Page Two articles are from our reader community.

An Introductory Scenario

It’s 2037! You’re coming in from getting on Link at SeaTac (or Tacoma or along MLK) and are headed to Capitol Hill, UW, North Seattle, Shoreline or Snohomish County. You have luggage. How will Link get you there? Unlike today, you will have to change Link trains. ST proposes having the Green line be the line for SeaTac (as well as SE Seattle, South King County and Tacoma), while the Red line will connect West Seattle with UW and Snohomish County.

Then the next obvious decision you have is this: Where will you change trains? Will you have to go up and down escalators (or worse yet, elevators)?

You’ll have some choices to make. As current plans show, you will be able to transfer at three or four stations – Westlake, International District-Chinatown, Stadium and SODO. Which station is best?

The Initial SODO Station Transfer Design

Early ST documents shown here ( suggest that the SODO Red line (West Seattle Link) station will be separate from the original Link station. Sound Transit proposes a center platform for the Red line, but the only place that a rider can go from the new platform is to return in the opposite direction on the Red line.

Designing the station this way will mean that anyone transferring between the two lines much change elevation to get to the other line’s platform. No level, cross-platform transfers will be available.

SODO Station as the Best Transfer Station Location

With multiple Link lines, how ST designs the tens of thousands of daily transfers is fundamental to its user friendliness. As noted in the above scenario, there will be three or four stations to transfer between train lines. Of these four stations, two (Westlake and International District-Chinatown) will be completely underground or below the street, so that designing for transfers is extremely costly and difficult because it involves tunneling. At the possible Stadium Station transfer point, the two lines will be close to overpasses and the elevated East Link tracks, so that designing tracks for cross-platform transfers here is also potentially complicated and costly; in fact, current concepts appear to have one of the lines skipping this station altogether! That leaves SODO as clearly the easiest and least expensive station in which to design for this transfer between these two lines.

Prioritizing Transfers at SODO Station

The current plans propose that any transfer between the two lines would always require exiting a platform and going up or down to a new platform level. Having lived in places with cross-platform transfers, I can tell you from personal experience that simply walking across a center platform in a few seconds from one train to the other (no elevation change) is by far the best way to transfer – especially with luggage or a stroller or a bicycle or a wheelchair! It’s also much easier and attractive than even having two train lines on the same track, because you have to get off one train and wait for the other one to pull in.

In fact, many systems take it one step further, scheduling timed train-train transfers (especially at off-peak hours) so that the time penalty for changing trains is fairly minimal. MacArthur Station works in this way for BART, for example.

Two Alternative Configuration Options

To do this, Sound Transit would need to reconfigure how the planned platforms are built. There are a few different options to accomplish this.

  • The new center, elevated platform could be designed to serve one direction (such as southbound) of both train lines. The current surface platforms could then serve both lines headed in the other direction. (Perhaps the current surface line could be redesigned to have a center platform by shifting one of the tracks in a later phase — noting that having multiple tracks available would make construction phasing easier). That would mean that only people transferring between West Seattle and the southern portions of the Green Line would have to change a level, and everyone else could have a train at the same level.
  • The entire station could be elevated above the street with two center platforms serving four tracks – inside tracks for one branch (like the Green line) and outside tracks for the other (like the Red Line). With that arrangement, transfers between the two lines would also be quite easy. Of course, transferring between West Seattle to the southern portions of the Green line would require two level changes in this configuration.

A comment on East Link trains: Obviously this station doesn’t allow for transfers to East Link; those riders would have to change trains in Downtown Seattle. However, having a lower frequency of trains to West Seattle at SODO Station (compared to the combination of Red and Blue line trains further north) would actually make it more operationally feasible to have a timed-transfer at this station. It also would allow for the station design at International District-Chinatown to prioritize connecting east and south direction train transfers in that design. For example, the new southbound Green line platform at International District/Chinatown could be built just east of the current northbound Red/Blue line platform at this station so that riders heading to SeaTac from Bellevue would have a same-level transfer.

A final point is that ST will need to turn around trains to and from West Seattle in SODO for several years until the Downtown tunnel opens. During that interim period, a cross-platform transfer could significantly reduce the transfer hassle for riders. Imagine if every West Seattle shuttle train rider had a longer-distance Link train to board waiting at the same platform for them (and vice versa)!

Why We Must Act Now

Sound Transit is now initiating studies on how operations will work after the opening of the West Seattle segment (2030 in ST3 materials), and in the new configuration (2035 in ST3 materials). ST hasn’t yet presented about on how many people will transfer between the two lines. I think it’s worthy to transit advocates to get Sound Transit to rethink the initial station track plan at SODO, and instead prioritize a Red/Green line cross-platform design objective into the design. If the current SODO station plans get built and this is ignored, we will be dooming thousands of riders each day for decades to changing levels to continue their light rail trips. Let’s get Sound Transit to design an easy transfer now to prevent this hassle or an expensive fix later!

46 Replies to “Providing SODO Cross-Platform Transferring”

  1. They’re going to rebuild Intl Dist and Westlake stations anyway as transfer stations, so I don’t see why they won’t be suitable for the more common SeaTac to Capitol Hill/UW/north trips. SeaTac to West Seattle will be less common, and that’s about the only combination that a SODO transfer will be preferable. But Metro will also have an Express route on West Seattle – Burien – SeaTac – Kent, and that may draw some people. In the interim between 2030 and 2035 it sounds like the West Seattle line will terminate at SODO and people will just have to transfer no matter what the quality is. So definitely we should push for the best transfer conditions possible, but I don’t think we have to worry much about south-to-north transfers that are probably better downtown.

    1. You’re saying we don’t need to worry about five years of forced transfers?

      And why would International District and Westlake be better, when their design is more forced and the platforms will be farther apart?

    2. It looks like transfers at Westlake and IDC Stations will require using escalators or elevators no matter how the stations are redesigned. This would not.

      It also seems like a good thing to do because it doesn’t appear to have huge cost implications.

      I think it’s a great idea!

    3. “You’re saying we don’t need to worry about five years of forced transfers? And why would International District and Westlake be better, when their design is more forced and the platforms will be farther apart?”

      We don’t know what the design of any of these stations will be, so it’s unclear how far apart the platforms will be. What we know is that we’ll have to transfer for five years from West Seattle, and I’d rather wait until we get some indication of what the transfer will be like before worrying it will be bad. Richard Bullington has a great idea below, but unfortunately we don’t know whether ST will do the sensible thing and put both lines at the same station or whether it will leave the Red Line alone with a surface station because it’s already there. That would create the situation of West Seattle having higher-quality transit south of downtown than Rainier Valley does, and unfortunately I could see that happening.

      1. Mike,. I’m already worried that the transfer will be bad. That’s based on the bus/link connections that ST has already done. It really doesn’t seem like transfer convenience has been a priority for them.

      2. I see ST getting gradually better over time, with ST2 stations better than ST1, and ST realizing it has to be pro-TOD and pro-density. The worst problem in ST2 is the gap between UW Station and the Stevens Way buses, but that’s a temporary stopgap, and it would be better if the UW allowed it to be.

      3. Now I’m getting confused with the phases. I mean the 2016 stations. That’s essentially halfway between ST1 and ST2, since they were going to happen eventually with or without ST2.

  2. Great post. I agree with all of your points. I think this is a very important aspect of our system, and details matter.

    As far as transfers go, here is my guess. I’ll ignore East Link for now:

    1) Lots more people will ride from Rainier Valley/SeaTac than from West Seattle.
    2) Lots more people will ride from Capitol Hill/UW than from Ballard.
    3) While reversing directions will be a relatively high portion of the trips from West Seattle, it still won’t be a huge number. The 50 goes through SoDo, but it doesn’t pick up many riders. Those headed from West Seattle to the airport, meanwhile, will likely take a direct bus.
    4) Reversing directions for a north end trip will happen at Westlake, not SoDo. Even if SoDo was designed for reversing directions between the two lines, it wouldn’t make sense for someone to go through half a dozen stops just to avoid an escalator or two.
    5) Therefore, the most common transfer at SoDo is same direction, from the east side of both lines (e. g. SeaTac to UW).

    East Link complicates things, but not in a huge way. An East Link train will head to Capitol Hill, the UW (and beyond) which will probably represent the bulk of riders that stay past downtown. I wouldn’t expect large numbers of people making the transfer to the other line, but if they do, they can make it at Westlake. Westlake transfers will be very common, but as you said, fixing that problem will be expensive. There will be a mix of same direction transfers as well as reverse direction transfers (e. g. Uptown/South Lake Union to Capitol Hill).

    I don’t know if that assessment is helpful in the least, but I find the topic very interesting. If we were building the system to avoid transfers, then we would connect Rainier Valley/SeaTac to UW (as it is now) but since that would mean building the longest light rail line in the world, we split it in two. So same direction transfers in SoDo will be extremely common. So too will reverse direction transfers at Westlake, since there is no good way for someone to get from South Lake Union or Lower Queen Anne to Capitol Hill. Until they build the Metro 8 subway, folks will either ride a slow bus, or transfer at Westlake. We may be stuck with a messy solution at Westlake (because it would be very expensive to fix) but we should at least fix the situation at SoDo, as you have written.

    1. “The 50 goes through SoDo, but it doesn’t pick up many riders.”

      That may be because of its infrequency and slowness. It makes riding between West Seattle and Columbia City possible but not enough to get people out of their cars. People don’t go between West Seattle and southeast Seattle much because the barriers are so great one after the other. But if there were significantly better bus service between them, or Link going halfway transferring to another train, that would probably make people think more about going from one to the other. asdf2 said that after U-Link opened, he started going to Capitol Hill a lot more often because it was more accessible from north of U-Village. Some people who aren’t on the 50 are probably taking the C instead and transferring downtown because it’s a higher-quality and more frequent trip, but if Link went from WSJ to SODO they’d be on it. When I go to Alki I have to take the 50, and sometimes I take it from SODO, but more often nowadays I take the C to WSJ and transfer there, because it’s a more pleasant place to wait for a 30-60 minute bus and I can stop at Bakery Nouveau if I have excess time. With a 10-minute train-to-train transfer in SODO it wouldn’t be as much of a transfer to avoid.

      1. I agree with your point about frequency. Half hour buses are bound to have low ridership.

        But I don’t by the argument that this bus is slow. It spends much of its time on the freeway, for heaven’s sake. There are plenty of very slow buses (the 44, the 8) that have high ridership. When Link gets there, I don’t see it being that much faster. For some trips, it definitely will be (e. g. the junction to Columbia City) but many others, you are talking about a four seat ride, instead of a one seat ride (e. g. Alki to the V. A.).

        Just by looking at the areas involved, it doesn’t look like there would be huge numbers of people making the trip. Neither West Seattle or Rainier Valley is really high density and they lack big destinations. There are no major employers or schools by any of the stops. You can make transfers, but then you are pretty much back to where you are now, in that people will just drive. I’m all for acknowledging the importance of community to community transit. I think it has been neglected when it comes to building out the system. I also agree that the geography of the two areas forces you to go around to the north anyway. But I just don’t see that kind of travel being anywhere close to the number of people going from Rainier Valley (and all of the stops to the south) to Capitol Hill, UW (and all of the places to the north). Just Rainier Valley to Northgate will probably exceed Rainier Valley to West Seattle, and that will pale in comparison to Rainier Valley to UW.

        >> after U-Link opened, he started going to Capitol Hill a lot more often

        Yeah, and I will definitely do the same when Link gets to Northgate. But that’s Capitol Hill. It is a great example of what a subway system should look like. It is very urban, making it attractive for both shopping and nightlife. There is a college there. There are lots of people who live there. Alternatives to getting there are also difficult. It is a pain to drive there, and when you do, parking is very difficult to find. If you catch a bus, it is slow, and often involves a transfer downtown even when downtown is out of the way (e. g. from Northgate to Capitol Hill). The Junction just doesn’t have any of those qualities. It is nice, but not “I’ll leave my neighborhood to visit every week” nice. Driving is easy, parking is easy, so if I’m visiting a friend there, I’ll just drive. I just don’t think you are going to see a huge rise in trips from West Seattle to Rainier Valley (or vice versa) when Link gets there.

      2. “I just don’t think you are going to see a huge rise in trips from West Seattle to Rainier Valley (or vice versa) when Link gets there.”

        That mostly depends on what destinations emerge in the future.

        “Neither West Seattle or Rainier Valley is really high density and they lack big destinations. There are no major employers or schools by any of the stops.”

        North Seattle has a lot of people criscrossing it, and they aren’t all going to UW. South Seattle is beset by physical barriers that form long narrow isolated neighborhoods, sometimes only a few blocks wide. And there are an extraordinary number of them: California, 35th, Delridge, 16th, SODO, Beacon, Rainier. (West of California I don’t know well enough to characterize it.) I believe this is what led to the lack of large destinations emerging: the immediate neighborhood was too small to support it, it was so difficult to go east-west that people were reluctant to, whereas it was easier to just go flat north to downtown so people did, and thus south Seattle is more downtown-centric than north Seattle.

        The question is, what should we do about it? What is south Seattle’s potential? Can we hope that more of a grid and citywide destinations might eventually emerge to make it more like north Seattle, so that it can become more equally a source/destinstion rather than mostly a bedroom community? Could better east-west transit help it significantly? Is it worth making it into a more even source/destination, both as a whole and the subsets of West Seattle and southeast Seattle? Or should we just leave it as is?

      3. I think it should be noted that east-west demand is a different issue than the Route 50 alignment.

        Crossing SODO is a decent distance and that puts some disincentive to go between the two areas. That dampens east-west travel interest.

        However, Route 50 is structurally flawed as it barely serves anything but residential areas. It is a string of neighborhood circulator routes tied together at an inconvenient frequency. It also suffers reliability problems due to SODO delays in the middle of the route because of traffic and trains crossing the route. Metro should probably reexamine SE Seattle restructuring in 2023 when Judkins Park Station opens.

        Finally, I think it is very likely that West Seattle trips to and from SeaTac will be happening with a SODO transfer once Link opens. More frequent, smoother, undelayed light rail is a powerful draw for that. Regardless, the transfers going fully northbound or fully southbound will probably be the dominant transfer need at SODO.

      4. >> However, Route 50 is structurally flawed as it barely serves anything but residential areas.

        Yes, but that is largely all there is within the two areas, which is my point. But it does serve Alki, which is relatively popular in the summer, as well as the V. A., which is a major employer and likely a bigger destination than anything in West Seattle. Yet Link serves neither, which is my greater point. I am not saying that the 50 is a great route, but I’m saying that West Seattle Link isn’t either.

        >> Finally, I think it is very likely that West Seattle trips to and from SeaTac will be happening with a SODO transfer once Link opens.

        Sure they will, but very few. SeaTac is a fairly popular station, but still not huge (less than 5,000, now that it isn’t a terminus). My guess is most of those riders are coming from downtown or Rainier Valley, while the rest of the city just drives to the airport.

        But the bigger weakness is West Seattle. There is just very low density there. Two out of the three stations are feeder stations. This means that folks will take a bus north to get on the train. For many headed to SeaTac, it will make way more sense to take a bus south, and transfer in Burien. That’s two seats, instead of three. So you are mainly talking about folks who live close to the junction, and there are simply not that many people there. Yes, I know it is growing, but it isn’t Belltown. My guess is that this will be the third most common transfer, just form West Seattle. It will lag both the I.D. transfer to get to Bellevue, and the same direction transfer to get to the Ballard line. The Ballard line will serve places that are destinations. Not only South Lake Union, but the Seattle Center (which may have a basketball and hockey team by the time this opens) as well as Expedia (which is overrated as a stop, but still something) and Ballard itself. Ballard is an employment center, with the hospitals, and is rapidly adding office space (in some cases, right next to the station). That set of destinations is bigger than what lies to the south (which is largely just SeaTac, unless you want to, again, transfer to the V. A.). For both heading north, and heading east, there is a clear advantage to taking Link, but for heading south, there are other options (e. g. from High Point to Rainier Valley you can go the other way).

        Stepping back a bit, there will be a lot of transfers within the system. My prediction in terms of transfers is this (in order):

        1) Same direction at SoDo. Lots of people from the south end, transferring to get to Capitol Hill, the UW and to a lesser degree, Northgate. Add in those from West Seattle to South Lake Union, Ballard, and all the places in between.

        2) Reverse direction at Westlake. This may exceed number one (hard to say). A lot depends on how much we improve the bus system. Someone from South Lake Union headed to Capitol Hill might be better off with a more direct bus, but not unless they substantially improve the 8. Likewise, South Lake Union to the U-District (and Roosevelt) may improve, but that project (now labeled Roosevelt HCT) has some major flaws, from what I can tell. Still, even if they make things great from a bus standpoint, you are going to get lots of people from Lower Queen Anne and South Lake Union headed to the UW (and places north).

        3) Same direction transfers at I. D. Downtown Bellevue and Redmond are major employment centers, so it attracts people from all over. The geography is such that Link is often the best option. Even a trip from Interbay to Bellevue makes sense. There aren’t a lot of people in Magnolia, but taking Link is by far the best way to get to a job in Bellevue or Redmond. Meanwhile, you have plenty of people in Ballard, Lower Queen Anne and South Lake Union headed the same way. Then you have folks on the East Side headed to Ballard, South Lake Union, Expedia or the Seattle Center (hopefully to see the Sonics).

        4) Reverse direction at I. D. This will probably be a lot less than the top three. From West Seattle, this will be an excellent way to get to the East Side, even if it involves a three seat ride (i. e.. a transfer next to the bridge). For those in Mercer island, this is the best way to get to the airport. For those in Bellevue, it is hard to say. It may make more sense to catch a bus, but if they don’t offer it, Link makes sense. For those in Rainier Valley, it is hard to say. If you are close to the (new, improved, RapidRide+) 7, then it probably makes sense to take that bus, and transfer at Judkins Park. But if you are close to one of the stations, or coming from Beacon Hill, then taking Link, and making this transfer, is the way to go.

        5) The transfer we were focusing on (West Seattle to Rainier Valley/SeaTac).

        6) Same direction transfers at Westlake. I’m not sure how much sense this makes. A lot depends on where they put the Madison stop, as well as how easy it is to transfer. If they put the stop substantially away from the other two stops, then this could easily jump to number five on the list.

        Anyway, that’s my take. I think it will be the least or second least common transfer, substantially trailing the top three. By all means we should accommodate it, But it is far more important to improve the other transfers.

      5. I make a Beacon Hill/Rainier Valley to West Seattle Junction round trip about once a week. It’s a 10 minute walk to the 50 stop (no problem for me) and then about a 30 minute bus ride to the Junction. That’s about 90 minutes of time invested in public transportation. Driving my own car, it’s about 10 minutes each way and there’s always plenty of free parking available at the Junction. So, driving my car wins about 95% of the time. The 50 is too slow and the 30 minute headways create too much risk of just missing a bus and wasting another 30 minutes.

        On trips I have taken between Beacon Hill and West Seattle Junction (via route 50) it usually seems that I’m the only rider who is using it as a BH>WSJ bus. There seems to be a complete turnover of riders in SODO. As Al S. points out, the 50 is basically a long string of winding neighborhood routes connected at poor frequency.

      6. Driving my own car, it’s about 10 minutes each way and there’s always plenty of free parking available at the Junction.

        On trips I have taken between Beacon Hill and West Seattle Junction (via route 50) it usually seems that I’m the only rider who is using it as a BH>WSJ bus.

        Thanks, Guy, that confirms everything I thought. There just aren’t that many people making the BH>WSJ trip. Yes, the bus is slow and infrequent, yet that doesn’t stop “a complete turnover of riders in SODO”. People are putting up with the slow speeds and very bad frequency for other trips (even when they involve a transfer), but there just aren’t that many people going from Beacon Hill to the junction. I don’t think that is going to change, even when the train makes that trip faster and more frequent.

      7. It’s surprising that many people are transferring at SODO because it’s such an unpleasant place to wait — all concrete and asphalt, nobody there, nothing to go to. Transferring to Link is OK because you’re only waiting a few minutes, but transferring to the 50 is worse.

      8. The 50’s SODO stop is mostly useless, no matter where you’re coming. Coming from West Seattle, you probably want the C-line to go downtown. Coming from Beacon Hill or Ranier Valley, you probably want Link (or, possibly, another bus, such as the 7 or 36) to go downtown. In fact, about the only time taking the 50 to SODO makes any sort of sense is for trips like West Seattle->Renton, where you’re transferring to another bus. These kinds of trips are, by nature, so hopelessly slow compared to driving that nobody with access to a vehicle is going to go do it. And, enabling these trips comes at the cost of making Beacon Hill->West Seattle trips slow enough to the point where it’s not worth riding the 50 for those either. I don’t even own a car, but if I lived in Beacon Hill and had to travel to West Seattle once a week, I’d likely use Car2Go.

        My personal opinion is that the 50 is too circuitous spends too much time duplicating other routes, such as the 21 and Link. In some ways, the 50 is even duplicative with the C-line – what’s the point of transferring to the 50 in SODO to go from downtown to Alki when you can take the C-line and transfer at the junction instead?

        The one-seat ride from Seward Park to Alki looks great on paper, but is utterly worthless in practice.

      9. “The 50’s SODO stop is mostly useless, no matter where you’re coming.”

        “There seems to be a complete turnover of riders in SODO.”

        ContradictionError: does not compute. Please specify which one is reality and restart program.

      10. I’d be interested to know how many people “complete turnover” means. Is it all of 10 people getting off at SODO, or all of 3? And, is this rush hour or all day?

        My guess is that rush hour, the 50 would get some people commuting to jobs in SODO, but outside commute hours, there is little reason to visit SODO on a bus (e.g. you can’t use a bus to go Cosco shopping), and better downtown options exist on the C-line or Link.

  3. Just combine the two lines at the SoDo station and then separate them. Chicago runs a hell of a lot more lines and frequency than Link will. Transferring on the loop is frequently just a matter of getting off a train and boarding another at the same platform because many of the lines are combined at the Loop.

    Ballard-SeaTac is limited by Rainier Valley and West Seattle is limited by trains from the east side, so there should be enough capacity at SoDo to combine the two lines.

    1. Glenn, I agree completely. I hadn’t noticed your post when I said essentially the same thing below.

    2. The Belmont station platforms in North Chicago are a good comparison to the service concept proposed here.

      CTA is even wanting to spend about $700M to fix the crossover track design omission from its original design 100 years ago (done to lower the capital costs back then). Delays have plagued northside trains for decades. How much forethought that ST can give now to designing for operations can prevent costly mistakes like this later.

      1. Al, YES, Please, ST, build no level crossing junctions except for deadhead runs. And for those, widen the right of way to allow for a train going out of service by crossing the opposing track to shift off the in-direction running track, come to a halt without fouling the turnout Into the middle track, and wait for the opposing track to be clear of traffic, if not completely then for the shortest period possible.

        IOW, separate the turnout and the crossing by about a train length and a half.

    3. I agree guys, that makes so much sense. Same direction transfers will be quite common, and the only place where that makes sense is at SoDo.

      Being a devil’s advocate, I can’t think of too many drawbacks, but I’ll go ahead and list them. First, this creates a single point of failure. If a train breaks down in that stretch, the entire system comes to a halt. But as Richard proposed, this would actually be elevated, so at least this won’t be the result of some idiot driver. It also happens to be very close to the train yard, which means that if any mess can be cleaned up quickly, it is here. Hell, you might be able to keep the old tracks just for redundancy (run surface in case there is a problem with the elevated tracks) but even without that, I don’t see this as being a major problem. I’m sure there are much bigger systems (BART comes to mind) with similar choke points, and they operate just fine. In those cases the problem isn’t reliability, but lack of throughput for a key section, which really doesn’t apply here, for the reasons Glenn mentioned.

      I think the bigger problem is if headways improve in North Link, but not in Rainier Valley. I could easily see this happening. If there is a capacity problem with our system, it is between downtown and the UW. This would likely happen in the evening, as lots of commuters head home, at the same time there is an event at the UW (and Capitol Hill for that matter). The extra line really doesn’t help the situation — it actually makes it worse. Someone heading home from South Lake Union to Lynnwood, for example, would put stress on North Link (starting at Westlake).

      So, with that in mind, ST might want to increase headways to two minutes. They have said doing so is possible, but makes the system less reliable. Fair enough, but if the trains merge at SoDo, you run into problems.

      To explain why is complicated, so here is a chart of what it would look like with the three minute headways. The time is on the left, and is basically when a train passes through a shared point (e. g. Westlake). I chose destinations as opposed to lines, but hopefully it is pretty clear. So here is a “normal” set of runs:

      8:00 West Seattle to UW (via Sodo)
      8:03 Bellevue to UW
      8:03 SeaTac to Ballard (via Sodo)
      8:06 West Seattle to UW (via Sodo)

      Everything works out great. The trains run by SoDo every three minutes. You have six minute headways from West Seattle, Ballard and SeaTac, along with three minute headways to the UW.

      But now look at what happens if North Link is bumped up to every two minutes. Assume that the trains are split between West Seattle and Bellevue:

      00 West Seattle to UW (via Sodo)
      02 Bellevue to UW
      04 West Seattle to UW (via Sodo)
      06 Bellevue to UW

      So that means the train is running every 4 minutes by SoDo, just from trips to West Seattle. This means that the trips from SeaTac either have to run every 4 minutes as well, or every 8 minutes. The former is currently unacceptable, while the second is a big degradation.

      All that being said, I don’t think this should stop us from doing exactly what you suggest. It isn’t clear whether capacity will be a huge issue. If it does become one, then we could work on improving the headways to Rainier Valley (by burying the line, or building under or overpasses). That would then not only enable four minute headways to Rainier Valley, but also the split that Richard alluded to. Run the trains every four minutes down Rainier Valley, but then split, with eight minute service to SeaTac and Renton. Works for me.

      But that is a long, long way off. For now, we should simply have the trains share the same platform in SoDo.

      1. As far as long way off discussions go, having four platform boarding areas rather than two facilitates operations for things like a Duwamish bypass or a second Ballard-SODO line (to serve anticipated high demand through downtown). Having four tracks also can help when a track must be closed for whatever reason.

      2. “ST might want to increase headways to two minutes. They have said doing so is possible, but makes the system less reliable.”

        Going below three minutes requires capital upgrades to the DSTT, which is not in ST3. There was a project for this but it wasn’t selected when they decided to go for the second tunnel. With these upgrades the DSTT would support 90 seconds. So if there is a crunch between Westlake and Capitol Hill even with twenty four-car trains an hour, it doesn’t look like ST will be able to address it until after 2040.

      3. @Mike — Good point. ST was a lot less definite about the issue, in the interview they gave Martin ( He wrote “Furthermore, it would likely require additional investment in Traction Power Substations.”. Based on what you wrote, I think we could remove the word “likely” from that sentence. This will require an investment. However, I don’t think we have to wait until 2040. We could easily pass something before then, and put the money into improving headways almost immediately. Unlike a lot of other projects, it wouldn’t require a huge amount of planning (it is all just engineering).

        In any event, I think we are in agreement that such a project is a long ways off. Likewise, I don’t think a Duwamish bypass or another end to end downtown tunnel will ever get built. Neither make sense (although we have built things that don’t make sense in the past, so I could be wrong). I could see a Metro 8 subway curving around and serving Belltown (like so — Doing that would send yet another train towards SoDo.

        Even before that, you have the issue that Al mentioned below. This, to me, is the strongest argument against simply sharing the line at SoDo. As he mentioned, it is quite possible that you will need better than six minute headways on Ballard Link. If that is the case, there is no place to put the train. It is headed to SoDo, and now you have a problem, even if you run the trains every three minutes.

        Looking at a schedule again, here is what I’m talking about:

        00 UW to West Seattle(via Sodo)
        03 UW to to Bellevue
        03 Ballard to SeaTac (via Sodo)
        06 UW to West Seattle (via Sodo)
        ?? Ballard to SoDo

        If you run trains from Ballard every three minutes, then you have a conflict in SoDo, unless you stagger the times a bit. But now we are getting to the issue of headways on this new shared section. If it is 90 seconds, then everything is fine, like so:

        00:00 UW to West Seattle(via Sodo)
        01:30 Ballard to SoDo
        03:00 UW to to Bellevue
        04:30 Ballard to SeaTac (via Sodo)
        06:00 UW to West Seattle (via Sodo)

        But if this new shared section can’t handle 90 second headways, then you can’t send another train down to SoDo, unless you have separate tracks down there. That is really the strongest argument for having two sets of tracks. Or, at the very least, the issue should be considered. We need to be confident that we can have 90 second headways on the shared section, or we have to build two different sets of tracks (or we just put up with a crowded Ballard Line).

      4. Because trains must have at least 30 seconds to stop at a station, getting trains to pull into a platform every 90 seconds becomes easily problematic. It only takes a slight delay of 60 seconds one one train to throw a string of trains off schedule for a good period of time.

        With there being existing tracks and two platforms and an intention by ST to add two platforms, why not just build the four platform boarding zones and two sets of tracks but do as the first option suggests and have the old tracks and platforms serve one direction and the new ones serve the other? The Lander crossing would remain, but the trains could easily be operated in a way to often have them cross Lander simultaneously.

      5. From the maps I was under the impression that the Bellevue line splits off north of SoDo and south of ID so that the International District Station would be the primary transfer point to-from that line. Thus, those trains would be out of the picture at SoDo.

      6. The way to handle the capacity in the U-Link and Downtown tunnels is to continue Downtown-Ballard on north into Shoreline and eventually build north fron a junction at Gates Foundation (please, please, please stack that station, ST!) through Dexter, Fremont, Wallingford, the West U-District, U-District, East Campus/U Village and north to Lake City and Kenmore along either 25th or 35th. Such a plan gives a “Seattle-centric” urban system overlaid on the Interurban-flavored “spine”.

        This would not happen until ST4 of course, which will come about should Seattle and especially King and Snohomish Counties continue to grow with “Urbanists” and the Republicans suffer shellackings in the coming years sufficient to bring them to their senses about the needs of the State’s economic engine.

        If Fremont and EastCampus/U Village were also built stacked, extensions to the west and east could make Ballard-UW-Kirkland_Redmond in ST5.

        Plan for the future.

      7. @Richard – I agree planning for / building in a junction at the Harrison station (i.e. “Gates”) is a great idea, but wouldn’t that be a Y-junction? Why would it need to be stacked?

        Whether it’s a Fremont-UW-Lake City line or a rail “upgrade” of the E-line, either way I can’t imagine needing to build a 3rd tunnel through downtown – the ST3 tunnel should have plenty of capacity for both the Ballard-Downtown and this new line to run at 6-minute headways each.

      8. AJ, yes it would be a “Y” (or “wye”) in lay terms, though not strictly so to a railroader. Such a facility wouldn’t have the third leg of a true “wye” so technically it would be a “junction”. I mean that the station must be stacked to that the junction at its northwest end can be non-crossing. If the station (and therefore the tunnels approaching it) is stacked there are a pair simple turnout to the diverging and from the joining tunnel one above the other. It’s the same logic as why SoDo should be stacked: no crossing junctions to and from revenue trackage.

        Junctions to segments of “service” trackage (e.g. the entrances and exits from MF’s and passages between lines used only for deadhead runs) are fine.

        You are absolutely correct that the second tunnel should be sufficient for any reasonable future. Downtown Seattle will never be Manhattan with several north-south trunk lines, regardless how large the Central Puget Sound Region grows. Bellevue, Kent and Lynnwood would become major centers long before that would happen.

      9. Ross, your idea of a Metro 8 that curves into downtown is less than optimal. Rather than joining the Green Line tunnel it should continue on west, probably using the Battery Street right of way (DO NOT FILL THAT TUNNEL WashDOT!) to a station at Western and Battery. There is plenty of quick service from the Broadway corridor to downtown Seattle via five bus lines, and it’s getting better. Better to add regional service to the very high density residential cluster in Belltown.

        I’m going to up the ante on the proposal at the end of the page (with all the bold face). Build the system as defined in it with the shared segment between Holgate and just south of Lander for ST3, but build the stacked elevated and particularly the new SoDo station with the ability to add a second set of stacked elevated tracks over the existing ROW after the existing at-grade ones are removed. Then you would have the best of all possible worlds in terms of north-south capacity for the future by simply building that second elevated section and tracks on the other side of the platform at SoDo.

        Let’s assume that the structure is built as described below with the southbound track on the upper level in order to make the departure from SoDo for Green Line trains as quick as possible during the period of shared trackage use. At the north end of the shared trackage there would be a turnout from the lower northbound track somewhere south of Holgate which would stay level across that street and then descend to grade and continue down into the new tunnel. The southbound track would presumably come out of the ground directly adjacent to the southbound entrance, but it would rise across Holgate and continue doing so at a constant gradient. When it reached sufficient height would swing over to the parallel Red line southbound track at the same elevation, either just to the north or right over the northbound turnout. The design would be such that it would be high enough to clear the northbound turnout at the point at which it turns to join the shared section.

        The same sort of thing would be true at the south end. The northbound tracks would be lower and the turnouts would be at the same north-south point.

        What that means is that in the future it would be possible to build a structure alongside the new shared one for just the trains to the Rainier Valley, but the structures for the original joining operations could become cross-overs! That is, a train which ran through the existing tunnel could visit Stadium and SoDo then turn left onto the structure into the Beacon Hill tunnel and descend to the existing structure right about the MF or continue straight. A train which ran southbound through the new tunnel could take the cross-over to the “Red Line” track to West Seattle or continue straight on the upper deck of the “Green Line” track to Beacon Hill and the Rainier Valley.

        In the other direction a train from West Seattle could visit SoDo and continue straight ahead at the Holgate turn-out to pass through the DSTT or take the lower “cross-over” into the new tunnel. A train from Beacon Hill-Rainier Valley could take the lower “cross-over” to the northbound “Red Line” track to continue through the DSTT.

        Grant, access to the maintenance facility from the north would probably have to be made at its south end. But there will surely be an entrance into the outer loop from the new Red Line tracks along Horton or Hinds. That could be used. And in fact, since the turnout from the Red Line to the Green line south of SoDo (later the cross-over) could continue high to pass over the existing facing-point cross-over for northbound-to-maintenance facility moves before descending to the level of the tunnel portal. I’m not certain that allowed gradients would be sufficient to do this, but it should be investigated.

        If there is ever to be a Duwamish Bypass this would be an excellent opportunity. The junction to the bypass could be just immediately south of the turnouts to the Beacon Hill tunnel allowing enough time for the tracks to flatten before crossing Spokane. The line could jog over to Airport Way using the median of Industrial Way and use the existing derelict trackway to pass through Georgetown inexpensively at grade. A train from the Bypass could then continue on to the existing tunnel OR use the north cross-over to the new tunnel with its likely greater capacity.

        A reasonable operating pattern for the new tunnel using two sets of tracks between IDS and SoDo would be Tacoma-Sea Tac-Bypass-Seattle (New Tunnel)-Ballard, Renton-Rainier Valley-Seattle (New Tunnel)-Ballard-Shoreline, and Sea-Tac-Rainier Valley-Seattle (New Tunnel)-Kenmore. West Seattle-Everett and Redmond-Lynnwood would continue in the DSTT. New service from Lynnwood to Sea-Tac via the Bypass could be added should 3 minute headways in the DSTT become inadequate.

        Even absent the Lynnwood-Sea-Tac booster, this would continue UW’s one-seat ride to the airport, albeit with three more stops than today and swapping East Campus for HSS, a loss for the medical staff. It would also add direct service to the airport for Wallingford, Lake City and Kenmore.

        Build for the future. It may not come, but the money to ensure that it can be served is a tiny fraction of what would be required to amend poorly designed junctions.

      10. Mike, yes it is, but it’s not growing right now as fast as is Seattle. It has had a GREAT boom, no doubt. But the cranes are mostly in Seattle.

      11. In the long post I’d like to make a couple of clarifications.

        In the fifth paragraph the sentence beginning “That is” should be “That is, a train which ran through the existing tunnel could visit Stadium and SoDo then turn left onto the then existing structure into the Beacon Hill tunnel (becoming at that time a cross-over) and descend to the existing structure right about the MF or continue straight to West Seattle.

        Just following that it should read “A train which ran southbound through the new tunnel could take the northern cross-over to the “Red Line” track to West Seattle or continue straight on the upper deck of the “Green Line” track to Beacon Hill and the Rainier Valley.”.

        The last sentence in the following paragraph should be: “A train from Beacon Hill-Rainier Valley could take the then-existing structure which would have become the southern lower “cross-over” to the northbound “Red Line” track to continue through the DSTT.

        I hope this makes sense with these clarifications. Basically by doing these things you’d have a long “double-crossover” (but not a “scissors”) with a station in the middle of it.

      12. With four tracks, there are certainly lots of different train line combinations that could be introduced, Richard. It’s good that you can vision these kinds of options!

        I kind of like an option that has the entrance on the ground level, southbound trains on the next level up, and northbound trains above that. I’m not sure with all of the structures south of the station and the proposed new tunnel less than a mile north of the station how the elevations work out, but a stacked station design option would seem to be the best of all possible worlds. That option could get expensive though.

        One additional important benefit of the all-southbound/all-northbound cross-platform transfer concepts is that it would let ST phase in track relocations as needed under all kinds of scenarios. The initial plan is for the two lines to not share any active tracks south of SODO, so that it will be very difficult to keep operations going if they want to switch tracks later.

        If ST built the center platform with two side boarding areas early (finishing completion in 2027), they could route SeaTac trains through there once completed and have a few more years to build a second center platform with two more boarding areas in anticipation of the opening of the West Seattle line in 2030 and well before 2035.

        The real issue I have is this: How can any SODO track and platform design change be raised to be an important of enough priority to modify the station concept directed by ST’s design leaders? The engineering firms are simply going to design what the ST design leaders tell them to design. Without putting some pressure on elected board members and the design leaders in the next year, this pretty low-cost opportunity may be summarily ignored.

      13. Al, you may be right about having the northbound trains on top. I don’t think it changes much and might make the interface to the existing trackage at the curve to the south easier. However, it would make the climb up from the surface on the northbound entrance track quite a bit longer. Anyway, these are things that can be worked out with the engineers, as you say.

        The point is, as you also say, to get ST thinking about merging the services initially in order to open up these future options. An added bonus is that the Green Line immediately becomes more reliable and a bit faster by not crossing Lander and Holgate at grade.

      14. Ross, your idea of a Metro 8 that curves into downtown is less than optimal. Rather than joining the Green Line tunnel it should continue on west, probably using the Battery Street right of way (DO NOT FILL THAT TUNNEL WashDOT!) to a station at Western and Battery.

        I put the Belltown Station at 2nd and Vine. You put it a couple block away. Fine by me. Other than those two blocks, i don’t see the difference. I may be missing what you are suggesting. Would you still have the same connection points for the Metro 8 Subway?

        Just to be clear, I wouldn’t call this an ideal layout. I would have run the Ballard line through Belltown (keeping everything to the west). Then the Metro 8 subway doesn’t connect to Belltown, it simply cuts across, and connects to Lower Queen Anne. To get to South Lake Union, you make a transfer. That has its disadvantages, but makes for a much cheaper, straightforward set of lines.

        Anyway, the key point is that we all agree that we probably won’t build another downtown tunnel. Likewise, we can all see how the Ballard train could run every 3 minutes, instead of 6. Maybe it runs every 3 minutes because there is a split, and you have 6 minute service to Ballard, and 6 minute service to Fremont (and on to the UW) as you suggested. Maybe you have 3 minute service because one train is going to Ballard, and the other train is going to Belltown, then looping around towards South Lake Union and the C. D. Or maybe you just have 3 minute service because you have lots of riders, and really good bus integration. No matter what, we can all envision the desire to have 3 minute service on the new tunnel long before we need to build a new tunnel.

        Which means that sharing a line is probably not a great idea. It locks us in to 6 minute service for the Ballard line. If I thought that was only a long term issue (i. e. only an issue when we made an expansion) then I wouldn’t worry about it. But I could easily see this creep up a lot sooner. That is why I’ve changed my mind, and think that we should build something like what Joseph proposed. We need to enable transfers at SoDo — especially same direction transfers — while simultaneously allowing for 3 minute headways on each track.

      15. Ross,

        My apologies. I thought you were advocating the Seattle Subway “fishhook”. I’ll agree that Vine is a superior location than Battery; though I’d push it one more block west to First. The reason I chose Western and Battery is that there is already a tunnel to it! Yes, yes, “it needs seismic strengthening”, blah-blah. That’s a whole lot easier than digging a new trench!

        Now I get it that you would “stub-end” the Metro 8 at Westlake — or actually a block or so west, perhaps at 2nd and Pine? I think that’s a whole lot of money for not much value, honestly. If you’re going to take the thing that far south, it makes sense to push it on to a midtown station. You can’t go all the way through town because of the spaghetti of tunnels at the foot of Second Avenue. But you could put a station at Spring or Madison. That would serve the part of downtown that’s at the bottom of a big hill fairly well.

        That said, it’s a lot of money; I’d prefer the Metro 8 to be an inner ring and punch a couple of blocks farther west. If DOT does fill the Battery Street tunnel, then I agree that Vine is a better location.

      16. Two comments:

        1. A design problem with only one platform in each direction is that the southbound track to Beacon Hill will have to cross the northbound track from West Seattle. I think it would take 3000 feet or maybe more to get one track at an elevation to allow for a separated crossover.

        2. Rather than build an elaborate tunnel for Belltown, a “subway lite technology” that stops at Denny Park station, uses the Battery St tunnel, stops at First and Battery, emerges out of the tunnel and stops just underneath at Steinbrueck Park, and terminates near the Second Avenue entrance of University Street Station. It could be smaller cars that could turn tightly and maybe be rubber-tired like the train shuttles in the Airport. The only extra tunnel digging would be minimal. It could have two reversible trains that leave each end at the same time, even as frequent as one train leaving every 6 minutes.

        Belltown is a very wormy issue. However let’s get back to the core point of the post, which is to get relatively low-cost and useful cross-platforms at SODO.

      17. Al, tracks don’t “cross”; they are stacked one above the other. That way junctions are all simple turnouts, not a turnout and a four-frog crossing. It does not take 3000 feet of lineal distance to gain twenty feet. That is a 2/3 degree gradient. LRT vehicles can manage 3% at speed in revenue service. That’s “three feet in a hundred”, or seven hundred feet. With the vertical curvature at both ends the entire separation would probably take eight hundred, but there is considerably more than that between the SoDo busway and the Beacon Hill tunnel.

        And if you make northbound on top you’re going downhill so you can probably entertain a slightly steeper grade.

        Grant that it makes entering the MF more difficult.

  4. It may be too late, but it seems to me the best thing is to have both Red and Green Lines share elevated trackage between SoDo and just south of Stadium. Have SoDo be “stacked” so that the turnouts at both ends would be non-crossing. Central Link is already elevated just south of SoDo and if the southbound track were the higher platform it would be a “downhill” run to get to the existing trackway.

    By doing this, every “in-direction” transfer can be made at the same platform! True, Sea-Tac, Rainier Valley and we can hope one day, Renton to or from West Seattle would require changing levels. But as pointed out by others, this is a fairly rare transfer and the current proposal requires it anyway.

    As close to Holgate as is possible to accommodate the gradient changes, another set of stacked turnouts would separate the Red Line for its run to Stadium and on to the existing IDS from the Green Line. Once across Holgate itself, it would descend to the existing grade level line for the run to the tunnel portal for the new IDS platforms. By placing the junction close to Holgate only a few hundred yards of “duplicate elevated” would be required before the Green Line descends into its tunnel.

    Notice that in the diagram THERE IS NO STADIUM ON THE GREEN LINE! That may be a drafting error, but I think not. Please notice that the “at grade” symbol for the Green Line changes to the “tunnel” symbol right next to “New Stadium”.


    Notice also that the new Red Line elevated tracks will transition to surface just north the Holgate (Holgate is right at the curve on SR99 and the change from elevated to surface is right north of there on the diagram). In fact the tracks need be elevated only as far as Holgate to eliminate the two most troublesome grade crossings. The Red Line will still need to cross lower Royal Brougham, but perhaps it can be made bus-only between Fourth South and Sixth South.

    By sharing trackage for one station and roughly 2/3 of a mile of elevated structure ST would be able to eliminate grade crossings on the Green Line north of Walden Street at a cost only of roughly one-half mile of elevated structure alongside the Red Line elevated and stacking the station at Lander and the tracks just to the north and south of it.

    This makes so much sense I’m surprised it has not already been announced as a possible enhancement.

  5. A few other advantages to two directional platforms proposed here:

    The second downtown tunnel demand will require higher frequencies than MLK can effectively sustain. Trains will likely have to reverse somewhere south of Downtown. Directional platforms make it less disruptive to operations.

    If ST leaves a surface station but reconfigures it to having trains in one direction and the trains have a timed-transfers like one of the options mentioned, it would allow for both trains to cross Lander at the same time.

  6. Another design option would be to have one center northbound track, two center platforms (northbound train doors opening on both sides), and two southbound tracks and platforms (one for West Seattle and one one for Seatac) on each side. That would provide level transfers in all directions but provide the added advantage of a three track station.

    1. On second thought, the better operation of a three track platform would be for each side track and center platform be for one direction, and ST to have the option to use the center track as needed — even leaving a non-operable train on the center track that can be used for walking through at times (doors on both sides open).

Comments are closed.