Seattle Subway LogoStill wondering if transit advocacy can have an actual impact on public policy? Last Thursday, when the Sound Transit board voted to adopt the new Long Range Plan the answer was made crystal clear: transit advocacy works.

Proof that our voices were heard came in the form of Sound Transit Long Range Plan Supplemental Environmental Impact Statement Final (SEIS) which individually responded to the hundreds of stakeholder, organization, and reader comments to the SEIS, and included a section with summary responses, the majority of which are directed at issues that were brought up by both Seattle Subway and, via our guest posts, readers of Seattle Transit Blog.

The biggest news for us in these responses to our comments is an about-face on policies related to driverless trains, a pillar of what Seattle Subway advocates for since our inception three years ago. We support driverless trains because they can lead to a self-sustaining operating system (as in Vancouver), which frees up money for even more transit. Until this response, Sound Transit solidly refused to study the technology, publicly declaring they would not consider driverless trains. However, their latest response says that driverless trains can be considered as long as it does not interline with the spine section or have at-grade sections. We continue to remain confident in the value of driverless trains (we’re getting close to the era of driverless cars, after all!) and believe that by the time an interlining opportunity comes up, Sound Transit’s perception will likely change.

Many of you echoed our championing of building out West Seattle and the Ballard Spur. The Sound Transit responses to West Seattle and Ballard were similar: they will not make routing decisions until the next phase (system planning) which is resource-constrained. Our contention is that the data given to the Sound Transit Board during system planning must contain ideal options in order to be considered, which is exactly why we began to give detailed feedback at this stage. We will continue to monitor the situation and will give feedback as opportunities arise during the system planning phase.

The response to our Better Eastside Rail article was similar to West Seattle and the Ballard Spur: specific routing decisions will be made as part of the system planning process. However, we break the Eastside out separate from West Seattle and Ballard because we think Eastside options need further development. Due to the scale of the rework needed to bring good study options to the table, we have concerns about timing. Any options presented to the Board should be worth voting for in the ST3 package. Remember, the Eastside voted for ST2, but heavily against King County Proposition 1.

A surprising number of you (60) echoed our thoughts about the PSRC population numbers. Initially our stance didn’t seen to be well received, but people in a position to fix the problem read it and the point was well taken. Sound Transit’s response consisted of a mostly technical explanation of the model that objectively doesn’t work and concludes with discussion of making changes to the model used for ST3. However, since we wrote this article, we have met with PSRC staff and had some very productive discussions. Matthew Johnson will publish a more in-depth PSRC article soon, but the bottom-line issue appears to be related to the way population models treat designated regional growth centers (Ballard isn’t one). Despite the appearance of doubletalk, we do think ST and the PSRC are aware of the issue and are dedicated to making sure they get the best numbers possible in the version of these estimates that really matter—the ones that determine funding priorities and the service plan for ST3.

We have long supported an infill Graham Street Station, and many of you agree, evidenced by a petition with 776 co-signers submitted to Sound Transit. Their response has proved very interesting. Apparently Sound Transit’s board can decide to build the Graham Street or Boeing Access Road stations at any point, so long as funding becomes available in the North King Subarea.  As you likely know, U-Link is finishing up $150M or more under budget and we wonder if that money could be used to add infill stations.

When we recently found out that the Sand Point corridor we referenced in our Let’s Build a Sand Point Crossing! article was in danger of being removed from the LRP for study we issued an action alert and a lot of you responded.  Due to the the leadership of Councilmember O’Brien and Mayor Murray and with the help of ST staff, the corridor is now explicitly included – it will be studied when an additional Lake Washington crossing is considered. We think that study will show that it is a better rail crossing than 520.

We want to thank Sound Transit staff and board for listening to the public and thank the members of this community for substantive policy discussions. Your efforts supporting Seattle Subway truly have real world dividends. Thank you.

If you would like to help Seattle Subway or there are things you think we could be doing better we invite you to come get involved.  We are an all volunteer organization and strive to be non-exclusive.  Come help us make a difference for the future of Seattle.

88 Replies to “Sound Transit Responds to Seattle Subway, Transit Advocates”

  1. I heard an amendment to the LRP about a second downtown tunnel was rejected – does that mean there’s no way one could be included in ST3?

    1. Alex – no. Our understanding is that this was just the first step in what will be a multi step discussion of who might fund a new DSTT (subare wise.). The new tunnel was initially studied as part of the West Seattle study and is included in the LRP.

  2. Perhaps you should consider whether any cost savings from U-Link should be applied to a NE 130th St. station over a Graham St. station or BAR station. If one one could be financed with that money, which would have the best value?

    1. Absolutely. 130th would probably be the priority, but we think both can happen without additional funding. We are actively involved in the fight to get 130th and will have more info specifically realted to that battle coming soon.

    2. Who would use a Boeing Access Road station? Where are they travelling from and to? How would it be useful in light of Sounder’s severely limited schedule and risk of missing the train and having a long wait? Is ridership dependent on rerouting buses (e.g., the 150 and 101) to the station?

      1. I guess it depends on what the intent is.

        Is the point to allow folks riding the south sounder like to transfer to Seatac airport?

        Is the point to have a special line there that will take people to Boeing field?

      2. This is the first I’ve heard of a Boeing Access road station. Where exactly would this be? If it was next to the freeway, it might lead to better Link to bus access. But I see two problems with that. First, it would more expensive than the other stations (since to make it work properly you would have to add bus only ramps and exits). The second is that, from what I’ve heard, getting to the Rainier Beach station isn’t a problem. The problem is that no one wants to cut their express ride to downtown short, by getting on at the slowest part of Link.

        So, any way you cut it, the other two stations make a lot more sense to me. I would guess that the Graham Street station is cheaper, but the 130th station would be more valuable (because it would lead to much better bus to train transfers).

      3. It would be where Link crosses the Sounder track south of Rainier Beach station. The primary purpose is for transfers to/from Sounder, but people have also talked about buses terminating there and people taking shuttles to Boeing.

        I don’t see enough room for a good bus station, and it’s in the middle of nowhere so not a good transfer point. Plus I don’t think these Link+Sounder transfers would really materialize, because for downtown and beyond International District is just as good (Kent-UW, Tacoma-UW). Short of downtown there’s only Rainier Valley, Beacon Hill, and SODO, which aren’t pulling in enough Puyallup trips to justify a transfer station. And anywhere south of TIB is backtracking and has more direct bus routes (Kent-SeaTac, Puyallup-SeaTac). Hence my question: are there any significant travel patterns I’m missing?

      4. Why would people take sounder to get to the airport? Sounder is every 20 minutes link is every 7.5 minutes plus transfer time. So not arguing against the station just not a good rail to rail connection.

      5. @John Slyfield

        Its unclear as to whether or not it would actually be used, but the folks who might use Sounder to LRT to the airport are folks coming from Tacoma or other points on the south end of the Sounder line, not from Seattle.

        The other potential ridership from a station here would be a shuttle from this station to Boeing field.

        Both of these might not be enough to justify the station at this point though. I suspect 130th and Graham are the more likely winners for infill on the existing planned line for the foreseeable future.

      6. BAR station might have a better transfer experience for Sounder passengers from the south, compared to using Tukwila and RR-F.

        There might be rare passengers, again from the south, that would prefer a more convenient transfer to Link at BAR station and a slower overall journey than transferring at King Street station/ID station.

      7. The boeing access road was discussed a few times, and always rejected as a waste of resources.

        The better choice between Rainier and Seatac was in Tukwila – 133rd ave?

        Since that site wasn’t originally studied, I’d assume there are more hurdles, but at least it wasn’t a 15 minute walk from every single destination.

      8. Thanks for the clarification. I think it would add some value, but not a lot. I think you have the trip pairs right, Mike. I agree and think there just wouldn’t be that many transfers here. I think it makes a lot more sense to add a station south of Angle Lake as a freeway bus station (with bus lanes leading to it and away from it). That way a Tacoma bus could make one stop on the way to Seattle, and let off folks trying to get to SeaTac. That doesn’t do much to connect Link with Sound Transit, but you could maybe run a frequent bus from the Kent Sounder station to that particular Link station.

        You could, maybe, go from that station (by the freeway) to the Kent Sounder station, and end Link with one more station just to the east of there. That particular area is actually more densely populated than anything in Tacoma (believe it or not). Regardless, I think that would be way more cost effective than pushing Link further south. The big challenge, though, is getting light rail to make a sharp turn and drop off the hill (and cross the river). That wouldn’t be cheap or easy I imagine, but I’m not an engineer, so I don’t know. But if you did that, you would serve some decent locations, while connecting the freeway, Sounder and Link.

      9. And a sounder to link transfer point would require two new stations be built. Huge investment considering the ridership involved.

      10. Ross,

        The place to have it is along East Marginal Way South just north of the Duwamish bridge. Then add bus only ramps from EMWay to SR599. Make the left turn absolute priority; there aren’t enough cars to make a difference if one or two are forced to wait twenty seconds.

        There is already special access to the southbound HOV lanes which serve both I-405 eastbound and I-5 southbound from SR599. At this time there is no preference for HOV’s from I-405, but the interchange lane enters I-5 from the right and SR599 departs from the same side about 3/4 of a mile to the north. The Southcenter way on-ramp would have to “merge through it” but there’s no reason that an additional right hand HOV only lane could not be added from the I-405 merge to the SR599 departure.

        I know you’re an advocate for a truncated Link with better bus collectors. This would be a way to get the buses to Link with much less traffic interference than MLK and since the station would be elevated it would be a more pleasant transfer experience.

      11. @Anandakos — I’m definitely an advocate of truncated bus routes with good Link connections. But in some cases (and this is probably one of them) I think it would prove to be very unpopular. I would have to do the math, but as I said earlier, the biggest problem I’ve heard with buses going to Rainier Beach is not the bus getting to Rainier Beach, but Link being slow once it gets there. This means the bus would still keep going to downtown, just because riders would demand it. This is similar to Tacoma buses that get off at my mythical Kent freeway station. The buses keep going to downtown Seattle, but folks who want to get to SeaTac can just get off at that station. The same thing could happen here, but I think there is a bit more of a time penalty. To be honest, I’m not sure which buses would be re-routed to serve this station. Can you clarify a bit what a typical bus route would look like?

      12. I would think that having a very large park-and-ride garage at the edge of Link with direct I-5 access could add significantly to Central Link ridership. More strategically, it might provide an incentive for Renton residents to support ST3 because they would get some added benefit. With notification of the travel time to Downtown Seattle, drivers could easily pull off the freeway, and pay for parking and transit fare and come into Downtown or the stadia or UW. Add to that a freeway link to ST buses and for Renton-Seattle buses, a frequent circulator bus to Southcenter and perhaps a redevelopment opportunity and a station could be quite useful. Without a strategic station access and development strategy, it would be of little value.

        I would think that either BAR or 133rd St S could work well for this.

        I also agree with others that if an I-5 station is developed between Angle Lake and Federal Way, that could also serve some of these functions.

  3. Yay for SoundTransit for being responsive, but driverless trains item seems like a Pyrrhic victory. Driverless trains are ok unless they are at-grade or part of the spine (i.e. downtown Seattle?). I think that is just another way of saying no driverless trains. Where are you going to have a completely grade separated line?

    1. DP’s version of UW-Ballard?
      Seattle Subway’s version of Ballard – Downtown and Downtown – WS Junction?

      Just two off the top of my head.

      1. UW to Ballard should be grade separated, but I don’t think it matters if it is driverless or not. It should be grade separated because the alternative is terrible. Not only would it be extremely slow, but it would make transfers very difficult. As I say below, consider a trip from Ballard to Bellevue. Now assume that the trains from Lynnwood go every four minutes; they split, with half of them headed to SeaTac, and the other half to Bellevue. If timed correctly, this makes for a pretty sweet commute for a Ballard rider. Get on the train, ride it for five minutes to the UW, after a thirty second wait, get on a different train that is headed to Bellevue.

        But that only works if the trains can be timed properly. A missed connection is not a four minute penalty (in this case) but an eight minute one. Now add on a bus trip to get to the Ballard train, plus enough extra time for that transfer (and needing to get to work by a specific time) and that sweet commute just became really sour.

        Of course, it gets worse. One of the big worries (from ST) about a Ballard to UW line is that it will crush load the main line. I disagree, but if it is an issue, then an inconsistent ride from Ballard to the UW will compound the problem. In my example, everyone who expected a quick transfer to the east side simply waits for the next train. Meanwhile, there are plenty of other passengers (who walked to the station or took a bus) headed to the same location. This could make for a very crowded train. There are other scenarios that could be a mess as well.

        The long and short of it is that a Ballard to the UW line needs to be grade separated and needs a very good headway. It is arguably the second most important section (behind UW to downtown). It needs to be integrated really well with the other line, otherwise it could cost us quite a bit in the long run.

      2. I agree that grade separation is crucial for the entire Ballard/UW line. Disagree that transfers are going to cost that much for Ballard to Downtown riders.

        Ballard riders transferring would have their choice of SeaTac trains or Bellevue trains, all of which go through downtown first. So Ballard to downtown is faster than the current bus options even without counting for bus traffic. Ballard to Bellevue suffers from that possible 4 minutes extra wait time, but it is still going to be faster than a bus.

      3. Ross – Driverless matters for $ reasons every bit as much as it matters for operational reasons. When Eastlink opens they are talking about 4 minute headways at the transfer – so no issue there.

    2. Everything in Seattle that isnt part of ST2 will meet this criteria. Essentially everything that might be built in Seattle in a the next 25 years.

  4. Thanks to the Sound Transit board for being responsive to these issues.

    “driverless trains can be considered as long as it does not interline with the spine section”

    That is technically understandable but it presents a long-term dilemma: the spine is where we most need frequent trains. It will look backward and inside-out if, for instance, 45th-Ballard-West Seattle has driverless, 5-minute, 55 mph service while The Spine is limited to 10 or 15 minutes and 35 mph. That could lead to some people avoiding the spine or transferring from the spine, and make visitors wonder why Seattle put its best service on secondary corridors. Of course, this all points to the folly of having at-grade segements on the spine in the first place.

    1. Replying to Jason Schindler: “I think that is just another way of saying no driverless trains.”

      It’s the closest to yes they can do without retrofitting the spine. The situation raises a “What’s the point?” issue if short-distance lines in Seattle, Tacoma, and the Eastside are driverless while the Spine is driverful. What’s the benefit in that? Why not just use interchangeable techonogy at that point, to lower maintenance costs? The ultimate problem is not ST’s lukewarm answer to this question; it’s the surface segments in the spine. And those were decided in the 1990s and early 00’s, so the die was cast then.

      “Where are you going to have a completely grade separated line?”

      That is looking more positive. ST2 has much more grade separation than ST1, and ST3 will probably follow suit. The public rode the ST1 line after it opened and decided they wanted something better in the extensions, and ST did it. The tragedy is that people could have demanded the same thing earlier for ST1, and ST could have done it then. But the public was not ready to pay the capital costs for grade-separating MLK and SODO, and ST was afraid of penny-pinching voters rejecting ST1 if it was that expensive. We did get grade separation in Tukwila because Tukwila refused to allow surface Link on Tunwila International Blvd, but ironically that’s not where we most needed it.

      Facts: grade-separated from International District to Lynnwood and presumably Everett. And from International District to 120th (Bellevue), and thereafter only a tiny surface segment and a couple road crossings to Redmond. (East Link was going to be fully grade-separated until Bellevue’s infamous subway required cost-cutting.) And from Rainier Beach to Kent-Des Moines and presumably Tacoma. So where are the surface segments? Rainier Valley (open) and SODO (with railroad gates).

      1. Note: the I-5 segment in north Seattle/Shoreline is technically at-grade because it’s on the ground, but I’m calling it grade-separated because it will have 100% overpasses like the freeway does.

    2. If you define the spine as everything from Lynnwood to Tukwila (and beyond) then I don’t think anyone would say that the whole thing could be considered the “primary corridor”. The primary corridor — the real spine, if you will — is downtown to the U-District. It is essential that we get frequent service between those locations, as it connects the two most important areas in the state, and includes Capitol Hill, which some would argue is the third most important. What the trains do after going through this essential corridor is not critical, in my opinion. Ideally you would have fast, frequent service to the East Side as well as Rainier Valley, but I don’t think that is possible. But between the two, it does mean fairly frequent service in the core. If Tukwila train runs every 8 minutes, and the Bellevue train does the same, then you can have four minute frequency from downtown to the U-District. That is key, and would enable fast, frequent service from Ballard to the UW, even it means a transfer.

      Now, as luck would have it, light rail is grade separated from (beyond) Northgate to downtown, which means consistent (and frequent) service. The trains will alternate, and the trains headed to the UW should be timed so that they arrive in time for everyone to make the transfer. This would allow someone to get to the east side (or SeaTac) with ease, by simply picking the right train from Ballard. That is why it is essential that the Ballard to UW line be grade separated. It also makes sense to make this transfer as easy and quick as possible. Going the other direction, it is much harder to time. But since the UW is the end of the line for the Ballard to UW line, then the trains can simply wait until the other train arrives. This would mean that trains headed east would be consistent, but those headed west would not. It is essential, then, to build in fast headways, so this can be possible.

      Full grade separation throughout the entire line would be nice, but I doubt it would be worth the money. The graded section is not ideal, but it isn’t terrible, either. I can think of lots of more important tunnels to build before building that one.

      One example is a second tunnel through downtown, which would be essential for Ballard to downtown service. But I don’t think it is essential to make the whole thing grade separated. For example, if the modified C proposal ( was significantly cheaper than similar proposals, I think the time penalty (one minute) would be worth it. Likewise, the cost of having a driver is minimal as well. I would much rather build good systems, even if they contain a bit of shared grade, then build grade separated runs with very low ridership.

      It is less than ideal, of course, but I think worrying about maximum speed and driverless technology is the least of our worries, and could, in many cases, lead to poor decision making.

      1. I agree with most of this until you say full grade separation won’t be worth it. It will matter in the short term that the trains to Ballard are limited to 35 MPH. Speed matters. But in the long term that time penalty will be excruciating.

      2. @Keith — Did you read the blog link, and the comments after it? The difference is less than a minute. Put it this way, how much is that minute worth? Would you rather have the minute, instead of stops at NW 65th and NW 85th (for that route) as well as stops at 24th NW (for the Ballard to UW route)? Oh, and while we are at it, why not throw in a grade separated spur line out to Lake City. Is that minute still worth all that?

        By the way, I think that Bruce was being conservative on his timing. 56 seconds was the difference, but that assumes instant acceleration. In real life the difference would probably be more like 35 seconds, by my estimation. We could save that much time by simply shrinking our dwell times. As he said, “Lots of people worry about the top speed of transit service, but it’s not very important for in-city services (say, typical trips of less than ten miles, stops about every half-mile), because even a fully grade-separated train spends much of its time accelerating or decelerating for stations; frequency and reliability matter most.”

        Not to suggest that his alignment is perfect, but I think the biggest trade-off is not the cost in speed, but the trade-off in stations. His alignment doesn’t get that close to Uptown (it stays closer to the water). I think that would hurt ridership a bit. But probably the thing that would kill it is the disruption caused by cut and cover. I think that would be politically unpopular. Although, with Bertha’s problems, folks might be a bit more interested in a less tunneling.

      3. @Ross

        Its not just a time saving. Its a time saving plus a huge cost saving. IMHO, driverless trains are the one “no-brainer” business decision in front of ST. You get faster times, the marginal cost for increasing service is WAY lower, and your operating costs go down by a lot.

    3. One of out points is that, by the time we get close to interlining old with new the argument that the current line cant be driverless will go from wrong (what is it drivers can do that driverless can’t in the RV or in the bus tunnel?) to just silly when driverless cars are prevalent.

  5. The best driverless car system is only as safe as the worst tire on it. Not a winning example to lead off with here. Same for so much uncritical enthusiasm for a smaller payroll. A computer is only as good as the worst thing its dumbest programmer tells it.

    Meet you half way on this one- let a representative team of operating people pick the individual drivers replaced by automation. And finalists do a classic contest out of tunneling history, like in the John Henry song: “Before I let your computer beat me down, I’ll die with that controller in my hand!” Any flesh and blood winner gets to drive the subway.

    But Market Street to Phinney Ridge and Sand Point Way east of the U-District illustrate a specific point. In addition to the expense for tunneling or elevating the flat distances east and west, none of the rest of the line can open until the whole thing is complete. True it’s a good bet that the years ’til construction begins will improve time and cost. Maybe even before Bertha gets dug out.

    So even a line firmly intended to be automated might best be designed to be operated as driver controlled light rail until the project is “built out.” Considering the years starting with Forward Thrust, I’m weary and wary of the fast and flawless future being used to excuse a slow, lousy and endless present.

    Mark Dublin

    1. The best driverless car system is only as safe as the worst tire on it. Not a winning example to lead off with here. Same for so much uncritical enthusiasm for a smaller payroll. A computer is only as good as the worst thing its dumbest programmer tells it.

      Driverless systems exist, right? So there’s no reason to rely on platitudes like this to make assessments about their safety; just compare the safety records of such systems to ones with drivers.

      1. There are driverless systems that are decades old. There is one 180 miles north of Seattle even. They have a better safety record than trains with drivers.

      2. Let’s get one thing straight, guys. I mistakenly thought that the fact that Quebec City is only rubber tired subway capable of being automated, the main type of automated system I’d consider dangerous is one involving private automobiles.

        Where critical maintenance of thousands of individual privately owned vehicles is literally a life and death matter, since something as minor as a blown tire at a crowded time on a freeway full of automated vehicles at high speed would wreck more car than one.

        I’ve been on Skytrain many times, and have seen no safety problem whatever with it. However, the last time I rode it, maybe fifteen years ago, trains did have roving customer assistance agents aboard who were trained to drive the trains manually in case of a breakdown.

        Over the time it will take to complete our own automated project, Seattle and environs may very well have changed so much that like San Francisco, both driver-operated light rail and automated heavy rail will both be in use.

        But having watched my favorite transit related equine veterinary problem develop over many years, I’m very much interested in the mechanics of the transitional period that while it may not occur in my lifetime, it will definitely occupy the attention of most commenters here.

        And like with any new technology here, which like everywhere else has its own conditions, everybody knows what questions to ask at every stage. Main hope is that I’m setting a good example.


    2. Confused about this attitude toward automated metro systems. They have been around since the 80’s and really aren’t new technology anymore. Actually for completely grade-separated transit, they are probably the default. There are numerous manufacturers that would be happy to sell an off-the-shelf system to Sound Transit. Vancouver has two separate systems, and they work just fine.

      The real advantage with an automated system is not just replacing drivers per se, but being able to run trains extremely frequently. A frequency that would cost too much with drivers on each train, and a frequency that human drivers have difficulty maintaining without bunching. Automated systems are able to adjust the timing and spacing of trains all along the system to respond instantly to one person holding the door on one train at one station. The delay is evened out and made up automatically. For this reason it does not make sense to think of the automated drivers replacing the human drivers individually as you allude to above.

      When Skytrain was new, in the 80’s, there were attendants on the trains because the transit agency thought people would be spooked by driverless trains. But now, there is hardly one attendant per 10 trains. Yes they are trained to drive the trains manually, but that is not what they are there for.

      And very frequent service brings with it the benefit of very frequent service, but also the ability to run shorter trains into shorter stations. That I suspect is the big cost saver in automatic systems. And cost savings is certainly something that all infrastructure projects in North America need.

  6. Has anyone published a back-of-the-envelope calculation or analysis of what it would take to grade separate the sections of Link that didn’t get it in ST1? I would think that building overpasses for the cross streets would be technically simpler and allow the line to keep in service (as opposed to building new tracks elevated).

    Just wondering how much money we’d be talking about. How many crossings are there? Just eyeballing this on a map, I’d think you’d want crossings at Alaska, Orcas, Graham, Othell, Cloverdale and Henderson, so as not to disrupt the grid too much. Are there other, less busy street crossings you’d want to include? What would six overpasses and attendant disruption reaching back into the neighborhoods, cost?

    1. Excellent idea, kpt. CentralLINK will be our airport line for a very long time to come. I doubt that others don’t take my wife’s orders forbidding a LINK ride to an international flight.

      So I really think that elevating over or cutting under major cross-streets, and eliminating cross traffic the whole surface route, would increase passenger peace-of-mind, and with it, ridership.

      Mark Dublin

    2. Not to mention, you would deal with the bothersome frequency issue.

      The SoDo section would be somewhat hard I think.

    3. I think overpasses (for the train or the roads) would be politically difficult. This was an issue when Link was first being talked about. Tunneling was considered too expensive and elevated was considered too unpopular, so they put it on the ground.

      1. Dig a pit and put the cross streets under both M L King and Link. It makes the interchanges more complicated but makes all traffic (road, Link, and pedestrian traffic to the stations) flow a bit better.

    4. It will be very difficult to make the change at this point. Its part of why its so important to make long term decisions in the first place.

      1. Plate tectonics and weather are examples of problems very hard to cope with, but impossible to eliminate. Which would have very bad results if we figured out how.

        But it’s a like avoiding combat by a self-inflicted wound to use politics as an excuse to avoid necessary, beneficial, and highly possible measures- like bridging and undercutting arterials along an existing mainline rail right of way.

        However, precisely because of our weather, any undercut will have to be carefully designed and drained to avoid being flooded out of service. It’s happened other places.


        In addition, in the five years since LINK has been operating, the MLK corridor has been experiencing much change in business activity and population- whose arithmetic can often reverse previous elections and majority preferences.

      2. I think one benefit of an approach like this (whether it’s overpass or underpass) is that each individual project has benefits for the neighborhood and for link. They don’t all have to be done at once. If each is a 100M level project, and you only have 200M, you just do two – pick the two that would have the highest benefit. Then, when you raise another 100M, you do another. You don’t get to driverless trains any time soon (and there are other non-grade separated sections anyway, as another commenter points out), but each time you remove a bottleneck, the trains get a little more reliable. (Maybe not enough so to make it worth it to do it, unless you get all the way to driverless.)

        But, mainly, I’m curious if this is really that level of project – are we talking 50M for an intersection fix? 100M? 500M? I’ve never actually seen an estimate for what it would take to fix. Maybe it’s so much that it would just never even be evaluated.

      3. I agree. But let’s face it, the main reason we built a line out to the airport was because we wanted to show that we could build something, people would like it and that it would serve the suburbs. It was built out there for political, not practical reasons. Otherwise, we would have built a line from downtown to the other side of the UW first (with several stops along the way, instead of one). The last thing we wanted to do was overbuild a low priority line. It was impossible, politically, to build the line elevated through the Rainier Valley. It would have been way too expensive (both politically and practically) to tunnel under Rainier Valley. We probably could have eliminated many of the cross street issues by digging trenches for those streets, but compared to the other failures of our system, this is small potatoes. For example, why is there no station at the intersection of 520 in Montlake? There isn’t even a flat spot for a future station. If we are going to spend hundreds of millions on tunneling to solve previous mistakes, I would rather spend it there (correcting that problem) than spend it on a line that will, over time, prove to be one of our least productive.

    5. I was about to ask this too. We should at least get an estimate of what it would cost to finish the grade separation. But I would not bother with overpasses.Do it elevated or do it underground. North Link was going to surface around 70th but ST found that going under and over the other roads and up and down was more expensive than just staying underground till 95th. The only surface line I’ve seen with several arching overpasses is the LA Blue Line where it goes over freeway interchanges and major boulevards, but those are probably farther apart than MLK.

      1. I think trying to make it above ground would prove to be just as unpopular as it was originally, if not more so. I would like to get estimates for tunneling as well as overpasses/underpasses.

        I would also like to get time estimates, in terms of how much time it would save (with either solution). My guess is that overpasses (eliminating all intersections) would be almost as fast as a completely grade separated line. The speed limit doesn’t play a huge factor on this stretch, because unlike a lot of this line, this part has decent stop spacing (and will have better stop spacing if Graham is added). I think the difference between a top speed of 35 versus 55 is probably less than a minute for this stretch. I think we could save more time by just shrinking the dwell time.

        Also, if we built overpasses on all the crossing streets, would that enable much higher headways? That would be nice.

      1. You mean lower the track and rebuild the cross streets over it? That’s an idea. It would cost at least part of a cut-and-cover tunnel. And it may be unappealing empty space, making the roadway wide even though there’s nothing there unless you look down.

      2. It would also mean shutting the line down until the project’s finished, though. That’d be completely infeasible unless we’ve already built a Duwamish Bypass or a West Seattle line extending down to the airport.

    6. Physically it would probably be easier to put the line in a trench and extend streets over top. The trench could them be covered at some time in the future. But another possibility would be to use plenty of remote sensing and moveable physical barriers that came up along the line at intersection. The cameras and radar would see that the line was clear and that there were no people nor cars making a left across the tracks and then gates would rise from the street blocking vehicle and pedestrian access to the line while the train crossed. This would involve a new application of this technology, but in the time it would take for Sound Transit to consider this, it might be ready for prime time.

  7. Is driverless an all-or-nothing thing? I’m thinking in terms of an autopilot in the car. Wouldn’t you want the ability to put the train on autopilot from say, Everett to the ID, then have a driver come in, take the driver’s seat in the lead car at the ID and drive south?

    1. Something to think about, Basselle. But first, if you ride LINK through a driver-change, do a stop-watch check on how much time this takes. When you get to the headways planned, a short stop can give you a long problem back up the line.

      But one idea: Up in Vancouver, aren’t there people in sport jackets rather than operators’ uniforms who ride the trains as customer assistants? Not exactly classified, but every train has folded-away manual controls, which these assistants are trained to operate.

      It would definitely boost the efficiency of the system to have people who aren’t also driving on hand to deal with passengers’ questions and problems, especially through a long subway.

      So we could think about having one of these “train agents” take the controls while the train is still on automatic, so there’s already a driver in the seat by the time the train arrives IDS southbound.

      Worth checking out.

      Mark Dublin

      1. Driverless trains can be run on our current system. ST is just saying they won’t consider it. Lets not get into a technology argument on this point. There are cars that already can do driverless from highway entrance to exit – a far greater feat than a driverless train and its limited ROW issues.

    2. The benefit of driverless trains is you can run them very frequently full time, because the marginal cost of power and maintenance is small compared to the cost of drivers. This allows high-quality service and transfers even when ridership is low, as in the Vancouver Skytrain (5 minutes until midnight) and the airport subway (2 minutes).

      1. Understood, Mike. But a couple of questions I think would add a lot to this discussion.

        First, how many years are we talking about before we even break ground for any driverless train system, let alone start operations?

        And second, what is the average training and break-in time for a system in a place which has never had heavy-rail transit before, let alone automated?

        And the most important practical point that Baselle raises is to the means and extent by which the driver-operated light rail service of the present and near-future can be transitioned or integrated into the long-term automated system.

        The BART and the streetcar-evolved MUNI metro systems share two separate levels of the same Tunnel under Market Street in SF- but could never, and were never intended to be integrated.

        So could further comments on this posting present some thoughts and suggestions on this question. Because this matter will have very far-reaching consequences for the development of both systems here.

        Mark Dublin

      2. Mark – a driverless system is a marginal cost on a project like this. Trains already have complex signalling systems. Salaries are also the biggest cost in operations. I doubt the ROI is more than a year.

      3. Mark, Don’t we have a fully automated system already? Rubber tire I believe, maybe with some sort of track? At Sea-Tac airport…

  8. I would love to see the Ballard-West Seattle line use similar technology as the Copenhagen Metro. Its a smaller-capacity system, but has automated trains that can run 24 hours a day at a high (2-4 minute) frequency. This would be a perfect capacity for a secondary market like this.

    It’s too late to fix some of the problems with the existing Central Link corridor without an expensive grade-separation project. That said, if a second tunnel through downtown is needed, we would now help to rectify this by building a smaller tunnel and limiting future staff costs. One only needs to take a look at the recent BART labor issues to see the benefits and cost savings of driverless trains.

    1. Actually, a larger downtown tunnel would be better. We will need more capacity for future lines and you can only build so many tunnels. Whatever tunnel we build for Ballard to West Seattle also needs capacity for at least the SLU/Aurora corridor.

      1. I will never understand the argument that the growth in transit use will be so massive as to require trains every couple of minutes… but that we also need massive, exorbitant, inconvenient-to-access stations for all the people.

        Do overbuilder-advocates think riders will be hanging out on the megaplatforms for the fun of it?

    2. You planning on using contract temp labor to build, wire, and maintain your railroad, Ryan? Careful. Good chance there’ll be a democratic revolution in China, and another one at McDonald’s before this one hits the drafting board. Unions have their reasons, Ryan, and loss of them has its consequences.

      Might also want to check your assumptions about which side of the BART dispute is responsible for current problems. Because however small your workforce, if your system is run by arrogant incompetents regardless of their number, results will be the same regardless of control technology.


    1. Haha – If we’re waiting for the end of complaining to build things we’ll be on track to build nothing, ever. :)

    2. Ballard hasn’t stopped complaining yet. Hasn’t stopped the building either.

      Growth is going to happen whether we want it or not. I say we welcome it with open arms, especially in neighborhoods that want better service.

      Here in Northgate I would love to see more development… soon. We need a grocery store or two and I would venture to guess a number of folks wouldn’t mind a bunch of new apartments if they came with walkable grocery stores.

      1. Target and QFC aren’t good enough for you? :)

        I guess a lot depends on which part of Northgate you are on. I think it would be great if there was a Trader Joe’s inside the main mall or a grocery store towards the south end of Northgate (closer to the transit center). The QFC is out on the fringes, and just feels that way. This is in contrast to Lake City’s Fred Meyer, which I’m sure is just as far for a lot of people in the area, but it happens to be where streets narrow, which makes it feel a lot more convenient (and is a more pleasant walk).

      2. If combined with a cross I-5 bridge, a grocery store within 5 blocks of either side of I-5 would significantly increase the walkability of the apartments near the transit center as well as make the core of the Licton Springs neighborhood a lot more walkable.

        This would help quite a bit to reduce short distance car trips… right now the poor pedestrian experience at either the Northgate Way or 92nd crossing of I-5 has folks climbing into their cars for even minor amounts of shopping.

  9. If Sound Transit need a little local blog to tell it how to manage its agency and plan its future, what does that say about Sound Transit?

    1. Haha – Sam. If you need to be incessently negative on said blog – what does it say about you?

    2. So, I find all of this a bit disconcerting. I would assume that Sound Transit has experts, and those experts do the math, and have figured out whether driverless technology makes sense or not. Which really leaves two possibilities: either we are wasting their time with these type of suggestions, or they really don’t know what they are doing. The first isn’t too bad (ask any planner and they will tell you that citizen input is often whacky) but the latter is really disturbing.

      Don’t get me wrong. I love citizen input, and I think it has its place with regards to the governing of Sound Transit. There are things they miss (just as the city council missed the fact that the monorail didn’t accept ORCA). There are plenty of things that Sound Transit missed (130th station, 520 station for U-Link, etc.) but in general, I would assume that most input is designed to steer the agency towards a different philosophy, as opposed to different technology. For example, we can ask Sound Transit to focus on more of an urban, rather than a suburban system. Or place more emphasis on bus to rail interaction. Or try and build stations that are more convenient for people in the neighborhood, as opposed to grand works of art. All of these things make sense. and involve trade-offs. But the choice between technologies, especially these technologies, seems so obvious. Either it saves enough money to justify the extra initial cost or it doesn’t. Is Sound Transit that incompetent, or are are we that arrogant?

      1. OK – So I’ll give a real response.

        Ross: Neither.

        The people at Sound Transit in planning are not incompetent. They are smart people who very much know what they are doing. They are working in a extremely political environment and many times are working within parameters that are outside of their control.

        What Seattle Subway does is change what those parameters are by building public support for higher quality transit and making noise about it.

        Keep in mind that Sound Transit is an extremely conservative organization after a rough start, are not a monolithic “they”, and have three counties worth of constituents and politicians to contend with – most of those constituents are suburban. They would never take on something as “controversial” as driverless trains without someone(s) in the community making noise about it. Despite the technology being prevalent all over the world for decades – other than Airports and the private monorail in Vegas – there are currently zero driverless train sets in the US. Hawaii will be the first when it opens in a few years.

        Seattle Subway are a group of fairly reasonable people. Despite how some would like to frame us, we are not rail extremist or egomaniacs. We fought for (both) Prop 1s and have some policy opinions coming up that will very much prove the rail point. Not to say we don’t have strong opinions about things and like rail quite a lot. We do.

        We are attempting to work within the political parameters that are laid out in front of us in the region and look for opportunities to expand upon what Sound Transit can do within Seattle.

        Part 1 of that sentence often puts us off with commenters here. Part 2 of that sentence puts us off with the political establishment – but we’re working on that.

      2. You’re mistaking Sound Transit for an agency like in Canada or Germany that can just hire Jarrett Walker to design the network and then build it, and sometimes has the same authority over roads and highways. Sound Transit is under the authority of political forces (created by the legislature, and subject to county/city politicians and voters, many of whom are suburban-minded). Seattle has only 1/5 clout in terms of population. Everyone recognizes the city needs more lines than the suburbs do and downtown Seattle should be the primary transfer point, but that doesn’t mean they go as far as supporting all the City Rail (tm.) and City First ideas. If you want to build a first-class city network from scratch, you don’t ask Sound Transit to do it, you do something like the monorail did. But the time to really plan it right was in the late 80s and early 90s when Sound Transit was being created and structured. We could have done something city-focused either outside or inside Sound Transit, either light rail or heavy rail. But not only did the suburbanites have most of the power, the city politicians and activists were not focused on that. Seattle Subway and STB did not exist. The almost universal goal was (light) rail to the U-District, SeaTac, and eventually Northgate, Bellevue, Everett, and Tacoma.

        So you ask how Sound Transit “missed” 130th (Pinehurst) station, or Summit/15th/23rd/Montlake stations, and the answer is, they were not goal. The goal was not to replace the 43 or the 520-downtown buses or the 522 or the 150 or 101. It was to replace the 712/72/73X, 41, 550, 510/511/512, 301, 4xx, 194, 7X, 9X and similar routes; and to replace (which it may fail at) the 57x and 59x; and to supplement (not replace) the 7, 8S, 36, 60, 66N, etc.

        As to the mode, that was driven by the desire to be street-running compatible, and that in turn was because of its low capital cost. All previous light rails in the US were 100% or 98% surface. Street-running is incompatible wth driverless, so that’s why driverless was thrown out. If Sound Transit had instead made its primary goal 100% grade separation and 55 mph minimum (as it should have), then it would probably have chosen a different technology.

  10. “Driverless trains can be run on our current system. ST is just saying they won’t consider it. Lets not get into a technology argument on this point.”

    Kyle, with exception of one or two pathetic Mark Twain imitations, I think I’ve made it clear that I’m willing to accept automated trains when conditions warrant and their planning, construction, and operation assures they’ll be safe, reliable, and efficient.

    I’ve used San Francisco’s simultaneous use of light rail subway-surface and easily-automation convertible BART as an example I think would work very well here. I’m also really curious about either transitional or permanent use of the modes on the same train- as Basell suggests we consider.

    But local prevalence of Karaoke makes me suspect that on this matter, Sound Transit is only lip-synching its risk management department, with mike-feedback heavy metal sound effects from the wailing of its attorneys.

    Especially if “our current system” includes the grade-crossed section of MLK. With right-of-way so incurably intrusion-prone that even above grade separation won’t be sufficient to keep your trains from being taken down by the buzzard droppings shorting out your pantographs.

    Except for a brief driver-operated trip through heavily crowded and jaywalked crowds in downtown Gothenburg, with an instructor pushing my hand forward to keep normal human flinch-reflexes from slowing the train, my driving days include only machines with relatively short stopping distance that can be steered around blockages.

    But friends of mine who presently drive CentralLINK have had enough non-preventable car-train collisions on MLK that I’d strongly advise you to run the paragraph quoted above past a ready-room full of train operators and give us a Sunday morning video of the meeting. It will also drive every single cat video off of YouTube.

    Since I still respect Ben Schiendelman, who introduced me to Seattle Subways, I’ll let every comment above be evidence that the I consider automated operation of a future system absolutely reasonable. But if I wanted to put public transit in the hands of anyone who considers technological discussion an irritation, I’d be handing out Monorail petitions.

    At least somebody would have to hit the accelerator like Elliot Blues going for the drawbridge to have a train hit or get hit by a car. And the impossibility of texting while climbing would prevent the next girl from walking into the side of a moving railcar.

    Mark Dublin

    1. Haha – Mark. I don’t mean it as an insult. Just making a point about whats possible. We know that driverless trains are possible with grade crossings.

      There are nowhere near the variables that exist for a car in traffic and it can react to random things in the limited ROW has to deal with faster than a person can.

      The only question is whether or not we choose to go down that path. My point is that – by the time ST would actually consider running a driverless train where there is a grade crossing (20+ years before any new lines might interline) – this will no longer be even remotely a point of contention.

    2. It is my understanding that BART trains are (or at least were at project rollout in the 1972) indeed “automated” , but carry an operator for safety/backup reasons. The Automated Train Control system was the subject of engineering safety concerns and a managerial/political coverup. The ATC system failed during an incident where a train over ran a station.

      1. The BART design, copied by Washington DC Metro, was *not fail-safe*. It was a bad design which should never have been permitted. Quite shocking, really — but that’s what you get when you have a bunch of airplane engineers try to design a rail signalling system without ever asking railway signalling engineers how it works.

        A properly designed system is designed to “fail safe” — when a component fails, the system moves into a safe mode. The classic example is that trains *do not move* unless the signal system is assertively sending a “you may move” signal.

        Believe it or not, the original BART design does not do this — the signal system has to assertively send a “stop” signal or the train keeps moving. Grade A design idiocy. This was patched up in the 1970s for BART… and just a couple of years ago for DC Metro.

        (Incidentally, the majority of model railways use non-fail-safe signalling. I was thinking of designing a demo system for a model railway just to teach kids how failsafe signalling actually works, but it’s a lot of work.)

  11. And like real authors say at the front or end of all their books, all typographic and spelling errors are my responsibility. Sorry I left of the last letter, Basselle.

    And Keith, since this isn’t the movies or the non-passive aggressive part of the West, last-name address wasn’t intentional. And not even official correspondence uses last name and “Mr.”

    But aside from that, Keith, I won’t move a step off anything I wrote here tonight. Every time and every place this topic comes in front of me.


    1. You do not mean to say that you refuse to learn things, right, Mark? That you are unwilling to acknowledge new information that might lead to new insights, that might lead to a natural evolution in your understanding and your prescriptions and views?

      Because that is what you just said.

      I couldn’t care less about this driverless debate, as even Seattle’s best proposals lack the demand for unsupportable-by-human-drivers super-frequency. Many of the applicable segments are so short that they could easily be driven by 2-3 drivers at any hour of the day. It is far more important to plan infrastructure that is right-sized to need and to whatever amount of money is rational to spend in order to meet that need. If that demands limited (well-designed and prioritized) surface segments, so be it.

      But I would caution you, Mark, about viewing every choice through the outmoded lens of “make-work” pork. Mandating and maximizing unproductive labor has never been a sustainable strategy, nor an especially progressive or ethical one — why subject your populace to unneeded and unnecessary functions, when there is so much better work that needs to be done? Even organized labor has come around to rejecting this strategy as hurtful to its cause, and as a factor in the its losing political image war with blue-collar populations in sectors it ignored for too long. Even social-democratic states have discovered that over-hiring mandates and forced inefficiencies are unsustainable without the expectation perpetual growth fed by limitless natural resources.

      The domination of economic policy by a quasi-libertarian plutocracy has not been good for the general welfare, and as you know, cities like Seattle have seen some quite unintended adverse outcomes from our moment of inequitably-distributed e-commerce largess. But nothing good or just has ever come from Luddism. Moving backward has never been the answer.

  12. I hope they do fix the PSRC model. Most population projection models are fairly poor, but the PSRC model seems to be exceptionally crummy. (I attribute this to over-complexity; you can get a good idea of what’s going on with a really simple model and lose it completely with a model with too many bells and whistles).

Comments are closed.