Page Two articles are from our reader community.

Sound Transit plans to expand the Link system with new lines to Ballard and West Seattle and extensions to existing lines. I like many of these ideas, but I also have ideas for other lines, as well as some changes to the already planned lines. I will divide this post into different lines. Here is a map of my proposed Link system:

West Seattle Line

The West Seattle Line will run south of Downtown with the following stations:

Alaska Junction
Morgan Junction
Fauntleroy Ferry Terminal
Westwood Village
White Center
Burien TC (Maybe, but there should probably also be some infill stations)

Northwest (Ballard) Line

The Northwest Line will run north of Downtown with the following stations:

South Lake Union
Seattle Center
Smith Cove
Market St
NW 65th St
Crown Hill (NW 85th St/15th Ave NW)

Fremont Line

I think there should be a line running to Fremont. I would prefer this over a 40 RapidRide. In fact, I wish the Ballard line would go through Fremont, but ST3 already has the Ballard line going through Interbay.

The Fremont Line will run north of Downtown with the following stations:

South Lake Union
Queen Anne (if it’s physically possible)

Northeast Line

The Northeast Line is the line that currently goes to University of Washington and will be later extended to Northgate and Lynnwood.

East Line

The East Line is the line that will go to Bellevue, Overlake, and Redmond. I like the plan Sound Transit has.

SeaTac Line

I think the Tacoma extension would make sense, though if it goes north of Downtown Seattle, it would result in a line about 40 miles long. That wouldn’t be too bad though. It could be through-routed with my Fremont Line.

Route Names

A while back there was a post by Joseph Story on a possible naming scheme for the Link system. I find his ideas pretty interesting, so you guys should read his post.

In my map, I have color-coded the lines. I avoided green and blue because those are being used for the Swift BRT. However, I think there should be some other kind of naming system for Link, maybe by direction, such as Northeast-Southwest.

11 Replies to “Expanding Link”

  1. Directions don’t come easily to some people. Ideally all the agencies should agree to a common numbering and lettering scheme.

    Examples: 1. Put a letter in front of any service that isn’t a bus. Link would be L1, L2, L3 and so on. Streetcars would be S1, S2, S3 and so on. Sounder would be T1 and T2. Ferries would be F1, F2, F3 and so on. BRT is perhaps more blurry; Median BRT might be B1, B2 and so on, while RapidRide and similar limited-stop routes could be R1, R2 and so on and express buse routes would begin with X. That would tell a rider what kind of vehicle and ride to expect.

    Regardless of identity, listing the end station is always important to indicate directionality. Otherwise people can get on a train headed the opposite direction!

    If agencies can’t agree on a system, the proper names for Link lines is a great way to go. They can be alphabetized like Story’s scheme or not. They should have an obvious color to help orient people. Because there aren’t many light rail lines, people are less likely to confuse them.

    1. Germany uses letters for modes. U1 is a subway line, which in a few large cities is heavy rail but in smaller cities is surface light rail with a downtown tunnel. S1 is an S-Bahn line (commuter rail). The buses just have numbers. I haven’t seen a letter for all-surface light rail; the signs have the full name “Stadtbahn” or “Tram”.

      Russia uses letter modes more intensively. M1 is metro line 1. T1 (with a Cyrillic 3-legged T) is tram (streetcar) line 1. T1 (with a 1-legged T) is trolleybus line 1. A1 is bus line 1 (autobus). The same city has all these so it’s important to know the mode. Commuter rail (elektrichka) and run-when-full van routes (marshrutnaya taksi) don’t follow this system. Bus and van routes exist only in the outer edges, at the last Metro station or where a future extension Metro is planned.

  2. Generally speaking, light rail investments are expensive. It only makes sense to build them where they can do the most good. There are trade-offs, of course. Sometimes it is a lot cheaper to build something say, close to the freeway, instead of where you would pick up more people. But in general, light rail is very expensive, and you want to spend the money where it can do the most good.

    Here are some things that you want your light rail line to do:

    1) Cover densely populated or very popular areas.

    2) Be significantly faster than driving, or an express bus.

    3) Work well with buses. This means not only complementary bus service, but also fast complementary bus service. Ideally your buses and trains both run fairly fast.

    4) Provide a network that allows for fast trips in multiple directions (using the above system). A trip might include a transfer (involving buses and/or trains) but you should be able to get around a city without having to go way out of your way.

    Unfortunately, your proposal fails on all counts. Before I get into the reasons why, let me start by saying that the ST3 map of the Ballard line is misleading. Normally, this wouldn’t matter (it isn’t the focus of your proposal) but you are talking about adding a new line next to a new line. There is some wiggle room, but my understanding is that Ballard Link will add stops at Denny & Westlake and Aurora & Mercer. This is a good map of the ST3 proposal: For a line heading basically the same direction (towards Westlake) you add very little if you squeeze in a stop between the freeway and the Ballard Link stops in South Lake Union. A stop at say, Fairview & Harrison would add some value, but not that much, as a lot of people would be just as well off walking to one of the Ballard Line stops. It gets worse as you head north, and are pinched in with the edge of the water. It isn’t clear whether you would even want to start the line off at Westlake, or make it a branch of the Ballard line, starting at Aurora. Branches have their issues, but with so little added between Mercer and downtown, that might make sense. I will, however, assume that you are building a second line, right from Westlake (since that is what you proposed).

    Now, to the particulars:

    1) This is a density map of Seattle: Your West Seattle proposal goes through very low density areas.

    The Fremont proposal is a little bit better in terms of density, but still has some major problems. As mentioned, any stops south of Mercer would be very close to existing stops, and thus redundant. Realistically, your first valuable stop is a little bit north of Mercer on Westlake (say, Highland & Dexter). Population density and office development is pretty decent there. But this still isn’t a great stop, because the water and hillside dramatically cut into ridership. It is difficult in many cases to walk down the hill (from the east side of Queen Anne) to a stop here. Many would simply catch the E. In the other direction, you have the water, and fish don’t ride subways. From there to Fremont is very low density and there is no point in adding stations. So you are basically talking about three stops: The first one is redundant (very close to the existing stop), the second one is mediocre (good density — poor access), and the third is OK (Fremont). To build this would require a new tunnel, and a new crossing of the ship canal. Even if you saved some money by going on the surface on Westlake, it is a very expensive line, for only a couple decent — not outstanding — stops.

    2) As you might expect, you can drive fast in West Seattle along the route you propose. The area is low density and was built with the automobile in mind. A train will rarely, if ever, be significantly faster than a bus, let alone faster than driving.

    It is a bit different with the Fremont proposal. Congestion is really bad in this area, especially as you get close to Mercer. But in the middle of the day, Westlake is a fairly fast road. The bridge is an obstacle, but for a lot less money, you could build a brand new bridge over 3rd Ave West, and make the Fremont bridge transit and bike only.

    3) Neither proposal does anything for buses. That is arguably the worst part about it (or any poorly designed transit plan for Seattle). Just to back up a bit, Seattle is not a city of very high density islands surrounded by nothingness. It is an area with lots and lots of medium density areas. It is unrealistic to assume that we will ever have a subway line connecting all of those medium density areas. Most of the city will be dependent on good bus service (hopefully connecting them quickly to their destination, or Link). In the case of West Seattle, you can’t possibly make a grid consisting of east-west bus lines serving a long north-south line (as you propose). There simply aren’t enough streets that go east-west. That means that even after you have extended the line, the buses would still do what they have always done: head towards the junction or the West Seattle Freeway.

    The same is true for the Fremont line. By running this north-south, you prevent any sort of bus to rail integration. It is literally impossible, for example, for a bus to serve Queen Anne, and then drop down to Westlake without going well out of your way ( There is no way a bus would do that, which means that despite being less than a mile away from Queen Anne, this subway line is irrelevant. You can’t have bus trips from the other direction (there is a lake in the way). This means that you can’t have any bus integration at all until you get to Fremont. For Fremont, this adds very little. You’ve ignored one of the key elements of bus integration — having buses and trains running fast. Getting to Fremont from the UW is not very fast, nor is it easy to get to Fremont from Aurora. Again, for the most part, this line becomes irrelevant from a bus standpoint.

    4) You’ve added nothing in terms of new trips. Fremont to the UW is as slow as ever, despite a brand new rail line added there. There is no way someone would go all the way downtown and then turn around and head back. Same with Ballard to Fremont. The buses are very slow, but the distances involved with taking a subway are so great that you are better off slogging with the bus.

    This proposal either assumes that everyone is headed downtown, or that east-west bus travel is fast. Maybe that’s the problem. I have no idea if you’ve ridden buses like the 44 or the 8, but they are horribly slow. This is not intuitive. You would assume that the slowest trips are those that involve heading downtown, but it is often the opposite. North of the ship canal, heading east-west is much slower than heading north-south, even though heading south means heading towards downtown.

    I’m sorry to be so negative. I really like these types of discussions, but this proposal is just not good at all. It might sound arrogant of me to say so, but I assume that it springs from lack of familiarity. Again, maybe you weren’t aware of the density or traffic patterns in Seattle. If so, then I hope my comments helped.

    As far as what I would build next, it is pretty clear: Ballard to UW rail and the Metro 8 subway (probably in that order). I’ve made the case (using many of these same arguments) that Ballard to the UW should have been the first proposal for ST3: The Urbanist made a solid case that Ballard to UW rail along with the Metro 8 subway would be a much better value than what was proposed for ST3: In both cases, the lines do a better job on all counts. Higher density areas served; much faster than driving (even at noon); much better bus integration (fast buses and trains); much better network making a lot more trips significantly faster.

    The route for a Ballard to UW rail line is fairly obvious. The only controversial part is whether you serve upper Fremont (for a better connection to the E) or lower Fremont (to better serve the cultural center of Fremont). Other than that, the only thing that people argue about is whether it should be extended further into Ballard (to 24th) or east of the UW (to Children’s).

    The Metro 8 proposal is a lot trickier. The Urbanist basically followed the current 8, which would make sense if you did a cut and cover operation for much of the way. But if you didn’t, there are a number of other possibilities. You could include First Hill, for example. Then there is the question of where the Metro 8 ends. Do you just merge with the Ballard Line, or find a way to serve the most densely populated part of Seattle (Belltown)? A Metro 8 proposal — given the changes that are coming — is long overdue, really. I’ve played around with a few maps, but haven’t come up with anything satisfying. This is something you might consider, if you felt like it.

    1. I see what you mean. Yeah, I’m not super familiar with neighborhood densities or anything. However, I am aware that there are some areas that can be very congested in terms of traffic, the best example being South Lake Union. I have been on the 44 and 8 many times, and I have seen the kind of traffic they get into. I agree that a Ballard-UW line would really improve some travel, but I don’t really have an opinion on the Metro 8 subway.

      I was hoping that my Fremont line could somewhat simplify travel to neighborhoods like Fremont, Wallingford, and Phinney Ridge. For example, many routes run express on Aurora, like the 5 and 26. I was thinking that since Aurora is already served by the E Line, the other expresses can be truncated in Fremont and provide efficient transfers to the Link line. Also, my Queen Anne proposal wasn’t supposed to go back onto Westlake, it was supposed to go straight to Fremont.

      I was also thinking that a Link line to Westwood Village/White Center could improve transit in that part of Seattle, but that area is definitely very tricky for a Link line. In fact, I do agree that the ST3 idea for West Seattle Link is stupid, but I just decided to integrate it into my proposal because ST3 is pretty certain. Though unlikely, hopefully ST3 is more flexible, and the West Seattle and Ballard lines can be exchanged for a Ballard-UW line.

  3. Cool.

    Yeah, I understand what you mean with West Seattle. Since they will build it, you might as well extend it, and get more out of it. You still won’t get a lot though, which is why I have a hard time spending any money on that thing. Ridership won’t increase much, if any, although a handful of people will avoid making a transfer. Maybe someday, after everything else is built, we could build something like what David Lawson proposed way back when:

    I was hoping that my Fremont line could somewhat simplify travel to neighborhoods like Fremont, Wallingford, and Phinney Ridge.

    Fremont maybe, but not Wallingford and Phinney Ridge. It becomes similar to the West Seattle situation. You are riding a bus, and have made your share of stops along the way. Just as the ride is about to become very fast (as you get on the freeway) you are asked to make a transfer and wait for a train. In this case, just dropping down to Fremont from Phinney Ridge or Wallingford takes a while. Unless there is a huge problem on Aurora, you are better off using it.

    You could argue the same thing for Link from the U-District to downtown, but there is one important difference: You stop twice along the way. Taking a train from one end of campus to the other is a reasonable trip, and serving Capitol Hill is huge. My guess is there are a lot of riders who just do that, right now (take the train from the UW to Capitol Hill). But with the Fremont line, there really is very little until you get to South Lake Union, and the Aurora buses stop at South Lake Union.

    As far as the Metro 8 subway is concerned, at some point I will write up a Page 2 post. I actually started a map, and didn’t realize it was pretty close to being done. Here it is:

    It is busy, to say the least. But there are various layers that you can turn on or off. You may want to turn off the BRT layer at first.

    There are also layers representing different options for turning the corner from 23rd to CHS. To view them, you need to select one, then deselect the other. For example, try selecting “First Hill Variation” and then deselecting “Garfield High School” variation.

    I prefer the “Garfield High School” variation (it should be the default view), but I don’t feel that strongly about it. I even left out the version that the Urbanist proposed, just because I don’t feel like going that far to the northeast is necessary (although an argument could be made for it). So there really are four decent variations, with one of them implied.

    Anyway, the key element with this map is that you have very urban stops. Every one of these stops is in a relatively high density area. You finally cover the highest density area in Seattle (Belltown) and connect it to South Lake Union. It is not perfect by any means — there are common combinations that just don’t make sense via the train, and thus will still be done by bus. For example, at 23rd and Jackson, if you want to go to the I. D., you are going to take a bus. But if you are headed to the other end of downtown (which now includes South Lake Union) then this train would save you a considerable amount of time. More importantly, it connects you to the rest of the city extremely well. You can get from anywhere in the greater Capitol Hill/Central District area to anywhere Link goes, very quickly. The UW, Ballard, East Side, or South end. This is an important change for the most densely populated area in the state.

  4. A Queen Anne tunnel and station is possible but extremely expensive, especailly the station. A station at Galer or Boston Street would be as deep or deeper than Beacon Hill or UW. One of the ST3 alternatives was a downtown-QA-Fremont-Ballard line, and that was rejected because of higher cost, the low-hanging fruit of the 15th Ave W expressway and railroad land next to it, and the less-interesting desire to serve Interbay and the future Expedia campus.

    There’s also a difference in need level between Ballard-downtown and Fremont downtown. A bus or streetcar can get from Denny Way to Fremont in ten minutes if it has adequate priority, and that’s good enough for Fremont’s size because it’s just a 10-15 minute overhead to regional trips. Ballard is twice as far so it’s a 20+ minute overhead, which takes up a good chunk of an hour on every round trip. That’s why Ballard and northwest Seattle is not adequately served by just ST2 Link and a priority direct bus, and needed a grade-separated solution.

    Your West Seattle line sounds like ST’s, just extended and with a transfer at Stadium. ST wants to get to Burien, it just couldn’t afford it in ST3. The best routing would be south from Morgan Junction to Westwood Village, bypassing Fauntleroy. If the ferry ran every 10 minutes it would be a different sory, but the ferry runs every 40-60 minutes, so 75%+ of the trains would not meet a ferry and the Lincoln Park/residential riders are not that numerous. Metro plans an all-day Fauntleroy-downtown express with stops on Fauntleroy way, so that can serve as a “second Link line” for areas Link doesn’t directly reach.

    Your terminus at Stadium also raises the question of the second downtown tunnel. PSRC says downtown will run out of transit capacity in a decade if we don’t do all this 3rd Avenue RapidRide-ization and second tunnel or something similar. So the second tunnel is highly worthwhile, I assume your Ballard and Fremont lines will have it, so why not put the West Seattle line in it too.

    RossB, what do you think about Seattle Subway’s idea for a long-term all-elevated Aurora-downtown-Madison line paid by the monorail authority? It wouldn’t serve all SLU-Capitol Hill trips but it would serve some of them that are close to Madison.

    1. RossB, what do you think about Seattle Subway’s idea for a long-term all-elevated Aurora-downtown-Madison line paid by the monorail authority? It wouldn’t serve all SLU-Capitol Hill trips but it would serve some of them that are close to Madison.

      I think you mean Doug Trumm’s idea. He posted it here: I think the idea was conflated with the Seattle Subway proposal, in part because Bruce Englehardt conflated it (see the Twitter references in the Urbanist article).

      To answer your question, I think Doug’s idea has merit — it makes a lot more sense than most of what Seattle Subway came up with (e. g. a crossing of Lake Washington between Sand Point and Kirkland). It is a relatively complicated proposal, and to make matters worse, Bruce suggested something a bit different. So let me just break it down piece by piece, starting to the east.

      Doug is pretty vague with the eastern end, for good reason. There are a lot of options, all of which sound good. However, I doubt any of them could be done cheaply. Running elevated from downtown to First Hill or the Central Area just doesn’t seem plausible. To quote his article:

      the elevated route could either take a direct line toward First Hill on Union or stop near Denny Way and Broadway to offer transfer to Central Link at Capitol Hill Station. An elevated Central Area line could potentially stop at Boren Ave and 16th Ave and 23rd Ave on E Cherry St before turning toward 23rd and Union or Mount Baker.

      None of those ideas seem physically, let alone politically possible. It just seems too steep, and some of of the streets too narrow. Union also has buildings that would have to be removed ( Cherry is the same issue — too steep and too much in the way. I think you would have to go underground through there.

      Would it be worth it? Maybe. But unfortunately you are making the connection to First Hill via Madison, which we will already have fairly soon (via BRT). It would be better, but not that much better. Either you spend a fortune (adding lots of stops) or you have fewer stops than the BRT, and you haven’t added much value at all. A lot of people would prefer the bus, just because it doesn’t involve a deep hole, and gets them closer to their destination. Making a connection between the Metro 8 subway and the rest of Link is a good thing, but that should occur via CHS (something the Seattle Subway plan inexplicably does not do). Connecting via the south end of downtown has its advantages, but is still not a great connection, because not every train line serves it. If you connect to Madison, then First Hill is a three seat ride from Capitol Hill, UW or Northgate. If you connect it to Pioneer Square (as proposed in Bruce’s map), then it is a three seat ride to First Hill from South Lake Union or Ballard. Most of those trips just won’t happen, which means that the line just won’t be worth it. Personally I would rather focus our energies on making sure that Jackson is a BRT street (we need to take those lanes as soon as possible) and building a good Metro 8 subway in the area. A subway 8 line could dip a bit into First Hill (on my map I have it as an optional layer entitled “First Hill Variation”).

      The next part of the proposal serves Belltown, which is long overdue. But instead of connecting Belltown with Westlake and the rest of downtown, it builds a bypass around the two tunnels. The idea is that it is elevated (and thus cheaper). At least for this section, it appears physically possible (the terrain is not too steep). But I doubt it would be cheap or easy from a political standpoint. Even if it works, you have a very fragmented line in a system that is already way too fragmented. It is basically the (old) monorail, but instead of connecting at Westlake (where all the trains meet) it connects to only half the lines. A trip from Westlake to Belltown is a two seat ride, involving a minimum of four stops. This is Belltown we are talking about — the most densely populated part of the state. Again, it just seems to me that the thing to do is connect Belltown to both South Lake Union and Westlake via a tunnel, as part of the Metro 8 subway.

      North of South Lake Union it is certainly possible to run an elevated line. But that would essentially mean building right next to the Aurora bridge, which gets expensive. You also have a subway line that is not that much better than the bus line it replaced. Unlike Metro 8, or Ballard to UW, you don’t have a huge speed advantage over bus service, let alone driving. If it can be built cheaply, then that would be great. My guess is, though, it couldn’t be built cheaply.

      If you can’t build it cheaply, then it probably makes sense to use Bruce’s approach, and go through Queen Anne. As you said, that is extremely expensive, but at least you add some brand new, very hard to get to stops. The fastest way to various parts of Queen Anne would be to take the subway. Even at noon it would be faster than driving. That leads to very high transit ridership. I like the line — everyone does — but it is extremely expensive, and probably not worth building until much later.

      Heading north again, either way you reconnect with Aurora, at 46th (by the zoo). From there to Green Lake, there is little value added with a train (because the bus is just as fast). But after that, an elevated train starts making sense. You avoid congestion and traffic lights and it does look relatively cheap to build.

      Which makes the whole thing a big conundrum. The only section that is clearly a good a value is the northern part. But unless you connect it to the rest of the system, it just doesn’t make much sense. Running a line along Aurora and then ending at 46th would be a weird thing. There is very little there. To be fair, ST will be forcing transfers in West Seattle where there is even less, so there will be precedent. But as bad as those transfers are, they are at least bus to train, not the other way around. You could have a whole train full of people disembark, then wait for express buses into downtown. They could ride the Ballard to UW train, but that would be slower. Making matters worse, ST probably won’t add a spur at the U-District, meaning the fastest train ride would involve a transfer.

      One possibility would be start up on Aurora, head south, then go underground somewhere in Woodland Park and hook up to the station at the zoo headed east. It would then go on towards the UW (more or less like this: To get to downtown would require a transfer, unless Link built a spur ( With the Ballard to Interbay line, you don’t need to run the Ballard to UW line every three minutes. Six should do it. That means you should be able to interline just fine by the zoo.

      The only way that would make sense is if you could build it really cheaply. That would involve a lot of elevated rail and only one tunnel, with no new underground stations. It is less than a mile of digging, but that still would probably cost a lot, making me skeptical of the merits of it. In general this sounds like one of those projects that only makes sense if Seattle continues to boom, and we still have money to spend on big transit projects after building the Ballard to UW and Metro 8 subway.

      1. I don’t think that there’s enough dirt above Second Avenue to go as far south as Pioneer Square. That’s especially true since the new tunnel would have to miss both the DSTT and the BNSF tunnels. They’re very close together there. If there’s to be a diagonal between Second Avenue and First Hill, it should connect to Link at University Street. There’s plenty of vertical clearance there.

      2. Oh, sorry. This was supposed to be a reply to Mike’s remark about the Hot Pink alternate in the original article.

    2. I’m referring to something Keith Kyle told me. He didn’t mention a specific alignment so I don’t know how closely it relates to Doug Trumm’s proposal, which I didn’t know about until afterward. I don’t know whether Seattle Subway has adopted it or is just considering it. But the idea is it would come after ST3, and we can outline it now, and then in the 20s or 30s or 40s decide whether to go ahead and build it.

      The most interesting issue, and why I asked you, is the question of how much the presence of a Madison line would affect the need and use of a Denny Way line. ST itself was not sure whether Madison RapidRide would be sufficient long-term; that’s why it left the Madison HCT corridor in the long-range plan in 2013. I live at one of the worst points for a Madison line, where the 8 and 11 are close but the 12 is a way’s away, so I know it would help me only modestly. But those further east where the routes are closer together, and those coming from elsewhere in the region (not the Denny Way area) , it would be fine for them. So those riders alone may make it worthwhile, and it may pull some riders from a Denny Way bus or train, but there would still be the other half of riders whom it would be too out of the way.

  5. Operating rail lines requires analyzing loads. If a segment is too crowded, no rider jade room to board a train. If a segment is too empty for a long stretch, it’s a huge drain on system productivity.

    With that in mind, how should light rail operations on these lines work? Should we have two branches per line , rather than single long ones? Should there be a mix of longer and shorter lines on a track, with turn-around sidings?

    For example, couldn’t the crowded Northwest line split with on gpingvnorth and the other going east to Fremont? Should West Seattle have two lines at lower frequency (making at-grade options more reasonable) rather than one at higher frequency?

Comments are closed.