in too large of a deviation, so I sent them to Husky Stadium Station instead. The point of the 75/45 through-route was more to reduce the number of buses in the UW area rather than to provide a new connection, but the good part is that NW 85th residents would have a one-seat ride to Children’s Hospital.

4. >> Northgate is not just the transit center, it’s the entire urban center which is the largest jobs/retail/education/service center in far north Seattle.

   Oh come on. First of all, we are talking about the transit center, not just the neighborhood. Without the bridge, this puts you on the opposite side of the college. There are a handful of medical buildings in the area, but nothing like Northwest Hospital. In terms of density, Lake City still has more people. The area by the transit center really is a minor location from a destination standpoint.

   The main advantage of Northgate Transit Center is that it is next to express lanes which enable a very fast ride to downtown. That advantage goes away, and other locations (like Roosevelt) become comparable. Of course the main destination in the north end is the UW, and all other destinations are puny in comparison.

   The Northgate Transit Center, meanwhile, has several big disadvantages. It is a dead end. You can’t head east from there, making it a terrible location for a bus. So much of the walkshare is taken up by the freeway (even when they add the bridge). Northgate Way and Roosevelt or even Northgate Way and 5th would be a decent location. But the TC is in a terrible dead end, and would never be considered for a light rail stop if it wasn’t dirt cheap to add one there.

   As for 15th being barren, that simply isn’t true. For much of it, sure, but not all of it. Besides, if that is the problem, there is a simple trade-off — use Roosevelt Way. Roosevelt Way puts you right next to the action, from the U-District to as far north as you want to go. Eventually you make the jog over to 15th (via Pinehurst Way) and that works well as a means of getting you by the commercial and residential center of the various areas (i. e. places where people want to go). That leaves a common and fairly easy choice. Do you go right in the heart of things (the heart of Roosevelt, Maple Leaf, Pinehurst) or do you skirt them all, and ask people to walk a few blocks? Doing the latter would mean a faster bus, but more walking. I really don’t have a strong preference, but I prefer either over taking a bus which makes several turns before it finally comes to the transit center.

   >> North Seattle has the largest rectangular area in the city where a grid can work, but the distribution of non-residential destinations is still clustered in islands,

   That’s really not the problem. The problem is geography. You can’t run an east-west line from, say, 95th. As it is, most commercial corridors are covered by buses, just not buses that form a grid. Lake City Way and Roosevelt Way/Pinehurst/15th have non-residential destinations, and bus service. Northgate Way has bus service as well, as does Ingraham and Nathan Hale. It is just that none of those places form a grid. Instead they have buses that curve around, and make getting from neighborhood to neighborhood difficult. From Lake City to Ingraham, for example, is a joke. The same is true for getting from Lake City over to anywhere on Northgate Way east of the freeway. But the worst problem, of course, is that making a transfer to anyplace on say, Aurora, involves going way out of your way to the Northgate Transit Center (not Northgate Way) which is just a bad design. I understand why it evolved that way, but it still isn’t a good system. If you think it is frustrating trying to get over to Aurora, listen to someone who is flabbergasted by the trip from Ballard to Lake City.

   We can do better, and we should do better. It is probably too late to have a bus that goes right along Northgate Way, from Ballard to Lake City. The Transit Center (and now Link) is just too much of a pull. But a bus that connects Bitter Lake and Lake City (and thus serves Ingraham High School, Pinehurst and lots of other destinations) is definitely a possibility. It will happen once a Link station is added at NE 130, but it is definitely reasonable to add it before then.

   1. I agree that the role of Northgate Transit Center will fade once Lynnwood Link is opened.

      With a high-speed, high-frequency line making multiple stops in North Seattle and Shoreline, trip-makers will seek to get to the nearest station –which isn’t necessarily Northgate.

      The popularity of Northgate evolved because it was at the end of the express lanes. It increased when commuter parking became an option. While those facilities remain, their roles will be significant changed in light of Link.

      Its shortcomings of good east-west connectivity make it really undesirable for many to want to reach. It’s also on the edge of a high -activity district and not in the middle.

      Three restructures from now, 185th and U-District stations will likely emerge as more important for north-south bus connectivity because they either are at the end of service areas and make a good layover spot (assuming the Metro remains serving King County only) for 185th, or the trip making funnel in North Seattle combined with density of trips in the station area will likely make the U-District station more popular (unless traffic congestion pushes riders to an adjacent station).

      Although the situation is not identical, it’s notable that Mount Baker Station today has fewer boardings than either Beacon Hill or Columbia City do. (https://m.soundtransit.org/sites/default/files/2017-q2-service-delivery-performance-report.pdf) I don’t think that there were many long-time transit advocates who would have expected that!

      1. I think the Mount Baker situation is a great example of why details matters. If you look on a map and are told that the 7 is an extremely popular bus, you would assume that ridership at that connection point would be huge. But it isn’t, and part of the reason is that the transfer is horrible (https://seattletransitblog.wpcomstaging.com/2012/04/18/the-awfulness-of-mt-baker-station/). For a lot of people, it just isn’t worth the transfer.

         I don’t think Northgate will be that bad, but there are some similarities. Let’s say I’m on a bus headed south on 15th NE and want to get to the UW, As the bus approaches Northgate Way, it can do a couple of things:

         1) Head over to Northgate.
         2) Just keep going towards the UW.

         Most of the time, I would pick the second option. I think most people are that way. Even if I was headed to Link, the Roosevelt station is only two minutes further than Northgate. This means that if I’m headed downtown, option 2 is just as fast. The only people who would benefit from going to Northgate would be those headed to Northgate, or those taking a connecting bus. I suppose you could take Link north, but that really doesn’t make sense. Even if you add in the people trying to make a transfer to buses that might still serve the Northgate TC (e. g. the 40) it is still way fewer people than are headed to the UW (which is also a major transfer point).

         This is very different than today. Today, the number of people heading downtown exceeds those headed to the UW. Since taking the 41 is a great way to get downtown (much faster than heading south through town) sending buses that direction makes sense. That dynamic changes dramatically once Northgate Link is built, and the bus system should reflect that change.

         As I said, it is not at all intuitive. You would assume that you want to send more buses to Northgate, because they will soon have a train station. But they aren’t the only ones getting one, and Roosevelt Station makes more sense as a transfer point, because it really is “on the way”.

5. What’s the reasoning behind swapping the 40 & D?

   1. It straightens out the routes and the grid. The D continues going north as 15th bends into Holman Road. The 40’s terminus on Holman Road is a holding pattern until Metro does the sensible thing and sends it east on 85th to Greenwood, and possibly north to Northgate from there.

6. Any recommended web apps for armchair transit planners?

   I’m imagining something where with one click, it loads your city’s current network, then gives you the freedom to click around make tweaks, all while generating cool maps that show you changes, but with all the routes at once, nicely color-coded (perhaps even advanced features like an interactive map, trip planner support under the hypothetical network, isochrome generation, or estimates as to what impact the change would have on the agency’s operating budget).

   Viewing a change like this under such an app would be really cool (if it existed).

   1. There used to be Remix but they pivoted to professionals in agencies instead of the amateurs that they practically used as beta testers. Metro used it to develop the Long Range Plan. The original might still be open source.

      There’s a newer app called Enmodal but right now it seems geared toward rail networks, not buses.

   2. A challenge with doing this is that designing a route also has to consider union rules about breaks, so allocating buses to a route is “lumpy”. Each additional bus requires up to $1M a year in operations cost when fully allocated, because of the added cost of buying, driving and maintaining an additional vehicle. Depending on what gets negotiated, even an existing route may need more or less buses or schedule changes if a work rule changes.

      Thus, in some cases, extending a route a half-mile can cost negligibly more and in some cases it can cost significantly more.

   3. I wish there was something easier. I tend to just draw on Google Maps, and it can be time consuming. I often decide to change things half way through, which really messes it up. I’ve learned to “sketch it in” at first, not really following the road carefully. Then, when I am pretty happy with it, I’ll try and make the lines fit within the street. There is no good way I know of making it “lock in” the way that directions do. Speaking of which, Anthony takes a different approach, which is just make each route a direction. This makes for a nice, easy to read map of a particular route, but I don’t think there is a way to combine those routes onto one map.

      Wait — I just discovered something. On Google “My Maps”, you can have multiple directions (each one on its own layer). This is something you should consider, Anthony. I don’t think you have the ability to change the colors on each route, so in some cases it might be confusing, but it might in general make it easier to get the big picture. It is also possible that you can hack in different colors (by importing the KML file and then uploading it) which is something I might play around with.

      1. OK, I discovered something very interesting. As it turns out, all you need to do is export the file as a KML, then import it again. So basically, this is how you can produce a nice looking set of routes very quickly:

         1) Create your own Google Map. If you have a Google account, this is pretty easy. Open a map, and choose “My Places”. Then Select “Maps”. At the bottom of the list (which will be empty if you’ve never done this before) you will see a link for “Create Map”.

         2) On the new Google Map, create directions, each one of which is a proposed bus route.

         3) Using the menu on the top right, choose “Export to KML”.

         4) Select the “Export to a .KML file” checkbox. I’m not sure if this will work otherwise.

         5) Save the file by choosing “Download”.

         6) Create a brand new map, with nothing on it.

         7) On the main layer (which says “Untitled layer”) select “Import”.

         8) Drag (or select) the file you downloaded earlier, based on the first map.

         9) All the lines should be back on the map, but with more flexibility.

         Basically, it no longer treats these directions as directions, but as lines (and icons) on a map. My guess is that importing and exporting as KML stripped some data that is unnecessary. Now you can easily delete the icons, and change the color (and thickness) of the map. This means that you can create a map like this fairly easily: https://drive.google.com/open?id=1KDWI_ZGvzkoGxA6Z9DPeQr9EtQw&usp=sharing

         I wish I knew this before, as it would have saved me a bunch of time.

      2. OK, I decided to go ahead and copy (as best I can) this proposal. The hard part was reproducing the directions. Downloading, uploading, and then modifying it actually really easy. The only drawback is that once you do that, you are back to manually editing (clicking on lines) as opposed to changing directions (which is easier). Anyway, here it is:
         https://drive.google.com/open?id=1dRZzzjLlpnJphXJ6LevM9czWmNQ&usp=sharing

         You can see the big picture (or at least the bigger picture) with this map. You can also highlight a particular route, or even remove routes from view.

Comments are closed.