Last week Sound Transit released the Draft Environmental Impact Statement (DEIS) for the West Seattle/Ballard Link Extension (WSBLE), the long-awaited Link line from Alaska Junction to 15th and Market, via Sodo, Downtown, South Lake Union, Uptown, and 15th. The comment period began today. View the online open house and comment here.

Really Deep Stations

Both Doug Trumm and Mike Lindblom ($) have already explained the very deep stations proposed along underground segments downtown quite thoroughly. But briefly: typically, the deepest stations in a system are among the lower-ridership ones (like Washington Park in Portland). Westlake, Midtown, and Chinatown would be among the deepest in North America but also among the busiest in the system. Stations this deep mostly depend on elevators, which limit throughput, have an (ahem) spotty maintenance record, and increase the length of every trip that uses the station.

Worse, there is no engineering reason for Chinatown to be deep. It’s a potential concession to a neighborhood that is tired of decades of disruptive construction projects nearby. There is also no appetite to cheaply and shallowly cut-and-cover 5th Avenue downtown for similar reasons. Hopefully, early media attention will help politicians stand up for future riders, and engineers to get a little more creative.


A DEIS is largely about “impacts;” things, sometimes bad and always disruptive, that will happen to the surroundings. Politicians care a lot about impacts on businesses and voters, so there will be much more focus on the next 15 years of construction than the infinity of future riders. All the impacts are there in the report for you to read, but here at STB we care about riders.

They say that all models are wrong, but some are useful. This model says that the alternatives under consideration don’t really differ in ridership at all. Most of the segment alternatives have literally no difference. The exceptions:

  • An elevated Fauntleroy Way station (kinda near Trader Joe’s) actually adds 100 more riders (out of 7,800 in the segment) vs. all the other options while also negating various NIMBY impulses around the Junction.
  • Moving Delridge station north of Andover St, and the walkshed further into the Delridge/bridge interchange, costs 200 riders (out of 5,800) in the segment, with no real cost savings.
  • In Downtown and SLU, running on 5th and Harrison nets 5,000 more riders (to 163,700) than 6th and Mercer while also being slightly cheaper.

You might ask if it makes sense that adding several minutes to platform access at Chinatown would have no impact on ridership at all. Sound Transit could not promptly answer my question as to whether station ingress/egress time was part of the ridership model. But, officially, there’s no difference. Perhaps this is something that could be considered in the final EIS? That said, detailed forensic analysis on the results that don’t match your priors, and easy acceptance of those that do, is a cognitive error I’m disinclined to make here.

Regardless, given the results as they are, it’s hard for future-rider advocates to be super-energized about such small differences, except that (1) the cheaper option is more likely to be built promptly, and (2) even if the number of the riders is the same, wasting an extra 6 minutes or so of their time on every trip forever is a shortsighted decision.

Cheap and Shallow, Like a Reality Show

So what does this future-oriented best alignment look like? From South to North:

  • An elevated station at Fauntleroy Way (WSJ-2, this is “preferred” but, comically, 4 of 6 options are preferred; let no faction be left behind)
  • A “low-height” elevated option along Genessee St. with the Delridge Station actually a block off Delridge (DEL-2a, one of two preferred, 8 total)
  • Hugging the highway bridge over the Duwamish (DUW-1a, preferred)
  • At-grade through Sodo with separated crossings at Lander and Holgate (SODO-1a, preferred)
  • The “5th Avenue Shallow” option at Chinatown (CID-2a, no preference given)
  • 5th Avenue through downtown and Harrison St. through Uptown (DT-1, no preference given)
  • Through Interbay, a stop at Galer St and then elevated along the railroad tracks instead of 15th (SIB-1, preferred)
  • In Ballard, a 14th Ave Tunnel (IBB-2a) or 15th Ave Elevated (IBB-3). 2a is one of three preferred options; although the numbers show no difference, my lyin’ eyes say to pick 15th.

No doubt, some of these will not happen as protests about “impacts” emerge and pressure Sound Transit into picking more expensive, riskier options that don’t improve rider experience.

Environmental review runs into 2023. Then there’s a few years of design, and earth starts turning on the Sodo/West Seattle leg in 2026.

263 Replies to “Comment on the WSBLE”

  1. Gee, there is so much to unpack in the DEIS. Maybe this should be a separate post for every segment?

    My broad takeaways:

    1. Unsuitable transfers. The early transfer messaging was that it would be a quick level change. No we know it’s much more effort. Was this known beforehand? The alternatives development in 2017-2020 would have been very different had this been disclosed.

    In light of the several now-disclosed extra minutes to transfer, it really suggests even more strongly for same direction cross-platform transfers at SODO. Why would we relegate every Rainier Valley and SeaTac Airport rider to a deep station to transfer? The transfers are horrible! Just horrible!!!

    A second option would be to mix and match Ballard and UW/Northgate/ Snohomish branches with West Seattle and RV/Seatac/Tacoma branches, creating four lines rather than just two at 12 minutes each rather than six. It’s bad enough having a deep platform — but to have tens of thousands of additional transferring riders using elevators or long escalators is just plain nuts! The time it takes to transfer would add 6 minutes of time (possibly lots more of the peak loads are heavy) — the time it would take to just wait on a platform for the next train going to the other line if there were four alternating lines rather than the proposed two!

    2. Do we need three stations in West Seattle? Some of the revised ridership numbers suggest that two would be plenty. In particular, building a tunnel station for a relatively small number of riders is not cost effective. It may even ruin the project’s FTA New Starts rating!

    3. I could be wrong, but I don’t think the access time to reach a deep station is in the forecast algebra — and it should be. Thousands of Downtown riders will not use Link for short trips (say 1 to 3 stations) because it will take so long to get in and out of stations.

    1. 10 minutes is the minimum headway for a good rapid transit system. The US always shortchanges transit and then wonders why it has low mode share. If we must consider 12 minutes, which segments are we inflicting it on?

      1. A rider would have the choice. In a worst case, the rider would hop a train within 6 minutes and transfer (taking another up to 6 minutes) or wait the full 12 minutes for a direct train.

        In other words, the wait for a train would never be less than 6 minutes (or 10 minutes off peak).

      2. If we went with the proposal I have down below, every station gets 6 minute frequency during rush hour. The only drawback is that the trains might get delayed a little bit during peak. This is a minor problem, and outweighed from a user standpoint by the additional depth of the station, let alone the cost.

        The rest of the day, frequency is dependent upon the whims of Sound Transit. I would set a baseline of 7.5 minutes. If money is tight, trains would turn back at Federal Way and Lynnwood, for 15 minute frequency on the outer edges. This is normal for lines like this.

        As I wrote below, in the long term, automation should reduce the cost of operations, which might allow the trains to run more often. There is still the cost of security and fare enforcement, but you can always reduce that on the outer edges while still running the trains. The only significant cost becomes wear and tear.

      3. “If money is tight, trains would turn back at Federal Way and Lynnwood, for 15 minute frequency on the outer edges.”

        This! This is revealing one of the unrealistic expectations of ST3 — six minute trains peak and ten minute trains daylong all the way to Everett and Tacoma. It sells well as an utopian idea — but it’s not fiscally practical to run mostly empty trains (with drivers) that frequently north of Mariner or south of Federal Way.

        Ironically, it does become more fiscally practical if trains become driverless.

      4. Ross, I have to agree that the 7.5 minute three line concept (24 trains an hour total) seems reasonable. It’s worth noting that the East Link EIS assumed 8 trains an hour in the peak and 6 for most of the day.

        Looking at the demand for West Seattle Link and Everett Link north of Mariner, 8 trains an hour would be more than generous and that line could be done easily with 4-6 trains an hour at peak.

        ST3 proposed 20 trains in ST3 in today’s DSTT. 8+8+5 for the three lines (21 trains total) seems effortless as a starting point. Frankly, I don’t think the 2 Line will need more than 6 trains an hour heading across Lake Washington and the other two trains an hour were needed to get the 20 trains between Downtown and UW. With a third line that extra carrying capacity is provided.

        So the base system could be:
        1 Line – Tacoma to Northgate
        2 Line – Mariner to Redmond
        3 Line – West Seattle to Everett

        Then Ballard to Westlake could be a stand-alone automated Line 4.

        By the way, driverless trains could be driven to the Central OMF if an access track just for access was placed between Westlake and University St stations. ST operations could benefit from opening up Third Ave at Pine for a new scissor track crossover point as well as this access track . Then the new track could turn at Virginia or Stewart and back into the second Westlake platforms under Fifth between Pine and Olive perpendicular to the current ones. It’s just speculation on my part, but seems doable. Still, it may be easier to just add a small OMF for the vehicles in Interbay.

        Why do this?

        1. It saves billions by not needing a second tunnel between Westlake and ID.

        2. No ID/Chinatown subway construction at all! All downtown tunneling would be Pine St northward.

        3. New Westlake platforms would not need to go deeper to get under the existing tracks.

        4. Westlake transfers could be level between the old and new tracks in one direction. No stairs or escalators!

        Really the only added downside is the double transfer for South Sounder riders to get to SLU. But with trains in the DSTT so frequent and with Line 4 (or 5) trains automated and frequent, this seems like a modest sacrifice compared to the deep underground mole holes in the current scheme.

        Then the remaining incompatibility with ST3 would be no Midtown station. That’s easily remedied with a Jefferson Street above ground funicular from Third/ Jefferson to Harborview with an intermediate stop just east of Fifth into one of the county buildings there. That would not only fulfill the ST3 Midtown Station promise but would also fulfill the original First Hill promise in Sound Moves. It could be Line 5 or 6 or Line F!

        Ok … back to reality … the only way this would happen is if some powerful corporation forced it in a backroom meeting. Sometimes I think the only way out if the ST3 debacle would be to convince Bezos to make it his personal crusade. The Board doesn’t listen to little creative people with solutions.

      5. “but it’s not fiscally practical to run mostly empty trains (with drivers) that frequently north of Mariner or south of Federal Way.”

        Is it? Much of the day, the 512 is already running every 10 minutes. It’s not that much of a frequency upgrade from the 512 to the planned Link level of service, especially since the 6-minute peaks include not just the 512, but also other buses, such as the 510/511/513/etc.

        The only way I see this really becoming a problem is if each service hour costs significantly more with a train vs. a bus, but I’m not sure that’s actually the case. Remember, the tracks, stations, everything is already built, so the only variable cost is drivers, energy, and maintenance. Considering that drivers get paid the same either way, running a train every 10 minutes may not really cost anymore than running a bus every 10 minutes. The only difference is that, due to the differences in vehicle size, a bus with 20 people on it looks mostly full, whereas a train with 20 people on it looks empty. But, if frequencies are set based on operating budgets, not capacity needs (e.g. divide rider demand per hour by the capacity of each vehicle), that doesn’t matter.

      6. Oh I would agree that it’s a drop in the bucket to run trains all the way. In fact, I think a more likely scenario is to leave the base frequency alone.

        It’s the peak overcrowding that raises the need for short trips that turn around before the end station . That’s a capacity issue. Why should Fife boarding riders get their pick of empty seats, but Rainier Valley riders not even be able to board the same train when it’s standing-room only? The same goes for Everett riders compared to ones in North Seattle. A short-turning train would address the overcrowding problem.

        Anyway, the discussion is in the context of having three lines between Northgate and Downtown. It’s not mandatory to have an even 7.5 frequency on each line. If one branch or segment gets too crowded, it may be possible to merely shift a train or two an hour from an underutilized branch to the branch that needs the capacity. Once a train is coming every 10 minutes on each branch, it seems more strategic to add additional trains (shortening wait times) based on capacity rather than promised peak headways. A cost-effective rail operation should actually have this kind of flexibility to maximize comfort for everyone and better serve surges with special events like games at various stadiums.

        A base three-line operation at 10 minute frequencies would mean a train every 3.3 minutes (18 an hour) in a single Downtown tunnel (and ST3 assumes a train every 3.0 minutes in a two-line tunnel or 20 trains an hour at peak times). Rather than 10 trains to West Seattle and another 10 to Redmond, a three-line operation may have 6 trains hourly to West Seattle and 6 to Redmond with 8 to Tacoma. It’s still 20 trains an hour. If the signal system can push through 24 trains an hour (the combined 2.5 minute trunk frequency suggested by Ross and others), the allocation of those additional four trains can vary by branch — and some could be short-turned to address overcrowding.

      7. “the 512 is already running every 10 minutes.”

        The 512 has become the most-frequent, highest-ridership ST Express line. It overtook the 550 a few years ago when the 550 lost DSTT2, the South Kirkland P&R, and the Rainier freeway station. But 512 ridership keeps climbing. It remained high throughout the pandemic because Snohomish County has a lot of lower-income essential workers, and the distances aren’t as far as in South King and Pierce Counties and there are fewer highway alternatives, so transit is a more favorable choice. The 512 is now 10 minutes weekdays and Saturdays, 20 minutes Sundays, and 15-20 minutes evenings.

      8. The only way I see this really becoming a problem is if each service hour costs significantly more with a train vs. a bus

        It is significantly higher. I haven’t seen recent numbers but back when a bus was ~$130/hr (don’t recall if that was revenue hr or platform hr) the cost for Link was ~$250/hr. The kicker though is ST reports Link cost per LRV so you have to multiply that $250 X2 or X4. Which seems wacky since I doubt the incremental cost of a 2 car train doubles by adding extra capacity.

        I tried to find recent info and the ST Q1 2021 report is baffling to me. Compared to 2020 they operated about half as many trips. OK, Covid, no surprise there. In 2020 there were 62 boardings per revenue hr and 143 boardings per trip. So, simpleton math says a trip is (143/62) or just over two hours. 2021 the boardings per revenue hr was 90. OK, less frequency so more people pile onto the limited number of trains. I think I can follow that. But boardings per trip dropped to 83??? How can that be unless the trips are now shorter (less than an hour). Can someone explain the new math please. And why did cost per boarding go from $6 to $26? That seems like the opposite of cost savings.

      9. There is actual cost and leveraged cost. I bet they published leveraged cost. As ridership and miles dropped during covid but Sound Transit staff levels did not, all the overhead had to be spread over less miles/riders and therefore shot up.

      10. It all depends on how you do the accounting. What really matters here is the marginal cost to provide each additional service hour. Not the total cost to operate the system (including a bunch of fixed costs that have nothing to do with how often the trains run) divided by the system’s current service hours.

        Inferring the cost per service hour from cost per boarding and riders per hour could potentially result in inflated numbers, if the accounting is based on total cost, not incremental cost.

        The only time incremental cost really gets expensive is if it requires purchasing new trains (and, possibly purchasing new land to store the new trains).

      11. Bernie is right. It costs more to run a train than a bus. The train makes up for it by carrying a lot more people, which means it can be a better value in terms of cost per rider. But an empty train costs more to run than an empty bus. A train with 50 people on it costs more to run than a bus with 50 people on it. It is only when you exceed the capacity of the bus that the train becomes cheaper to operate.

        This begs the question: In the middle of the day, will you exceed the capacity of a bus between Everett and Lynnwood? I seriously doubt it, no matter how frequent it runs. The 512 got around 30/40 people per trip, and most of those riders never went further than Lynnwood. Link will follow a different path, but I don’t see a huge increase in midday ridership. You add three, maybe four stops, but one of those (Boeing) is clearly peak oriented. You also lose South Everett. If there were a lot more stops, then I could imagine this providing decent ridership along a busy corridor (after years of unprecedented and unexpected growth in Everett) but that isn’t the plan.

        It does seem rather harsh to leave the northern section with 20 minute frequency, but plenty of agencies have done the same thing with similar systems. It is tough on riders, but as we’ve seen, ST does not prioritize frequency. I think 7.5 minute baseline, with 15 minute on the extremities makes a lot more sense, but they don’t seem to be interested in the former, so they may not care about the latter either. It is likely that buses will complement the train service for a lot of the trips.

      12. I’ll plop a link to this article here:
        As best as I can tell, ridership north of Lynnwood winds up being about half of that south of Lynnwood.

        That’s why I propose only half of the trains go north of Lynnwood.

        If there’s more demand than that, you could add a bit more by extending East Link for those trips, but mostly the peak frequency just isn’t needed up there, even in 2040 estimates.

      13. I still don’t get how the boardings per revenue hr and boardings per trip can be apples to apples between 2020 Q1 and 2021 Q1. One has a greater number of boardings per trip than boardings per hour and the other is reversed. As I do the math it would mean the trip time went from 2 hr to 30 minutes. Hard to believe that less frequent trains created a 4X improvement in trip time. Seems either a misprint or substantial change in the way numbers were calculated.

      14. “What really matters here is the marginal cost to provide each additional service hour.”

        What really matters is access to places and passenger convenience. We aren’t grudgingly providing as little transit as possible; we’re trying to make a robust transit corridor because it’s a basic feature of a metropolitan area.

        “It costs more to run a train than a bus.”

        It costs even more to support an SOV on a freeway.

        “It is only when you exceed the capacity of the bus that the train becomes cheaper to operate.”

        But you have to start building the rail ten years before the bus reaches capacity rather than waiting until the last minute. And the train may be worthwhile for other reasons even if it costs more to operate.

        I’d be fine with truncating Link at Lynnwood or Ash Way or Mariner and having some combination of BRT and/or ST Express beyond that. (E.g., BRT to Everett, ST Express to Seaway Transit Center and the Mukilteo ferry terminal.) What bothers me is the attitude that high-quality transit isn’t good in itself. Subway stations should not have 15- or 20-minute service daytime or early evenings. If cities and counties and voters are willing to go further out than transit fans would, we should celebrate that rather than obstructing it. Even if it’s unnecessarily far or costs more than a bus, it will give more mobility to people riding it, and help essential workers who often have to live and work in the worst locations and travel long distances, and will improve the metro’s transit circulation options overall. The problem is that the US severely shortchanged transit for fifty years and wasted money on eight-lane highways and unwalkable big-box stores and gas-station mini-marts instead of building on what we had. If we occasionally longchange transit in our attempt to retrofit what we should have had from the beginning, that’s not the end of the world. The biggest problem is not too much transit in the suburbs, it’s too little transit in the city. Truncating Everett will not get us more lines in Seattle.

      15. Well said, Mike.

        Generally, running rail along a freeway only makes sense if the rail capacity is needed. But Link projects, even freeway running TDLE, do more than just create capacity. A truly apples to apples bus service would require either hundreds of millions in investment in bus freeway infrastructure and/or running buses far more often than the train (to make up for inferior reliability). Replacing a Link project with an ST Express bus running every 15 minutes is a clear and obvious inferior projects (which is very different than saying “just running a bus is a more cost effective solution”)

        A bus might have a lower platform cost/hour, but along most routes it will higher a higher cost/mile because of slower speeds. East Link will certainly be cheaper to run on a cost/passenger-mile basis, even with flat ridership, because the 550 required many more buses to support the same headway.

      16. Connect 2020 was in operation much of 2020 Q1. Service was substantially reduced.
        Yes, that’s clear in the report. Service was essentially cut in half. But if boardings per hr went up (possible since everyone had to get on fewer trains) then how can boardings per trip go down? Link is the only mode where these two metrics are inversely related (ST Report).

      17. Al, I agree with everything you said about a stub-end at Westlake, except the scissor track between Pine and University Street. You simply cannot do that! Those are compression-ring bored tubes you want to break into and what’s more, do it under traffic!

        No, just no!

        The only way into the DSTT is through the retaining walls in one of the station boxes or at the Pine/Third curve if you count that as a separate vault. That’s it.

        If the connection is only to be a MF access facility, a cross-over within University Street and a connection to the southbound track at the curve would work well. Most non-revenue moves to the MF could be undertaken during periods of long headways at the start and end of daily service. Put a long siding just south of Dravus to store trains which need no service during the mid-day base period. Send trains out from the MF during that time only under the greatest duress.

        Use Stewart for the connection because the angle to Third Avenue is not 90 degrees as it would be at Virginia, but more like 70 or maybe even 65. The angle into Westlake is the same from either street.

        Doing this with the same-level stub would be a decision never to add a new tunnel through downtown, and that might be a show-stopper for ST.

        So far as the frequency discussion, yes, Lynnwood is a natural turnback point. I have no idea what ST thinks it gains by going to Mariner.

        And, Mike, you’re wrong. There is very little additional cost to add another SOV to a freeway, once the road is already there. The road surface abrasion is caused by heavy vehicles (including buses), not passenger cars or even pickup trucks. Running a train is a perpetual operating cost and required a capital expenditure by the agency every fifteen to twenty-five years.

        The only reasons to run trains are capacity, natural barriers and access to traffic constrained activity centers; they only rarely “pay for themselves” through lower operating costs.

      18. If it’s not feasible to break the tubes, I have little issue with moving the switch and connecting track to a station vault.

        The gist of the idea is to have Ballard-Downtown as a stand-alone line. How that line terminates in Downtown Seattle would need further study. The best way to have an OMF needs further study.

        I was actually wondering if we even need to keep University St Station open. It’s so close to Westlake. Maybe it could be closed for a few years while tracks and platforms are reconfigured.

        That reminds me that DSTT2 would have similar issues if connects were ever needed. That includes the Aurora and CD branches that Seattle Subway suggests.

      19. I don’t think operating Ballard-Westlake as a stub will work, for the same reasons a Ballard-UW stub doesn’t work – you cannot empty a 4-car train onto a line that is already full. Westlake is much better than UW (because many of the stub riders will exit the station rather than transfer), but the crowding onto the DSTT1 (running only 2 lines) is likely prohibitive. Those trains from Ballard need to run through downtown, no simply to downtown.

        ” I have no idea what ST thinks it gains by going to Mariner.” It gains better frequency for riders traveling between Lynnwood and Mariner.

      20. I don’t think operating Ballard-Westlake as a stub will work, for the same reasons a Ballard-UW stub doesn’t work – you cannot empty a 4-car train onto a line that is already full.

        That is absurd. I don’t think you understand the rider dynamic or the transit geography in the area.

        First of all, this is a subway inside an urban area. It is not a commuter train. On a commuter train it is common for a train to arrive at a station and have very few people exit, while a lot of people enter. In Puyallup, for example, 1,500 people a day board the train headed to Seattle, while less than 80 exit. Many of our commuter buses work the same way. At Ash Way, for example, a southbound 512 will pick up 330 riders a day, while it drops off less than 40.

        All of this changes once a bus gets into the city. In the case of South Sounder, it happens right outside the city. More people get off a northbound train at Tukwila than get on. This means that the train actually gets less crowded before it reaches Seattle. In the case of the 512, way more people get off the bus in the U-District than get on. I realize there are other reasons for this, but the point is, even an inconvenient-as-hell stop next to the freeway still had close to 300 riders arriving from the north end. The UW is a major rush hour destination, which explains why Community Transit used to run peak buses there.

        You are just ignoring the dynamics of an urban transit system. A train from Ballard to UW might be full, but it would not be full of people all waiting to transfer to go downtown — quite the opposite. A few would be headed south on Link; others would be headed north; some would be transferring to a bus somewhere else; a lot would be headed to the UW, and of course, plenty of people got off the train before it reaches the UW. During rush hour very few would be transferring to go downtown — not when express buses can get them there faster. They might be headed to Capitol Hill, but that is more of an all-day destination, not a peak destination.

        The idea that four car trains running every 3 minutes would be overwhelmed at the UW ignores the rider dynamic present there. It also ignores how big these trains are, and how frequent 3 minutes is. Just look around at various systems in other parts of the world, and it is easy to find plenty of systems with trains the same size running far less often with much bigger demand.

        What is true of the UW is true of Westlake. A southbound train on the main line will let off plenty of people at Westlake (or before). Even with UW Link, which only had two stations north of Westlake, more people got off at Westlake than at University Street and Pioneer Square. Likewise, a train coming from Ballard will have riders who get off before Westlake. They will get off the train at Uptown, and the two South Lake Union stops. Then of course, you have plenty of people who will get off at Westlake, and walk to their destination, or transfer to a bus.

        In both cases you are ignoring where crowding is most likely to occur. This is the part that Tom gets right (even if he underestimates the ability of the trains to handle it). The most crowded part of our system will be between Westlake and Capitol Hill, northbound, in the evening. This will include commuters to the north end, but also people going out in the evening (to Capitol Hill and the UW). Whether the Ballard train is a stub or connects to the main line has little to do with it. Riders from the Ballard train who contribute to the crowding (e. g. South Lake Union to Northgate) will have to transfer either way. From a potential crowding standpoint, having a stub is very similar to having a second tunnel. You free the main potential choke point (Westlake to Capitol Hill) to have more trains (every two minutes instead of every three). It wouldn’t contribute to the crowding problem, it would help alleviate it.

        The main drawback to a stub is that it would be an inconvenience for some riders. Whether spending an enormous sum alleviating that problem (with a handful of very deep stations) is a different issue, which deserves its own comment. The main point is, a stub line wouldn’t alter the crowding dynamics one bit — if anything, it would help them (as much as a second line would).

      21. The only way into the DSTT is through the retaining walls in one of the station boxes or at the Pine/Third curve if you count that as a separate vault. That’s it.

        This kind of comment reminds me of those you will find on sports blogs. It is common for people to talk about football, even though they don’t have any experience coaching a team, or playing at a high level. Somehow they can judge a quarterback, even though they can’t recognize a cover two defense, let alone the variations.

        The simple fact is, you don’t know where a split would occur, what it would entail, or how much it would cost. No one does. If you talk to a real professional — someone who does this for a living — and ask them this question, they will simply say “I would have to get back to you”. Any other answer is just speculation. It is like asking an NFL quarterback whether we should have handed the ball to Beast Mode. He has to look at the tape (and review the personnel in the game at the time).

        But as long as we are speculating, I might as well join in the fun. My guess (which has just as much merit as anyone else’s) is that the southbound track would meet up with the southbound track on the main line around Westlake station. This is how most maps would draw the connection, as it is intuitive. The northbound track would do the opposite initially — it would veer right (east) away from the two mainline tracks (but without crossing them). It would then head downhill, dropping below the height of the other track, which I assume is gaining altitude at that point*. Then the track would go under the tracks, and meet up before or at the Denny Station. That means two tunnels connecting the Denny Station with Westlake. The northbound track would be curvy, but this is light rail, and the trains will be going slow anyway. I have no idea if this is any easier or cheaper than any alternative. No one here does, and if a professional has done some initial research, I would love to read about it. Speaking of which:

        * I don’t know of a 3 Dimensional rendering of the light rail line. It would be interesting to see, especially when it comes to discussions like this. At the end of the day, though, it is a pointless discussion. Unlike larger, more conception issues (like whether a station at 130th makes sense) it is best left to professionals. People like us should be coming up with the ideas (since Sound Transit does not employ professionals tasked with the job and the amateurs have done poorly). But engineers should research those ideas, to see how much they would cost, and how disruptive they would be. At that point, we could have a public debate as to whether it would be worth it. Simply put, the concept of sharing the tunnel between the various lines should be studied, in the same way that they’ve studied a new tunnel.

      22. A truly apples to apples bus service would require either hundreds of millions in investment in bus freeway infrastructure and/or running buses far more often than the train (to make up for inferior reliability).

        OK, folks seem to have missed the point. What we were speculating on was whether ST would run trains to Everett frequently. My point is that agencies are reluctant to do that, simply because it is very expensive. A bus that carries 50 people costs less to operate than a train that carries 50 people. In the middle of the day, a train between Everett and Lynnwood won’t carry that many people, which means it will be extremely expensive to operate. Thus it is highly likely that trains will turn back at Lynnwood. What that would actually mean for riders depends on ST’s approach to frequency. Given ST’s history, it doesn’t look good. If ST is comfortable giving riders in Rainier Valley crappy frequency, then giving (far fewer) riders in Everett frequency that is twice as crappy seems likely.

        As to whether it would take an enormous sum in bus infrastructure or service to match the quality of light rail service along the freeway misses the whole point. No one would do that. You don’t try and mimic a terrible value by building something similar. If your spouse wants to a buy a Hummer even though they never go off road, you don’t try and convince them to buy a Jeep. Both are crappy cars, inappropriate for the driving that actually occurs.

        If you want to improve transit in the suburbs, it is pretty simple:

        1) Have a good terminus station, with HOV ramps connecting to it. Ideally this is a destination in itself, but it doesn’t have to be. Lynnwood and Federal Way are fine for this, and arguably overkill.

        2) Spend money improving the local bus system. There are savings that come from building a rail system, but they are minor compared to the costs. East Link, which is definitely justified in its own right, and comes with very big savings from truncation, will not result in great improvements in bus service on the East Side. Quite the contrary. It isn’t good now, and it will continue to be poor in the future. There is just no substitute for spending money on bus service in those areas.

        3) Run the express buses from various neighborhoods to the terminus. The main advantage of a subway line versus an express bus is that it serves places along the way. People from Lynnwood used to take express buses to the UW. Now they take express buses to Northgate, which means that it is much easier for them to get to Northgate, Roosevelt and Capitol Hill (as well as all of the buses that serve those stations). The same is true of riders of the old 41. It takes longer to get downtown, but now they can get to Roosevelt, the UW and Capitol Hill much faster. The same is true with service from Lynnwood to Everett. The problem is, there are very few people trying to get to places like Ash Way.

        4) Change the HOV 2 signs to HOV 3. The fixation on The Spine has allowed this simple improvement to be ignored.

        It is worth noting that this is what people do all over the world! This really is a uniquely American approach (running subway lines next to the freeway for miles and miles) and like many uniquely American approaches to transit, it has failed, miserably. Studies prove this. You simply aren’t going to save that many people that much time, despite the enormous cost.

        This is why everyone else in the rest of the world just runs commuter trains (on *existing tracks*, at very low cost) or they run buses out to distant suburbs and cities. This is not only what other countries do, but it is what the most successful transit cities do here. It is crazy to think that our rail system will soon be one of the biggest in North America, yet get so few riders. This is because it will be significantly different than most systems. It will cover very little of the urban core (with very few stations and lines) and yet extend an extremely long distance to low density neighborhoods in low population cities. It will extend farther from the center of the city than massive transit systems covering mega-cities (like in New York, Toronto or Chicago). Yet it will barely move the needle when it comes to transit mode share. To do that would require spending money on more cost effective projects (like rail in the urban center, and buses everywhere else).

        No, this isn’t an apples to apples trade using a different mode. It is applying best practices — hell, even just standard practices — from around the world to our situation.

      23. “This is because it will be significantly different than most systems. It will cover very little of the urban core (with very few stations and lines) and yet extend an extremely long distance to low density neighborhoods in low population cities. It will extend farther from the center of the city than massive transit systems covering mega-cities (like in New York, Toronto or Chicago). Yet it will barely move the needle when it comes to transit mode share. To do that would require spending money on more cost effective projects (like rail in the urban center, and buses everywhere else).”

        Very true, and to be fair Ross has been saying this for a long time. But I am not sure what can be done now. Most of the spine is built, and as built I have my doubts East Link will have the ridership to prove its worth.

        Rather than begin at the urban core(s) and work out ST and the spine started at the outer limits and worked in, but ran out of money when they got into the core. This includes a tunnel under Bellevue Way.

        If you want to transition folks from cars to transit you need to realize they don’t need transit in suburban/exurban areas; they need it in urban areas, and want easy feeder service, whether buses or park and rides, to get to the urban core. Transit works to urban cores because it can be easier, faster, and cheaper depending on traffic congestion and parking costs.

        Feeder bus service — whether it is half a mile or ten miles from a rail station — will determine the success of light rail because both need a second wait and transfer even though the total length of the trip is the same either way, and budgets don’t look good.

        Maybe the only area I would add to Ross’s post is I think park and rides are very effective in suburban areas, although park and rides depend on traffic congestion and mostly peak commutes, because otherwise that driver simply drives to their ultimate destination. (I also don’t think it makes sense upzoning the same outer suburban residential areas if providing transit service to them is so difficult).

        Providing any kind of frequent bus coverage in suburban areas like East King Co. is just too expensive and slow, and the one place those citizens will take transit to is an urban core. No one will take light rail from Issaquah to S. Kirkland, because driving is easier, and you don’t need transit to either terminus.

        Generally you get one transfer if you want folks to take transit, and that first mile has to be EASY. If the bus in an area like East King Co. is not easy because lack of density, size, and cost, then be realistic and find another first mile access, most likely a park and ride. You will never make those folks hate cars, but you can make them ride transit to an urban core if that first mile is as easy as …. well, driving.

      24. Ross, I wish there was an easy way to search prior comment threads, because I’ve already addressed this with a direct link to a source document in a prior thread (someone else provided the PDF link so I don’t have it): ST’s estimates for Link ridership heading south of Westlake during peak exceed the carrying capacity of a single tunnel.

      25. I wish there was an easy way to search prior comment threads

        Google/Yahoo with at the end of your search string.

      26. AJ, because this is such a long thread, I provided a link to the PDF I think you were wanting as a new thread at the bottom.

      27. “If you want to transition folks from cars to transit you need to realize they don’t need transit in suburban/exurban area”

        You need to realize it was the SUBURBS who pushed for the Spine. It wasn’t something ST dreamed up and imposed on the suburbs because it doesn’t know what a city is. It was the suburbs and counties who insisted on the 60-mile model as a condition of them agreeing to ST1/2/3. Your beef is with Everett, Tacoma, issaquah, Bellevue (partly), Redmond, King County, Snohomish County, and Pierce County. Those are the ones that pushed the long spine and convinced the legislature to structure ST so that it would prioritize the spine. (One 3-county tax district, subarea equity, all that.) It wasn’t Seattle or urbanists or Seattle Subway who championed that at the beginning or in early ST3 planning, it was the suburbs, the ones you say are too low-density for Link.

        It was Everett and Tacoma and their counties who pushed for the long spine. It was Issaquah’s mayor who pushed for the Issaquah-South Kirkland line.

        ” Thus it is highly likely that trains will turn back at Lynnwood.”

        ST already has a framework for that. Line 1 (West Seattle) will go all the way to Everett Station. Line 2 (Eastside) was originally going to go to Lynnwood peak hours, Northgate off peak. Then ST extended all Line 2 trains to Lynnwood thinking it would need the capacity. Now it plans to extend it to Mariner for the same reason. It could easily be re-truncated back to Ash Way or Lynnwood if need be. Not even ST is proposing to extend Line 2 all the way to Everett Station.

      28. Ross, your design would be ideal, but, unfortunately, Pine Street is not wide enough to put a new track outside the existing tunnel without seriously threatening the foundations of the buildings on the south side of the street.

        If Metro or whoever designed the cut-and-cover section east of Pine had kept the center passing lane all the way up the hill to the Convention Center Station turn it would work. The center between the two tracks, which isn’t load-bearing to any significant degree could be diamond sawed, the slabs removed, and then a trench dug out for a plunging junction at the curve. That would have been great!!!!. It could even have curved into a station a couple of blocks north of the Convention Center in the middle of all those tall buildings along Minor just south of Denny Way.


        But the tracks narrow immediately east of the Westlake platform and there is a very important load-bearing wall to the south of them which can’t be disrupted for more than a few yards at a time. It’s what holds the street decking up. Look at this diagram:

        That bearing wall looks to me like it is underneath the sidewalk, directly abutting the foundations of the buildings. You can’t belly out to the south.


      29. TT, do you think you could dive up or down between University Station and Westlake in the center lane for the North bound track and use the curve to tie in the South track?

      30. Martin, that is exactly what I have been advocating if through revenue service to the south through DSTT1 rather than a Westlake stub is chosen for Ballard. I am less confident that it would work than I was, because the tubes bulge into the space between the tracks quite a way. The connection would certainly have to be cut-and-covered; there is not enough room for a third TBM-bored tube between the other two at the north wall.

        If the available width between the tubes is too narrow to put a track it might be possible to place the turnout right at the south wall and begin the plunge within the station box. The stations are about 400 feet long, so if fifty feet were consumed by a tight turnout, and the gradient were 6%, the trackway could have descended by about fifteen feet by the north wall. At least in theory that should be sufficient for the cars to clear the fat sides of the tubes and the pans to pass between them. The south approach to University Street actually “pinches down” toward the platform, so the northbound track is angled toward the center of the station box. The turnout could consist of a straight branch into the center lane rather than an “S”. That would allow the vertical curve toward the down-ramp to begin earlier.

        Unfortunately, however, University Street straddles the BNSF tunnel and there is only fifteen feet of vertical clearance between the floor of the station box and the old tunnel’s ceiling. I have not been able to determine exactly where the centers of the two tunnels cross one another, but if it’s anywhere north of the southernmost portion of the station box, even this option probably can’t work.

        However, the idea still works for a stub Ballard-Downtown “non-revenue” connection, because that only needs to be single-track. It’s fine to run trains coming from the MF headed to the interconnection track “out of direction” on the southbound track between the curve at Third and Pine and the station. That won’t happen often and could always be quick.

        Sadly, though, I now think that my idea of using USS as the junction point for “through” revenue trains to Ballard is likely not “doable”. Any such connection will have to be on the surface, probably using the CCC trackage. Doing that would make Ballard-Downtown much less valuable, though, because of the hideous climb between First and the Financial District center at 4th and Madison. Yes, people can switch to the BRT line, but that makes two transfers for most trips between there and Ballard.

        Things are not looking good for alternatives to a deep DSTT2. The best alternative, a cut-and-cover tunnel under Fifth Avenue, is politically difficult, verging on the impossible.

    2. The suburbs have 3/4 of the metro’s population and 3/4 of the ST boardmembers. Dow Constantine, the King County executive, represents the suburbs at least as much as he does Seattle, and is sympathetic to suburbanites’ demands of lines to all their cities, lines parallel to highways, many large P&Rs, and not upzoning single-family areas because it would anger voters like DT. The “parallel to highways” part is key, even if it’s not on the highway but a mile away, because many suburbanites see the primary problem as highway congestion and want an alternative to it. They don’t think about first/last mile access because they’re thinking about highways. This is all what led to such a long spine, too few stations in Seattle, not enough lines in Seattle, and not enough resources in Seattle. It’s because suburbanites are suburban-minded and outvote Seattlites. And it’s suburbanites who want the long spine. They aren’t saying, “Don’t build it in my area because we’re too undense”, they’re saying, “Build it in my area now because we need it.” If you suggest shifting the resources to Seattle because they need it more and will ride it more, the suburbanites say, “No, build it in my area now, because we really need it.” Otherwise the project wouldn’t exist or be cancelled. It was suburbanites who got these lower-density extensions onto ST’s radar and prioritized. So if you want to cancel them, it’s your suburban neighbors and politicians you have to convince. Ten people on Mercer Island may want to cancel East Link, but hundreds of thousands of voters/constiuents outside North King wants those non-North King extensions.

      1. Mike, you are confused about the design of Link.

        First without subarea equity there would have been no ST, and I for one am damn glad for subarea equity. If you look at the total cost of East Link — around $5.5 billion to Redmond which is close to $5 billion less than just ST 2 will raise for the eastside subarea –it was an economical light rail line using public rights of way and surface lines and stations. It was ST, not the subarea, that rejected a tunnel under Bellevue Way, which will badly hurt East Link, although East Link was never going to be well used. I also think the Link designs are comparatively efficient and cost effective in every other subarea, except N. King Co.

        WSBLE is only a N. King Co. subarea problem, and I have news for you: Pierce Co., S. King Co. and SnoCo don’t have the $275 million each ST 3 requires them to contribute to DSTT2. DSTT2 is going to end up a 100% NKC subarea project, with maybe $275 million from the eastside subarea.

        You missed the point of Ross’s post: he wasn’t complaining about subarea equity, he was complaining about mode, specifically demanding rail for 90 miles, and the cost of rail when ridership does not support it for many areas outside the Seattle core, and likely never will. Transit advocates somehow think rail will be the change agent to get folks out of their cars onto transit, but really it was congestion that did that and the forced work commute WHICH IS AN URBAN PROBLEM, and the pandemic has alleviated congestion, a huge concern for ST if the commuter does not come back.

        What really drove these expensive decisions was something AJ repeats: the claim that this region will someday have millions more residents and millions more transit riders, when current growth rates for Seattle and the County are at or below 1%, and transit ridership has plummeted, but not miles driven. Both ST and the PSRC have been very dishonest about population and ridership projections.

        If you want to see the future when it comes to ridership on Link look at ridership today. That is the future. The pandemic will slowly fade, but the work commuter will not return to anything like pre-pandemic. So plan for the future based on ridership numbers today, not dishonest projects from ST. What I think Ross was getting at is those numbers suggest we built too much light rail, not too little.

        I can’t believe anyone would believe ST’s future ridership estimates today. We built a 90-mile spine through three huge counties that would need at least 10 million residents to pan out (in a pre-pandemic hub and spoke world leading to downtown Seattle), so we are building a light rail system for millions of residents who are not here now, and won’t be. By the time this region reaches 10 million (which I doubt it ever will) transportation will be completely transformed, and “fixed transit” like rail will be obsolete.

        There really isn’t an option whether “to take a break” on WSBLE. The subarea does not have the money for any of the alternatives, and the “stakeholders” are demanding the most expensive option. Other subareas are ok because they did what you complain of: built reasonably priced light rail lines along the interstate because their budgets demanded it because light rail is so enormously expensive per mile, and avoided $12 billion projects to stay within their budgets. The eastside is an anomaly because of the huge growth in ST revenue, but the relatively modest cost of East Link.

        What the pause is really needed for is not to gather more money for WSBLE, or to convince the stakeholders surface lines and stations will be acceptable, because the folks on this blog make a terrible mistake: for the 95% of non-transit nuts transit and light rail are not the end all and be all of a community, and no one wants a surface rail line and station through the heart of their community. I thought East Link being moved to 112th proved that. Even Seattle wants fewer buses along the 3rd Ave. “transit mall”.

        The pause is necessary to determine whether the entire paradigm of transit and light rail makes sense anymore, without the commuter, and with a deurbanization of the workforce. Rogoff’s recent statements about the huge drop in farebox recovery rates should be a wakeup call. Those riders are not coming back, so plan for that.

        Even if the N. King Co. subarea had the $20 billion it will likely take to complete WSBLE (and based on Rogoff’s $11.5 billion deficit admission last January when WSBLE was still estimated to cost $12 billion my guess is the subarea has close to zero after 130th and Graham St. stations) I would still recommend a pause, because I think ST and this region is pursuing a very expensive project at the end of a gilded age for this kind of transit.

        Transportation is going to change dramatically over the next 20 years, and my guess is it is going to become much more personal, like driverless car fleets, driverless shuttles, not very expensive trains that need deadheading with very few riders to far flung areas, especially during non-peak hours, if there are any peak hours post pandemic.

        The advantages rail has are: 1. congestion in urban areas; and 2. it can move more folks than buses. Well, capacity is really only an issue during peak hours, and no matter what we are looking at a 20% to 40% reduction in peak hour commuters post pandemic, probably more today based on current ridership on Link. And if your trip ends anywhere in suburbia or exurbia you don’t need transit, or want it. Who in the hell would take Link to Federal Way or the Tacoma Dome if I-5 has no congestion? What do you do when you get there? Transfer?

        ST can finish most of the ST 2 projects, and will add stations at Graham St. and 130th. It will open East Link, Federal Way Link, and Lynnwood Link. According to the Times article ST does not have close to the money to run those operations unless ridership increases dramatically, and I don’t think it will. Let’s first see if we can afford to run the light rail we have built, and whether light rail is the future of transportation, the really bigger issue.

        WSBLE will get paused because the subarea does not have the money for any of the alternatives let alone the one the stakeholders want, but my guess is the stakeholders won’t miss light rail in reality, and by the time we revisit the issue in the 2030’s transportation will be fundamentally different because work and housing will fundamentally be different, and new technology will serve that new paradigm.

        It wasn’t subarea equity that brought us here, it was a pandemic, a transit mode that is too expensive and on the cusp of being irrelevant, an agency and its sycophants like the PSRC who sold everyone on population and ridership numbers that were simply false, and a demand for tunnels and underground stations and lines among the stakeholders, or no rail, which is their call.

      2. “First without subarea equity there would have been no ST, and I for one am damn glad for subarea equity.”

        Subarea equity was for the spine. It was to prevent suburban money from going to Seattle first and then the spine would be delayed or never completed. It’s the opposite of what metros with the most successful transit networks have, like say Vancouver or Germany. where they tax the entire regional unit but build rail where the pedestrian concentrations are.

        “It was ST, not the subarea, that rejected a tunnel under Bellevue Way,”

        It was Kemper Freeman who spearheaded it so it wouldn’t go near Bellevue Square. He’s part of the subarea, one of the most influential parts.

        “really it was congestion that did that and the forced work commute WHICH IS AN URBAN PROBLEM”

        There are two ends of a work commute, and frequently the house and/or workplace are in the suburbs. Some of the worst congestion is on I-5 in Lynnwood and Everett, 405, I-5 around JBLM, and 522 in LFP and Kenmore. Those aren’t urban areas. Yet it’s why it’s so urgent to get Link to Lynnwood, one of the reasons Everett insists on Link (besides jobs), and why Stride is so important.

        “the claim that this region will someday have millions more residents [doubling to 10 million] and millions more transit riders”

        Neither the PSRC nor the politicians nor ST are saying that. The PSRC is saying a million people will come in ten years, and when that’s up it says another million is coming, etc. ST3 was originally timed for 25 years, so you can take that as one million or two million more people. Eventually it might reach 10 million, but only in fifty or sixty years, and the projects aren’t scaled for that long. They’re scaled for one 20-year spurt of growth now, another one maybe later.

        “What the pause is really needed for is not to gather more money for WSBLE, or to convince the stakeholders surface lines and stations will be acceptable,”

        What pause? ST is not pausing, it extended the construction period. Construction can’t fully start until ST2 is finished because 2/3 of the revenue is going to ST2. If you mean the kind of pause that third-party advocates are recommending, that’s still unofficial.

      3. Mike, the eastside subarea paid for 100% of Link including across the bridge, as well as 100% of the east-west-east buses until East Link opens, which will cost around $1 billion. Plus our park and rides were extended so WSBLE could meet the debt ceiling. So I don’t see how my subarea received a dime from another subarea.

        There is plenty of capacity in DSTT1 for East Link and Line 1, especially post pandemic, and the county as a whole paid for that. I don’t think the eastside should have to pay for DSTT2 for WSBLE when eastside ridership to either West Seattle or Ballard will be near zero. The other subareas never received a contribution from North King Co., and are supposed to contribute $275 million/each to DSTT2. IMO that is unfair because the capacity is for WSBLE. N. King Co. doesn’t own DSTT1. Bellevue and the eastside paid for its tunnel.

        When it comes to the spine I have always said I thought it was unwise for the N. King Co. subarea to spend so much to run rail to S. King Co. and SnoCo when there is so little there and in between, but that was not my decision, and Seattle thought the world ran through downtown (and what is the point of light rail without peak urban congestion). There is your WSBLE there, although the tunnels and underground stations to Northgate through Seattle were very expensive. But economically, running light rail from WS to Ballard is not unlike running it from Issaquah to S. Kirkland. Why?

        Ross’s point was buses should have fed light rail that began much closer to the urban core. N. King Co. should not have spent its revenue running light rail to the edges of its border, or to Lynnwood, when those riders could have taken buses to a light rail station much closer to Seattle, which would have funded more rail (subway) within the urban core. Very few riders will be from Everett or Tacoma all the way to Seattle.

        The main issue going forward however is likely future ridership won’t come close to covering operation costs, especially without the peak commuter and a current 30% non-payment rate on Link when ST never built a turnstile type enforcement system.

        The enormous costs of light rail per rider/mile were questionable pre-pandemic, which is why ST inflated future ridership estimates so much. Based on today’s actual ridership, which likely reflects actual future ridership, not only does the capital costs for light rail and exurban rail like WSBLE make no sense, but the costs to run that rail are well above revenue, both general fund and farebox recovery. The good news in one way is Link will have too much capacity, so DSTT2 is not necessary.

        Generally, if you can’t afford to operate the light rail you have you shouldn’t build more. ST will complete the EIS, the Board will select the least affordable alternative that is favored by the stakeholders because none of the “alternatives” are affordable, and then what?

        We will find out there is more than enough capacity in DSTT1 for lines 1 and 2, (actually too much), which is probably why ST is looking at something as stupid as beginning WSBLE with the West Seattle stub, unless Dow and the Board know the stub is it, and even the stub is affordable.

        I think the solution if peak hour capacity is not an issue is to go to 15-minute frequencies on Link, because 15 minutes will still handle post pandemic ridership levels just like it can handle ridership today, and hopefully balance operation costs closer to farebox recovery.

        I know some think such long frequencies discourage transit ridership, but I just don’t see much more ridership post pandemic on Link with shorter frequencies, unless it can truly be shown shorter frequencies result in greater net farebox recovery (which is number of riders times fares actually paid) less the extra operation costs of shorter frequencies.

        Operations always came down to number of riders who actually pay a fare, amount of fares, vs. the costs of operations which includes frequency. When ridership falls so should frequency, although ideally more riders should pay a fare.

      4. Random thought: what if highway maintenance and expansion was done at a subarea level? I’d sure prefer all that revenue from Seattle stay in the city to rebuild bridges and pave streets rather than pay for miles of highway upkeep in the middle of nowhere.

        Of course, there is a larger societal benefit to an effective roadway network. I make the point to illustrate how misguided strict adherence to “subarea equity” is.

      5. Al, I think a lot of high taxpayers agree with your point. I know a big point of contention on the eastside is how much of our taxes go to areas outside the eastside.

        Subarea equity did not disadvantage Seattle or N. King Co. That subarea simply chose very expensive projects with lots of tunnels and underground stations (at least in the white neighborhoods), and very, very expensive projects like WSBLE. The four other subareas are supposed to contribute $1.1 billion toward DSTT2 when none of them got tunnels (except Bellevue which it paid for). Talk about the king robbing from the poor. Imagine S. King Co. giving N. King Co. $275 million so NKC can have lavish tunnels and underground stations to remote residential neighborhoods.

        East Link proves you can build a pretty long line for a relatively reasonable sum, if you use public right of ways or green belts, run trains on the surface, build modest stations, and hug freeways as best you can. But if you insist on deep tunnels and underground stations in the downtown core or in the white neighborhoods, and long rail lines to small residential communities like Ballard and West Seattle over water, then yes, you are going to need a lot of money.

      6. And saddle Seattle with cost of rebuilding I5 through the city? Seems like a loser for the Seattle subarea.

      7. Daniel, North King did NOT make the decision to “spend so much to run rail to S. King Co. and SnoCo when there is so little there and in between”. What an effyouseeking Republican lie. It was the Sound Transit BOARD, dominated by suburban King County and non-King County members that made that decision.

        You make a lot of excellent points about the costs of these various projects and the foolishness of building in the suburbs, and THEN you blame Seattleites for that foolishness. How stupid can you be? [That was rhetorical: the simple answer is “He’s not stupid. He’s a classist prick!”]

      8. Tom, I am not a big fan of the victim card, especially when the victim is Seattle, or N. King Co. This idea S. King Co., Pierce Co., and SnoCo (the latter two still don’t have light rail 30 years later) made Seattle do it, or were bullies on the Board, is silly and hollow. Or that somehow S. King Co. and Pierce Co. are the “classists” in this play, and Seattle the downtrodden victim despite the fact 90% of all rail completed has been in N. King Co. Come on.

        What the other subareas — especially East King Co. — insisted on was subarea equity. Looking back, using their transit funds for non-light rail projects probably made better sense and would have gone much further. Light rail and subways and underground stations benefit an urban area, not these four other subareas that now have blown their transit wad on rail, and don’t have the funding for good feeder service.

        Seattle progressives and the transit nuts sold the region on the spine because they thought it would eliminate cars, create a dense urban zone from Everett to Tacoma, solve global warming, and most importantly lead all work and commerce to downtown Seattle.

        To be fair, how could they know ST had manipulated nearly all the numbers, light rail is too slow for a spine, it would skip First Hill and run out of money for WSBLE, or Seattle would upzone SLU without a rail station, or the city council would systematically drive business to Bellevue and make the downtown core the least attractive area in the city

        I suppose if the cost estimates and ridership estimates had been remotely accurate, all the promised stations built, and Seattle was still the one true hub, and first/last mile access that understood suburbia had been created, the spine might have made sense in concentrating urbanism in the core, the one area our meager population could support true urbanism, and protecting the residential neighborhoods, the entire point of light rail.

        But then the same progressives decided to upzone all of Seattle, which further hollowed out the downtown core, even though it is virtually impossible — logistically and financially — to provide feeder service to these remote areas, and the city council started a war on business which moved to the eastside.

        The fundamental problem IMO with Link is it was designed to bring suburban and exurban folks into the urban core, for work and play. Unfortunately, it was designed by urban planners.

        Had it been designed by suburban planners it would look today much like what Ross has described: a vibrant downtown core with easy (and mostly underground) transit from one corner to the next, with feeder service (buses and park and rides) at the margins from Bellevue (not Redmond), Northgate (not Everett) and the airport and then Sodo (not Tacoma), with yes some one seat buses for longer trips from Tacoma or Everett to Seattle or beyond. There just is almost no point in transit in suburbia, certainly expensive transit like light rail.

        What the urban planners never understood is how does someone get from a SFH in surburbia to a transit station, rail or bus, to get to the urban core (because no one in suburbia takes transit to suburbia) when it is too expensive for micro transit, and too large for feeder buses. So now they think let’s eliminate the SFH because they must be classist, when that is not the problem anymore because that commuter is not coming back to the urban core.

        You are correct: probably 90% of the eastside is “classist” to some degree. They don’t read blogs like this one. When they no longer need transit — like today — it means nothing to them. Their lives have nothing to do with transit, and when it did they hated it. Unfortunately, ST designed a system in which these same classists were expected to pay for light rail with fares and levies, and it looks like the classists have decided against that idea. You blame them. At least when they rode transit they paid their fare.

      9. North King built Link to Rainier Beach, as it should have. If the spine vision didn’t exist, well, there wouldn’t have been Sound Transit at all, because the impetus to create it was inter-county expresses that the county-based agencies were structurally unable to provide (because they kept being prioritized last after neighborhood service). Neither Seattle nor Metro had enough tax authority on their own for light rail like Link. The Monorail’s Ballard-West Seattle plan was predicated on Link serving southeast and northeast Seattle. Without Sound Transit or Link. Seattle might have gone to the legislature for light rail in eastern Seattle, but there’s no guarantee the legislature would have allowed it like it did for ST. If it had to do the Monorail without Link, it couldn’t have served Ballard and West Seattle and left the U-District unserved, because that’s the highest-volume corridor and the one the express buses were most melting down on because they couldn’t keep up with capacity demand and I-5/Eastlake congestion.

        If there was no Sound Transit or no spine vision, and Seattle somehow had tax authority for light rail like Link, then maybe Seattle might have a completely different light rail network without southeast Seattle (so nothing for South King to latch onto), but that’s so many hypotheticals it’s not worth considering.

        The spine vision came first. The spine vision came from all the suburbs. The spine vision convinced legislators to create a regional transit agency and give it enough tax authority for Link and Sounder. The suburbs insisted on subarea equity to force ST to focus on the spine first. North King wanted light rail in Seattle. It did not push for the spine, the suburbs did. The suburbs have 3/4 of the population, 3/4 of the voters, 3/4 of the ST boardmembers, and the legislators are most sympathetic to them (because suburban legislators identify with suburban drivers, and Spokane/Yakima/Vancouver/Pullman/Omak legislators identify with them too).

      10. Over and over you have blamed North King for building light rail “to the edges of the subarea” instead of building urban subways, as if it had control of the process. And then you turn right around and JEALOUSLY criticize North King for having “subways in the white areas”.

        Seattle showed what its vision for a subway system was in 1972 with Forward Thrust: a heavy rail subway serving the three of the four quadrants of the city, cross lake commuting to Overlake and Renton. It even has BRT to West Seattle when “BRT” wasn’t a thing.

        It was you dog biscuit eaters in the ‘burbs who shot it down and sent the money to Hotlanta.

    3. “creating four lines rather than just two at 12 minutes each rather than six. ” I think that would result in many riders just waiting 12 minutes for their specific train, like in the Muni tunnel. The best design is for most riders to want to take the first available train. Mimicking the Muni’s multiple branches is not best practice.

      Personally, I think the best ST3 operating pattern would be to simply send 100% of the UW trains to Bellevue (boosting frequency on Seattle-Bellevue), as UW-Seattle downtown-Bellevue-Redmond should be our core urban axis, and then send both WS and Rainier trains to Ballard (boosting frequency on Ballard-ID). This would require a larger fleet and a 5th OMF (or a major expansion of an existing OMF).

  2. I find the UW Stadium to be abhorrent at 95 ft and costing 100s of millions of dollars. So now what is ST doing, trying to out do themselves with 100+ foot and costlier stations — sheer stupidity. It seems to me the only reason they added some of those stations was to make a Ballard station palatable. I say give Ballard their damn station and proceed their line to South Lake Union followed by the elimination of the next 3 stations and then popping up with a shallow station at ID.
    Streetcar is planned for SLU to ID, just reconfigure it if need be for a fraction of the cost.

    1. UW Station is deep because it’s next to the Ship Canal and light rail can’t climb steep inclines. The station should have been further north like at the HUB but UW wouldn’t allow it. The downtown tunnel is deep because it’s going between DSTT1, the BNSF tunnel, utility lines, building foundations, and soil conditions. You can’t just wave those away and make it shallow. The most feasible way for a shallow tunnel is cut-and-cover under a street (5th Avenue). But a large part of the community and voters oppose cut-and-cover because of construction disruption. They’re more concerned about businesses failing during the five years of construction than about businesses thriving during the hundred years it operates. Unfortunately that’s a strong political faction that has clout. If we could find enough RossB’s to fill the city council and mayorship and somehow get them elected, then this could change. We haven’t found a way to do that yet.

      1. voters oppose cut-and-cover because of construction disruption. They’re more concerned about businesses failing

        Can’t they do cut & cover one block at a time? I realize that still kills throughput for the street but business should be more concerned about access than “throughput”. Building on 1st or 5th is a lot less impactful than building the original bus tunnel on 3rd. Anyway, it would be helpful to have a dollar cost for such a tunnel. If it doesn’t save much then we can toss it out knowing why.

  3. Remind me again why these three projects (Ballard Link, West Seattle Link and the second downtown tunnel) are being treated as one for the purpose of this DEIS. This isn’t how the projects were presented in the 2015 templates that were developed (using 2014$) for the initial and subsequent iterations of the ST3 package. And the final ST3 proposal certainly was not sold to voters in this fashion.

    My gut is telling me that it’s being packaged this way now because of the weakness of the arguments for the very expensive second tunnel (and its duplicative stations) has been exposed.

    1. I think it’s combined partly because West Seattle Link with three stations cannot compete well for an FTA New Starts grant if it stands alone.

  4. I’ll admit, I never had high hopes for ST3. Even the best rail project (this one) didn’t look that good. Unlike so much of what we have built, and are about to finish, it didn’t pass the “driving at noon” test (at noon, how does a subway trip compare with driving). Too much of it follows an expressway. This is a great way to cut costs (which seems comical at this point) but not a great way to add value. Unlike the existing system, much of this adds nothing along the way. The old 41 was a fast express from Northgate to downtown — often faster than Northgate Link. But Link adds very popular stops (Roosevelt, UW, Capitol Hill). Some riders get a slower trip to downtown, but lots more get dramatically faster trips to other, very popular destinations. From West Seattle, riders will transfer so that they can get to … SoDo?

    So my expectations were low. Yet somehow, they didn’t even meet those. This is much worse than what voters approved, and yet it is much more expensive. Even the best station at the end of the West Seattle line is farther away from the cultural center of the area (the Junction). The preferred alternative for Ballard is to move the station away from the heart of Ballard, towards an industrial area. In downtown, the stations will be extremely deep — so deep they won’t be worth bothering with much of the time. A trip from South Lake Union to the International District will just be done on a bus (which would also be significantly more frequent).

    More than anything, this shows a lack of imagination. The board just doesn’t understand mass transit. They think of it the way you would a freeway. It is geared towards covering distance, and not connecting densely populated areas. Details are ignored, as if a few extra blocks to the station is like driving a bit farther to the on-ramp. It isn’t surprising so many of the plans follow freeways, even though all the literature says that is a bad idea. That is just the mind set of the board, which simply doesn’t know any better. There is no transit expertise on the board. The only person who knows anything about transportation works in Olympia.

    Locally, this will go down as the worst public transit fiasco every built, making a mockery of all the success (from a cost perspective) we’ve had over the years. Even nationally — where we have plenty of competition — ST3 will be looked at as a very poor value. From an international standpoint, it will be laughably bad. A lot of money will be spent, and very few will benefit. Remember, Ballard was by far the most cost effective rail section with ST3, and now prices have skyrocketed, and ridership will go down.

    There are ways to fix this, or at least mitigate the problems, but again, they require imagination — something the board clearly lacks. I’ll mention those ideas anyway, in a separate comment. They aren’t anything knew — people on this blog either came up with them, or have been discussing them for quite some time. But the board should seriously consider them, before it is too late.

    1. I think ST3, as proposed, is still better than some of the crazy extensions that BART has done into tiny suburbs in the far east, while leaving the vast majority of Silicon Valley unserved. The Puget Sound equivalent would have been to run Link to Orting, while Northgate to Lynnwood requires a transfer to a separate heavy-rail train that runs once once per hour.

      1. That’s because San Mateo and Santa Clara Counties voted BART down in the 1950s. They said Caltrain was sufficient, and Santa Clara County wanted to use those federal funds to build expressways instead. An expressway in Santa Clara County terminology is a sub-freeway: it has a limited number of level crossings. If you took 15th Ave W in Interbay and made the exits level, that would be an appoximation.

        So they built those expressways, and the house I lived in until age 6 now has the Lawrence Expressway near it. But now the county has high congestion and much of it is car-dependent and transit is limited, so then they had to build the eastern BART extension anyway,.

      2. Uh… East Contra Costa County now has over 250,000 people! That’s not “tiny”.

        Plus, there is only one BART station (two if you count the transfer station) actually using a classic BART train with the other stations using a cheaper self-propelled standard gauge train running at grade. Finally, the local share to pay for the line was approved by Contra Costa voters and taxes didn’t change for any other voters in the region.

        While I was initially skeptical of not using BART trains there, I have to say it works ok. If we traded 55 mph light rail for 70 or 80 mph self propelled units south of Federal Way or north of Mariner, the rider travel speeds from those places would be faster even with the transfer time and effort.

      3. I double checked the 2020 census. The East Contra Costa cities and unincorporated areas now total about 340-350K. Granted it’s low density — but it’s not tiny.

    2. “Even nationally — where we have plenty of competition — ST3 will be looked at as a very poor value. From an international standpoint, it will be laughably bad.”

      Except one thing: it’s better transit than most of the US has. I can’t think of any network besides NYC that serves more communities and cities with such high frequency. That makes it more like a European network. International people won’t laugh at that, they’ll say “FINALLY!” They might say it should have been a two-level network: inner light rail/metro and outer S-Bahn rather than a hybrid between the two. But at least people will be able to get around, which is the point of a rail network. People in Bellevue and Tacoma won’t have to wait half an hour for the 550 on Sundays or 594 every day and have it get caught in congestion whenever traffic is heavy or there’s a car collision or breakdown. People in Ballard won’t have a half-hour overhead to get to rest of the region (Westlake or U-District access points).

      The costworthiness of a network isn’t just the raw number of butts in seats, it’s also the potential mobility it provides. Even if you don’t use it today, it’s there when you need it, and that makes it more competitive with driving so more people are willing to use it. If we have a good network that goes many places and has high frequency, long-term ridership may be higher than the official estimates. Because the official estimates don’t take into account future upzoning, increasing willingness to use transit, increasing frustration with driving (especially in traffic), and increasing concern about climate change and local resilience.

      1. First off, I was referring to the costs. This is extremely expensive compared to most U. S. projects, and ridiculously expensive compared to the rest of the world. This is going to cost a huge amount of money and frankly, won’t provide that much.

        I can’t think of any network besides NYC that serves more communities and cities with such high frequency.

        Uh, let’s see, there is Chicago, D. C., Boston, L. A. (fairly soon, if not already), the Bay Area, and those are just the ones in the U. S. In Canada you have Vancouver, Toronto, Montreal. Most of those invested in transit a long time ago, or are much bigger cities, so they aren’t a good comparison. But Vancouver is. We are spending way more than they will ever spend, and they will have a much better system, probably forever.

        That makes it more like a European network. International people won’t laugh at that, they’ll say “FINALLY!”

        Oh come on. No one from Europe, or anywhere else that has a decent transit system will look at what we are building and think it is great. They will wonder what the hell we were thinking. Why is there no rail to First Hill, the Central Area or Belltown, but there is a line from West Seattle? Why are there two tunnels in downtown, even though they have almost exactly the same stations? Speaking of stations, why are there so few, even in urban areas? What is up with the line from Issaquah to South Kirkland — what is the point of that? Why did we build an extremely expensive metro to distant cities when we haven’t even covered the urban core? (To be clear, European cities typically have good train service to towns outside the city, but those leverage old rail lines — i. e. they were very cheap to build — otherwise they just run buses. They just don’t build brand new metros that far out — even in really big cities). Oh, and why does this expensive metro stick to the envelope of the freeway so much; I get that Americans don’t want to copy best practices from the rest of the world, but can’t you even read your own studies? They will wonder why the one, single, solitary station in Ballard is nowhere near the heart of the community. They will take a trip from Ballard (one of the more densely populated communities a significant distance from downtown) to the UW (the second biggest destination in the state) on a bus that averages less than 10 miles an hour at noon — and you think they will be impressed?

        The only thing that is impressive about ST3 is the cost. When it is all done, the transit problems that hurt riders the most will still exist, although some other agency (SDOT) may alleviate them . This is a massive investment, with very little return.

        We fall short in this regard, even just compared to our nearest neighbors. When Vancouver builds the Broadway subway, there will be plenty of people saying “Finally!”. You’ve finished one of the key rail corridors, providing an excellent combination of rail and bus service throughout the entire city. When Portland finally builds a tunnel downtown, people could exclaim “Finally!” as well. Sure, its not the fastest system in the world, but it works fairly well, didn’t cost that much, and a tunnel downtown is definitely a major improvement. But when ST3 is built out, most people around the world will simply exclaim “What?!”.

  5. WS is actually more complicated as shallow Delridge station only works if you tunnel. That means you can’t do WSJ-2 and DEL-2a together. If you go elevated, you’re stuck with a 2+ mile guideway between 80-160 ft high across the Duwamish and up Genesee, one station design calls for 3 flights of escalators up to a station extending 110 ft into the air! Never seen an elevated light rail line like that around the world!

    1. A note about the station height data: The data appears to be listed for the entire station and not the platform — apparently for analyzing visual impacts. This would mean that the boarding platform would be about 80 feet off the ground. It’s still high but not as high as suggested.

      In contrast, the deep station platform depths appear to be fairly representative of the depths listed.

      1. yes, platform is at about 80 ft, but the top of the station is at 110 ft. which means it might be the tallest building in West Seattle.

  6. International District: Somebody in an earlier post suggested to reconsider where transfers happen and suggested a new station South of ID, in fact you could just move the Stadion station slightly North. I think that’s an alternative worth considering! Both existing line already intersect there, there is plenty of space between the tracks to add elevator and escalator and a mid-platform on the elevated Eastlink portion, you may have to move the tracks below a bit but those are at groundlevel. The new line could either be below or next to it. This part of town has been slated for redevelopment, you could turn it into a combined transit mall, business center, and highrise residential tower.

  7. “You might ask if it makes sense that adding several minutes to platform access at Chinatown would have no impact on ridership at all. Sound Transit could not promptly answer my question as to whether station ingress/egress time was part of the ridership model. But, officially, there’s no difference. Perhaps this is something that could be considered in the final EIS?”

    Transit Ridership Modeling is something I’ve been trying to wrap my head around, and since I’m not employed in that field, I haven’t thought to delve into the actual algorithm (procedural or actual software).

    However, during my time on the I-405 Citizens Committee I noticed that transit modeling, regardless of mode for a given segment, gave the same basic ridership numbers.

    This has led me to the conclusion that planners basically stuff various numbers in one end of their modeling software, and the process gives the ridership numbers on the other end.

    I found this paragraph in an article discussing how to improve transit ridership modeling accuracy:
    “2. Macro-Level transit ridership model

    Transit ridership models are mainly applied at meso- or micro-level. The ridership of a transit line is usually described as a function of variables which characterize the features surrounding transit station areas or along transit routes.
    These variables usually include land use variable (population, population density, employment, employment density, etc.), social economic variables (family income, car ownership, etc.), environmental variables (parking facilities, walking corridor density, etc.) and transit level of service (frequency, ticket price, etc.). “

    The basic approach appears to be “how many SOV’ers can we get to choose transit over their car?”

    From the basic description above, it appears the Transit User’s experience is not factored in. Transit Users aren’t considered important STAKEHOLDERS.

    Sound Transit might not have a real answer because 1) It’s not part of the modelling input, and 2) it’s a nuanced decision.

    For instance, I described a choice that a friend made to have her husband drive her to the airport during a snowstorm, rather than deal with one simple transfer (incorrectly assuming her not being a soft and squishy Mercer Island type) that would have safely gotten her there, and reduced her husband’s exposure to a possible accident.
    It points to the fact that the harder you make navigation within the system, the easily disoriented get lost…. and if you’re lost without the protection of your 4k lb. shell, you will feel vulnerable around the unwashed masses.

    When I was employed at Nordstrom, it was during their initial foray to the east coast. Hearing the Mr. First Names describe their store siting process, I was struck that they also did a flyover to evaluate the potential neighborhoods that shoppers would be coming from. Maybe they were looking for the Karen Hairdos?

    When I take flights out of SeaTac from the north end (Lynnwood area), I always use some combination of Link, Sounder, or bus (I prefer the train) .
    I am as much a stakeholder as the NIMBY’s along the route.

    I’m not a regular user of Link (given the nature and location of my job), but I do consider it a preferred alternative for various trips.

    I will be sure to comment on any obnoxious transfer situations downtown, that’s for certain.

    1. An analogy: Think of ridership forecast models like estimating water flows in pipes. The overall water demand is constant (based on assumed land use) no matter how the pipes are laid. It’s just how much water going into the several parallel pipes available where you see differences. If the pipes are laid out in paths close to each other, the amount of water won’t significantly vary.

      They also assume the same road system. It’s rare to see a transit study that gets to include reducing street or highway capacity or speed. That also means that there isn’t much impetus to shift from driving to rail.

      I’ll add that FTA won’t let an agency assume different land uses by alternative.

    2. I looked at transit modeling a little bit and it is easy to find flaws. My (very limited) research was geared around the streetcar. The estimates are based on the STOPS model ( What is clear when reading the documentation is that there is tremendous flexibility when making the estimates. It is obviously complicated, and so far as I know, there is no professional regulation regarding its use. If a lawyer engages in illegal or even extremely sloppy behavior, they can be disbarred. Actuaries that make risky bets with pension funds are fired and lose their fellowship. I don’t think there is any harm though, in making a sloppy ridership estimate.

      Worse yet, it is common for agencies to ignore previous failures. The ridership forecasts for the existing streetcars were way off. They expected a lot more riders. Yet the new forecast largely ignored that. To be clear, they recognized existing ridership (as well they should) but there was no attempt to prevent making the exact same mistake. The lack of a proper feedback mechanism suggests a lack of professionalism within the field.

      My biggest complaint about Sound Transit modeling is that so little is known about it. With STOPS, the model is known, and it is simply a matter of putting in the data. The results vary considerably based on how you define the input, but that too should be public information. In the case of Sound Transit, I don’t know what model they used, nor do I know how they defined various aspects of the system.

      Like the streetcar, previous estimates for ridership have been way off. Yet there is little being done about this. I understand why most people have no interest in getting to know the nitty-gritty of the modeling, but I would have more confidence in their numbers if they addressed previous failures (either by hiring better people, or explaining why the numbers are better now). There is plenty of evidence that distance to a stop matters. If so, then this should be reflected in the models. If a station at 14th (significantly farther for most potential riders) has ridership numbers that are the same as 15th, we deserve an explanation. We aren’t getting any, and numbers like this are just an excuse for the board (willfully or ignorantly) to make the wrong decision.

  8. As I wrote up above, this is a mess, but there are several ways it can be improved. Here are a few, ranked from quite unlikely to quite possible:

    1) Bus Tunnel. I only add this to be complete — obviously this won’t happen. But it would have added the most value (i. e. saved the most time for the most riders) for considerably less money then what they are planning on building. By the time this is fully built, we’ll probably have self-driving buses, so this would have been about the same cost from an operational standpoint.

    2) Smaller train sets. This has been discussed here for quite some time, but has made it to The Urbanist. The article focuses too much on automation itself, but the gist of it is this: smaller trains running more often. This leads to smaller stations, which can save a huge amount of money. It can also lead to better alignment, like in West Seattle (the train could run east-west, like originally planned, as a station could fit between the streets).

    Eventually all of Link will be automated*, but a West Seattle to Ballard line would be designed for it. Much smaller train sets could carry a sufficient number of riders if it ran more often. As the Urbanist article pointed out, this would require a change in the plans, which means fewer trains going to Everett. Given the expected ridership — and the long distance — this is fine (this might happen anyway, to cut costs). The Urbanist article suggested a turnaround at Northgate. I would go farther, to Lynnwood. So this means three lines:

    * Ballard to West Seattle
    * Everett to Redmond
    * Lynnwood to Tacoma

    Of course, if they invest in automating all of the lines, then they could pair them for longer trips (although I think this is adequate).

    3) Share the tunnel. With a little bit of work, the tunnel can handle trains running every two minutes (the signalling can actually handle trains every 90 seconds). At worst this could cause minor delays. But avoiding those delays is simply not worth the money, especially since it comes with deeper stations. Riders who use those stations would encounter a delay every single time, while delays caused by train bunching would be rare.

    As with the previous proposal, this would involve different pairing. I would pair Ballard with Redmond, given the geography. Bellevue or Redmond to the UW is faster on a bus, as is Bellevue to Lynnwood. In contrast, Link will be the fastest way to get from the East Side to anywhere on Ballard Link (South Lake Union, Uptown, etc.). During peak, trains would run through downtown every 2 minutes. Going north, there would be a 2, 2, 4 pattern. So basically, the northbound schedule for a downtown stop (e. g. Pioneer Square) would look like this:

    7:00 AM — Redmond to Ballard
    7:02 AM — Tacoma to Lynnwood
    7:04 AM — West Seattle to Everett
    7:06 AM — Redmond to Ballard

    Overall throughput would be the same everywhere. If there is crowding, it would occur after the 4 minute wait (in my example, the 7:02 train). Riders would simply wait a couple minutes for the next train.

    Outside of peak, the trains could run every 7.5 minutes, for 2.5 minutes downtown, and a 2.5, 2.5, 5 minute pattern to the UW. That isn’t as good as running every 4 minutes, but it still very good (much better than now). This type of schedule is easy to understand for every station.

    The last two options should be studied, as both are possible from a political and engineering standpoint. Both would be significantly cheaper than what is being planned, but how much cheaper would require some study.

    * If you can have self-driving cars, you can certainly have self-driving trains. At the very least, you could have semi-automated trains. The trains runs automatically most of the way, while a remote operator manages the relatively small sections that could encounter conflicts.

    1. In a rush to choose and then prepare the EIS, ST shut the door on studying these changes. The last four years have been wasted as the systemic problems of the overall scheme (deep stations with higher unplanned costs) were not revealed and so Ross’ alternatives were not assessed.

      I had long mentioned that the Downtown station details were not disclosed until now. They should have been revealed in 2018. Instead we got early diagrams for West Seattle and Ballard stations — and not the subway stations in the middle.

      Even the last year+ of “realignment” did not present and discuss the practicality of deep Downtown and SLU stations. It should have. It suggests to me some sort of “cover up”. If I was a Board member, I would be furious about this!

      1. “I had long mentioned that the Downtown station details were not disclosed until now. They should have been revealed in 2018.”

        It probably wasn’t known until now. ST has a habit of designing stations and transfers late in the process. We still don’t know how transfers would work from U-District station to a 45th line even though the latter was studied around 2015. You’d think they’d put a transfer stub into U-District station while they were designing it, but no.

    2. I think #2 holds great promise and could still fit within the ST3 framework (i.e. don’t need a fresh EIS nor need to go back to the voters)

      The Urbanist article made a key mistake by suggesting the 2-car operation would require a different mode and therefore an independent OMF within Seattle. The best outcome would be to simply use the same LRV technology as the rest of Link (automated by the time WBLE opens, as you note), but run 2-car trainset to unlock immense capital savings. That would allow for the 4th OMF to be built in Snohomish as planned.

      Another benefit of shifting to 2-car operations is this would be a better outcome for Ballard-UW as well, as that short urban subway would also benefit from small stations and higher frequency, and higher frequency would make branching more palatable (e.g. send 2-cars both north along 15th and east/west along the Ballard-UW corridor)

      1. Yes, I agree, the article in The Urbanist suggested that automated trains are completely different, which is silly. We aren’t talking a monorail here.

        Technically, I think any of the three could be done without going back to voters, by simply delaying the other parts. For example, build the tunnel first, but allow buses in it. Designing it for buses initially (with eventual conversion to trains) shouldn’t add to the costs that much. Spending money on a fleet of buses would probably be breaking the rules, so there would be some limitations. Anyway, that isn’t going to happen.

        On the other hand, simply building the tunnel last seems quite reasonable. The order they have now seems rather arbitrary. Just connect up the two pieces and then build the second downtown tunnel last. That could be deferred to 2060, for all I care. This would be a reasonable approach, given the value added. A train to Ballard adds value now — a second tunnel downtown only adds value if you want to run those trains a lot. I doubt that will ever be needed (and I’m not alone).

      2. I agree that the short automated train option is not a significantly different mode. The tracks are the same. The power source is generally the same. The Central OMF can be modestly redesigned. The communications system has to be changed regardless of mode when a line gets added. The big difference is that the platforms can be shorter and there may be some easier elevation changes.

    3. Driverless buses: How do you deal with ADA issues (i.e. people that need help boarding and/or securing a wheelchair)? With nobody to enforce rules what’s that going to do to the rider “experience”? And that includes people who ride infrequently or are on a route they don’t know. Drivers have more of a people oriented job than a delivery route.

      How does Vancouver deal with these issues on their light rail? I suppose with rail it’s easy to have real people at every station instead of one person for every vehicle.

      Finally, if driverless buses/taxis become a reality I see micro transit or a shared UberLyft model taking a significant portion of the ridership outside of peak commute. At that point it might be better to let private providers fight it out and just provide reduced fare benefits based on need.

      1. Driverless buses are still far off, I just can’t imagine how they can deal with complex traffic situations, construction, or accidents. It might work on fixed routes in less busy roads in the suburbs, but certainly not in urban situations.
        Trains are much easier as they have separate ROW. Platform screen doors help a lot and may be the best way to deal with them, but the fact that its all on level ground helps tremendously, with a bus you always have to lower the entrance to the ground somehow.

      2. With nobody to enforce rules what’s that going to do to the rider “experience”?

        This happens on subways all the time. The driver is simply responsible for driving, and has no interaction with the riders. There is no fundamental reason why buses can’t operate the same way.

        Finally, if driverless buses/taxis become a reality I see micro transit or a shared UberLyft model taking a significant portion of the ridership outside of peak commute.

        You still have the cost of vehicles, as well as maintenance. Sharing the ride is fundamentally much cheaper, and can provide better service. People think microtransit is great until they have to wait a half hour for a car, or the car is stuck in traffic. Transit scales — taxicabs do not. As Jarrett Walker likes to say, technology doesn’t change geometry.

      3. The SeaTac airport train between the terminals is already automated, and I don’t recall ADA being a particular problem there.

      4. with a bus you always have to lower the entrance to the ground somehow.

        Or you raise the height of the bus stop and have level boarding. The buses use what is called precision docking, and there are two types, vehicle-based and driver-based. The technology has been around a while (unlike self driving cars which still isn’t there yet). Note the date on this report:

        You are right though, it may be decades before vehicles can handle the complexities of a street without human intervention. A few years ago there was great optimism — now the pendulum has swung, and lots of people are pessimistic. At the same time, transit is well suited for the technological advancement that has happened. Self driving vehicles do better on the same stretch of road. It make more sense to manage a fleet of vehicles, rather than just one. Managing a fleet is similar to managing the self checkout lines at the grocery stores. There are still workers, just a lot less of them. They monitor things remotely, if alarms are set off (i. e. the robot doesn’t recognize something). This would mean a lot of false alarms, but it would be geared towards safety. So, for example, a bus stops because of a plastic bag on the road, and then a driver takes over remotely, drives over the bag, and the bus goes on the way. This, as opposed to the car doing the opposite, and driving into someone thinking they are a bag.

        A semi-automated approach would also make sense on long stretches of restricted road. For example, the 405 BRT route. It is only the tiny section in Renton and Bellevue where a driver would be needed. In this case, a driver could hop on, and make sure the robot is doing the right thing for this section, and then ride back the other way. With a bus tunnel it would be the same situation. A driver would drive (or at least personally manage the driving) of the 120 from Burien to the edge of the freeway. Then they would get off, and let the robot handle it until Ballard (or the very least, Elliot).

        But that could be a long way into the future. In contrast, having a robot (or remote users) handle Rainier Valley and that small section of Bellevue seems rather easy. This is something that Sound Transit should explore, given the enormous distance of our trains. The fact that pairing is influenced by the distance (and not rider experisence) is a very bad thing, and one more reason why the problem should be addressed.

    4. 1) Bus Tunnel.
      Never say never. A bus tunnel might bring in a wider base of support which means KC Metro (or KC taxes) might foot a greater portion of the cost. I know you’ve pointed out that we move a pathetic number of people currently through DT on the surface streets. Apologies if this has already been addressed but might it be one of the reasons for this is too many buses; in effect congestion limited? Seattle is wealthy and getting richer all the time. A new bus tunnel would improve the rider experience and lure more choice riders back to transit. And if congestion is reduced Metro saves money in operating costs.

      1. I wouldn’t say that the number of people moving through downtown is pathetic. I was referring to this report: (specifically page 7). We have a huge number of buses — way more than everyone else on that chart — but we are roughly in the middle in terms of moving people. I’m not sure why that is, or if anyone has tried to figure out why. I think the big issue is that the buses aren’t very full. If the problem was congestion, then there wouldn’t be so many buses per hour. It does seem odd, but maybe a lot of buses drop off a lot of their riders before they get to Third, thus dropping the numbers.

        I really doubt there is the political will to build another bus tunnel in this city. The politicians — and many of the transit advocates — are fixated on rail, even though the vast majority of transit trips take place by bus (and always will). It really makes the most sense for West Seattle, and I think it is too hard to steer them towards a bus tunnel, especially at this point. Once the rail work starts, it will be too late. The same is true of Ballard. Without West Seattle and Ballard the cost/benefit of a new bus tunnel isn’t great.

        That leaves pretty much just the Aurora corridor in the north, and the I-5 corridor in the south. You could maybe find a way to get a few more buses in there (e. g. the 131/132) but it doesn’t sound like that many. I suppose you could add entrances, like at Jackson (to get buses like the 7) or Mercer (to get the Queen Anne buses) but that gets really expensive. Then there is the 3/4, which intersects 3rd at James, or the 70, which comes in at around Olive. Third handles a lot of buses, and if you add entrances all over, it would really add up. If you don’t — if you focus on the longer distance buses, like the old tunnel — you wouldn’t get that many buses. A bus tunnel can handle a lot of buses (the old tunnel was nowhere near capacity). But a new bus tunnel would also be really expensive. It is hard to justify the investment without a lot of buses going through it. The combination of Ballard/Aurora and West Seattle was definitely worth it — take them out of the equation, and I doubt it is.

        And again, you have the politics. A bus tunnel for Aurora (as expensive as it would be) would still be a better value (and better for riders) than rail for Aurora. And yet there are more people pushing for rail for Aurora.

        I think the most cost effective solution is to improve Third, as that report suggested. Then address bigger problems, like congestion on James with the 3/4. Buses on Third move reasonably well, especially with the more recent improvements (that occurred after that report). Basically you just keep chipping away at the various problems all over the city. More off board payment, and in the case of Third, extend the bus lanes up to Denny. It is easy to assume that we aren’t making much progress in that regard, but we are, it is just that progress is inconsistent. The plans for the 40 are outstanding, and if implemented as suggested, will make a huge difference (for many buses, not just the 40). Meanwhile, buses like the 44 seem to have tiny improvements every so often. By the time this project is completely though, I would expect most buses to run fairly well, even the 3/4.

    5. I think automated trains are the way to go and they are becoming mainstream already and not much different from current LRV technology (see Hawaii).
      Now that we have large stations in the DSTT, could we even allow two short trains to board at the same time for even greater use of the existing tunnel?!? Then we may be able to get away without 2nd tunnel.

      1. If I understand you correctly, I don’t know if anyone does that, anywhere. Basically the trains would meet up, at say, SoDo, then then connect to each other (physically or electronically). Then the train goes through the tunnel together, and then splits.

        Even if this is possible, I don’t think it really gets you anything. The trains are limited to how often they can run down Rainier Valley and on the surface streets of Bellevue. Inside the tunnel, there are concerns that the limitations of the tunnel (90 second headways) would cause trains to delay. A train arrives, and another train is running late, so instead of being out of the train 30 second ago (as planned) it is still in there. Thus the train has to wait before entering. If the trains are supposed to meet outside the tunnel and go in together, then the delay would be similar, but worse. A train arriving early would have to wait — the effective cushion would be zero (not thirty seconds).

      2. The DSTT used to have different loading positions for buses. It worked because riders could form lines. That’s not needed with Link trains. Plus, the platforms may be too short for two two-car trains if a safety gap is factored in (even two-car trains are much longer than articulated buses)

        Park Street Station in Boston loads different westbound lines at different places inside the station. They all merge into a single track. I believe that riders can still walk across some of the tracks. I doubt that the DSTT station vaults have enough room to reconfigure the platforms and tracks to do something like Park Street. Still, it’s worth noting that such an idea operates somewhere — but with more tracks available in each direction.

      3. I’m trying to increase tunnel capacity as trains may arrive from Bellevue, RV, and WS in high-frequency intervall. It’s similar to how the bus tunnel used to work with platform A/B, except that you don’t allow for passing. The first train to arrive would stop at A while the 2nd would stop at B – it wouldn’t have to wait for 90sec. If the 2nd train was running a bit late, a 3rd train may have to wait until the whole platform clears, but that’s no different from the proposed timeslot operation.
        Anyways, this is an optimization to be considered if a single tunnel would create bottlenecks in the future…

      4. I’m trying to increase tunnel capacity as trains may arrive from Bellevue, RV, and WS in high-frequency intervall. It’s similar to how the bus tunnel used to work with platform A/B, except that you don’t allow for passing.

        Yeah, but unlike the buses, that isn’t the problem. Buses crowd up at the stations, but have excellent throughput on the open road. Trains are the opposite. A train can creep up to the station just fine. For a northbound train, it could be only a few seconds behind another train at Westlake, simply because it is going so slow. But once the train gets past Westlake, it can be at most 90 seconds behind the other train. Otherwise, if the first train has to stop in the middle, the second train would smash into it.

    6. Ross, you can’t have “2.5 minute headways”. You can have 2/3/2/3… Transit schedules and displays in North America don’t display seconds. They may in Germany….

      1. You can have 2.5 minute headways, regardless of how they are displayed on a schedule. My guess is they would be displayed as 2/3, just like 7.5 minute headways are often displayed as 7/8. But when they split (as these would) it becomes obvious that the train is actually running consistently (or at least trying to).

      2. Of course, it’s “trying”. But it can’t, because train timetables do not have decimal minute fragments or seconds. Transit trains are just not that reliable.

      3. Published timetables for the public are not published to that level of accuracy, but behind the scenes they can be.

        Once every 10 minutes is a base frequency. Split that in half and you get every 5 minutes. Split that in half again and you get every 2.5 minutes.

        The extra 0.5 minute can make a difference, because that’s about how long Link spends slowing down, stopped, opening and closing doors, and restarting at the downtown stations.

  9. The Doug Trumm (in the Urbanist) station details are an excellent source of the DEIS information.

    The three West Seattle stations have these non-bus-transfer (or walk and bike and auto?) boardings that I calculated using the data from the diagrams provided in the Urbanist:

    West Seattle = 3100
    Avalon = 800
    Delridge = 700

    The overall boardings are listed at:

    West Seattle = 6400
    Avalon = 1200
    Delridge = 5800

    In other words, there are only forecasted 13,400 boardings in all of West Seattle of which only 4,600 are NOT bus transfers and 8,600 ARE bus transfers.

    It seems to me to say that the most important consideration in the West Seattle is designing good bus transfers and the feeder bus system as the primary objective ! It’s also saying to me that West Seattle doesn’t need three stations! Finally it’s saying to me that the sites for the three stations can’t muster walksheds worthy of a subway and the removal of hundreds of homes.

    1. Finally it’s saying to me that the sites for the three stations can’t muster walksheds worthy of a subway and the removal of hundreds of homes.

      Agreed, not at these costs. If this was surface rail, then it would be one thing, but its not. West Seattle Link is expensive, which should be no surprise. Even before you get into the stations (which aren’t cheap) you have the huge bridge to get to the peninsula. Its really not an area appropriate for rail (not with that ratio of bus-transfer versus walk-up) let alone really expensive rail.

    2. I was surprised about the ridership as originally ST had published 25-27,000 boardings. By 2032 WSLink would stop at SODO and force a transfer. I wonder whether they reduced the ridership as some buses will continue to go downtown directly until trains go downtown and then expected ridership would go up to 25-27,000… Anyways, it’s still very low for a light rail connection and at $3.2 billion very expensive per rider.
      Yes, most of the ridership would come from bus transfers as plenty people live in other urban centers such as White Center, Westwood, and High Point. The main reason for an Avalon station is to serve the buses running on the 35th Ave, like from High Point. If the Fauntleroy station would be built, I think those buses could take a little detour to visit that station. But a Fauntleroy station would be farthest from the Junction!
      The WestSeattleSkyLink folks have suggested redirecting the WSLink funds to the Duwamish bypass and using gondola technology to serve the 3 committed (and possibly more) stations via gondola and also serve White Center / Westwood, may be even the Fauntleroy ferry terminal, with another gondola line:
      That way residents of those urban centers wouldn’t have to wait for a bus and then wait for a train at Delridge or Junction but take a gondola straight to Link.

      1. If you sim the three station boardings and assume a similar number getting off, you get 26,800 total average weekday ridership.

        It’s the expectation that about 2/3 of the users are bus transfers that is the more interesting detail that has now been revealed here.

      2. ah, good point about boardings vs ridership, that makes sense.
        Yes, 2/3 via bus is surprising. I wonder how many riders start in White Center, Westwood, or High Point… I suspect that’s where most riders come from but it would be good to verify that.

    3. The assumption of 5100 bus transfers at Delridge seems high to me, especially when the preferred stations are a block off Delridge and necessitate either crossing the street and walking a block (like at Mt Baker) or rerouting the 120 for a detour through narrow side streets. You probably could only get those numbers with full truncation of the 120. It really only would make sense to transfer to light rail if your destination is beyond downtown, as the 120 is quite fast for accessing downtown once it’s on the bridge and 99.

      On the other hand, the 400 Avalon station bus boardings seems absurdly low, since that’s where the 21 from High Point most conveniently intersects light rail. It makes me wonder if the transit modelers are making some weird assumption that Delridge becomes the transfer hub for the 21 as well.

      1. This just shows how difficult it is to model train ridership for stations that are highly dependent on buses. If Metro continues to run the 120, then very few people will transfer there.

        Before Covid, the 21 got 4,600 riders. Some of those likely rode within West Seattle or downtown. I would estimate that at around 600, which means that an Avalon station would likely pick up around 2,000, assuming all the riders switched. Maybe they assumed that the 21 would head west on Oregon, and serve the station at the Junction. I could see that (it could continue on California), but then the numbers for the Junction are too low. In any event, I could see the 21 going on Avalon, and continuing around to Alki (either by following the coast, or going up and over Admiral Way).

        It gets tricky estimating the numbers by station, but should be fairly easy overall, as long as you assume truncations. Then the only big question is the frequency of the buses (a big unknown) and the frequency of the train (which should be known). Given ten minute frequency of the trains, the transfer penalty is pretty high, so you will likely lose some ridership during the day. You will gain back some riders who walk to the Junction Station. Avalon walk-up ridership looks like a wash (the train is faster and continues to the north, but the combination of 21 and C provides better frequency) while the Delridge station likely won’t have many walk-up riders. I also don’t see much ridership between stations on the way to downtown. Overall, assuming all the buses get truncated, my guess is ridership for West Seattle will be very similar to what ridership is between West Seattle and downtown on the buses.

        I have no idea what that is, but that would probably be as good an overall estimate as what they came up with.

      2. The 120 is being upgraded to RapidRide H. It will go downtown. I assume riders going downtown will think twice about waiting ten minutes to transfer to Link. The longer the train segment and the greater the car congestion, the more worthwhile or justified a transfer is. Delridge-Westlake is in that short distance valley with highway speeds where transferring to Link is not as easy to justify as transferring from long-distance Northgate or congestion-prone U-District. And if you’re not going downtown but to Capitol Hill, U-District, North Seattle, Bellevue, or southeast Seattle or beyond, then transferring to Link looks even more desirable.

        I’m concerned about the 21 on 35th. Avalon Station is for it, and it will be truncated according to Metro’s 2016-2020 LRP. Delridge will have the H, the Junction may have the Fauntleroy-WSJ-SLU express Metro was considering, but 35th may have only a truncation at Avalon. That may seem not very good to them and drive them to their cars.

        Before Sam jumps up and down and says, “Why is a transfer at Avalon bad but a transfer at South Bellevue is good?”, it all depends on the distance of the train and bus segments, and the density and walkability and variety of businesses at the tail. Transferring at U-District is reasonable. If you’re going from Beacon Hill to 10th Ave E, it does not make sense to force people to transfer from the 36 to the streetcar to the 49 just to avoid overlap, because the overlap is only a mile or two. West Seattle is kind of near that boundary, and the short distance of a 21 trip makes the transfer overhead a significant part of the total trip. So I’m not sure whether we’re being fair to 35th, which has a lot of lower-income residents and high ridership. This is where multi-line BRT from downtown to 16th, Delridge, 35th, California, Fauntleroy, and Admiral would have been better.

      3. The Avalon station estimate is crazy low — pre-pandemic there were more than 1,200 daily bus boardings in the Avalon station walkshed. Then you include all of the additional housing units that are planned for that area — 500 at Alki Lumber plus several hundred additional along Avalon Way — plus numerous other parcels available for TOD. I wouldn’t be surprised if Avalon matches or exceeds the Delridge Station ridership in the long run. They are basically building a copy of the Mt. Baker station in Delridge and the bus transfer experience will be terrible.

        The Junction station is not going to be on Fauntleroy, they are almost certainly going to go with that “medium tunnel” option that gets the Junction station closer to the main business district, plus the retained cut Avalon Station that is visually more appealing and reduces property acquisitions.

      4. Joe, I think the experience in RV has been bad enough (fatal accidents) that Sound Transit is avoiding grade level lines though they have their advantages.
        I understand the medium tunnel is one of the least disruptive options (though a bad Delridge station), but for a WS line to have a chance of long term viability, you will need to extend it further south later. Once you get into a tunnel, where would you get out of it and at what cost?
        Instead of building two stations along Fauntleroy Way, may be build one in the middle?!? That would avoid acquiring / tearing down the large apartment building at the proposed Fauntleroy station, but how would you continue that line further south later? There isn’t an obvious way to turn east…
        Another possibility would be to go elevated from the Avalon station along 35th Ave to Barton to White Center. You could still serve the Junction from a station along 35th via some automated cable system (APM, funicular, gondola).

  10. Another idea that should be pursued is moving the Ballard Bridge to 14th, and the light rail to 15th. Again, this is an idea that has been discussed here (and may have started here) and has made its way to The Urbanist. Again, the Urbanist article spends too much time on side issues. It fails to emphasize the key benefits, such as:

    * Much easier implementation. It is difficult and expensive to put in a train next to a busy thoroughfare. Similarly, it is extremely difficult to rebuild a bridge without shutting down the old one. This solves the problem rather elegantly, in three easy steps:

    1) Build a new automobile bridge on 14th.
    2) Close the bridge on 15th.
    3) Build a new light rail bridge on 15th.

    * Cheaper alignment for 15th. With the bridge moving to 14th, the street could be one lane each direction. This allows a lot more room for the pylons.

    * Better station. Not only would this open up more flexibility with the station, but it would make it much easier to cross the street. If the station had entrances to the south of Market, crossing there would be trivial, unlike today.

    1. Everything RossB writes here and above is gold. Why are we not having a very serious and urgent regional conversation about reusing the DSTT, the tunnel we already have, for WSBLE? Any variation of the second downtown transit tunnel will be hugely expensive, disruptive to construct, and require lengthy, time-consuming, ridership-killing, accessibility-reducing deep station transfers. The rest is details, mostly unaffordable.

      Instead, this:

      New Ballard Bridge on 14th
      New light rail bridge on 15th
      Share the tunnel downtown
      Ballard to Redmond
      Lynnwood to Tacoma
      Everett to West Seattle

      For reasons of basic geography, UW to Redmond is fastest via SR 520, by far, with or without East Link, with or without any of the SR 520 expansions.

      For all the reasons RossB explains, this comprises the basis of the best we might hope to achieve for the long term public good in this region within the confines of the ST3 planning and budgeting framework. The Board needs to step up.

      Instead of focusing on how much we can’t afford what we shouldn’t build, we should start by making the most of what we have. And we should maximize planning synergy with everything else we need to build, like the new Ballard Bridge.

      This, btw, is an excellent piece, not to be missed:

      1. Obviously I agree. The one thing I would add is that these ideas are not new, nor are they my own. All of these ideas have been discussed here, in great detail. These aren’t radical proposals (no monorails, or Seattle Subway lines to Woodinville). These are common sense, value-engineered ideas based on consensus. The proponents of these ideas are very pro-transit, and yet this would involve spending *less* money on transit. Because at the end of the day, wasting money on a transit project means you get worse transit. There are much better alternatives that should be studied.

      2. “Why are we not having a very serious and urgent regional conversation about reusing the DSTT, the tunnel we already have, for WSBLE?”

        We’ve been trying to get the region to do that and other things for several years now but it falls on deaf ears. The reason we’re not is it’s not on the politicians’ radar.

      3. Because the technical analysis in 2016 indicated that the existing DSTT couldn’t handle the peak crowds, and because staff thought the resiliency of two tunnels would be valuable.

        Also, building a branch on the north side of the DSTT probably entails closing the lines for months, if not a full year. Given the political hesitancy to close 2 blocks of 5th Ave for a few months, I’m skeptical there is an appetite for that magnitude of disruption to transit operations.

      4. I agree that it would be difficult to tie a new tunnel into the existing without causing disruption and politically it would be tricky to eliminate a 2nd tunnel from ST3. TT or Mike had suggested at some point to put the new tunnel right under the existing one along 3rd Ave, does anybody have any idea what type of footings were built to support the tunnel? That should make transfers much easier. I like the suggestion to build a Jefferson funicular (or even a gondola to Broadway) instead of a Midtown station.
        Looking at the ID drawings, I see extensive footings under the existing ID cut to hold up the walls. That would not allow any tunnel underneath, but may be the stacked configuration along 5th Ave could be used (CID-2a).

      5. Martin, it wasn’t me who suggested a basement tunnel at least not for the full length of downtown. I did note the extreme challenges of doing it.

        Remember, DSTT1 crosses the BNSF tube twice, once under-running it at Washington and once over-passing it about Seneca. There is nowhere near enough clearance above BNSF to put another pair of tubes below DSTT1.

        What I have been suggesting is putting a facing-point turnout at the south end of University / Symphony and punching a hole in the north wall of the station box for a track to connect the two tunnels for Ballard-bound trains. It would plunge thirty feet as soon as it cleared the new portal and when it had done so move west to lie under the southbound DSTT1 track. It would continue north to Stewart where it would curve into the Stewart right-of-way and continue to Westlake where it would use the northbound SLU-Ballard plans. The southbound connection would be constructed one level above it and pass through the north wall of the big vault which hosts the DSTT1 curve between Pine and Third Avenue, connecting to the south bound track.

        A “basement” platform could be provided next to Macy’s for Ballard-bound passengers from the north end of the traditional CBD with underground connection to the southbound platform at Westlake.

        So far as those deep footings at IDS, they’re necessitated by the fill soil on which the entirety of the SoDo Industrial District is built. It may be a problem when TBM’s set out to chew through it. The good news is that it’s mostly Denny Hill and therefore not full of scrap steel like the waterfront. There will be rocks, but TBM’s are engineered to handle them.

    2. One addition I’ll make to this great set of suggestions is a cheap and easy way to effectively add passenger capacity to the existing light rail tunnel downtown, the Spanish solution:

      Basically, just add central platforms to the existing DSTT stations for people to exit left from the trains. Add up escalators, an elevator, and stairs. Keep the existing side platforms as the main waiting area served by down escalators. When trains arrive, open the exiting left doors 1-2 seconds before the entering right doors open, to create a pulse of people exiting the main desired direction.

      Given how the DSTT stations are already constructed, this would be pretty cheap. For the investment of concrete medians, escalators, stairs, and some extra wayfinding signs, you can significantly reduce dwell times and boost capacity.

      It also has the benefit of creating level transfers for reverse direction trains at ID and Westlake. Someone coming from the north on the Lynnwood/Everett line could simply exit at Westlake, continue to stand on the center platform, and wait for the train to Ballard. Ditto from West Seattle to the Eastside at the ID. Both transfer experiences are vastly more convenient than the currently planned 4-6 minutes of wandering through mezzanines and escalators.

      1. Sorry, but using the Spanish solution downtown would likely decrease capacity, at least in the context that matters. In this case, capacity is all about how many trains you can run in the tunnel at once. Maximum headways are an issue, as are dwell times. For stations like these, the Spanish Solution actually increases dwell times (the more doors are opened, the more likely to have a delay). It only makes sense when crowding at the station is the problem. It is especially appropriate if just about everyone wants to get off, while a lot of people want to get on. This happens at the Waterfront Station with SkyTrain. It is the last station downtown, and it connects with the ferry. Thus everyone gets off, while a big crowd gets on.

        That isn’t the problem here. The problem here is there are only so many trains you can run through the tunnel. ST can dramatically increase that number by investing in capital projects, but decided to build a new tunnel instead. Now it is going to cost a lot more money to build that tunnel, and the stations will be extremely deep, which is why a lot of people are questioning that decision (or at the very least suggesting that sharing the tunnel be studied).

  11. Wouldn’t the best transfer point between the Sea-Tac segment and the UW/Northgate/Everett segment be at SODO? It looks like SODO alternatives 1B and 2 have the 2 lines share a station complex, and preferred alternative 1A has them a block apart. Am I reading the maps correctly? Seems like 1B/2 would be the best transfers for everyone, right?

    1. You are right in that SODO would be the place where the least amount of vertical change would be needed.

      The problem with the current alternatives is that there isn’t a same-direction cross-platform track configuration at SODO. It’s all about how the tracks are configured. To make a northbound transfer, you would have to both go up and down to transfer in these alternatives.

      If the tracks were reconfigured to offer an all northbound platform between the two northbound tracks (and the opposite twin heading southbound), the the 16 doors of a northbound UW/ Northgate/ Everett train would be just 30 feet away from matching 16 doors of a SLU/ Ballard train. It would be an easy walk from one to the other. It wouldn’t take five or more minutes on one or two escalators, stairs or elevators — and instead would take about 10-20 seconds.

      Being out of a subway makes it much cheaper to build. It mainly would require changing the climbing curve just south of SODO. There is already a switch at the top of that curve so it may be possible to tie in a new structure to that.

      One other change would be to offer a pedestrian crossing under all of the tracks rather than force everyone to go above the catenary wires. The Lander auto crossing could remain a bridge — but the pedestrians wouldn’t need to climb up the viaduct to get to the station. Since the clearance need would drop from 50 to 20 steps, it’s much easier to get between t boarding locations.

      Finally, it would make it much easier to have crossover tracks between the two lines. That could be critical when service disruptions occur and a line has to be taken out of service.

    2. Larry, in 1A the two stations would be a bit staggered but still overlapping to avoid impact of the USPS parking garage, you could still arrive from WS and catch a train to Seatac fairly easily.

  12. We need to take a good, hard look at what it would take to get trains from Ballard into the existing DSTT. And, after we do, if sharing the downtown tunnel is deemed infeasible, then, really, we should not build Link to West Seattle at all. Link to West Seattle was a marginal project at its original, much lower cost. It doesn’t make sense to construct a second tunnel downtown only to serve West Seattle.

    Unlike Ballard, West Seattle has (or, will soon have again) a fixed span crossing on a limited access roadway with a bus pathway all the way to downtown. What if we invested a lot more money in the West Seattle bus network. How about supplemental direct service to SLU and the north end of downtown through the SR 99 tunnel. That would be faster than Link will ever be for those trips, and it could be done next year. Heck, those same buses could even continue from SLU to the Eastside on the new Mercer HOV connection to SR 520 that they are building now.

    The need for transfers, or worse, double transfers, is vastly reduced by retaining direct bus service to downtown from the major zones of West Seattle.

    And it certainly doesn’t make sense to build Link to West Seattle first, ahead of the more urgent and cost-effective subway to SLU and beyond. That’s what we should be prioritizing, because in 2019 before the pandemic SLU was gridlocked and since then, it’s remained a forest of cranes. 2037 is a long wait for that district. Our priorities are backwards.

    Link is really not the best mode for West Seattle. I admit to an irrational bias, love, or preference, in favor of trains, but I believe we are far better off giving West Seattle a five star bus system than the best Link system we’d ever build there. The primary Link transfer could happen downtown, maybe Pioneer Square?

    If some Ballard trains turned around at SODO, would that be OK? If sharing the tunnel simply can’t work, what if Ballard trains turned around at or near Westlake, temporarily or permanently? Or – and here I’m dreaming – somehow looped back through Belltown or First Hill? A lot of this is less ambitious than a tunnel to West Seattle or Ballard, and we could afford a lot if we didn’t have to bring Link all the way to West Seattle.

    I don’t know, but we’re short of funding for a mediocre plan, and it’s time to get creative.

    1. If it were me waving a magic wand, I guess I would take a look at using the existing tunnel bell mouth at the convention center from the old tunnel. You’d wind up with one at grade track crossing, but there are really busy subway lines out there that have those. They can be made to work.

      Angle Lake to SoDo maximum is what? Every 6 minutes? Then Bellevue comes in at a train every what? 6 minutes?

      This is 20 trains per hour per direction. There are systems out there running 32 trains per hour per direction. West Seattle -Everett needs what? 4 trains per hour, at those ridership numbers?

      Hell, pre-pandemic TriMet was doing some 28 trains per hour per direction on the Banfield line (saving space for 4 trains per hour for the yellow line to join it at the Steel Bridge), and TriMet has no flying junctions at all, unlike Link.

      One thing TriMet does with its lines is carefully manage headways to match demand. Eg, the orange line is advertised as every 15 minutes. However, pre-pandemic an OHSU shift change would crowd one of those trips, so once a day it got a two minute headway by extending a green line train onto the orange line. Blue line tended to be every 7 minutes, but a few times in evening rush hour that dropped to 4 minutes.

      I would imagine that similar arrangements could be made with these lines. Run everything as a base frequency of every 8 minutes or whatever and toss a few extra trains here and there to eliminate those times when stuff gets overly crowded. Nobody complains if, every once in a while, a line runs a bit more frequently than advertised.

      1. Glenn, your computations elegantly accommodate the RV and Floating Bridge trains which, yes, are limited at this time by traffic and engineering constraints. But you haven’t said word one about the North End trains. If downtown comes back, there will have to be twenty-five to thirty trains per hour in the peak periods up there. You claim that ST can be convinced to go for the level crossing with that congestion. I doubt it.

        Now maybe it won’t “come back”, and with twenty trains per hour (3/3/4 headways) they might agree.

        Remember that Tri-Met’s two car trains foul a divergence only half as long as Link’s four car trains would. At twenty-five miles per hour — which is about the speed that trains make the curve into the Pine Street tunnel — it takes a Link train thirteen seconds to transit a fixed point, but a MAX train six and a half. But in many cases the second of a pair of trains occupying a junction sequentially will be starting from a dead stop or a crawl and take considerably more than six and a half or thirteen seconds to clear. Finally, the “envelope” around the train adds at a minimum another minute before a train occupies the plant during which neither approach can be used safely because of passenger-safe braking decelerations.

        Even with two-car trains as proposed by you and others doesn’t help a lot, because the majority of trains using the crossing will still be four cars in length.

        So increasing the transit time from six and a half to thirteen seconds may seem trivial, but in fact might be quite disruptive. In all of Tri-Met’s divergences except the closely linked and timed switches at either end of the Steel Bridge, there is a pocket track upstream of the diverging point which allows a follower taking the other route than that of a leader delayed by an opposing move through the divergence to pass the leader. That will emphatically not be the case with a divergence at the curve between the Pine Street tunnel and the Capitol Hill tubes. Any delay to a Ballard-bound train waiting for a southbound train from Capitol Hill to clear the plant will be for the longer elapsed time.

        In the case of the Steel Bridge the Green and Yellow line trains leaving downtown are at the bottom of the totem-pole, often waiting at the Greyhound depot several minutes for their slotmate from the opposite direction to be ready on the east side of the river.

      2. The current DSTT1 maximum is 3-minute headways (20 trains/hour). ST says increasing it to 1.5 minute headways (40 tph) would require capital improvements to avoid unreliability and train bunching. There was an ST3 candidate project to do that, but it was deselected when ST decided to go with DSTT2 instead. ST never said what these capital improvements would be, but they surely cost less than DSTT2.

        This runs into overwhelming political headwinds, however. From ST’s perspective, it was a hard-fought battle to get ST3 approved with the Mariner-Everett extension and Ballard line that were originally expected to be in ST4. It was very glad to get all that authorized so it wouldn’t have to go back to voters again for 25 years (per the original schedule). Replacing DSTT2 with an upgraded DSTT1 might contradict the ballot measure and require another vote to change it, depending on whether it’s considered “in scope” of the representative alignment. It would also presumably require shutting down DSTT1 for months. The largest fraction of the public and voters are against construction disruption — viz the International District controversy — so ST would be loth to shut down DSTT1 again. Especially as downtown is trying to recover from the pandemic and high crime and homelessness and the loss of Macy’s.

        “West Seattle -Everett needs what? 4 trains per hour, at those ridership numbers?”

        6 trains per hour minimum no matter the ridership. 10-minute minimum frequency is what distinguishes a good and effective rapid transit system that people easily turn to, vs a mediocre system that’s not as competitive with driving. In St Petersburg and Moscow, 10 minutes is what the Metro goes down to after 9pm.

      3. Regarding Everett headways, East Link was first going to terminate at Northgate (off-peak), then Lynnwood, and now Mariner. I could see it contracting the other way if ST decides it doesn’t need two lines to Mariner after all. Or it may go to Mariner peak only.

      4. ST’s own ridership numbers show a drastic reduction in ridership north of Lynnwood. Current 512 ridership reflects this as well. North of Lynnwood there are a lot of empty seats.

        Keep in mind, we are talking of light rail trains, that can be up to 4 cars long. I’m guessing due to the substantial ridership south of Lynnwood, they’ll need the 4 car trains.

        So, let’s say a base frequency of every 15 minutes. If necessary, peak period could extend a few East Link trains north of Lynnwood. Or all of them if you really want to.

        Either way, it seems to me there’s enough capacity in the downtown tunnel so that it can be pushed to the end of the line for construction.

        As far as convincing ST to use a track crossing, that’s what the magic wand prologue was about. It’s pretty obvious ST wants to build DSTT2 first.

      5. Actually, 3/3/4 separations is only 18 TPH, which is way too low. Three/three/three IS twenty TPH (exactly), but what a mess in the schedule!

    2. So far as turning trains at SoDo, it will be essential if downtown comes back to anything like pre-Pandemic ridership regardless whether a new tunnel is built or not. The RV and Floating Bridge lines simply cannot deliver enough trains to North Link or accept enough trains from it to handle expected peak loads. Trains will have to be turned at SoDo.

      This means that Lander and Holgate will have to overpass the busway and Lower Royal Brougham closed to through traffic at the trackway. Buses could still turn between LRB west of the busway and the busway, and Greyhound buses could come from Sixth South, but nothing could cross the tracks there.

      The cheap way to make such a turn-back it is to take a turn around the outer Maintenance Facility loop. There is ample place for parking a couple of trains in the southeast quadrant of it and a parallel track just inside it for originating trains to pass stopped ones already in service. Drivers could take a break there.

      No new trackage of any kind would be required for this use.

      1. So far as turning trains at SoDo, it will be essential if downtown comes back to anything like pre-Pandemic ridership regardless whether a new tunnel is built or not. The RV and Floating Bridge lines simply cannot deliver enough trains to North Link or accept enough trains from it to handle expected peak loads.

        So you are saying that running trains every three minutes to Lynnwood (or Everett) is not adequate? I seriously doubt that. Our trains are big, and running every three minutes should be more than enough. Cities with bigger demand handle things with less throughput than that.

        If this was an issue, then yes, we would need to run an extra train to SoDo (or somewhere). We also need to make the necessary improvements within the tunnel to lower the headways to two minutes. It also means that the tunnel would be necessary, as trains to Ballard couldn’t share the tunnel (all trains go to the north end).

        But again, I really don’t see it. If we ever get close to that level of crowding, then there are cheaper alternatives that are better for the riders, like running express buses to various parts of downtown. As it is, much of the ridership comes from *forcing* riders to take the train. The 41 is gone, which means a lot of people that would have gotten downtown five minutes faster take the train. If crowding was an issue, then buses like the 41 get resurrected.

        I really don’t see it though. If agencies run express buses it will be to give riders a premier experience, not deal with crowding on the train.

      2. What stops the Floating Bridge line from having higher frequency? There isn’t surface running until BelRed; I’d rather turn trains around at the OMFE, not SoDo. If there’s crowding downtown, then there’s probably also crowding crossing the Lake.

        Ross is right: 4-car trains at 3 minute headways should be more than sufficient for downtown, even at peak of peak. The reason we need a 2nd tunnel is because of the sheer induced demand of a Ballard-LQA-SLU subway, plus all of the forced transfers pulling riders off of 3rd Ave surface down into the 2nd tunnel. Without that new subway north of Westlake, and all the current buses running on 3rd from WS and Ballard, the DSTT should manage demand just fine when downtown ridership exceeds pre-pandemic demand.

      3. The reason we need a 2nd tunnel is because of the sheer induced demand of a Ballard-LQA-SLU subway, plus all of the forced transfers pulling riders off of 3rd Ave surface down into the 2nd tunnel.

        OK, I don’t follow you. The Ballard line is supposed to be paired with the Rainier Valley line, which means that it will run every six minutes, at most. It may induce a bunch of riders, but it will only run every six minutes.

        If the proper investment is made, then you still have the same number of trains going through downtown every hour (30), and the same number on each segment (10 for Ballard Link, East Link, Tacoma Link and West Seattle Link; 20 for the north end). The only change is the frequency to the north end. It would have the same throughput, just not at a consistent 3 minute interval. Instead it would alternate between 2 minutes and 4 minutes. It is unlikely anyone would be left behind with the train running every four minutes, and even if they were, the next train would be there 2 minutes later (and it wouldn’t be full).

        Train bunching is a bigger concern, but given the cost, solving that problem doesn’t seem worth it. The second tunnel idea makes sense if it added value (e. g. with a station at First Hill) or if it was relatively cheap. Neither is the case. Given current costs, this is the type of project that should be deferred indefinitely, while we build other things (like Ballard to UW).

      4. Oh, I don’t think it’s possible to run 4-car trains at 2 minute headways in the current tunnel. Perhaps if it was entirely autonomous (through downtown) and the stations were rebuild to handle greater crowds (Spanish solution?), but that would entail closing the tunnel for a long period of time (presumably also while the tunnel is closed to create the new junction to the north).

        Small autonomous systems run at that frequency, but where else in the world does a major urban system run trains at that frequency?

      5. Based on my stopwatch, you could run trains at 2.5 minute headways in the tunnel as it exists. That’s what it takes to get from one station to the other, open its doors and wait 20 seconds, plus some time for error.

        My assumption is ST wouldn’t want stuff as tightly packed as TriMet has, so I measured it last time I rode link.

        Spanish solution shouldnt involve too much disruption. All the work would be done away from existing platforms, and some stuff could be done at night.

        Closing the tunnel will have to be done at some point anyway. Rail eventually wears out and needs replacement. TriMet replaced the stuff between Lloyd Cener and Rose Quarter in a week, but surface running is easier to work with. Rail replacement in the tunnel will be more involved.

      6. AJ, SF Muni Metro light rail has regularly operated 30 trains per hour before. Granted that they use two-car trains and sometimes there are short delays where tracks merge or when there are lots of passengers but it’s possible to get close. There are plenty of other systems that can serve 25-30 trains per hour in a subway line.

        So let’s parse two minutes or 120 seconds:

        – pull into the station
        – hold the doors open
        – pull out of the station
        – provide an empty track before the next train approaches

        Allowing maximum 20 seconds to pull in and then pull out is more than sufficient. Holding the doors open for 30 seconds is a long time unless the train or platforms are too crowded. Even with these ample maximum times, that still leaves 50 seconds between trains.

        Really, it’s the time the train doors need to stay open that affects the ultimate throughput. For example, it can take lots of time to empty a train if it goes out of service or if riders are headed to or from a game. Even so, that delay wouldn’t linger for very long.

      7. Three minute headways accommodate 12,000 pph in the peak direction (20 trains x 4 cars per train x 150 passengers per car). That is quite a bit less than the projections for Westlake to Capitol Hill, well into the 30,000 zone.

        Grant, those projections were made pre-Pandemic, but if downtown does rebound significantly those two minute headways that can be bought rather cheaply will be necessary.

        Al, Muni Metro can run 30 tph because they double-berth the stations. The platforms for Muni are just as long as those for BART’s ten (shorter) car trains. Muni’s are at the most two cars so it’s really easy to double-berth.

      8. “Small autonomous systems run at that frequency, but where else in the world does a major urban system run trains at that frequency?”

        Would you consider the Victoria line on the London tube as “small”? I believe that line runs at 34-36 tph at peak periods.

      9. I’ll add an example to further enhance the Tlsgwm’s point: Almost every subway line in Manhattan has three branches. Look at their subway map!!!

        To the three-line naysayers, you are being ridiculous!

      10. Al, I don’t believe any of those branched routes run 30 tph in their trunks. Lexington (4+5+6) runs at 24 tph, I believe. Looks like the 7 runs at 33 tph, but I believe that includes significant 4-track segments and no branching.

        Thank you, Tlsgwm, yes it appears the Victoria line is a good example (36 tph), a highly successful urban line, though I will note it has no branching.

      11. Al, that’s not entirely accurate; the stretches of NYC’s subway with 3+ services are on triple- or quadruple-tracked sections. For example, along 6th Ave (orange), you have the B, D, F, and M trains. However, the B/D never share tracks with the F/M, so you generally only have one or two services per pair of tracks. The exceptions to this are the J/M/Z going across the Williamsburg Bridge (although the Z is really just a variation on the J), and N/R/W between 34th and Lex/59th

      12. AJ, the 7 is double-tracked from Hudson Yards to Queensboro Plaza (Manhattan and Long Island City in Queens), then triple-tracked to Flushing. During rush hours, there’s a mix of local and express trains on the triple-tracked section, resulting in fewer trains for the Queens stations as capacity is limited by the double-tracked section in the city

  13. Well, I’m going to take my time reading this EIS, before I say anything. Probably the last rapid transit line constructed here in my lifetime, after all.

  14. Whoa up, guys. While I like the flurry of ideas how to share DSTT1 and make Ballard-West Seattle the “through” urban subway, you must explain how you would join the Ballard end to the existing tunnel. No handwaving with a mumbled “The solution is left to the student” is allowed.

    Everyone is talking about running trains every two- to two-and-a-half minutes through the DSTT, so that means ST will categorically deep-six any level crossing in the tunnel as a threat to reliability, regardless of Glenn, d.p. and the Chicago El at Lake and Wells. ST simply isn’t going to build itself a level crossing for revenue trains in the single most congested segment of its main line.

    If the cut-and-cover section east of Westlake Station had been built the full three lane width of the Westlake vehicle area, it would be a tricky but relatively inexpensive matter to slip a burrowing junction for northbound Ballard trains in between the two existing tracks connecting to the lower level of a stacked tunnel under the Reversible Lanes Pine Street off-ramp. The upper level could simply merge with the southbound track at the Capitol Hill curve.

    But it wasn’t, and its placement under the center of Pine Street means that the burrowing junction can’t be sliced into the south half of the street. It would require the closure of the sidewalk on the south side of the street the entire distance to at least Ninth Avenue. There are far too many quality businesses between Sixth and Ninth to do that. And it’s very likely that digging that trench would hideously disrupt the supports for the road decking on the southside of the street.

    That means that some very tricky double level engineering north of University Street, probably requiring use of the bus-bypass lane in USS for the northbound connection, will have to be built. There’s really no other place to put it. The track would drop steeply to pass under the southbound track at the curve from Pine to Third west of Westlake.

    That means that Ballard-West Seattle trains would not be stopping at the University Street platform; it’s plausible to put a “rump” station along the east side of the northbound Ballard track as soon as it reaches “depth” and connect it via underground passageways to the Westlake platform and to the surface next to Macy’s or perhaps in the block between Pine and Pike. The southbound junction would be “upstairs” at the Pine Street curve. West Seattle trains would just “merge into traffic” by waiting in the tube north of the junction. They would stop at USS’s southbound platform.

    That would make USS somewhat less useful for Ballard bound riders than sharing the existing northbound platform, but they’re going to be up at Fifth and Spring or Sixth and Seneca in the present proposals, so they might not mind walking three blocks on the flat instead of two up steep hills.

    However, then you have to decide whether the Ballard end is “done” or just connected temporarily until the money to build a “real” tunnel through downtown is made available. If the former, then just turn the stacked tunnel right into the Stewart right of way and then left at Westlake and build whatever is best from there north. I would cut-and-cover this relatively narrow “trench” because it would make the “basement” platform by Macy’s somewhat shallower.

    In all honesty, I’d cut-and-cover all the way to the south end of the Denny Way station box, because Westlake is a BIG WIDE STREET that can be dug up a la Market Street and decked over during construction for use by transit. In fact, it would be the first step in turning it into a permanent “Transit Boulevard” with the streetcar and RapidRides running in three lanes above to allow the RR’s to pass the streetcar stops, like in the Portland Transit Mall and the subway downstairs. That would be two lanes narrower than the current street, allowing wide sidewalks and plantings. It would be like a “Grand Plaza” leading to downtown from the Lake Union district.

    In the distance between the curve and the south wall of the Denny Way station the upper southbound track would rise from bored depth and when it achieves the upper level depth, move east to above the southbound track, making the curve into Stewart easier and in a single track width trench.

    But if the connection here is to be temporary, and the desire to build a second tunnel merely “deferred”, then there are some constraints. The connection should be just the same, but it is probably necessary to bore the Westlake right-of-way between Denny Way and the junction because a New Westlake will still have to be underneath DSTT Westlake. There must be a junction box placed where the diversion will in the future be made between Virginia and Stewart under Westlake to allow for easy connection of the new tunnel.

    I like the Fauntleroy elevated station, but I don’t see why that nick has to be taken out of the east side of the street north of Alaska. Light Rail Trains can make curves. It’s what they do.

    So far as the Ballard end, I still think it makes much more sense to place the tracks at-grade on the north side of Elliott and replace the businesses there with mid-level buildings above and in front of the trackway. North of the Helix Bridge rise up to the level of the Magnolia Bridge flyover, take the bridge footprint and cross over to the west side of 15th West and descend to run adjacent to the BNSF tracks. I’d put a station just south of Armory for Whole Foods and the growing neighborhood there and straddle Dravus at BNSF level for the station there. The Urbanist’s “build a car bridge in the Fourteenth W/NW right of way, demolish the Ballard Bridge and replace it with a transit bridge” plan is a huge win for everyone, with the caveat that the “wiggle” between Fourteenth and Fifteenth means some buildings near Market Street have to come down.

    But, again, this is Light Rail. If you’re going to narrow Fifteenth NW and make half of it a transitway, put the tracks on the ground north of Leary. Have a station just before the tracks start to claim, somewhere about 50th.

    How hard is it to understand that Link uses Light Rail Technology which was explicitly created to take advantage of available surface rights-of-way.

    “Oh, but that’s two additional stations!!!!!” [including the one at Armory] Yes, it is. And two additional opportunities for dense transit-oriented development on what will likely be Seattle’s only “urban rail” line for thirty years. Put the track on the westside of Fifteenth, and to accommodate the needs of drivers from west of Fifteenth to go south, improve Seventeenth’s connections to Leary Way and Market Street.

    1. So, here’s one crazy, out-of-the-box idea. What if the Ballard train were to run tunneled through SLU and lower Queen Anne, but just run on the surface down 3rd Ave. through downtown. Heck, you could even run it in exclusive lanes down 1st, as a replacement for the CCC, if that’s what it takes to unlock the CCC money. The line would end at International District, with a nonrevenue connection to the mainline around Stadium Station to allow the trains access to the SODO O&M. Passengers continuing onward would exit at 3rd/Pine and transfer to the existing line in the Westlake tunnel station. West Seattle trains would just tie into the existing line at SODO station and run through the existing tunnel.

      It sounds crazy at first glance, but would save several billions of dollars in construction costs. Since DSTT2 is funded by all subareas, this could cover cost overruns throughout the system. As a consolation for Ballard, maybe some of the cost savings is used to give them a better station on the Ballard end, perhaps a tunnel under the ship canal to 17th or 20th, race than a station at 14th. I would argue, for a Ballard resident, the better station access on the “home” end would more than compensate for having to put up with a few stoplights on the downtown end, and if you get off the train at 3rd/Pine, you’d miss most of them, anyway. Even if downtown is surface running, you would still benefit immensely from the train having exclusive ROW through chronic congestion points, such as the ship canal crossing, Lower Queen Anne, and Belltown/SLU, so this would still be far superior to just building a cheap-ass streetcar.

      1. I think it has real possibilities. That’s exactly the trajectory that the Muni Metro took. The Twin Peaks and Buena Vista tunnels were built in the 1920’s. The trains passing through them used surface trackage on Market Street through downtown until post-war traffic demanded and BART construction provided a trench for them to run in.

        Again Westlake offers an “across the grid” roadway to close for the transition to the surface. I’d put it on Fifth, though, because it serves the downtown core better, especially important public buildings.

        A decision to use Fifth might also cause some unexpected “third- party funding” to be found in the cushions at the Downtown Seattle Association. Jes’ sayin’.

        What I find MOST attractive about this suggestion is that North King would suddenly have one of the smaller subarea project budgets. East King would be building the wasteful Kirkland-Issaquah “Train to Nowhere” while reimbursing North King for exceeding its share of the downsized package. MOTU Mortification, indeed.

      2. This is kind of what I was thinking above except I thought they would need a 2nd tunnel for future extensions (Aurora, Fremont, etc). But you’re right, any extensions could also ride the surface and transfer as well.

        You got my vote! Please submit idea to WSBLE

      3. I don’t see an especially good reason not to widen the existing streetcar lines a few inches, so that they could use compatible equipment with Link. That way, they could run on the surface line through downtown as well, serving both the CCC desire as well as the Link line.

        Routing the line through Belltown on 1st would add service to this area that otherwise is lacking.

        Previously I suggested doing the surface line first because it’s cheapest, as well as you can start the Ballard line with two car trains. The tunneling through Lower Queen Anne is going to take time to dig.

      4. “North King would suddenly have one of the smaller subarea project budgets.”

        Not quite true, since the ST3 had DSTT2 funded by all subareas, not just North King, since trains from all over would use it. Nevertheless, every subarea has cost overruns, and could definitely use the savings.


        It sounds like a tempting idea at first, but the more I think about, the more I see issues. First off, I’m not sure that the streetcars and Link even use the same track gauge or overhead wire specs. If they don’t, it’s a nonstarter.

        Even if they do, I’m still not sure it’s worth it. Ballard Link would add yet another travel option between SLU and downtown, which the CCC would largely duplicate, on top of 40, 70, and C-line buses (I’m assuming the C-line’s SLU tail gets tacked on to some other route if the C-line, itself, is replaced with West Seattle Link). You’d still have to tear up a few blocks of downtown and buy more streetcar trains. It would be worth at least finding out what the marginal cost of completing the CCC would be, but my gut feeling says, even with the 1st Ave. tracks already built, it would still not be worth it. I would rather spend the money giving Ballard a good station location to compensate for a surface-running train downtown.

      5. I wouldn’t consider Westlake’s ROW ‘available.’ During peak, there should be a bus or streetcar coming down Westlake every few minutes; there isn’t capacity to then layer on a frequent Link train, unless you want Link to be infrequent/slow or you want to block off cross traffic (like the 8).

        The most important part of dedicated ROW is in the downtown core. The only time it is OK to run surface running through an urban center is if there is no through traffic (e.g. Long Beach in LA or Karlsruhe in Germany). There is massive latent ridership that wants to use WSBLE to get through downtown; surface running would squander that ridership. The place to save money with surface running is in Interbay, Ballard, and West Seattle, not the urban core.

      6. We could just extend the SLU streetcar to Ballard as in ST3’s first system plan proposal. But the SLU streetcar is very slow: it stops for a light every single block between Stewart and Denny, in addition to stations every two blocks. The 40 does better than that.

      7. Extending the streetcar to Ballard would be a terrible investment.

        If ST really does end up hitting reboot on WSBLE, it can grant money to SDOT to fund frequency & ROW improvements for the D & 40. That would be a much better (and quicker) short term improvement while the city & region rethinks how to serve the corridor long term.

      8. There are two differences between Link and the streetcar: car width and overhead line voltage.

        Overhead line voltage has a lot of margin of error. For both systems it’s highly likely they’re insulated to the same voltage, because you don’t want a bunch of custom hardware. They like to see insulation rating in the 5,000 volt range, even though Link is only 1,500 nominal and the streetcar only 750. At some point in the future, the streetcars will wear out and need replacement. Probably around 2035 or 2040. At that time, streetcars with different static converters can be obtained, or just purchase new static converters for the several streetcars. They retail about $100,000 last I checked, and during a mid-life overhaul on the cars would be replaced anyway.

        Furthermore, there is at minimum some overlap of voltage. International agreed upon standards say that a 1,500 volt system be able to handle points in the overhead lines that drop as low as 1,000 volts as a permanent feature (not a transient voltage sag). The specifications for 750 volt equipment goes up to 1,000 volts. For safety, a margin of error is built in to the converters, so even the existing equipment should be able to share overhead lines, so long as the overhead voltage is designed to be right at the 1,000 volt spot the two specifications share.

        The track gauge is the same, but the width of the cars is somewhat narrower on the streetcar. This is probably the most difficult thing to overcome, as it will mean some actual careful movement of traffic lanes, parking spots, and platform space.

        Sharing the same track isn’t vital, but if you’re going to put in a surface rail line that’s good quality for Link trains, it seems to me you might as well get as much use out of that expense as possible and make the two systems compatible.

        As far as train frequency goes, what I proposed on the other thread was that surface trains are going to be two car trains. So, the route through Belltown would be a supplement to the trains through Lower Queen Anne. It would be a train every 7.5 minutes, alternating a two car train on the surface with a four car train that would serve the tunnel and continue south to the Rainier Valley, where frequency is limited anyway.

        Possibly, with the shorter surface trains, you could even build a surface line along Market and bring those trains into Ballard itself.

        I was still picturing Lower Queen Anne to Westlake in a new tunnel, but feeding into the existing one.

        As far as duplicating the 40 or other routes, currently there isn’t much of anything on 1st, which is why I’d put the surface route there. Any duplicating of bus routes is really more of an issue with the tangled streetcar plan.

      9. asdf2, I was assuming no West Seattle when I made the quip about a low-cost project list. That may be a pipe dream, but building those soaring stilt monstrosities is clearly not good value and should be chopped. Build a second lane on the outside of the clover leaf from the eastbound West Seattle Freeway to northbound SR 99 and give the buses the current inside one. Then they’d have an almost traffic-free path to Columbia.

      10. Glenn, are you actually proposing a fully parallel line between the curve at First and Stewart and “somewhere” along Elliott? You’d open Ballard without the LQA/SLU tunnel? How would you get from First and Denny to the junction point? You sure as heck can’t run on Denny and Elliott.

    2. you must explain how you would join the Ballard end to the existing tunnel.

      No, we don’t. We only have to make the case that it is likely to be significantly cheaper than a second tunnel, and thus worth studying. Sound Transit never told us how they would build a second tunnel. They only said it would be a good thing. OK, now we know the details. It turns out it will be extremely expensive, and all the stations will be really deep.

      Of course it will take a real study — by real professionals, not the folks here. Maybe they will find that there are way too many issues, and it turns out to be more expensive than building a new tunnel. I seriously doubt it. My guess is that it would be considerably cheaper.

      1. Of course it will be cheaper to build. The question is whether it will be more disruptive to build, less resilient to operate, and have less local & regional capacity than a 2nd tunnel.

      2. But sure, go ahead and study.

        What I’m more interested in studying is a shallow tunnel under 5th & Westlake. Yes, cut & cover would be immensely disruptive, but if it saves a billion dollars, then perhaps it’s worth the pain. The region dealt with 3rd Ave being closed for over a year in the 1980s; surely we can deal with a simillar closure on a less important avenue.

      3. I think both connecting Ballard line into the existing tunnel as well as doing a cut&cover 5th Ave alignment should be studied as everything presented is unacceptable.

      4. No we don’t

        Yes, you have to show that it’s at least feasible to get the politicians to listen at all. The consultants are going to defend their work as the only possible solution and call any such proposal “The Seattle Monorail reborn”. Count on it.

        We have to bring a solid proposal to the table with a consensus logical design, not seven different options of what “would be cool”. Of course we can’t do the detailed engineering, but any such proposal has to have a high degree of plausibility even to be investigated superficially.

  15. #1 All these alternative ideas seem eminently sensible, but the fact remains that unless there is some sort of revolution ST WILL dig the 2nd tunnel – even if it’s for no reason than “this is what the people voted for”.

    #2 The DEIS panders to certain, organized, interest groups. The best example of this is the preference given to the deep stations to avoid the short term disruption to the international district businesses.

    #3 The biggest interest group – the riders – is not as organized as it should, and could, be. Whether it’s STB, which likes to talk; Seattle Subway, which likes to dream; or the TRU which, well, I don’t know what about the TRU other than they really don’t seem to represent the bulk of those who would use the system; there is no real organized rider group that ST, or the City of Seattle will listen to.

    #4 People who are going to use the system, the riders all along the line from Everett to Tacoma, need to get angry, and organized to get ST and the city to build something that works. Please don’t let them build a line just for the sake of building a line.

    Lastly, it is a slim hope, but if the mayor, the new head of SDOT, and whoever the new head of ST will be, has a vision, and the will to lead, things might change. Yet my expectation is that these new people will be caretakers of the process, and unfortunately the project is likely now set to continue, as is, on inertia alone.

    1. “The biggest interest group – the riders – is not as organized as it should, and could, be.”

      Do you have a suggestion? The problem is that ST listens to “stakeholders”. Each city, county, the Port, and large employer is a stakeholder. STB and other transit advocacy groups and riders are all lumped together into one stakeholder.

      The cities and counties want the Link projects in their districts, and don’t want to hear about canceling DSTT2, converting West Seattle to BRT, truncating Everett at Mariner. The Port and Fisherman’s Terminal don’t want Link on 15th. The International District businesses who don’t want a shallow alternative have the city’s sympathy. West Seattle was prioritized over Ballard in the original ST3, and they insisted they should get nothing less than light rail. Whether the city might be willing to bend on West Seattle now that all these problems have come up, is still unclear to me. So far there’s been no indication of it. ST and the cities seem to prefer sticking to the original projects even if they take years longer to complete.

      1. The biggest process problem with WSBLE project is that there is not a Rider’s Committee in addition to a Stskeholders Committee and an Elected Officials Committee. All The ST Board had to do is add this new committee to review the DEIS and everything after that.

        This Committee should have at least 15 people on it and give substantive feedback on all the designs. It should have architects, engineers, planners and other types who can look at images and text and visualize how people will use the system. It should have riders who have lived in other cities and know what problems and opportunities can exist.

        The current process is akin to having a school planned without consulting teachers or students — or a park planned without park users reviewing the project. Instead, it’s planning geared to listening to only the wealthiest neighbors that we call “the stakeholders” although they have no intention of actually using the system on a daily basis.

      2. Mike, I would think that The Port and especially Fisherman’s Terminal would love moving the auto traffic to 14th and building a new high — or even a medium height one — at 15th. Properly designed, it might actually increase the size of The Terminal.

      3. “The biggest process problem with WSBLE project is that there is not a Rider’s Committee in addition to a Stskeholders Committee and an Elected Officials Committee. All The ST Board had to do is add this new committee to review the DEIS and everything after that.”

        That’s exactly what I said in my feedback to ST on the Alternatives Analysis and a couple other times. So far it hasn’t worked.

      4. I’m many ways, calling Seattle and Pacific each one stakeholder, regardless of vast differences in population reminds me of the U.S. Senate, where California and Wyoming are treated as equal stakeholders in the national government.

        The net effect is a bias against people who live in a dense city, and all belong to just one stakeholder.

  16. Where is the equity?

    West Seattle to North Seattle – fairly well-off neighborhoods – get their trains in the first tunnel with shallow stations.

    Federal Way through the Rainier Valley (somewhat poorer neighborhoods) to Ballard, get their trains in these deep, inconvenient stations in the second tunnel…

    Is institutional bias of some sort behind the design here?

    1. This is an important “privilege” bias that doesn’t get discussed. DSTT2 actually makes it worse for SE Seatttle riders on Link. ST does not propose any mitigation for this. Meanwhile, the WS segment serves very few lower income households.

      Of course, ST3 Link itself skips the lowest income areas in almost every subarea. Issaquah before Renton? South Federal away before Burien? WS Junction before White Center?

      And ST laughably says that ST3 is “equitable” because the construction doesn’t impact low income communities!

      1. Renton and Burien are served well before Issaquah and SFW, with Stride. You can critique Renton for picking the wrong mode (their decision, not ST’s), but Burien I believe is very happy with their early ST3 projects.

        How is DSTT2 worse for SE Seattle? Giving up a 1 seat ride to UW but gaining 1 seat rid to SLU & LQA feels like a wash. It’ll take longer to exit the station, but Westlake will be 2 stations closer (no Stadium or Pioneer Square stops) so that’s also a wash.

      2. “ How is DSTT2 worse for SE Seattle? Giving up a 1 seat ride to UW but gaining 1 seat rid to SLU & LQA feels like a wash. ”

        I disagree. First, SE Seattle residents lose closer direct access to the Courthouse, the Ferry Terminal, Pike Place and Capitol Hill in addition to UW. Then, using all of the new Downtown or SLU station platforms becomes challenging since the stations are planned to be much deeper than the current ones. The only advantage is getting to SLU or Seattle Center or LQA or Ballard more directly — and the latter two are more neighborhood hubs rather than big citywide destinations. ST3 also extends service closer to Tacoma — but that’s not a big draw from SE Seattle unless one is going to the casino.

        Even though Judkins Park becomes an attractive new transfer point for Route 7 riders using the existing DSTT in 2023, that’s in ST2 and not ST3 so ST3 doesn’t make it better..

      3. Giving up a 1 seat ride to UW but gaining 1 seat rid to SLU & LQA feels like a wash.

        Seriously? Come on, the UW is a much bigger destination. It is also worth looking at the entire set of stations (e. g. Capitol Hill, Roosevelt, Northgate, etc.). Do you really think that more people will ride the Ballard line than the train to the north end?

        Sorry, but Al is right. This is a degradation for South Link riders, and one of the flaws with the current alignment. It is crazy to think that Dow insisted on giving West Seattle riders a one seat ride to downtown (by avoiding First Hill) but lots of existing riders will have to transfer. I’m not saying a transfer is a big deal (it isn’t) but the pairing is less than ideal. It would make much more sense to pair the East Side with Ballard, since a lot of the trips from the East Side (Bellevue to UW, Bellevue to Lynnwood) will happen by bus anyway. There is also a business-to-business connection (e. g. Amazon Seattle to Amazon Bellevue) that doesn’t exist with West Seattle or the south end.

    2. The decision to build West Seattle first was in 2016. Strongly prioritizing equity started in 2020. So one interpretation is that the subsequent changes haven’t caught up to the West Seattle decision yet. Another interpretation is that Delridge, 35th, and 16th, are lower-income equity-emphasis areas.

      1. Focusing on “equity” for a long term project like ST3 feels like a fools’ errand. At a minimum, by the time any of this is built, all the neighborhoods that are getting Link will have long since gentrified. If we really cared about equity, most of the ST3 money would probably not be going into Link at all, but, rather, more bus service that spreads out better throughout the region.

      2. Mike, I totally agree. When I listened to Sound Transit and SDOT at the advisory group meetings, they mentioned equity and talked about placemaking: BIPOC need to feel at home at the stations. That’s not equity to me. Equity would be to revisit ST3 station location. WS stations and its anticipated extension would go through the least diverse areas of WS and not serve High Point, the most diverse part.
        The Duwamish bypass would have much higher equity value and would help increase capacity to Seatac and beyond. Gondolas to Kent and White Center could get the equity up further. And all this probably would cost about the same as the current WS plan.

  17. Question: What is the main reason to build a new tunnel? It doesn’t feel like the existing tunnel is at capacity. I’m sure there are reasons but it sounds unnecessary at first glance.

    1. ST wants to be able to run all its lines at 10 trains/hour (6-minute headways) at peak. Lines 1 and 2 (Central Link and East Link) inter-lined result in 3-minute headways through the tunnel. WSBLE will be Line 3, and attempting to run all three through the DSTT would require 2-minute headways, which ST says the tunnel tech can’t handle. Many have reported that the feasible limit for DSTT is 2.5 minutes, which is 24 trains per hour, or 8 trains per hour per line, which is 7.5 minute headways.

      DSTT2 should be a cut-and-cover from Chinatown/ID to the Midtown station along 5th Avenue. If they really need to launch TBMs to get under Westlake and the SR-99 Tunnel, they figure out how to launch them from the Midtown station box dig and catch them on the side of Seattle Center, where the cut-and-cover resumes along Republican.

      The alternatives north of Smith Cove are fine. During the Community Advisory Group meeting on Wednesday, I’m going to be asking how the high bridge and tunnel costs came to be within 10% of each other, and what it would take to consider the bridge-swap alternatives proposed by several on this blog.

      1. “ ST wants to be able to run all its lines at 10 trains/hour (6-minute headways) at peak. Lines 1 and 2 (Central Link and East Link) inter-lined result in 3-minute headways through the tunnel. WSBLE will be Line 3, and attempting to run all three through the DSTT would require 2-minute headways, which ST says the tunnel tech can’t handle. Many have reported that the feasible limit for DSTT is 2.5 minutes, which is 24 trains per hour, or 8 trains per hour per line, which is 7.5 minute headways.”

        Not every line will need 10 trains an hour. If three lines are in one tunnel, it could be 10-8-6 or 10-7-7 or even 10-6-6 or 9-8-6 or any number of combinations as long as the branches can handle the crowding.

        Looking at West Seattle to Everett, there is no segment carrying more than 16,000 daily riders in one direction in the places where there isn’t another overlapping line, it will probably never need six minute service (10 trains an hour) and that 6 trains an hour (every 10 minutes) seems more than adequate in those stand-alone segments. The higher frequency is needed for Line 3 to serve the crowded Downtown-UW segment and that need goes away if Line 1 isn’t moved over to DSTT2.

        So I see the question as whether something like 24 or 26 trains an hour would be operationally doable rather than a full 30

        One advantage about a Tacoma-Northgate line is that it also better guarantees a seat for North Seattle riders.

      2. attempting to run all three through the DSTT would require 2-minute headways, which ST says the tunnel tech can’t handle.

        They said they can’t handle it now, but they made clear that with a relatively small investment (in traction power substations) they could run the trains every 2 minutes. It is just that doing so “wouldn’t give our ridership as reliable a service.” Trains might occasionally be delayed (bunched). The signals are limited to 90 seconds, so that is likely the upper limit.

        Running trains every 7.5 minutes is just a suggestion for the middle of the day (as opposed to running every 10). It isn’t as easy to understand as running every 10 minutes, but it is better than 8, since it is the same every hour (and half hour). You can list the timetable for that hour, and everyone understands it is the same every hour, for that period. In contrast, it is a lot harder to look at the schedule right now. For example, let’s say I want to take the train from Othello to downtown around 4:00 — I have no idea when to expect the train (unless I do some tricky math):

        In any event, a second tunnel is better, it just isn’t worth the money. Every rider in the new tunnel would be delayed (because the stations are so deep) while relatively few could see a delay on the train caused by bunching. Lots of agencies simply put up with a little train bunching, while investing in more important things. We should do the same.

      3. It may very well be worth the money to have a generic second tunnel.

        *This* tunnel is not. Not with the vastly deep stations and consequential inconvenience to riders, with resulting low ridership.

        In the end, I think it would be a catastrophe for DSTT1, because I’m guessing a substantial number of passengers transfer at Westlake and SoDo to get closer to the surface, resulting in overloading DSTT1 anyway, only without enough trains to carry the added passengers.

        It’s also one of the reasons I suggest a surface line through Belltown to add coverage, as well as getting people closer to their destinations. Through passengers can take the deep tunnel, but if your destination is surface downtown Seattle you might prefer the surface line. It might reduce crowding a bit in DSTT1.

        If the problem is the substation capacity, they should look at line-side capacitor banks. Systems that use those have had a decrease as much as 60% in their energy bills, and you can get by with smaller substations.

  18. If we’re talking about a single tunnel, why do we need a double line through SODO? Can’t we just branch the RV line?
    RossB suggested Ballard to Redmond. That line would run through the DSTT2 (or single tunnel). Then DSTT1 could handle everything else.
    We would still need to remove car intersections but may be a single bridge (Holgate or Lander) would be sufficient.
    We wouldn’t lose the bus lane, wouldn’t have to move the power line, wouldn’t need to worry about the USPS site, should be a lot cheaper.

    1. In short, two sets of tracks wouldn’t be needed, but only if both Holgate and Lander are bridged and Lower Royal Brougham is closed at the train tracks.

      A cheap way to branch from a single set of tracks is to use the MF bypass track for the northbound connection. It already has a flying junction. The southbound track to Ballard would just junction at the top of the rise south of Lander.

    2. If we are truly capable of running all 3 lines through a single tunnel (as many argue here), then yes surely we can run 2 lines on a single pair of tracks from SoDo to ID?

      But if we are building the full 2nd tunnel, one of the key benefits is resiliency through redundancy, so therefore it makes sense to also have parallel tracks (with a few switches/crossovers) so if there is a issue on the line between SoDo and ID, it doesn’t impact the overall operations of the downtown trunks.

      1. This is also why it would be useful to have the Ballard line connect to the existing DSTT. At some point, even if DSTT2 is built, the DSTT2 will need to close from time to time for maintenance.

        Being able to feed the Ballard trains into either tunnel allows for a bit more flexibility when that happens.

        Witness the recent Chicago L red line rebuilding: red line trains were moved from the subway to the loop. Sure, it had limitations, but it sure was useful to have those connections on each side of downtown rather than try to figure out a bunch of bus bridges. The L has a fair number of connections between lines for this very reason. They may not be used much, but when they are used they prove pretty useful.

      2. Yes, Glenn, no matter what happens, the RV line needs to have full access to DSTT in addition to DSTT2. The same should be true for West Seattle.

        I don’t think reviewers fully get that not only is the line configuration changes per ST3 current plans, but the way in which the WS and RV tracks cross in the current plans make it hard for ST to run a train on a different track though Downtown.

        That said, I used to think that four tracks were needed at SODO. After looking at the ridership levels for West Seattle branch, I can now see that there could be 10 trains an hour headed towards the RV and just 6 headed towards West Seattle.

        That also means that it’s not really necessary to have a West Seattle stub segment. I suspect that the stub will be a major source of aggravation for West Seattle residents anyway. Plus, having a Tacoma train run north to Lynnwood after 2032-3 could be too long for one driver to do.

        So I see three lines in DSTT as an eventual operation driven by WS clamoring no matter what (at least until a DSTT2 would open). I could see 9-10 RV trains ending at Northgate, 7-8 Redmond trains ending at Lynnwood (eventually Mariner) and 6 West Seattle trains ending at Lynnwood (eventually Everett). Perhaps the riskiest scheduling problem is turning around a train at Northgate.

      3. Al, well said about Northgate. Pocket tracks are not really meant for turning trains in scheduled service. They’re for parking a few “overflow” trains that can sit in the pocket for several minutes while the operator reverses. It’s not immediate, which is why terminal stations have scissors “nearside” of the platforms. [I am shaking my head about that damn tail track in Ballard. What are they thinking?]

        You can reverse in a “farside” tail track by double-seating [e.g. an operator boards the last coach at the station, rides into the pocket while preparing their cab, and then takes the train out after control is passed from the operator who brought the train in. That operator then deboards at the platform for a break of a train or two and repeats the process.

        But that is kind of expensive: there must always be a ready operator on the platform when a train to be reversed arrives. I don’t think ST wants to do that long-term.

      4. Yes, both Ballard and WS should be able to use either tunnel for redundancy, but I’m not sure they need a double line through SODO for a while.
        If WS gets built it could connect to the existing line (and Holgate/Lander bridges built), it would not have to dead end in SODO, but could continue to Northgate providing extra capacity downtown until DSTT2 is complete.

  19. I think the biggest mistake from an advocacy perspective was dismissing at-grade options prior to alternatives scoping. By requiring grade separation, we locked in the idea that existing roadways would not be modified to accomodate light rail. Fauntleroy+the WS bridge ramp in West Seattle and 15th in Interbay/Ballard were both workable options for running light rail down the median. It could have been done with two or three at-grade crossings at most, or the road crossings could have been elevated over the train tracks.

    At-grade is the best option for bus transfers. No escalators. No deviations from the roadway.

    At-grade would have avoided many of the cost overruns caused by acquiring hundreds of properties to build a new right-of-way in established neighborhoods.

    At-grade would have used the smooth grade of the West Seattle bridge approach, as opposed to going up the steep hill on either Genesee or Yancy St.

    At-grade was the best opportunity to convert some of our worst streets into more bike and pedestrian friendly options. The mess of highway ramps in near the Ballard Bridge and where Delridge meets the West Seattle bridge could have both been reimagined.

    I’m not sure how much of this is actually going to get built. There are $2+ billion in cost overruns that remain unaccounted for. We know they are going to forge ahead with the West Seattle spur in order to get shovels in the ground by 2026. Just that section alone is loaded with engineering challenges and the potential for unforeseen problems. They’ve already set up the likelihood of terminating at Smith Cove, so it won’t be a big shock when Ballard gets punted to ST4.

    1. West Seattle is a loser through and through, and it really can’t be “fixed”. Buses. Nice buses, but buses.

      1. @Tom T
        This. Times ten.

        @Joe Z
        These are not “cost overuns”. The increased costs are a reflection of the dreadful job ST did with developing its capital cost estimates for the ST3 package. The final reports from Triunity confirmed that.

    2. I wouldn’t say that is the biggest mistake from an advocacy perspective, just one of many. To be fair to the transit advocates, there was only so much they could do. Somehow West Seattle jumped to the front of the line when it came to projects. Then in the early planning, their “BRT” plans had the bus taking longer to get through downtown than it does now (let alone with a new bus tunnel). I guess you can criticize the advocates for not being more critical, but the average person had no idea.

      From an advocacy standpoint, it is hard to pinpoint the problem — it depends on how big you want to go. The core of the problem is that planning is done by amateurs and advocates. The advocates often have it right, simply because the
      amateurs don’t have any idea what they are doing. But the best they can do is solve the obvious, relatively cheap problems (like adding a station at 130th).

      The root of the problem is that decisions are made by a board that knows nothing about transit (these are the amateurs). They are all politicians. I don’t mean that as an insult (my mom was a politician, and a damn good one) — it is just that these board members don’t know anything about transit, and they all have more important jobs. This would be fine if they delegated to a transit committee (full of Jarrett Walker type professionals tasked with coming up with the best way to spend transit money) but that’s not how they operate. The politicians simply ask each community what they want, and the politicians haggle their way to a proposal. But the communities (lead by local politicians) don’t know any better either. On the rare occasion that they do hire a transit firm to make a recommendation, the more powerful board simply shoots them down (thus, no BRT on the CKC).

      If you wonder how West Seattle jumped to the head of the line, you only need to know this: Dow Constantine was both the head of the board and the most powerful person in the three-county region (by a big margin). He lives in West Seattle.

      From a secondary perspective, the problem was not taking an iterative approach. It is tempting to “go big”, but this creates a tremendous amount of inflexibility, and financing problems. A lot of the huge delays are simply because the projects are so huge. It also amplifies the political problems. It is quite possible that West Seattle would have gotten some sort of transit improvement (disproportionate to their needs) but it would have been relatively cheap if the package was smaller. For example, they could have addressed the relatively minor congestion that occurs every morning at the West Seattle Bridge/SR 99 interchange. A couple hundred million or so, and buses would cruise right to downtown. This isn’t necessarily the best use of money, but nothing like the gigantic waste of money that will be West Seattle Link.

      From a public standpoint, a gigantic set of projects (with no consensus as to what they should be) reduces the opportunity to step back, and see what we’ve built so far (or are about to build) and ask ourselves what we should build next. If Lynnwood Link is just about done, what else do we need — an extension to Everett (via Paine Field) or good bus service in Snohomish County? Same with Tacoma. They are about to build a BRT line in Tacoma, and Federal Way Link is almost done (finally). How about a bunch of express buses through downtown Tacoma to Federal Way? Is it really any better to transfer at the Tacoma Dome than Federal Way if you are headed to SeaTac? Speaking of which, if they build Link to the Tacoma Dome, does that mean the express buses from Tacoma to Seattle go away (even though they will get riders to downtown much faster)? Maybe connecting the I-5 HOV lanes to the SoDo busway (a longstanding WSDOT project) would be a better value.

      From a Seattle project standpoint, the biggest mistake was not insisting on the most cost effective rail project, which was Ballard to the UW. This was by far the best project, by every measure. Riders per dollar spent, ridership time saved per dollar spent — you name it. It would have transformed transit in the north end (unlike Ballard Link, which will only effect a tiny corner). Hell, they probably could have put a station in Ballard (imagine that).

      The transit community simply accepted the idea that a line through Interbay was better, and haggled over how that would work. Originally the argument was that it would be cheaper, since so much of it could be on the surface. But then the community rejected any options on the surface, because they were slower. To be fair, Bruce Nourish proposed a great plan which consisted of surface rail starting from Ballard to Elliot, but underground through downtown, but that was never taken up by the board. The simple fact is that the community, in general, was in denial. A second crossing of the canal was bound to be extremely expensive, as was any new underground stations. Then somehow someone decided we had to build an entirely new tunnel (making this way more expensive) and we are stuck with this mess.

      Is there any way out? Probably not. At best we’ll tweak things around the edges (have a station at 15th, as if that is ideal, when it is simply what people voted for). This will be built, eventually, and yet it won’t address the biggest transit needs of the area, despite the enormous price tag. Ballard to West Seattle will only be one of the costly projects in one of the biggest transit boondoggles in the world, and by no means the worst.

  20. As a cut and cover option, I can’t help but wonder whether DSTT2 should be using the I-5 corridor for part of its route. For example, closing the express lane access from Cherry and/or Pike Street to build cut and cover seems possible at first glance. Between the ramps, the express lanes and the express lane ramps, there seems to be opportunities to close something to build (temporary) or add (permanent) DSTT2.

    1. Temporary, maybe. The I5 ROW is super useful, but I wouldn’t want to built a permanent Link ROW on top of very old & in need of replacement freeway foundations.

      A good example would be running elevated up 5th Ave; the monorail ROW would be very useful to use, but you wouldn’t want to simply run Link on top of those old monorail support structure, but instead would first remove the Monorail and then build a new elevated viaduct. I think running anything along I5 ROW would be the same – whether it’s Link, HSR, or whatever, I don’t think that should happen until I5 has been removed/replaced/rebuilt.

    1. I had long been in the camp that while WSBLE is unfortunately expensive, the value of a 2nd downtown tunnel and a subway through SLU was so enormous the region just needed to put its head down and just get the project built, but after reading that article, plus the comments here, I think I’ve changed my mind.

      To me, the primary reason for the 2nd tunnel is that the region, long term, needs that additional throughput capacity in its largest urban center. But that need is truly long term, for when there are literally millions more people living, working, and traveling through the region every day. In that context, delaying 5~10 years to get a much better project is very reasonable, as long as we have the capacity we need whenever the DSTT1 would be at capacity (2050??).

      If ST has to take a mulligan and go back to the voters in 2024 or ’28 with an entirely reimagined project (say, elevated through downtown, or a 2-car automated line, or whatever), we can still look at the ST3 vote in 2016 as accomplishing the following
      1. Expediting the completion of the ST2 system, including deferred segments to Redmond & Federal Way
      2. Delivering new wins, notably Stride and 130th Station
      3. Moving the ball forward on major projects elsewhere in the region, notably OMF-S and TDLE likely through design & ROW acquisition and Everett Link through EIS (another project that may merit a reboot in ST4)
      4. Banking a bunch of money for ST4

      If Seattle gets a better project in ST4, I think we can still look back at ST3 being a big success, even if it ended up primarily completing the ST2 vision and setting the table for a new ST4 vision.

    2. Mulligan seems to mean the opposite of what it sounds like, but I like the idea. Delay Ballard and West Seattle until cooler heads can prevail. That would necessatate delaying DSTT2 too, so we would effectively be in an ST2 Link network. Without DSTT2 the Everett and Tacoma extensions may not be able to proceed, but that may be a feature rather than a bug. If ST does this, it should advance some money for even larger RapidRide C and D improvements. And also maybe E, H, 21, 40, and 44.

      1. “Without DSTT2 the Everett and Tacoma extensions may not be able to proceed, but that may be a feature rather than a bug.”

        The spine concept has always been a seriously flawed idea. With that said, one cannot simply dismiss the prior commitments ST has made, including Motion 1 passed by the board way back in Dec 1994:


        “A MOTION of the Board of the Regional Transit Authority for the Pierce, King and Snohomish Counties region clarifying the RTA Board’s intent with respect to future light rail service to the City of Everett.

        “WHEREAS, on October 29, 1994, the Board of the Regional Transit Authority adopted a
        Regional Transit System Master Plan and a Phase I element of that Master Plan; and

        “WHEREAS, the completion of the Master Plan will require multiple phases; and

        “WHEREAS, the Master Plan on page 2-8 provides that priority shall be given in subsequent
        phases to linking the four major centers of Everett, Seattle, Tacoma and Bellevue;

        “WHEREAS, Phase I will complete light rail connections between Seattle, Tacoma and Bellevue; and commuter rail connections between Everett, Seattle, and Tacoma.


        “The Regional Transit Authority Board intends that, in order to complete the planned
        connections to the four major centers, extension of light rail service to Everett shall be a first priority in Phase II.

        “PASSED by the Board of the Regional Transit Authority for the Pierce, King and Snohomish
        Counties region at a regular meeting thereof held on the 2nd day of December 1994.”

        As far as I’m aware of, this motion has never been superceded by a subsequent board motion.

      2. Tlsgwm’s point is essential to understanding what has happened. In the very early days of Sound Transit the system was envisioned as a long distance commuter railroad and should have been built using BART-style technology.

        Instead of getting high-geared third-rail trainsets, though, the decision was made to buy the slowest rail technology: low-geared overhead power “light rail” trainsets. That’s presumably because the vehicles are (relatively) inexpensive. But they are slow.

        Then, instead of building the part of the system where there is no viable parallel “commuter rail pathway” — the north line — a decision to go to the airport first was made because the other was “too hard”. Since light rail vehicles were the chosen technology, the opportunity to build on the surface was taken through the Rainier Valley. If the line were just to go to the airport that might have been a defensible decision. Those willing to take transit to the airport probably don’t mind the extra ten minutes it takes. But on a line expected to be nearly forty miles long it is a catastrophe.

        The folly of the commitment to build commuter rail using updated “interurban” technology has been exposed.

        And now, to build a relatively short, inner city line where the primary strength of light rail technology — the ability to run at-grade without electrocuting people — might shine, ST cleaves to a “no at-grade running” philosophy at any cost.

        More folly.

      3. TDLE would be not impacted by DSTT2. Staff was concerned about operating length with Everett Link, but I’m deeply skeptical this is actually a problem; at worse ST needs to switch drivers mid-run until DSTT2 is built.

        Tom, I’m always perplexed by the complaint that the Spine is poorly suited as commuter rail, since none of the spine supporters have ever positioned it as commuter rail. The fact that Link is slow over long distances is a feature, not a bug, of the intended design, both within Seattle and within the other subareas. But yes, it is infuriating that ST seems to be abandoning one of the major advantages of LRV, at-grade operations and level crossing.

  21. Once the buses were banned from the DSTT, it seemed cavernously empty in there to me. Infrequent trains, hardly anybody waiting. These days it’s practically post-apocalyptic. One of the main reasons I voted against ST3 was that the mainstay was this second tunnel through downtown, virtually adjacent to the existing tunnel. What city has done that? I always thought if there was going to be a second tunnel it should actually go somewhere new.

    What’s the rush on deciding how to proceed with ST3 anyway? Heck, it’s been five years since the vote and the route isn’t even determined yet. Why not wait until East Link and Lynnwood Link come on line and some experience is gained with twice as many trains in the DSTT?

    1. Two parallel tunnels don’t make much sense to me either. How about running a tunnel along Boren and continue elevated along Rainier? It could connect with Eastlink at Judkins and potentially run through the Beacon Hill tunnel back towards West Seattle (or Duwamish Bypass).

    2. Underground rail requires long-term planning because of the ten-year construction time. Kicking buses out of the tunnel improved Link’s reliability and speed, because buses get thrown off-schedule by traffic congestion which leads to bus bunching, buses take longer to load wheelchairs, and internal-combustion vehicles break down more often than wired electric vehicles.

      The timing of the tunnel closure was driven by two things: Connect 2020 installed a maintenance turn track at Intl Dist that’s incompatible with buses, and the county wanted to get on with convention center construction that eliminated the northern bus entrance. (The reason for that was to get revenue from large conventions that the current convention center is too full for.)

      The pandemic rose right when Connect 2020 ended, and soon ridership plummeted as offices were closed and officials urged transit was for “essential trips only”.

      The 550 lost ridership for several reasons: losing the tunnel made it slower, losing the South Bellevue P&R made it infeasible for P&R drivers, and losing the Rainier freeway station made it infeasible for southeast Seattle. All these will come back with East Link.

      The downtown tunnel is now supporting Northgate riders again, and new Lyunnwood/Everett riders who previously didn’t use the tunnel. It will soon see the return of Bellevue and Issaquah riders, and new Redmond and Federal Way, and Issaquah riders. And Tacoma riders if the 57x and 59x are truncated in Federal Way as in ST’s 2016 planning scenarios.

      The decisions are being driven not by current ridership but by projected ridership in the 2030s (for ST2) and 2040s (for ST3). The population will continue to increase, highways will become further congested and slower, and Seattle and the suburbs are continuing to build multistory housing near stations, and extending Stride and RapidRide lines that feed into Link and generate ridership.

    3. The Chicago Blue and Red lines are 2 blocks apart. Manhattan has A/C/E, 1, F/PATH, N/R/W, 6 all within 4 blocks of each other. Boston’s Orange and Green lines are parallel to one another in downtown.

      1. Hey, thanks for pointing those examples out. I guess I would respond that when Seattle gets a metro population that is similar to New York and Chicago (19 million and 9 million, compared to Seattle’s 4 million), parallel lines downtown might be in order. In Boston don’t the Orange and Green lines share the same tunnel? Which is what some are advocating here.

        The metro I am most familiar with after Seattle’s is London’s, which has many lines that seem to go everywhere, in a city not hugely more dense than Seattle. The Tube tunnels are not particularly close to each other, but up to three lines sometimes share the same tunnels and stations.

      2. And San Francisco has two tunnels that are zero blocks apart! DC has 3 tunnels for 5 lines and has a need for an additional tunnel downtown (or at least another crossing of the river) to reduce sharing of tunnels to boost overall frequency and capacity. DC’s various downtown plans are very analogous to DSTT2 – there are some modest improvements with new downtown stations (e.g. a new station in Georgetown), but the primary value is in boosting downtown capacity and boosting frequency outside the core.

        Green is LRV and Orange is heavy rail, so I don’t think they could share a tunnel?

  22. I haven’t commented on this thread because I live in the East King Co. subarea, so the only issue for us is to cap our contribution to DSTT2 at $275 million, which should be pretty easy since three of the other subareas don’t have $275 million for DSTT2, let alone more than $275 million.

    For about two years I have been posting two things: 1. N. King Co. cannot afford DSTT2 and WSBLE with ST 3 revenue, even with the “realignment” that extends project completion along with taxes so it really raises no new revenue after inflation (which smarter folks than I on the eastside raised in 2016); and 2. even before the pandemic ST would never realize a 40% farebox recovery rate and would need an operations levy.

    Of course now with the pandemic and WFH these issues are only exacerbated and accelerated because the pandemic has shifted so much sales tax and business revenue outside Seattle and the N. King Co. subarea, and the peak commuter who tends to pay 100% of their fare since someone else is paying has disappeared, maybe forever.

    Tom Terrific has a point: so far on this thread there are about 100 different “value” options to WSBLE, which dilutes any kind of input. But the reality is it doesn’t matter because the N. King Co. subarea does not have the revenue for any of the alternatives. Some complain transit riders should be a stakeholder. The reason they are not is because they have no money, and in effect are getting a huge transit gift from the 90% of the citizens who don’t use transit, at a cost that has become obscene.

    What I do think is a mistake on this thread however is to assume ST is filled with idiots who don’t know about the capital revenue shortfall in ST 3 and WSBLE, or the operations shortfall. They knew all of this in 2016, but passing ST 3 to complete ST 2 was existential, so ST made some stuff up about project costs, general fund revenue, and operations revenue to pass ST 3, probably hoping for a ST 4 and not a pandemic. Hey, it worked. ST 3 passed.

    So why would ST/Board choose a DSTT2 configuration and design it knows it cannot afford and is the most expensive? Because ST knows none of the alternatives or “value engineering” are affordable.

    And because it is the only politically viable option. If you are the Board do you want a decade of litigation and political fights with Seattle, the downtown merchants, Ballard and West Seattle over “value engineering” alternatives that N. King Co. cannot afford, or do you want to adopt the design everyone wants but is unaffordable.

    The answer to that question (not unlike the BBB bill that came out of the house from a pro like Pelosi) is who does the Board want to give the bad news to the “stakeholders”. They want the voters — or someone else — to give the bad news, probably by rejecting ST 4.

    At the same time what does ST do about operational costs? Over and over I have raised this issue on this blog, only to have folks post no, there is plenty of operational revenue, or there will be TOD’s up and down I-5, or we will get rid of freeways and cars, we will make folks ride light rail if it kills us.

    If there is one thing Rogoff and ST are good at it is controlling public relations. That is why recently The Seattle Times had two articles: One basically adopting the very deep tunnel as the preferred alternative; and 2. Rogoff’s mea culpa that ST’s operational fund/model is inadequate, which he blamed on transit riders not honoring their fare and contribution, and the pandemic. Apparently non fare paying ridership is up to 30% (which might explain why other systems use a turnstile system).

    It was the Times however that pointed out that ST never came close to a 40% farebox recovery rate pre-pandemic, and the operations hole will only get more acute with the opening of East Link, Federal Way Link, and Lynnwood Link, let alone feeder bus systems.

    So my guess is the Board, seeing the post-pandemic future, knows two things: 1. WSBLE will never be affordable for N. King Co.; and 2. there will never be the ridership post pandemic to validate this crazy light rail system, or to afford operations.

    The loss of the commuter, who pays full fare and close to a 100% fare paying percentage especially when ST has no effective way to enforce fares, is not surmountable. Light rail was always based upon unrealistic project costs, general fund revenue estimates, and more than anything population increases (Seattle grew 0.7% last year), and net ridership increases over simply transferring riders from buses to rail which translated into farebox recovery.

    If anything, it is the absence of the commuter transit rider at the table that is most troubling, because they never liked riding transit, or commuting to work, so don’t care about transit anymore, but underpin the economic viability of Link, which is why Link is SO commuter oriented. Without the peak traffic congestion transit makes very little sense for someone with a car, and without the commuter ORCA card the system’s operation funding model does not pan out.

    There is only one alternative that is politically and financially viable: mothball WSBLE and DSTT2. Go to buses (and I wouldn’t get my hopes up West Seattle will agree to converting precious bridge lanes to transit). Ridership will never support Link post-pandemic. It was an unwise system financially to begin with, at the very end of a gilded age for this kind of public expenditure in this area.

    At this point, and at every point, the real issue is how to afford future operations for a light rail system with five different subareas when future ridership will be around 50% of pre-pandemic estimates used to sell the levies, and a ST 4 is very, very unlikely, otherwise there would have been a “realignment”.

    Forget about WSBLE and DSTT2: no alternative is affordable. Instead focus on how to fill in the 50% hole in operations revenue for the existing system, when I really doubt an operations levy will pass if the commuter is not interested in transit.

    Sometimes you have to rip the band aid off, although politicians hate to do that. Hopefully the Board will find a replacement for Rogoff outside ST who will come in and ask WTF were you people thinking, WSBLE and DSTT2 are not only unaffordable but unknowable, and we have a huge hole for future operations that needs backfilling. Rogoff will simply tell him we could have had this discussion in 2016 if ST 3 failed, so I did my job.

    But like I began this post, all I care about is my subarea is limited to $275 million for WSBLE and DSTT2 as set forth in ST 3, when I know the three other subareas don’t have the $275 million they are obligated to contribute to DSTT2. The great irony is I am sure N. King Co. thought it was being clever obligating the four other subareas to pay 1/2 of DSTT2 when DSTT1 has plenty of capacity for their lines, except relying on S. King Co., Pierce Co. and SnoCo to come up with $275 million was probably optimistic after N. King Co. paid the lion’s share to fun the spine to these areas.

    1. Think of it glass half full. They have $10 billion to spend on WSBLE and they are sitting on a completed DEIS after a half decade of scoping, engineering, etc. $10 billion buys you a ton of transit and at this point it’s a no brainer to get the EIS approved and get shovels in the ground. Politically, that HAS to happen. That was the whole point of realignment — getting a plan to move forward.

      Everyone knows the current preferred alternative is bad and is going to change. There are decent alternatives that can be swapped in. I have no doubt that construction in West Seattle will start on schedule in 2026. The Avalon/Fauntleroy/Alaska corridor in West Seattle is the #1 spot for TOD along WSBLE. Assuming the city gets its act together with upzoning, a Roosevelt-like transformation will occur there and it will look good for everyone involved. Link ridership is almost irrelevant compared with getting that neighborhood transformation.

      At that point they are still going to be sitting on $6-7 billion to construct DSTT2. I’m fine with the deep bore. If peak ridership is permanently reduced, then it’s not a problem to have elevators. All of the activity is going to be at Westlake. Who cares if Midtown has elevators, that station is never going to generate huge ridership.

      Ballard is the real loser in all of this. It’s not really that ‘WSBLE is unaffordable’. It’s the Smith Cove to Ballard section that is unaffordable. Or maybe it’s just that West Seattle has the right people advocating to be first in line.

      1. Link ridership is almost irrelevant compared with getting that neighborhood transformation.

        Ah, so the whole point of the light rail system is to develop neighborhoods. Here I thought it was to provide the city with a good public transportation system. Fair enough. The thing is, that development will happen regardless of public transit. As soon as you change the zoning, the neighborhood changes. You can see the growth in various neighborhoods around the city, and it has nothing to do with light rail. At most you can credit the light rail for altering the zoning, but that could — and has — happened by itself. Yes, Roosevelt is booming. So is Stone Way. Ballard is booming, but far away from the station. Fremont has seen a lot of development (and would see more if they allowed it). With few exceptions (lead by reluctant businesses who refuse to sell) areas are growing simply because they are allowed to grow. Public transportation has nothing to do with it.

        Even in West Seattle I don’t see this as being pro-development. By the particular stations, this is better, but not that much better. Frequency won’t be better (in some cases it will be worse) and the buses are plenty fast to downtown. The train will make it a lot easier to get to SoDo, but that is not much of a destination. If you had planned on transferring to a train heading north (to get to Capitol Hill, UW, etc.) then yes, this is better — for those that live close to the station.

        But for people close to bus lines in West Seattle, public transportation will be worse. A lot of people will have to transfer to a train that runs every ten minutes. For those who live in West Seattle, but not close to the stations, at best this is a symbolic improvement (Yay! We have light rail!). The people who ignore transit will continue to ignore it, and the handful who do use it will endure as they always have. The good news is, those other places in West Seattle don’t need an improvement in transit to see growth — they only need a change in zoning.

      2. All of your counter examples are desirable neighborhoods already well served by transit. Public transit absolutely supports development; it’s literally in the name “streetcar suburbs.” If KCM just cancelled all bus service to Ballard, rents would fall and development would likely dry up (or land prices would drop sufficiently to sustain development at lower rents). We saw that happen with rents in WS when the high bridge was closed.

        There are many neighborhoods in the region that have ambitious midrise or high-rise zoning, lack the market rents to support private development at this time, and expect to have robust development once rents rise in response to dramatically improved transit access. Those neighborhoods are not in Seattle.

    2. I don’t see WSBLE and ST3 getting cancelled. There are too many other ways that money is going to get spent on projects desired by other subareas.

      I see merely that the question is getting flipped from “how do we pay for the 2016 promised scheme?” to “how close to the 2016 promise can ST get given limited financial options?”

      The posters here are just thinking ahead. They understand that the shortfalls are not recoverable.

      I’m also not seeing any local elected official saying “we will build ST3 just like we planned in 2016 no matter the cost”. Instead, there is radio silence right now. With Rogoff leaving, there isn’t even a visible agency advocate to promote ST3 like Joni Earl did with Sound Moves. It seems to be running on sheer inertia here in 2022.

      New York is taking decades to get the Second Ave subway. The original DEIS for BART south of Fremont was completed in 1990 — with construction not really starting until 2011. We collectively think that things should work much faster like they did for ST2 and choose to ignore that the ST2 underground segments are a much smaller part of the cost and timeline. The more reasonable truth is to admit that there are really just two choices at this point for WSBLE — delay+add$ or delay+redesign.

      So I think delay is the next step here rather than abandonment.

      I have some skepticism that the voters will forgive the ST3 deceptions about costs and wild elevation differences and authorize substantially more funding. I could be wrong. Still a “no” campaign could be pretty brutal with attack ads that ST overpromises.

      I don’t think delay is a bad thing although others do. I see 2025 as the defining year of figuring out urban transit’s role in a future Puget Sound region. ST2 extensions will be open. The pandemic will be over. The impact of increasingly cheaper electric cars will be understood. Probably some sort of economic downturn will occur making the shortfall worse.

      There comes a point when a community quits looking at an investment that they can’t afford (“I will just extend my credit card payments.”) and turns their attention to something similar and cheaper that can meet most of their needs anyway. I’m not sure when or how that will occur but I think it will.

      1. I think “sheer inertia” is both accurate and wildly understates the good work the agency is doing. The vast majority of staff is focused on either operating the existing system or trying to get ST2 & ST3 early projects (Stride, 130th) completed. The number of people working on the WSBLE EIS is a pretty small group. The problem is that until there is clear direction given by the Board, the planning staff’s marching orders are to get the WSBLE EIS completed as quickly as possible; Rogoff was pretty clear that his takeaway from the ST3 public outreach was the priority was simply to get through the EIS ASAP, and no one has given guidance otherwise since then.

    3. “The Avalon/Fauntleroy/Alaska corridor in West Seattle is the #1 spot for TOD along WSBLE. Assuming the city gets its act together with upzoning, a Roosevelt-like transformation will occur there and it will look good for everyone involved. Link ridership is almost irrelevant compared with getting that neighborhood transformation.”

      Ballard is a larger urban village than West Seattle. West Seattle fiercely opposes upzones. It considers itself entitled to suburban-like density, and the city government complies with it because VIP politicans live in that area. The Junction area has already been upzoned modestly and is unlikely to go further. So you’re assuming something that’s unlikely to happen.

      “Everyone knows the current preferred alternative is bad and is going to change.”

      Some of the representative alignment is better than all the other alternatives.
      The RA has the Ballard station on 15th instead of 14th, closer to the pedestrian concentration between 17th and 24th. It has the International District 5th Avenue shallow station, which is both cheaper, higher-ridership, and more convenient for shoppers than the other alternatives. I don’t know as much about the West Seattle alternatives but the RA seems like one of the better ones. ST doesn’t have to do anything for those other than revert to the RA in the ballot measure. Whether it does or not is another question.

      “It’s the Smith Cove to Ballard section that is unaffordable”

      What? ST has never raised that segment as particularly risky or costly, so what are you basing that on?

      “Or maybe it’s just that West Seattle has the right people advocating to be first in line.”

      That’s obvious. Dow Constantine and other VIP politicians live in West Seattle. West Seattle considers itself entitled because it has so many suburban-like middle-class People Like Us. They point to the West Seattle Bridge as a bottleneck and a single point of failure, and that the few egresses to the south are similar bottlenecks. They want a high-capacity corridor to downtown that’s not dependent on the bridge or going around. They say they’re the only Seattle neighborhood that’s so isolated and has so few egresses. And DSTT2 is expensive and will take many years to complete. All this is why ST scheduled the West Seattle stub first. The stub to SODO is ridiculously useless though. The reason it’s scheduled first appears to be to show West Seattle an early deliverable in deference to its entitlement.

      “New York is taking decades to get the Second Ave subway.”

      That’s not a benchmark; it’s an abysmal failure. Pedestrian Observations has written volumes about how US rail projects are much more expensive and take longer than peer countries. It all has to do with US regulations, labor expectations, contracting process, outsourcing design and separating it from construction, EIS delays, and management unwilling to look at what works in other countries, especially non-English speaking countries. New York stopped expanding the subway in the 1930s and didn’t start again until decades later. Peer countries either continued expanding subways and regional rail throughout or started doing so again in the 1970s. London and Moscow start another project as soon as one is completed.

      “I don’t think delay is a bad thing although others do.”

      If we can just get ST2 done, the near-term expansions (240th-320th in Federal Way, Redmond Tech-Redmond downtown), and the three Stride lines, we’ll be doing well. We can live with that for a while.

      1. I’d add a pair of dedicated transit bridges to that. One for Ballard and one for West Seattle. Make them convertible to light rail, someday. Getting the C and D out of regular traffic for those short segments would really help them, especially at the Ballard drawbridge.

      2. @Glenn — That is one of the drawbacks to the current approach. It wouldn’t cost that much to improve transit along the West Seattle to Ballard corridor. Some of it has been done, or will be done soon. Buses from West Seattle will encounter a lot less traffic downtown; Uptown will have more bus lanes. But some projects aren’t likely to happen as long as light rail is expected to get here. ST was supposed to improve the C and D (as part of their early deliverables) but even that got delayed eventually, which means it will never happen.

        This means a bus stop underneath Dravus (which would allow the bus to avoid exiting 15th) probably isn’t going to happen. Likewise, building an extra lane so that buses can get up to the Ballard bridge when it is up (the way they will for the Montlake Bridge) is unlikely. So too is a bypass so that buses can avoid the congestion from the West Seattle freeway to SR 99. These aren’t horribly expensive projects, but they aren’t cheap either — which means it is unlikely they will happen unless there is a big change in attitude.

        Of course if the city gets an agreement with the feds to extend the no-opening time an extra hour, then that would help the Ballard Bridge problem. The other problems are not huge — I’m doubt they would make the top ten in terms of transit related bottlenecks.

      3. Yeah I’ve been perplexed why the “C&D Capital Improvements” project got functionally cancelled during realignment. Politically, 130th, Graham Street, and West Seattle all had their supporters, but there was never a Ballard advocate on the board. It’s a big chunk of money ($227M in 2014 dollars, minus contributions to Madison BRT), so it can do real improvements.

        But the project is still there, so if the Board decides to push pause on WSBLE, they have that legal authorization to throw million of dollars at D & C improvements and message these investments to the public as mitigation for delaying WSBLE.

        If ST finished the WSBLE EIS and picked 15th median running as the preferred alternative & aligned with SDOT on rail bridge over 15th & car bridge over 14th, I think ST could build the new ship canal bridge and rebuild Dravus bridge first, initially for bus operations, and then convert both to rail once the 2nd downtown tunnel is complete.

        I could even see SDOT get behind this approach if they corralled ST into paying for the car bridge over 14th as ‘mitigation’ for taking all that premium 15th Ave ROW. The politicians could then go to the voters and say, “look, we are going to use some of the initial WSBLE money to rebuild 15th in Interbay and Ballard such that the D will be dramatically improved and you’ll get a permanent replacement for the car bridge. This will be done by 2032. Then once we’ve raised enough money for the new downtown tunnel, we’ll be able to convert all that bus ROW into Link ROW, say around 2045.” And then to WS voters, say, “here’s $50M of early deliverables to make some big improvements to the C (say, a new ramp to bypass traffic), and then we commit to coordinate with SDOT to ensure that the WS Link bridge opens before the WS high bridge is decommissioned, target 2045.”

        Ballard & WS get better transit now and maintain a clear path to Link. The downtown tunnel is deferred for a better plan, and the ST3 spine can continue on schedule for Snohomish & Pierce. There’s a permanent solution for the Ballard ship canal crossing, and WS maintains its replacement solution for when the WS High Bridge reaches end of life.

      4. My last point about WS Bridge end of life I think is important. I roll my eyes at all the suggestion that WS just ditch Link and run buses in perpetuity; it’s a great idea for the next 20 years and a grossly inadequate idea for the next 40 years. If Seattle doesn’t spend $1B of its ST money on a rail bridge over the Duwamish, it’s going to need to spend even more than that on a new car bridge over the Duwamish to maintain that high quality bus service to downtown Seattle. Descoping the WS Link bridge doesn’t save Seattle any money, it just shifts the long term liability out of the funded ST3 Plan and into an unfunded SDOT project for the next generation to solve.

      5. Ballard: I would hate to use transit money on a Ballard car bridge, but this would be the perfect project to spend the Fed’s infrastructure money on and building this bridge on 14th is much easier than trying to replace the 15th Ave bridge in place! In the meantime we can run Link to Interbay and the buses can use the old or new car bridge until the Link bridge is done.
        WS: Why not just use a gondola?!? It’s much better suited to cross the water and go up hills and costs a fraction, still fully grade separated and not dependent on any car bridges which eventually needs replacement – therefore creates welcome redundancy. We could use the savings to accelerate Ballard or a Duwamish bypass.

  23. While I’m not a fan of two-car stations for capacity reasons, the largest cost of an underground line is apparently the stations. So if automated two-car trains are a political possibility (as if), what about excavating station vaults for 4-car trains but finishing only shorter two-car station? The initial segment had a preliminary plan with a station vault at Beacon Hill for a future potential station. This would be similar but with half-stations.

    1. I don’t think that helps – the cost of the station is primarily the excavation and creation of the station box. The savings of building a 2-car station within a 4-car station box are likely minimal in the context of WSBLE’s budget issues.

      The primary capital savings of automated 2-car trains is only needing to create 2-car station boxes.

    2. The idea with a two-car train is that it runs twice as often. It would have just as much capacity the day it opened.

      In the long term, it is theoretically possible that you would have crowding on that section. I seriously doubt it. There just aren’t that many stations. As with the rest of the system, any potential crowding can be dealt with quite easily, by simply restarting the express buses that we assume will be truncated.

  24. AJ, I believe this is the pdf that you are looking for:

    Please note that “peak volumes” are for a three-hour period of 3-6 PM. I’d guess about 40% are the actual peak hour.

    If you look at page 1, you will see that DSTT1 and DSTT2 northbound volumes are 28,700 added together. There is forecasted volumes of 30,100 between Westlake and Capitol Hill.

    The DSTT capacity problem appears to be southbound. North of ID, there are 41.100 PM peak riders headed southbound if both tunnel volumes are added together. Using Tom’s assumption of 600 rider per train and 40 percent during the peak hour, that’s 27 trains. Using 700 per train and 40 percent, that’s 23.5 trains. This suggests that the system would approach capacity and that

    Now it’s been pointed out elsewhere that eliminating some train cabs could add more capacity to a train. A whole new fleet for the 1 Line wouldn’t cost $4-5B and it may be possible to just remove one of the two cabs on each car.

    One other curious item: 3 Line between Midtown and ID forecasts 16,900 riders. Using the 600 per train and 40 percent, that’s 11.2 trains needed and the 3 Line is limited to 10 trains in the MLK median. So, it appears that this too will be a problem even if DSTT2 is built.

    Finally, I think it is worth pointing out that surface streetcars could offload these possibly overcrowded situations. It’s also worth mentioning that the forecasts did not assume the extra time to get to deep stations — and that some riders would move to a bus or maybe to the CCC if it gets built. So maybe building the CCC or a surface tram on Third Ave instead of the DSTT2 would be enough to solve the forecasted capacity problem.

    1. Yes, thank you that’s the PDF I was referring to. That analysis is central to my argument in favor of a 2nd tunnel. At some point (in that analysis 2040, maybe it won’t be until 2050 or 2060), throughput demand on the single DSTT will exceed its carrying capacity, even after fiddling on the margin with different vehicles design or boosting tph. A single tunnel with 3 lines is totally sufficient for 2030 demand, but I think that approach would be a very short sighted decision in the context of a generational investment that should handle growth through 2060 and beyond.

      If Seattle was in a slow growth region like the Great Lakes, then this would be less of a problem, but given the region’s geography, climate, recent economic history, and vast potential for infill growth within the UGA, I believe we need to plan for a region with millions more people. ST3 from the beginning only made sense in the context of Seattle & Region meeting & exceeding their long range growth forecasts.

      1. At some point (in that analysis 2040, maybe it won’t be until 2050 or 2060), throughput demand on the single DSTT will exceed its carrying capacity

        That is a wild assumption. To begin with, it was obvious *before* the pandemic that office work is moving towards a less rigid schedule. A 9-5, Monday through Friday schedule just doesn’t work for everyone. There will always be a peak (in the evening) but it won’t be as peaky as it is now. People will work a more flexible schedule. Some will sleep in, but arrive in time for the 10:00 AM meeting. Others will work an early schedule so they can be there for their kids in the afternoon. Meanwhile, there will be plenty of people who work from home most of the time, and only go into the office occasionally. This doesn’t mean the end of big office buildings, but it means that rush hour (to downtown anyway) won’t be as crowded.

        I have no idea what other assumptions go into those estimates, but bus truncations are likely one of them. Even today, with stations connecting the urban core, and a large percentage of riders walking to the station, ridership is being driven by bus truncations. Start up the 41 again, and peak ridership from Northgate plummets. Run the old 71/72/73, and suddenly the UW stations aren’t that busy. Snohomish County county is the same way. The buses are supplying the trains with much of its ridership. If this becomes a problem because of crowding, then it will stop (much to the delight of many riders).

        These trains just don’t serve that many walk-up riders. There aren’t that many stations, and many of them are in areas with little potential for walk-up riders (because they are close to the freeway). The trains are also big — they can handle 800 at standard capacity (with current seat arrangement) and 1,000 for a “crush load”*.

        Look, I get it. We all want to future-proof the system. The lack of tracks from the UW heading towards Ballard is a big loss. But there is a phrase in software engineering: You aren’t gonna need it. It is so common it has its acronym (YAGNI). It is very easy to just assume that every potential problem needs to be solved long before you’ve proven it is a problem, or explored alternative solutions. If this was a cheap addition, I would be fully in support of it (why not?). But it isn’t. This will cost a bunch of money now — there is little value in building it years before it will ever be needed (if it is ever needed at all).

        Nor is it likely to be our biggest need. It is one thing to assume the increased capacity might someday add value. It is another that it is the best thing to spend money on. There are agencies throughout the world that have capacity crunches, and I don’t know of any that are spending this much money building a secondary tunnel, right next to the first. They make improvements, to run the trains more often, or so they can carry more people (both of which we could do). Or they truncate the trains instead of interlining them, to improve reliability. More often, they just run a new line, somewhere else.

        The fundamental design of the downtown tunnel is flawed, since it doesn’t add any value other than increasing capacity. Not only are the stations too close to the existing ones, but they are *worse* than the existing ones. Riders from Rainier Valley who get off inside the existing tunnel will be worse off in the future. No one in their right mind will transfer from a trip on the existing tunnel to the new one. If we really do need to build a new tunnel to deal with crowding, it should include a stop on First Hill.

        But again, that is unlikely to be the biggest need in our system. Travel demand isn’t likely to be limited to growth along these corridors. We have other needs within the system that are likely to outweigh this in the future, by a wide margin. It should be clear to anyone who is paying attention that ST3 planning was not driven by need or value. The mindset that resulted in West Seattle and Issaquah-South Kirkland rail (over a Ballard to UW subway, or BRT on the CKC) also resulted in this additional downtown tunnel, with its stations so close to the first. This is just one small part of a poorly designed project. It just happens to be the easiest to fix, since it wouldn’t require disappointing voters. In fact, just the opposite.


      2. It is indeed wildly speculative, but I’m deeply skeptical that post-pandemic ridership will be less peaky. The peaks may no longer be 5 days a week, but the system still needs to be designed for peak ridership, even if that peak is only 2 or 3 days a week, not 5.

        “The fundamental design of the downtown tunnel is flawed, since it doesn’t add any value other than increasing capacity. ” I mean, yeah sure if you think a Ballard spur is plausible. But I’m arguing that a Ballard spur is not feasible, and therefore the increased capacity unlocks the ability to run a subway through Denny, SLU, and LQA. Whether a spur is feasible is the billion dollar question.

  25. This is a fun stream.

    I will ponder future comments to the WSBLE EIS. One focus will be station placement along the alignments. I agree with Martin that the minutes matter. The stations have to be along the alignments. The stations and alignments should connect with pedestrians and buses. So, the bus connections on 35th Avenue SW and Delridge Way SW are key. Should not the stations straddle the arterials so that the station infrastructure takes intending riders across the busy arterials? This was not done in the past and near future and riders have complained (e.g., Airport, Mt. Baker, NE 130th Street). The preferred stations in Interbay actually seem to hide from pedestrians and make bus pathways indirect. What would TransLink do? Would TransLink bury the station deep below 5th Avenue South?

    What is the ST3 advantage of opening the West Seattle stub line before the Ballard line? The phasing seems silly. Note that Seattle could provide transit priority to an awesome bus network in West Seattle with red paint and/or tolling.

    In earlier decades, the FTA wanted a TSM alternative. This comment stream made me think of that. The No Build assumes the ST3 stream of funding is not used at all. But what could it do? Suppose the dumb CCC Streetcar was canceled and SDOT used red paint and signage to provide buses priority on 1st Avenue. Several routes could be shifted to 1st Avenue from 3rd Avenue; this would allow other routes to shift to 3rd Avenue from 2nd and 4th avenues. The ST2 Link extensions will allow suburban bus routes to be truncated outside Seattle. ST3 funds could reduce headways on lines C, D, and E; it could elevate routes 7, 40, 44, and 48 to RR; all are delayed now. Further, it could elevate routes 5-21, 62, and 8 to RR. The E Line could be converted to double articulated electric trolley buses. With a few more buses with left side doors, another route could use the G Line pathway between 1st Avenue and Broadway and serve the Capitol Hill station via Broadway. 1st Avenue might have very short waits; how about routes 7, 14, 36, 124-24-33, 131-132. Yesler Way could have routes 27-28, 40, and 62. Routes 101 and 150 could be shifted to 3rd Avenue. 2nd and 4th avenues could have PBL and cars. ST3 capital could improve bus flow on the C, D, E, 5. 40, 44, 48, 62 lines. Route 5 could meet Link at NE 130th Street. Some Seattle couplets could be revised for better transit flow and shorter transfer walks: two-way transit on Pine Street, Queen Anne, and 12th Avenue NE; PBL on Pike Street, 1st Avenue North, and Roosevelt Way NE.

    The notion of a Ballard line being placed into the DSTT1 seems far fetched. The Pine Street stub tunnel was constructed in 2005-07 when the DSTT was closed. But as RossB asserts, it is for the engineering experts. Will the cost and difficult of DSTT2 overwhelm the fiscal capacity? ST can just delay until the fiscal stream is sufficient.

    The No Build data seems to be from 2019; that was a weak transit year with downtown congestion, buses booted from the DSTT, the AWV closed and the deep bore not yet open. It was before three Metro restructures in 2020 and 2021. But the No Build is just a planning artifice.

    One commenter asserted the DSTT2 is a done deal politically. It depends on the depth of the fiscal crisis. Sound Move was changed significantly.

    1. eddiew, of course the stations “should straddle the arterials”, but for some reason ST wants stations off to the side of the street. That guarantees that one direction of everyone’s trip will require crossing a major arterial. Blecch!

      But your passion to move a bunch of buses to First Avenue is going to run right into “THE HELL YOU WILL!” from the riders of those buses. Downtown is shifting up the hill, not down. It’s already starting to splash over the freeway to lower First Hill.

      Yes, there are lots of residences being built between First and Alaskan Way, so there should certainly be frequent service along First to take people to activities in Belltown, the Stadiums, SLU and Lower Queen Anne, preferably in bus lanes. But it will be local service — “downtown circulators” — and should not dump long-distance riders headed to the office core at the bottom of a five block 18-degree climb. People using the 13X’s coming to Seattle are (mostly) not headed for the Pike Place Market.

      First Avenue is a distinct market, not “just another pathway” for buses.

      Many of your other ideas make sense; this one doesn’t.

      1. I should have prefixed the “First Avenue is…” sentence with “Because of the topography of downtown Seattle, “

      2. Re: Shifting buses to First:

        The 10, 11 and 49 end abruptly at Pike/Pine. They don’t travel the main avenues at all. The new RapidRide G (the only BRT in the state) will go on Madison/Spring downtown, and loop around First. The 106 ends at 5th and Jackson. OMG, what do people do, if they want to go to the other end of downtown! They transfer, or walk.

        There are numerous buses that don’t travel the length of downtown. Many intersect somewhere in the middle, and end. Others pick a direction, and head that way. This means that if you want to get from the 2, 3, 4 or 13 to the International District, or the 70 up to the north end of Belltown, you will have to transfer, or walk. The Community Transit express buses run down Second, and up Fourth. So do a few Metro buses. This means every trip involves at least two blocks of steep walking.

        The biggest problem with our system, by far, is that the buses don’t run often enough. Spending money on a circulator would be silly. Put it this way: Ask a rider of the 27 if they would be OK with running their bus on First Avenue, but instead of the bus running every half hour, run every fifteen minutes. They would be thrilled.

        There is no perfect downtown corridor that will please everyone (the 2nd/4th combination guarantees you please no one). Buses could certainly run on First without creating a huge burden for riders; just like they could (and do) run on Fifth.

      3. Tom, thank you. I want to contrast the transit mobility that could be provided without the $100 million plus of the CCC Streetcar that would only provide 12 trips per hour per direction. Per RossB, some of the routes shifted to 1st Avenue would also serve South Jackson Street or Pine Street. There is a reasonable transfer with the DSTT via University Street. The G Line will serve Madison-Spring at midtown. Trust the grid.

        The agencies could seek the Goldilocks level of shift to 1st Avenue: just enough for short waits and to allow capacity on 3rd Avenue for the remaining trunk routes (e.g., routes 101, 150) and growth on lines C, D, E, H, and Route 40. In 2019, there were 130 trips per hour per direction on 3rd Avenue.

        Between 1998 and 2011, routes 15, 18, 21, 22, and 56 were on 1st Avenue, along with routes 10 and 12. All performed well.

        The one-way avenues on the hill have their own awkward features.

  26. Can’t we go back in time to a prior strategy? It would be interesting to perform and compare cost analysis with what we’ve learned from ST3.

    “The monorail project initially attracted two bids led by Hitachi and Bombardier, but both pulled out in April 2004 over cost concerns and the availability of local contractors.[219] The project was stymied by tax revenue that was lower than expected and design changes to keep construction costs within the proposed budget and open by 2009—a two-year delay from the original plan”

  27. Looking more closely at Ballard, it looks like tunneling is surprisingly feasible. 14th is obviously a terrible choice, but I do like the shallower 65′ station depth. Is there any reason 15th needs to be 20′ deeper? I get that there will be a pedestrian pathway, but surely that doesn’t necessitate going deeper. It seems that some creative engineering could raise the station depth.

    I would really like to have seen 20th Ave NW as an alternative studied during the DEIS phase. The additional cost to shift from 14th or 15th to 20th would not be excessive, and would most definitely be worth the far superior station location.

    1. 1500 NW Market St. Has a underground parking lot, two levels deep, possible source of obstruction. Not sure if 1448 NW Market St. has one or not, but its also a brand new building.

      Of course the tunnel could go straight up 15th St, assuming there are no sewage pipes in the way.

  28. Tisgwm, I guess I would ask why “[W]ithout DSTT2 the Everett and Tacoma extensions may not be able to proceed, but that may be a feature rather than a bug.” (which is not your quote). What evidence is there for this statement?

    How many Tacoma and Everett residents will travel to and through Seattle in the future on light rail, especially during peak hours, compared to those who do it today on buses or bus/rail. My guess is that ridership number will decline — by around 20% to 40% — post pandemic from pre-pandemic numbers.

    I don’t see the Link extensions to Everett and Tacoma (Dome) increasing ridership on light rail to and through Seattle over a bus intercept to Northgate or 145th before Everett Link opens. Those are very long commutes on light rail. Before then riders will continue to take a bus to light rail to downtown Seattle.

    The reality — and one I think Rogoff was driving at — is current ridership on light rail — including bus intercepts — will likely be the ridership post pandemic, which will determine farebox recovery. The work commuter just isn’t going to come rushing back in pre-pandemic numbers, and those riders are 100% fare paying riders. Without the peak commuter there are not capacity issues. Not even close.

    We spent years on Mercer Island chasing ST’s dishonest ridership estimates for East Link, until we — and Metro — learned they were all lies to inflate farebox recovery rates to lower general tax increases in ST 2 and 3 in order to sell those levies, and to convince the eastside subarea to contribute to DSTT2 to meet our “capacity” based on ST’s ridership estimates, which were lies before the pandemic, and today are just fantasy.

    The real issue Rogoff raised in the Seattle Times article was not too many Link riders, or capacity, but too few riders for such a long and expensive system, and the operation deficits from reduced farebox recovery. I would guess Rogoff would be thrilled if he thought the future was too many riders and too little capacity, and all that farebox recovery, but that is not the future.

    1. “What evidence is there for this statement?”
      You’ll have to ask commenter Mike Orr that question. I’m not convinced that DSTT2 is necessary for capacity reasons either.

      The purpose of my comment above (that you’ve replied to here) is to suggest that any proposal NOT to run light rail to Everett proper will be met with strong opposition from Snohomish County leaders and representatives.

    2. Some comments about the farebox recovery problem recently out on the table:

      1. This is not news. Most systems across the US saw this before the pandemic. ST was spared because U-Link opened, which is probably the most productive subway segment that ST could build. On top of that, restructuring and Metro buses leaving the DSTT put more riders on Link. There was this odd local belief that Seattle was different in growing transit ridership when what really happened was a huge decrease in transit travel time and less Metro competition making the ST system more productive.

      2. Now that he is leaving, Rogoff is no longer having to be complicit in the politics of the Board. Politically, the Board seems to me to be fixated on two things — a regional map showing more destinations and a design that isn’t considered a nuisance to nearby property interests. I think they trusted Rogoff to take care of the “pesky details” like capital cost estimating and farebox revovery. The problem I’ve seen all along is that Rogoff was never a builder nor an cost-effective operations guru; he was an FTA management type whose major strength was understanding FTA. In fact, bring fiscally prudent now is not only presenting a more honest Rogoff, but it’s stemming the eventual blame he is going to get when the next ST head takes over. In Rogoff’s defense, I will note that he was “late to the ST3 party” and the venue, music and food spread were already mostly decided when he arrived plus he had little credibility to challenge things during that initial time.

      3. The core problem all along is too much Board fixating on maps and not enough on productivity. No one in 2016 dated ask to do a cross-comparison of the ST3 projects nor did they probe about unrealistic cost estimates and low contingencies. The did not probe into vertical elevation challenges. The number of expected riders was even not making much impact on the package assembly discussions. The most expensive per mile element of ST3 was always the DSTT2 running through LQA, SLU, Downtown and the ID — yet the prior studies never examined the ST3 alignment as a whole (Ballard and West Seattle were not combined for study until after the 2016 packaging, and the SLU+Interbay path was not an alternative in the earlier study .

      4. The farebox issue has ramifications with Federal grants along with operating budgeting . Fares are part of the required “local match”. There are some vocal free fare advocates that think that eliminating fares just takes the waving of a magic wand. Failure to achieve farebox recovery and make expansion decisions on associated productivity measures are good ways to get lower on the grant awards list. ST3 can tout six minute frequencies for Link all day — but at some point the time-consuming tails at the unproductive ends (as much as 20 or 30 minutes on some of these tails) will mean that not every train will run to the unproductive end once the pencils get sharpened a year or two after opening day.

      5. The problem with a long spine approach is that a line can become overcrowded in the middle and unproductive at the tails — all at the same time! This is why most long urban rail systems have generally had branches like SF Muni or Boston’s Green line or the NYC Subway or Portland’s MAX or DART. Even two of the three DC Metro subway tunnels have two lines through the city’s core. Single line short urban rail makes sense and branched long urban rail lines make sense for productivity much more than a single long line does.

      1. As a Tacoma rider, but not commuter, the 594 was a revelation when I moved here. A 45 minute 1-seat ride off-peak to and from sporting events or just an evening out in Seattle. It’s fantastic.

        If it turns into a 2 to 3 seat transfer and/or the estimated hour-plus meander through Rainier Valley, that is a serious downgrade. Sadly, I’ll likely be driving to SODO and hopping on the train, or riding my bike from there, if the 594 goes away.

      2. “Sadly, I’ll likely be driving to SODO and hopping on the train, or riding my bike from there, if the 594 goes away.”

        Maybe this is the real reason for a station and parking garage at Boeing Access Road: it will be the main park-and-ride intercept for people coming north on on I-5 who don’t want to drive into downtown.

        Let’s just disregard the fact that the number of parking spaces will be fixed, and that more often than not when there is some sort of event going on those spaces will be full – which reduces the utility of the station for everyone…

      3. There are also P&Rs at TIB, Angle Lake, KDM, and Federal Way. They all have the same problem a P&R at BAR would have: they’re at the south end of the south Link bottleneck, which is between International District and Rainier Beach where the train is slowed down by surface running, crossings, and going east to go west. North Link will be competitive with ST Express and Sounder to Lynnwood and Everett, but south Link won’t. Everything south of TIB has the same advantages as the north end (grade separation and relatively straight;l it’s the slowdown in Seattle that makes it less competitive. But there was an alternative, a Georgetown bypass route. That was in ST’s long-range plan in 2012, but it deleted it as not worthwhile, and none of the South King or Pierce boardmembers defended it.

      4. 1. Seattle and Houston were indeed unique. They both proved that if you make transit better, you can boost ridership.

        5. I agree that the size of our system makes branching ideal, and ST2 took advantage of this by branching at ID. ST3 doesn’t do this; there should be a branch somewhere in Snohomish, and then Ballard Link should have a branch somewhere, either downtown (to Fremont/Aurora) or in Ballard itself.

        However, I disagree on you comment about tails; yes, the most productive part will be in the core, but ST3 is well design in that all the termi are at regional destination, ending in dense* neighborhoods and not P&R as the suburban fringe.

        *or at least intended to be dense in the future, like around the Dome or Central Issaquah, or is clearly not a permanent terminus like S Kirkland.

      5. My choice would be 272nd if I needed to park. There won’t be the traffic like 320th. As the second station there will probably be a seat. Finally, the hill climb on I-5 up to 188th is pretty slow for a few hours each day, so skipping that bottleneck is advantageous.

        Federal Way Link opens in 2024 and BAR won’t open until 2032 or later. That’s eight years of trying different garages.

        Probably half of the BAR parking will be Renton residents., East King should probably chip in for the garage.

      6. If the tails don’t branch out, at least we should pick up as many riders along the way as we can. I’m still surprised Everett is making a detour to airport and Boeing but then go back to I-5 instead of having one more station along Everett Way/99 to pick up rides there and serve airport and the Boeing transit center more directly with an APM or gondola.
        There are many other opportunities for such links: Edmonds (incl. college), Shoreline college, Kent (when Sounder isn’t running), Southcenter/Sounder South, Factoria/Eastgate – all those are urban center which could bring additional ridership to the main line.

      7. East King Co. can’t chip in for a park and ride for Renton because our park and rides have been “extended” to meet the debt ceiling for WSBLE even though the subarea has tons of money.

        This concept in a strange irony is a combination of arguments Ross and I have made: expensive light rail is for urban cores. So you take folks to the outskirts — by bus or car — where they catch light rail because it is so expensive per mile and slow over long distances and then let light rail take them to — and around — the urban core.

        The only difference I have with Ross on this is I believe park and rides are a very effective and cost effective way to get folks to the edge of the urban core to catch light rail to the urban core because the driver pays for the car and driver (themself). That basically has been the first mike access on the Eastside for years.

        You spend your money on expensive rail and subways in the urban core because that is where it can compete. Outside the urban core light rail can never compete with a car or even a one seat bus, especially over long distances.

        Much of what has corrupted Link is a belief it would eliminate cars, which was foolish, and that the current and future population could support more than a very small urban core in the three county taxing district. Or is anyone suggesting subways in Everett.

      8. Park and ride lots are a terrible investment because they really don’t generate that many riders for the money and land taken.

        Down here, Clackamas County estimates that each new housing unit generates about 7 car trips a day.

        A park and ride lot mostly generates two trips a day; to work and back. Furthermore, those are peak period trips only, when the transit system already has more than enough riders. Furthermore, as you have already noted, everyone is going to work from home anyway, so commuters won’t exist, rendering park and ride lots pointless.

        Putting a large scale mixed use development next to a station would generate far more ridership than the equivalent park and ride lot, and those riders happen at all times of the day and night.

        At least one part of one TriMet park and ride lot is being turned into residential, as it’s a far better investment. The worst performing MAX routes are the ones with the most park and ride lots.

        This isn’t to say that parking shouldn’t exist, but to think that 300 parking spots taking up an entire city block will do much to encourage ridership ignores everything that actually generates ridership. 300 parking spaces next to a station on a line getting 10 minute train service means what? Maybe 20 passengers per train during 2.5 hours of peak ridership, and nothing the rest of the day.

        Any equivalent space of Ballard generates dozens more. There’s housing, restaurants, a movie theatre, regular offices, medical and dental offices, and several popular tourist attractions. Just look at how many passengers are on the buses! I’ve been on a 15 (before the D) going into Seattle in the evening and not been able to get off at Dravus because the aisles were too packed to move to a door. Reverse of peak travel direction yet still packed.

        That’s the type of ridership that needs something better than buses. A park and ride lot in the middle of nowhere would do fine with a few peak period buses only.

      9. AJ, there is another good reason to short turn trains: overcrowding.

        It’s not fair to make every RV stand or even wait a few trains before being able to board. It’s very frustrating to support a system that you can’t board. That’s why some systems short turn some trains — to give the close on riders the chance to get on a train.

        I realize that Link riders are not routinely subjected to this, but the further the train go the more this can be a problem.

      10. RV is an odd example for short turning because any operating plan very likely is running at maximum frequency in the RV during peak. Short turning may boost frequency elsewhere, but for the RV it will do nothing.

        The best spot to branch is where the system would otherwise short turn.

        P&Rs aren’t about productivity or maximizing ridership/dollar. They are a form of coverage service, providing access to the system for riders who would be difficult/expensive to serve otherwise. This is why it is important to charge for parking, to ensure P&R users tend to be riders who otherwise could not walk/bike/catch a bus to the station.

      11. WSBLE might end up with branches on both sides after ST4 or beyond. There should be a Ballard-UW-Sand Point branch (while a second branch continues north) and a Delridge-White Center-Burien branch (the Alaska Junction branch would stay on top of the hill and either never extend or terminate at Morgan Junction/High Point).

      12. Glenn, you make a valid point: if there are no peak commuters park and ride lots probably don’t make sense, and in fact today many are empty.

        If the commuter is permanently gone then I imagine the park and ride lots can be repurposed depending on the zone (some could be made into parks) although some like the Tacoma Dome serve dual purposes, but that scenario is catastrophic for transit. Not only is the commuter a major percentage of transit riders, they tend to be 100% fare paying. So when comparing total farebox recovery, the commuter is gold when it comes to operation budgets (and levies since commuters not commuting generally don’t care about transit).

        However I would distinguish a park and ride in an urban area and one in a suburban or exurban area, (and I am not sure why anyone would build a park and ride in an urban area). It is practically impossible to provide adequate bus coverage and frequency in a suburban area like East King Co., and financially impossible to provide micro transit, to get commuters and suburban riders to the bus/rail stop.

        That is the rub for those who imagine the end of cars, and serving suburban and exurban areas with transit as first/last mile access. Without the work commute and traffic congestion these folks will almost always drive to their ultimate destination, and that does not benefit transit.

        Whether a large mixed-use project will result in increased transit use depends. For example, the high-end developments in the Spring District will likely not result in more transit use, even next to light rail, because these wealthy folks tend to drive, which is why the developments have so much onsite parking. Wealthy folks will not be coerced into taking transit if they don’t want to by eliminating their parking.

        For a “large mixed-use” development to create transit ridership you need: 1. fairly low-income tenants who need to take transit, not exactly what developers build for without public subsidies; 2. zoning and cheap land (i.e. next to an interstate); 3. merchants who want to be part of a large, low-income mixed-use development; 4. tenants who work and so commute; and 5. traffic congestion. And even then you don’t create many new transit riders because the tenants probably rode transit in their last abode.

        I think it is folly to think TOD will manufacture the ridership ST promised to sell its levies and to justify the cost of light rail in most areas, certainly outside the one “urban” core, downtown Seattle (Bellevue is so car concentric moving East Link to 112th and running a shuttle to Bellevue Way won’t make much difference: Ross is correct, development and vibrancy precede light rail). After all, most don’t want to live next to a train station or rail line or interstate, which is why so many Seattle neighborhoods want the stations and lines underground, and why on the eastside they run along highways or suburban arterials, not in neighborhoods (except ironically Mercer Island to the north of the station which is part of the current fight over Inslee’s zoning proposals).

        I don’t think there is much point in discussing WSBLE anymore because the process is mostly politics today: ST does not have the money for really any of the alternatives (except maybe a foolish stub to West Seattle that is pure politics until Dow moves to Olympia) and the stakeholders will demand the most expensive (underground) options and the Board will adopt that figuring someone else can break the bad news down the road when they are not on the Board.

        But what is relevant is ridership, and more importantly farebox recovery, because a rider who doesn’t pay doesn’t help cover operations. This is what we and ST need to focus on. Just about every post on this blog for the last 2.5 years has assumed ridership would return post pandemic, but now I am not so sure. Ridership levels today could be future ridership levels, and don’t wait for another million residents to move here.

        How does ST manufacture the ridership among those who pay that comes close to their fantastic pre-pandemic estimates used to assume operation budgets? A low-income TOD on a light rail system that depends on honor when it comes to paying fares, and today has around 30% nonpaying riders, is probably a worse investment than a commuter park and ride that generates the same number of riders back and forth with a 100% fare paying percentage, and full fare.

        If those commuters don’t come back the house of cards that is ST’s assumptions to fund operations falls apart. It seems silly to me to debate building WSBLE when Rogoff made it pretty clear there isn’t the money to operate the light rail we have based on current ridership and farebox recovery (fare paying riders), especially with East Link, Federal Way and Lynnwood Link coming online with the same ridership and farebox recovery issues and inflated ridership estimates pre-pandemic that will exacerbate the shortfall.

        Those empty park and rides in surburbia are terrifying for Rogoff, and humorously ST once considered charging for them.

      13. The bigger point is that the ST3 process was like a bunch of decision makers participating in an elementary school color book exercise. “What part of the page needs a colored line?”

        If a company wants to expand, they may begin with a map and ask “where do we need to put a store?” but will very quickly move into practical quantitative analyses that look at potential sales per store as well as potential costs to build and operate a store there.

        The ST3 process began with a set of corridor studies in pre-defined corridors (the ST2 “losing” corridors by the way) that did do some early quantitative assessments between 2013 and 2015. However, they did a lousy job on estimating capital costs. A few alternatives per corridor were assumed and studied, but there was never the opportunity step of asking something like “would branching optimize ridership?”

        This additional question was never asked in 2016: What’s the most cost effective or productive ridership operations scenario for the region? Instead, the ST3 package building didn’t focus on ridership at all. It turned into a potluck dinner where each dish was what a subarea wanted mostly on a whim given a dollar amount to spend. It wasn’t ever planned to be a nutritious meal for the region. Each subarea was left to mostly make whatever choice they could agree to and that is what got put on the potluck table.

        In decades past when public transit was profitable, the transit companies deliberately built systems to maximize their expected profit per rider or what we might call productivity. As a public agency, this has not been the primary objective of ST. However, it should still be one of the several objectives that gets assessed — and it seems to be mostly ignored.

        The unsustainable fare revenue problem is the most direct result — and that’s not only before ST3 but before most of ST2! The fare revenue problem is real — and even ST2 may make it worse.

        What I think is needed now is a realistic increased focus on productivity. That includes being realistic about all the elements of productivity — TOD potential, pedestrian access, station passenger experiences, possible overcrowding, full assessment of parking garage benefits and opportunity costs and things like that.

        I understand that it may not be as politically appealing as a colored diagram. But a “productivity strategy” about Link’s or ST’s future operation is badly needed. If anything, it should have happened in 2017 before the ST3 details began development. We are way overdue for this perspective!

      14. There was this odd local belief that Seattle was different in growing transit ridership when what really happened was a huge decrease in transit travel time and less Metro competition making the ST system more productive.

        It is worth noting that transit ridership overall increased — that was the surprise (not that Link ridership increased). Some of that was Link related, but a lot of it was due to the increase in transit frequency within Seattle. The third reason was increased urbanization.

        The point being that often the most cost effective way to improve transit ridership is simply to pay for more service, and allow for more density.

      15. Park and ride lots close to a station are usually very expensive per rider. They also reduce the opportunity for development near the station, reducing the net increase in ridership even more. It is luxury transit for a relatively small number. It is often pushed because people in the suburbs see themselves using it. But it doesn’t scale, and is just not worth it.

        I call it luxury transit because the alternative requires making a transfer. But feeder buses *do* scale, and often provide additional value. For example, the buses to Northgate provide value in themselves, as Northgate is a destination in its own right. Sometimes they can work well for reverse commuters, or people making trips in the middle of the day. For example, Mercer Island is not a big destination, but a bus along the I-90 corridor could provide people on Mercer Island with a good connection to Eastgate (and BCC) along with Issaquah.

        I’m not opposed to park and ride lots, but they should be cheap, and in the neighborhoods. Quite often these can be leased from churches. Drivers have a shorter distance to drive to the parking lot, while shuttle buses pick up riders from the lot, and riders from the neighborhood. Sometimes these become places for bikes as well. Neighborhood bus stops often act as de-facto park and ride lots. Feeder buses being fed by small park and ride lots are a good way to solve the “last mile” problem, as travel distance to transit becomes only a mile (or less) not five to ten miles, like the distance to Link park and ride lots.

      16. It really is noteworthy that most ends of Link trips are by walking yet ST3 gives only $100M to better walk connections and $0 to adding more entrances or vertical devices at existing stations. ST could do wonders for better station access and attracting riders by cancelling just one garage!

      17. But there was an alternative, a Georgetown bypass route.

        No, the alternative is to just keep running the 594. A bypass just doesn’t make sense. Yes, you do save a little time by skipping a bunch of stops in Rainier Valley, but not that much time. Meanwhile, you lose a tremendous amount of ridership, which means that you can’t justify running the train often. There is a reason why commuter trains don’t run often, even when they own the tracks. Running trains every half hour from Tacoma would be a big degradation compared to running a bus every fifteen minutes (let alone ten) especially since the bus could run right through downtown Tacoma (which the train won’t do).

        You can criticize the main line for making all of those stops along the way, but that’s what subway lines do. The problem is pretending that they make sense for long distance travel. For that there is commuter rail (using existing tracks) or express buses.

      18. “There is a reason why commuter trains don’t run often, even when they own the tracks.”

        No, there isn’t. The Duesseldorf S-Bahns ran every half hour 24 hours when I was there, and probably have a more frequent periods now. Nationwide trains in Switzerland to cities large and small are half-hour.y. Caltrain plans to be half-hourly all day. PATH runs every 10 minutes daytime, 40 minutes night owl. (And it was more like 10-20 minutes when I was there so it may be a covid reduction.) Metra, LIRR, and Metro-North are hourly. New Jersey Transit is 1-2 hours.

        As has been said about Switzerland’s referendums on everything and Obamacare-like insurance, it works because they’re Swiss. The reason commuter trains don’t run often in the US is that the US ignores transit’s potential. American cities see commuter rail as a kind of coverage service or premium service for 9-5 workers rather than as a backbone of the transportation network, which it could be with the right service.

      19. @RossB, on bypass: I would never advocate shutting down the RV line but if a bypass would be built to increase capacity to Seatac, the RV line could be extended through Skyway to Renton and hook up with the planned RapidRide I line serving Kent/Auburn and other bus lines serving Maple Valley etc.

      20. I’d prefer the term niche transit, not luxury transit. Niche still makes it clear it doesn’t scale and it intended for a small subset of riders, but is less pejorative. There are many legitimate reasons why P&R is the best station access for a small subset of riders, even if they live somewhere an abled bodied person could walk to the station or a person with different commitments could take the bus.

        SE Redmond is perhaps an example of a “good” P&R. It’s near the end of the line, intended to serve riders that would be difficult/expensive to serve with feeder buses, and it’s placed above the station next to a freeway, so the TOD impact is minimized. Yes, Sammamish should also be served with surface lots and feeder buses, but given Sammamish is a suburban edge city that shouldn’t be absorbing much future growth, there isn’t much opportunity to leverage feeder buses to build all-day ridership over time, so the midday/evening/weekend frequency of those feeder buses is likely to be either poor or excessively expensive in the long run.

      21. P&Rs should be outside urban centers, even if it means adding a station for them. The Bellevue TC/South Bellevue P&R pair work well. People who drive to P&Rs aren’t going to take a local bus a couple miles, they’re going to take an express bus or train five miles or more. So the P&R only needs to serve the trunk line, and the city center can be the transfer point to other bus routes. That way people can do an errand while they’re transferring, or stop at a cafe if they want refreshment or a newspaper, and it’s the same station they go to for downtown destinations. Lynnwood, Federal Way, Renton, and Burien P&Rs are in bad locations because they eliminate a significant part of the walkshed.

        “Sammamish is a suburban edge city”

        That’s being too kind to it. Bellevue and Kirkland were edge cities when they grew. But Issaquah is so far out it’s exurban, and anything beyond Redmond is exurban. I haven’t been to Sammamish because its transit is so skeletal, but I picture it as almost all houses and a supermarket plaza. That’s hardly a city even if Washington defines a city as anything with an incorporated government.

        “that shouldn’t be absorbing much future growth”

        I don’t know about shouldn’t but it’s definitely unwilling to. If it won’t accept density, then P&Rs outside it like SE Redmond and the Issaquah Highlands is a reasonable compromise. Just not in city centers. That’s where walkable retail and apartments belong, not hundreds of cars, especially cars that aren’t shopping there.


    Unpacking the idea here, keen for y’alls thoughts.

    1. Accelerate the C&D improvements, focusing on improving the C, in return to deferring WS Link until DSTT2 is complete.
    2. Complete WSBLE EIS with median 15th Ave as the preferred alternative.
    3. Using WSBLE money, rebuild Ballard car bridge over 14th, build a transit bridge over 15th, and rebuild Dravus bridge/interchange to support an in-line bus stop. Latter two projects will be forward compatible with rail.
    4. Take a mulligan on WSBLE and come back to the voters with a better plan.

    #3 not only moves the ball forward on Ballard Link and creates strong improvements for bus operations until Link arrives, it also takes two bridge projects off of SDOT’s plate, one major (Ballard ship canal) and one minor (Dravus)

    1. I’m assuming you mean the “WSBLE money” minus the other 4 four subareas’ contributions for DSTT2, correct?

      1. Yes. I would assume those fund would remain outstanding until the Westlake-ID segment was built.

    2. I still have serious issues with so many megadeep stations, especially where transfers occur.

      I like the idea though. I simply would suggest that an aerial alternative over Mercer St be considered after a short tunnel at Seattle Center, that Capitol Hill Station be assessed as the transfer station instead, and that this be the “moved Denny Station” with maybe an infill station just east of Fairview with a portal at I-5, that a station be considered near Boren and Seneca, that Midtown platforms be brought closer to the surface and that the ID connection options be expanded (like exiting the ground near Fourth/Washington and running above the UP tracks past the stadiums.

      I’m not wedded to this specific alignment but I really feel strongly that and DSTT2 plan has to look for less drastic elevation changes as well as cost savings. There seems to be general consensus that the vertical profile is unacceptable which would mean that the specific alignment would need to shift.

      1. Going elevated along Mercer St towards Capitol Hill station is intriguing. Why do you suggest a tunnel at Seattle Center, Al? How would you continue from CHS, towards ID? Or along 12th to Judkins and back through the Beacon Hill tunnel towards WS?

      2. It’s more illustrative than anything.

        It’s clear that there is an elevation change between Elliott and Fifth so I was thinking that Seattle Ceter could accommodate a short cut and cover segment.

        I’m not familiar with the tunnels near CHS but it seems like it may be possible to place a crossing above the current one.

        I’m frankly not sure what the best way to lay tracks south of CHS would be. It’s all about the grade. It’s hard to get close to ID and King St stations unless the air above the UP tracks are used or there is a deeper subway.

        My overall example is proposing what train tracks have done historically when it comes to steep grades. That is to not follow a straight line when elevation changes are needed. The straight lines of ST3 through Downtown Seattle are problematic because of the grade changes. (This seems to have been ignored in drawing the 2016 alignment map.) The plan has always been presented as two dimensional rather than three dimensional. Straight line tracks + steep grades = deep or high stations. It’s a geometric reality.

        There is one way to scale steep slopes with rail. That’s with cable pulled systems like funiculars or diagonal elevators. If automated and running frequently, it doesn’t feel any more like a transfer than using a regular elevator does. Consider that using a deep or high station is effectively “adding a transfer” to a rider anyway when only elevators get used.

      3. Yes, the Elliott connection may require some creative thinking, one possibility might be that the elevated line along Elliott may slope down more gradually.
        Yes, Seattle has a ton of hills causing all kinds of problems for light rail. Ever since we ran into the issues on First Hill, I would think we should have switched to 3D views, many planners still seem to use maps. People complaint about Seattle’s challenge with East/West connections, those are mostly driven by the fact that we have many North/South hills you need to cross which light rail is not good at, gondolas are so much better at this.
        Cutting through Capitol Hill and connecting at CHS, traversing the hill and then coming out on 12th may be an option…

  30. On peak traffic: Even Peter Rogoff reported end of last year that he doesn’t expect peak traffic to come back as commutes will decrease, but there are plenty of other people who rely on transit as explains. This may mean we can live longer with a single tunnel and fully automated systems (rail or gondola/APM) could provide consistent service not only throughout the day, but into the night., could we just go elevated along Mercer? Is Westlake wide enough to make a turn or still tunnel under it but come out of the tunnel on Mercer?

  31. Why is WSJ-2 better for the future? Wouldn’t being closer to the urban village be a good way to increase ridership and create a greater sense of place in the area?

    1. WSJ-1 would even be closer to the Junction, but both these stations would require tearing down huge housing complexes, WSJ-2 would take out a huge apartment building which was just built. I wish Sound Transit had looked at more options which would avoid these losses, it seems it should be possible to move the station a bit further South or North and may be connected via funicular to the Junction.
      To justify bringing light rail up into West Seattle, it needs to be extended further South to meet Westwood, Greenbridge, White Center. None of the proposed stations make this easy as there are no large arteries in West Seattle.

Comments are closed.