It turns out the Eastside Rail Corridor is good for at most 5,000 daily riders as a commuter rail line, which is about one rider for every blog post and op-ed that’s been written about it since BNSF decided to sell it a few years back.

That’s one of many things we learn from Sound Transit’s study of high-capacity corridors on the Eastside released last week. The study looked at two corridors in particular, the former BNSF corridor and I-405. As you can see from the chart below, the ERC doesn’t perform nearly as well as I-405 does.


There are several reasons for this. The less expensive version of the ERC alternative is single-tracked, which means trains could only run as often as once every 20 minutes. Ripping out the tracks and installing BRT or light rail improves ridership, but at greater cost (and encounters some issues with a narrow right-of-way).

But perhaps the biggest take-away from the 405 vs. ERC bake-off is the importance of path dependence. BRT on 405 is cost effective mainly because Sound Transit and WSDOT have already made massive investments in the form of HOV lanes and park-and-rides. ST has already invested $350M in the corridor and WSDOT is likely to spend as much as another $1.4B adding HOT lanes to improve bus and carpool reliability. So BRT ends up looking cheap, but only if you ignore the billions already spent on widening I-405 over the last decade.

Nevertheless, the cost-effectiveness means this is probably the end for the ERC as a viable Eastside transit option. It was a compelling idea – a nearly unbroken right-of-way through the Eastside just waiting to be redeveloped. But compared to the alternatives, it just doesn’t pencil.

Which leaves us with 405, and BRT. Should WSDOT move forward with the express toll lane project, BRT would have great reliability and decent ridership at relatively low cost. Sound Transit looked at two BRT options on 405. The first is a single line going from Lynwood to SeaTac – essentially all of 405. Buses would run every 10 minutes at peak. The other option is an intriguing trunk-and-branch setup, where feeder lines throughout the eastside would combine for 5-7 minute headways in the central Eastside. The second option would add another 4,000 riders, but cost nearly 50% more to operate.

The sad part about any kind of transit on 405, of course, is the abysmal walkshed. With a few exceptions (Bellevue Transit Center), the walk shed around most of the proposed stations makes Mt. Baker seem like a walker’s paradise. Great for park-and-rides, but not much else. But again, you build BRT with the infrastructure you have.

The full presentation is below.

  STCentralEast Level2Pres ERC405 043014 MASTER (1)

97 Replies to “Path Dependence: HCT on the Eastside”

  1. Looking at the map, it appears that the intent for routing LRT/Commuter Rail through Bellevue using the ESR corridor, does not really use the ESR Corridor, and instead uses the Vision line alignment. This is understandable as a new rail bridge would have to be created over I405 at the site of the Wilburton Tunnel, and for LRT, the Wilburton Trestle would be a pain to address.

    For LRT anyway, why not turn west at I-90 and connect to East Link through DT Bellevue at the South Bellevue Station? This would also be a step in the direction of an Issaquah extension which has been frequently mentioned.

    1. An Issaquah – Bellevue – Kirkland line is being studied separately, and I presume it would most likely share the East Link track between South Bellevue and Hospital.

      For a north-south line, also doing that is an intriguing idea. Except that ST seems to have ruled out LRT on 405. It probably found ridership lacking. Thus we’re back to BRT, which can’t use East Link’s tracks.

      1. Mike,

        If ridership is lacking in the I-405 corridor — a parking lot many days — how then will Issaquah – Bellevue – Kirkland pass the necessary threshold for LRT?

        I don’t see it. Are the planners expecting BCC students to walk to Eastgate Station (nee “Eastgate P’n’R”) to get the train to downtown Bellevue and Kirlkland? Because if they’re not, who exactly is going to ride this whiffy new LRT?

    2. Not having direct service into Downtown Bellevue in a the rail alternatives that are this expensive is such an omission that it makes me wonder if the alternative is being deliberately sand-bagged.

  2. Redmond calls it the Central Connector. Kirkland calls it the Cross Kirkland Corridor. Tracks have been ripped out in many places. Trains haven’t run on it in years, and probably never will again. Instead of continuing to call it the Eastside Rail Corridor, can we call it something more fitting, like the Eastside Central Corridor Connector Trail and Greenbelt?

    1. Or how about the Once Continuous Great Northern ROW around Lake Washington that has been cut into so many small pieces, even Humpty Dumpty has a better chance of being made whole.
      But I guess there was never any demand for travel in the corridor – just look at the pathetic usage of I-405 on any given day, or never ending plans to widen it to LA Freeway standards.
      Jeeesh! This region is pathetic on long range planning. Now they are finally talking about capacity of rail in the Green R. Valley, with a future passenger only track. That should have happened on day one of Commuter Rail planning, rather than pay $50 million a pop every time ST wants to add another train to the schedule.

      1. Thanks – was that the electrified section or the Ceder River section to Stampede Pass?

      2. The electrification was on the Milwaukee Road line from Tacoma through Renton (now the UP line), from Renton through the Snoqualmie tunnel and onward to Ellensburg and east to Othello (now mostly a *trail*, sigh). Also from Black River Junction near Renton to Seattle (now UP line).

        The Northern Pacific had an entire inland mainline from Vancouver, BC, which would be quite useful in these days of rising sea levels and collapsing bluffs. It starts at Mission City, BC, and heads south (this part still exists) to Wickersham, then passes east of Mount Vernon via McMurray (this part was ripped out from roughly Mt. Vernon to Arlington), parallels the Great Northern line through Marysville and Everett, then heads east again along what we call the “Eastside Rail Corridor” all the way to Renton, where it rejoined the Great Northern. For Seattle service, a branch followed the other shore of Lake Washington from the north, what’s now the “Burke-Gilman Trail”.

        For some reason BNSF and its predecessors, and also Amtrak, had a preference for the Great Northern lines over the Northern Pacific and Milwaukee Road lines. The Great Northern has a tendency to run underneath unstable bluffs and along flooding river valleys, and the other two lines reached more cities.

      3. Nathanael,

        VERY well done with one slight correction. The main NP line did not pass through Everett and Marysville but instead paralleled what is now SR 9. It is now the “Machias” and “Centennial” Trails. You can see it on Google Maps.

        I have an old 1930’s period NP map and it shows service to Everett on branches from Snohomish and Arlington but I believe that the Snohomish-Everett service was rights on the GN; there’s an old wye just south of Snohomish where the NP crossed the GN.

        The branch from the north starts just south of Arlington and survives. It merges with the old GN line a bit north of the enormous casino. I do not know if it was an Arlington branch owned by the GN or was owned by the NP. The former I’d imagine.

  3. I’m really disappointed by this study. It’s almost like they intentionally want this to fail. Compare this to the west Seattle one where they actually came up with different options that had truly different routes/walk sheds/characteristics. Here, they stuck almost religiously to one of 2 corridors that we know are not ideal: the eastside rail corridor or 405.

    If we ever want light rail or high quality BRT on the eastside it will almost certainly have to leverage portions of these ROW because they are there and cheap. However, it would be silly to stick to it. I think the highest ridership would probably be accomplished by:
    *From the Landing (connecting to the LRT there from Renton, Southcenter, etc…), go north on the rail corridor or 405 to coal creek
    *Jog east slightly (to either 124th ave or factoria) to serve factoria mall, t
    mobile etc…
    *Turn west along I 90 to hit S bellevue PR (this is important because it will connect here to east link and a future issaquah link)
    *North along the east link corridor back to the esr corridor once you’re past Hospital station
    *Stops at S Kirkland PR, and Carillon point (connection to hypothetical SR 520 HCT too)
    *Deviate west on either: 68th > State or 6th ave > Kirkland way
    * Actually serve downtown Kirkland/Kirkland TC
    * Travel east along Central Way/85th back to ESR
    * North to 405, with one or two stations at Totem lake
    * North to SR 522 where it goes west, with station in DT Bothell, and a connection to woodenville-bound HCT
    * Return back to 405 either by Beardslee, or (more expensively) Bothell-Everett hwy.
    * Station at Canyon Park w/ connection to SWIFT II
    * North along 405 to Lynnwood TC

    Now, that would be more expensive due to deviations from ESR or 405 — but it is at least worth studying the difference and relative advantages, disadvantages of such a routing. I hope that ST does some better investigations here.

    1. Stephen, low ridership doesn’t deter you from wanting north/south rail on the eastside?

      Also, please explain how train tracks are going to get to Carillon Point? Are you going to run a train down Lake Wash. Blvd?

      1. My point is we don’t actually know how low ridership is because they didn’t even study a routing that actually hits the all major nodes of the Eastside. Also, I don’t care so much about between LRT and BRT — in my mind the routing would be the same, it’s just a choice of rubber versus rails.

        For Carillon point, the ESR corridor is less than 1/4 mile away. That’s close enough IMHO, you just need to invest in some good pedestrian connections.

    2. If the project is approved, there will be a more detailed Alternatives Analysis as required by the EIS, and it can consider these deviations. Its budget would just have to be high enough to allow them.

      1. The project = 405 BRT. It’s not tied 100% to the freeway; that’s just the starting point for discussion. But the other project, a route primarily on the rail corridor, seems unfeasable now.

    3. You lay out a sensible route, if money was no object and we were looking to build ridership over a long time, while taking a huge hit to the budget in the meantime. For me personally, something like this plus East Link would allow me to finally get to work by transit (Bothell to Redmond) faster than riding my bike. But it would be a horribly bad return on investment in any realistic timeframe.

      A smarter strategy would be a full BRT build-out (the most expensive and highest-ridership option), along with the possibility of a start to the LRT long-term vision from Issaquah to Kirkland. After getting to downtown Redmond from Overlake, connecting downtown Kirkland seems like the highest-priority LRT project on the Eastside since there’s actually some density and walkability there. I’m just not sure where else you’d go on that line. I guess maybe Issaquah via Factoria and Eastgate. But everything between South Bellevue and Renton is 5-6 miles of low-density sprawl, and I can’t see LRT through there ever making sense.

    4. Stephen– The studies were exactly what the ST Board asked to be studied. The 405 BRT study was included in the ST2 Plan and scoped narrowly to respond to previous decisions about the preferred transit option in that corridor. The ERC study was directed by the Board last year after the corridor was acquired and ST3 planning was green-lighted.

    5. “If we ever want light rail or high quality BRT on the eastside it will almost certainly have to leverage portions of these ROW because they are there and cheap. However, it would be silly to stick to it. I think the highest ridership would probably be accomplished by:
      *From the Landing (connecting to the LRT there from Renton, Southcenter, etc…), go north on the rail corridor or 405 to coal creek.

      This study was done, during the I-405 Corridor Program completed back in 2001.
      Except for the part I specifically re-posted above, since people of enough influence nixed studying the ERC in that segment.
      Sound Transit penciled out the cost at ~$4.5 Billion.
      Ridership numbers didn’t justify the ‘travel time savings’ for the 30 year horizon, which is why it never went into more detail after the C/B analysis was complete.

      Face it, Freeway-BRT is what will be on the ballot, unless there is some grass-roots support by the citizens of the Eastside to have rail in the ERC, regardless of what any Cost/Benefit analysis shows.

      Not In My Back Yard.

  4. Seems like it’s pointless to invest in transit from Renton to Bellevue. If we ditch the I-405 idea and focus on Bellevue, Kirkland, Bothell, and Woodinville, I imagine that would be considerably better. If these are the options on the table, I’d rather we skip the N-S investment, it ain’t worth it.

      1. This is a different kind of report than the ones you’re probably thinking about. It’s is an unfinished progress report to the board, to show them the scope of what’s coming. It’s not a formal draft for public hearings and board decisions. That will come later this year. That’s where you can mix and match between the alternatives or say that none of them get close enough to activity centers. If the project is approved, there will be a more detailed Alternatives Analysis, which can study additional features.

        Also, this and the other studies are less extensive than Ballard-downtown and will have only one round of hearings each, because nobody put extra money into them like Seattle did for Ballard-downtown.

    1. Note that this study is separate from a forthcoming study to look at U-District to Kirkand and Redmond, and other options along that corridor. Ridership should look very different.

      1. Agreed. This has been my major problem with the ERC work done to date which has assumed it will be used to connect to Bellevue, but anyone who has used transit service in the Kirkland area know the ridership is going to Seattle.

    2. Renton to Bellevue may not pencil but what about Redmond to Seatac? Connecting major global businesses to an airport seems beneficial to the political class that decides these things, no? It also has value in that a lot of voters who aren’t able or interested in using transit daily might try it and like it when flying out of town, and then be more supportive.South of Bellevue is the most difficult part of I-405 to expand – a choke point that should make it easy to advantage transit. LRT could double frequency between Redmond and Bellevue without additional infrastructure cost, then serve Factoria and tunnel through the hill with maybe one stop between there and Renton. A few stops in Renton, Tukwila and then you’re on the main line headed south through Seatac. It would connect to a future Issaquah – Seattle or Issaquah-Bellevue-Kirkland….. route.
      Bellevue to Lynnwood can be BRT for a while. The biggest issues there are location of existing stations (outside lanes of the freeway), a lack of stations between Canyon Park and Lynnwood and the lack of HOV connections from I-405 to I-5 which make it impossible to use the direct access ramps at LTC or Ash Way. It would be easy to build a freeway station at Swamp Creek which then provides a great route – connecting ST 522 at UWBothell – Swift 2 (Canyon Park), 164th St (frequent bus to Ash Way and LTC… Swift 3?) and Swift 1 (SR99 – Lincoln Way). If the line keeps going to Boeing then there’s no need to divert the LRT mainline from Seattle to Everett. It could even hook around on SR 526 and terminate at Everett Station. Given the time the 532 currently takes just to get in and out of Ash Way, splitting it into one route serving Everett Station and SEFS and another taking 405/SR525 would improve travel times from SnoCo to the Eastside. It would also move eastside parkers out of Ash Way and over to Swamp Creek where there is unused capacity. If Ash Way absolutely must have eastside service then the NB ST 535 could come off SR525 at Alderwood Mall Blvd. go north, turn right at 164th, left at Ash Way and use the direct access ramps to go south to LTC.

      1. Even without an I-405 line, Redmond would have decent service to the airport via routes 542 or East Link over to Central Link. Given the sub-par ridership of the 560 and 566 today, I’m not sure that a separate route along 405 would offer much bang for the buck.

        It should also be noted that serving Renton is a huge time sink. Just getting off the freeway, waiting at all the stoplights, then waiting again in the long line of cars to get on the freeway adds a good 15-20 minutes minimum vs. just staying in the 405 HOV lane all the way. And this assumes negligible dwell time at at the Renton bus stops themselves, an assumption that would cease to be true if Renton had enough ridership to make those stops worthwhile. This is why Sound Transit recently moved some of the 566 trips into a separate route 567 that skips Renton.

  5. How much impact would the TIBS / Renton Landing segment of the South Seattle study, impact the numbers on the Renton Landing / Bellevue Segment of the eastside study.

    These numbers should be included in this report, as at least an addendum.

    1. Why did they only look at Full commuter rail build out for the Commuter rail option. Wouldn’t a DMU be a better option in this corridor?

      1. That would be something a region would start with if they wanted to grow their rail based infrastructure.

      2. Very valid points, Lor. I particularly note that the Renton-Burien proposal is for only BRT or LRT and completely ignores a through-routed DMU option. While Sound Transit can study everything, it does leave me wondering why the two parts of this corridor is studied as two different rail technologies – even though the studies are simultaneous. I wonder if the de facto reason is not to optimize a rail alternative, but to make ridership look at low as possible. That could also be why the commuter rail option doesn’t actually stop in Downtown Bellevue.

  6. There are a number of examples of reasonably well run single track passenger railroads. As a sampling. try this introduction to a route in Germany:
    Then, there are such operations as the New Jersey RiverLINE (essentially diesel light rail). For that matter, the first MAX line in Portland operated trains once every 15 minutes over a single track line with passing sidings, on the section that had bee adapted from a freight railroad.

    1. The New Jersey River Line is “reasonably” useless and empty. Except by the standards of the Northeastern Megalopolis, by which it is extraordinarily useless and empty.

      1. Agreed a DMU commuter line would be a huge waste of money. Far better to invest in proper BRT and express bus service. Invest in eliminating choke points with HOV direct access ramps, bus lanes, and signal priority.

        The same applies to West Seattle and South King County. Rather than build LRT any further south than SODO in ST3 it makes far more sense to invest in BRT and express bus infrastructure.

        The East Link extension to downtown Redmond and completing the “spine” to Tacoma have a lot of political support which means they will very likely be in ST3. In the case of the former the line isn’t particularly bad in terms of cost-effectiveness either. But let’s not squander billions on lightly used rail lines when there are far more cost-effective transit investments to be made.

      2. The RiverLine in NJ specifically serves “underserved communities”.
        Camden and Trenton appreciated it. The bedroom communities in between were the worst sort of auto-sprawl, but they’ve had decent commuter uptake of the line. I actually know a few people who commute on it daily.

        There are certainly better routes, but there probably aren’t better routes *in South Jersey*. Politically this was very much a matter of “geographical equity” when it got built.

      3. That’s nice. It still cost more than $1 billion for 4,875 daily users, it still has a 7% farebox recovery ratio, and it still (somehow) manages to be useless to both event attendees in Camden and government employees in Trenton!

        And it doesn’t even particularly serve the sprawl. Half its trip is through polluted riverfront wetlands, industrial sites, and absolute nothingness!

        But thanks for reminding us, again, that you have “a couple of [internet] friends” on every dumb rail project in America, which for you trumps any desire to be smart about transit investments ever.

  7. Serving Downtown Kirkland/Google is a huge omission, especially since the study area includes the Rail Corridor which skirts tantalizingly close. I guess the Sound Transit board isn’t interested in whacking another Surrey Downs-like hornets nest. Can’t blame them.

    Given the massive ongoing investments in 405, BRT seems like a good choice here. The best we can hope for is a well executed transfer point at 85th connecting bus service along that corridor. The Totem Lake freeway station is a good example of what’s possible, although the pedestrian crossings are still inadequate. (Transferring from a SB 311/532/535 to any route heading East on 128th is a PITA)

    1. The ERC study does appear to include a stop as near to downtown Kirkland as the ERC goes. I actually think that’s near enough to be useful to a lot of people.

      The 405 BRT map shown appears to include a stop roughly lined up with Kirkland. I can’t tell whether it’s 85th or 70th. My guess is 70th. Pedestrian conditions at NE 85th are even worse than at NE 8th in Bellevue — 85th doesn’t have sidewalks from 6th St. until just east of the interchange, and much of that is on a bridge, so making it even barely adequate would require lots of capital improvements. 70th has an existing freeway stop and P&R with paths to connecting buses (as well as more frequent service than 85th), though I doubt it’s anyone’s favorite transfer (I actually don’t know whether the 245 or 238 deviate into the P&R).

      Capital improvements to the 70th St. stop could improve transfer conditions a little. But the distance between either stop and the actual destinations of Kirkland severely limits its impact. Some retail businesses in the older parts of Kirkland have real employee parking shortages (particularly as the city is more concerned with keeping parking available for short-term users) but basically anyone outside of there or in a newer building is committed to providing free employee parking forever. Transit that provided a quick hop between Bellevue and the core of Kirkland could have a real impact, but the opportunity there is fairly narrow — driving on the eastside is convenient and cheap enough that a freewayside transfer (or the slowness of the 234/235) makes transit hopelessly uncompetitive.

    2. “Serving Downtown Kirkland/Google is a huge omission”

      That is exactly the kind of feedback we can give ST. “It doesn’t go close enough to major activity centers.” It’s the same issue as routing Link on Eastlake vs Broadway, or I-5/65th vs Roosevelt. If enough people tell ST that any alternative must serve downtown Kirkland even if it costs more and increases travel time, then it’ll either be serve Kirkland or nothing.

      What would the travel time penalty be of a route that exited 405 at 70th, went up State Street past downtown Kirkland and back on Central Way to the freeway? With only two or three stops (e.g., 68th, Google, downtown) similar to south Bellevue way. How much could street improvements or new on-ramps mitigate the time penalty?

      1. Google’s a small employer. Maybe 600 employees in Kirkland. The majority of them drive. A fraction take transit. It would be idiotic to divert 405 transit to Google.

        Thousands of employees at Southcenter. The majority take transit. Why no call for 1-5 routes to detour off the freeway to serve the mall? Stop kissing techie ass.

      2. I’m assuming Google will continue expanding, and other employers will locate on State Street too. The main issue is Kirkland Transit Center, not Google. But if you’re going so close to 68th and Google anyway, it makes sense to stop there.

      3. Southcenter is outside this study area, which is really between Renton and Lynnwood. The Burien-Renton Link or BRT would have a Southcenter station. Eastsiders working in Southcenter may have to transfer in Renton, but that may be acceptable, and it wouldn’t be much of a burden if they’re both frequent. A one-seat Renton-Lynnwood route seems to fit more of the Eastsiders’ trip patterns better than a one-seat Bothell-Southcenter route.

      4. Kirkland is just too much of a deviation for a bus route going all the way from Lynnwood to Renton. It’s too much of a deviation for today’s bus routes going from Lynnwood and Everett to Bothell, even! Traffic on 70th/68th and Central in particular is really slow during rush hour, with no room for bus lanes. And that’s after you’ve slogged through the general-purpose lanes of 405 to get to the exits.

      5. This is where targeted investment in eliminating choke points can do some good. Would accessing downtown Kirkland be as much of a problem if there were well places direct access ramps to the HOV lanes in 405? Heck, let’s go even bigger, how about a short transit tunnel through downtown Kirkland? While it would be expensive it still would be cheaper than building light rail between Kirkland and Bellevue.

  8. If we can run LINK on the I-90 floating bridge roadway lane, and on Rainier Avenue, why can’t it be run on a BRT lane on a highway?

    1. Because moron SOV drivers will be using the “Express Toll Lanes” (aka “HOT lanes”). The presentation states explicitly that this is the plan, and that the lanes will be dynamically tolled in order to achieve 45 mile per hour operation for the BRT buses. While LRT trains can brake much more effectively than heavy rail commuter trains, rubber tired vehicles can decelerate much more rapidly than they can.

      While it is true that relatively narrow LRT trains operated in mixed traffic in the Upper Noe, Oceanside and Sunset neighborhoods of San Francisco and single cars run in several stretches of street in Philadelphia and Cleveland, in all these instances speeds are 25 miles per hour and slower.

      The fact that you asked the question clearly puts you in the company of those drivers, John.


        Not sure about theoretical deceleration capacity- anybody driving LINK want to weight in here? But long ago rode on the Norristown line out of Philadelphia.

        Specially designed for awesome rates of acceleration and deceleration, trying to run an extremely fast car with closer spaced stops than usual for fast service.

        Original cars were designed using a wind tunnel in the 1930’s. Newer ones now.

        But to keep this on-topic, the Norristown line is 13 miles long. Very heavy rail right-of-way, mostly straight track if memory serves.

        So can anybody think of any similar distance on the east side that fits above description? Bothell to Stevens Pass via 520 and SR 2 longer and steeper- but would definitely mean less car lanes needed.

        Mark Dublin

      2. I’m not sure of the relevance, but Mercer Islandd P&R to Issaquah Highlands P&R is about 12 miles. But there’s not really any suitable ROW other than I-90.

      3. Mark,

        John asked if LRT could run “in a BRT lane”. But that isn’t the proper question to ask, since the study states unequivocally that the BRT buses, should they come, will run in the “Express Toll Lanes”. Those aren’t “BRT lanes”; they’re tolled transit lanes open to SOV traffic that pays the toll.

        So, the answer is not just “No”, but “Of course not!” Regardless how fast the Norristown High-Speed Line cars can stop, rubber tired SOV’s can stop lots faster. Steel on steel is never going to have the friction of rubber on concrete or even asphalt so the LRT’s will inevitably rear-end cars when accordion accidents happen. Or, the LRT drivers will be ordered to maintain a long stopping separation from the car closest in front and drivers will use the slip ramps to jump in front of the train. And that’s not even addressing the problems inherent in having untrained SOV drivers bouncing back and forth across the slippery and doubtless slightly recessed LRT tracks.

        It. Simply. Can’t. Be. Done. Safely.

    2. It’s not clear to me whether I-90 will be recessed rails like the DSTT, or traditional raised rails like in the spur tunnel branching off from the DSTT. Buses can drive in the former but not in the latter.

      1. What’s a plinth?

        Webster: “1a. The lowest member of a base. b. a block on which the moldings of an architrave or trim are stopped at the bottom. 2. A usually square block serving as a base. 3. A cours of stones forming a continuous foundation or base course.”

        I don’t even understand the definitions.

  9. Given the low ridership projections, a more incremental approach to extending LINK on eastside may be the way to go. How about just do slide 18: LRT from Bellevue to Kirkland/Totem Lake? The cost estimate still seems a bit high for ridership. But, it does connect with and naturally extend the LINK network. For example, it ties Kirkland with all the destinations that will be served by EastLink like downtown Bellevue, Overlake (MS campus), as well as Seattle destinations served by Central and U-Link. It also could connect actual or potential urban destinations and not just highway exits.

    1. I’d just point out that Downtown Bellevue is not a stop on this Totem Lake alternative. Since transfers are being forced at the Hospital Station on the other side of 405, the potential ridership forecast is likely quite low. Actually, that appears true for all the rail alternatives presented here.

  10. Problem of population centers so far from nearest freeway seems bad enough it’s worth some serious money to fix. Even average park and anywhere near an Interstate costs huge service hours fighting traffic.

    Lynnwood Transit Center is an exception- and still takes too long to access. Two solutions, depending on location:

    1. Elevated fully-reserved flyovers.

    2. Like Freeway Park downtown, structures built over freeways. Could be problem for residential re: air quality, but for other uses, wonder if it would be worse than average street?

    Saved operating time would pay for a fair amount of construction.

    Mark Dublin

  11. There are three Eastside studies going on. (A) Diagonal on Issaquah – Bellevue – Kirkland (LRT, maybe sharing East Link stations between South Bellevue and Hospital). (B) North-south on Renton – Bellevue – Kirkland – Lynnwood (BRT probably). (C) East west on Redmond – Kirkland – UW (LRT, maybe truncating or deferring the lake crossing).

    Lines A and B are parallel between South Bellevue and Kirkland. That raises the issue of one cannibalizing ridership from the other, and which one would have higher ridership. My instinct is that line A (Issaquah – Kirkland) would have higher ridership, especially if it shares East Link’s stations between South Bellevue and Hospital. So I would start with that line, and then make line B (north-south BRT) complement it.

    That implies three things: (1) don’t worry too much about the BRT routing between Bellevue and Kirkland because that’s just the inner gap to support the outer segments, and the line will inevitably be lower ridership in any case. (2) The Kirkland LRT terminus can later be extended many ways, north to Totem Lake or Lynnwood, east to Redmond, or west to UW. That would supercede the north part of line A or all of line C. (3) The South Bellevue LRT transfer point can later be extended south to Renton, superceding the south part of line B.

    So maybe we just want a minimal, cheap north-south BRT for the interim, without too much worry about its Kirkland routing?

  12. Assuming LRT for all of the below.
    Assuming that the state gives ST the additional taxing capacity that would be needed for ST3.
    Assuming that ST3 will not pass in Snohomish unless Link reaches Everett.
    Assuming that ST3 will not pass in Pierce, unless Link reaches Tacoma.
    Then it would be safe to assume that East King would have far more than enough to do The MSFT – Redmond extension, and Kirkland – Issaquah utilizing ESR North of Bellevue, and Eastlink in Bellevue.

    This leaves a probably large chunk of money on the table for East King to start working on another route.
    I do not see 520 being affordable, unless it is the only project that east king supports, and my understanding is that East King would end up shouldering the full cost.
    This leaves something either south of Bellevue towards Renton, or North of Kirkland toward Bothell/Woodinville/Lynwood?
    I do not see Eastside to Lynwood until ST4 (unless Snohomish ends up with extra money to run a spur down 405 to Canyon park.)
    Bellevue to TIBS via Renton is probably a far more realistic solution for the excess income within the ST East King subarea.

  13. Question.
    North King subarea only pays for the track up to it’s last station, and the other subareas pay for everything within North King from the last North King stop onward into their subareas.

    How will this work with other subareas?
    lets say a Lynwood to Kirkland extension is built.
    East king will pay for everything through UW Bothell.
    Snohomish will pay for everything Lynwood to Canyon Park.
    Who pays for Canyon Park to UW Bothell? (assuming a I405 alignment)

    1. The reason North King doesn’t pay for the edges or ST Express is that Seattle is overwhelmingly a destination rather than an origin. The 550 is the only route that approaches equal bidirectionality. So the question would be whether Bellevue plays a similar role for Snohomish and South King. I suspect it does a bit but not enough to be decisive. There’s also UW Bothell in King County though. I suspect East King and Snohomish would each pay half rather than getting into a trivial dispute over it. It’s a tiny portion of the total cost.

  14. Thinking about this some more, here’s a long-term Eastside vision.

    405 BRT is the spine. It doesn’t have the ridership to need trains, but it could be upgraded to such in the future if it ever becomes necessary.

    Instead of adding any north-south LRT corridors, plan for East-West LRT lines. They would be:

    * Revised East Link: This is East Link to Hospital station, but instead of going to Overlake/Redmond from there it takes the ERC to downtown Kirkland (with a deviation near the end to get to actual downtown Kirkland).
    * 520 Link. Extend the Ballard to UW line across the 520 bridge to Hospital station (possibly adding a northern approach to Bellevue that could have another Bellevue station) and then take on to the portion of East Link from Hospital to downtown Redmond . The big problem there is how to get around possible tunnel capacity problems through the U-District and its stations. The alternative is another lake crossing (via Sand Point/Magnuson Park) but that seems unlikely.
    * Lake City/Bothell Link: This has been discussed. If ridership supports it, you could extend to Woodinville eventually.

    To serve Issaquah/Eastgate/Factoria you split off some East Link trips or run a shuttle to South Bellevue. Alternatively, you could have an all-Eastside line running Issaquah to South Bellevue, then from there along the current East Link alignment to Redmond.

    405 BRT or eventual LRT would be much more useful with these east-west connections. In combination pretty much every dense(ish) area but Crossroads is served.

      1. That’s what the plans are.
        Lack of funding is the only reason it isn’t happening now.
        Lots of flow improvements are what you are seeing, but the big plan is 4 GP lanes added to the Corridor.

      2. 4 new GP lanes?! WTF?! In what sort of world does this make any sense?

        As much as we debate the cost of transit projects here, 4 new GP lanes on 405 isn’t going to come cheap. You can build a lot of rail and grade separated BRT for what adding to this car-sewer is going to cost.

        Lets not forget the environmental costs, especially with the melting of the West Antarctic ice sheet.

      3. Because the ‘problem’ that is being solved is traffic congestion.
        The COST is the $ amount for infrastructure.
        The BENEFIT is in what’s called ‘Travel Time Savings’, which makes the assumption that commuting is a non-productive endeavor.

        Adding a lane to a congested freeway immediately draws (roughly) 2500 people (assumed SOV’ers), away from the current lanes.

        No need to change where you live, or change the way you commute. No need to convince anyone.

        When a cost benefit analysis is done, the horizon year is an important consideration.

        For the I-405 study, back in 2001, 2030 was that year.

        A cost/benefit ratio of less than 1 is what is desirable.

        The 4 GP lane + BRT is what was chosen.

        Just putting only LRT in the corridor had a dismal C/B ratio. (Don’t ask what the actual numbers were, I haven’t looked at that documentation in ages, but I’m sure it’s online, somewhere.)

        Adding 2 GP lanes, + LRT was above the threshold, but what was interesting was that the 6 GP lane with no transit improvements option had the same ratio.

        The problem is that larger infrastructure investments (regardless of mode) don’t pay back in the 30-yr time frame.

        It’s why I was suspect of the North Corridor Link study (the time span).

      4. Everyone realizes that the Transportaton Package in Olympia will provide (at least a portion) of the funding for the lane additions on 405 if the gas tax increase is put in place.

      5. Adding 4 general purpose lanes adds next to zero car-moving capacity. By the time you have that many lanes, weaving movements eat up all the capacity.

      6. Let’s not forget adding additional lane causes induced demand. Adding 4 lanes just means 4 more lanes of cars crawling at the same slow pace as today.

        As Nathan said if there is any additional capacity the weavers will just eat it up.

      7. Maybe I’m missing something, but I can’t find any documentation of 4 additional GP lanes. Looking at the WSDOT project pages, I see 4 additional lanes vs. the 2005 baseline. Two of those lanes were built around 2007 or so. The other two are under construction. The two new lanes plus the existing HOV lanes become the new toll lanes.

  15. This is another example of doing studies backwards. It’s the “we have a track, so can we use it?” question, rather than the “where do we need transit connections?”. It does provide some valuable initial assessment of costs and concepts, but it really lacks a more strategic Eastside discussion with all of the cities involved at the table before a consensus alternative can be developed.

    1. @ Al S.
      It was studied.

      It’s called the I-405 Corridor Program study.

      Metro and the cities looked at it in 1992, and during the above study, it was brought up again.

      At that time the preliminary analysis showed commuter rail in the ERC with a direct connection to (what is now) the Tukwila Sounder station and up to Woodinville at roughly 3100 riders a day.
      It was not as extensive a study as the later PSRC/Sound Transit one.

      It never made it to the Cost/Benefit stage, and if you dig deep enough you’ll see why.

      Renton and the Kennydale Neighborhood Association sent letters to the program’s Executive Committee asking that using the corridor for any future transportation considerations be taken out of the study.

      End of story.

      1. I don’t think Downtown Bellevue had nearly the density in 1990 that it does today and parking costs were probably much cheaper. I don’t know how bad the I-405 traffic was in 1992, but I’d imagine that it’s more congested today, too. That changes minds.

      2. If funding was in place, the extra lanes would already have been added by now.

  16. We have well over half of LRT track built or under construction to provide a Renton-Bellevue LRT service and we don’t have to go up the Eastside Rail Corridor. All if would take would be adding a line from Renton to Central Link, run up the Rainier Valley portion of Central Link, build a new connection near the Mt. Baker Transit Center up to East Link, and then connect with the East Link line to points on the Eastside. I always find it to be an obvious omission that this concept is never evaluated in any I-405 corridor studies. I think this option should at least be assessed in this study.

    1. Great idea. Absolutely should be studied.
      and +1 on asking the wrong question. I learned that during grad school in Madison, WI. 20…now 30+ years of studies, millions of dollars asking, How can we convince the feds to buy us LRT ? – on this rusty old, single track, freight line limited to 10mph. Instead of: How can we improve transit to make it a viable option for everyone in this relatively small, compact city with an overrepresentation of students and 0 car HH’s. Certain persons insisted on trains, not buses because the tracks were already there and the “development potential” but the same people showed up to council meetings to oppose every new building over 4 stories. The mantra was “oh but we need trains because we have transit ridership typical of a city 3x our size.” My reply: “but we don’t have the tax base of a city 3x our size!”

      1. Madison needs trains, but what it needs is *intercity* trains. It’s small enough that bus lanes could probably handle the local traffic…

    2. Your shortcut would also make Eastside to Sea-Tac service possible.


      it will be impossible to add the interface because it hasn’t been designed into the Rainier Valley Station or the Mt. Baker Station. Had both those stations been “stacked” it could be done, but it’s too late,

      The same error is happening in the U-District. Both stations should have been stacked so that a Ballard-Bellevue (or Kirkland) line could share platforms with North Link. We’ve had this discussion before and everyone called me stupid for advocating a non-crossing junction for a proposed “Ballard Spur”, but a crossing junction immediately adjacent to a busy station on a line with four minute headways is going to be a nightmare.

      Now if the Ballard-Cross Lake line takes its own independent path east of 45th and Brooklyn, it’s a moot point. But that means a Magnuson to Kirkland alignment becomes mandatory and forces a bypass of the main CBD on the east side.

      1. I’d agree that it’s costly, but not impossible. Given the scale of the costs on the presented alternatives, it might be feasible or at least in the ballpark.

        I’d also suggest a much less costly variation of the concept, which is to have an above ground split south of Mt. Baker station that proceeds up the MLK median, and then turns west on the I-90 lid to end at the Rainier/23rd Link station. The line could run from Renton, connect into Central Link through Rainier Valley then split again near Mt. Baker. This would mean that Renton-Bellevue travelers would have to transfer at the Rainier/23rd station. The advantages are that it doubles the frequency of Rainier Valley service, and provides more direct Eastside service to all of Southeast Seattle as well as Renton, and creates a Renton rail line that doesn’t need to use the DSTT.

      2. I have been concerned about U-District station and a Ballard line for quite some time. I’ve asked ST to include in the station design potential expansion to a future crosswise platform, but the answer I got was, “No east-west line has been approved by voters yet, much less decided where it would go, so it would be premature to design that.” That seems to me backward, because this is the most critical transfer point in north Seattle.

        But there’s another issue that may interfere with it even if ST did try to design an expansion. Namely, the obstructions on both east and west (the UW Tower foundation and whatever’s on the other side), and the historic apartment building south. The station box has only like five feet of leeway. So there may not be room for a second platform in any case.

        That potentially leads to the very worst scenario, a separate station a block away, where you have to come up to the surface, wait for a traffic light and cross a street, and go down to the other station. That would dissuade half the transferring ridership right there.

      3. Given the potential for an Eastside light rail line at some point, I’d add the Downtown Bellevue station decision to build two side platforms rather than a center platform as another design flaw that will haunt system expansion for decades. That pretty much precludes a second LRT line from ending at this station, like the one studied here that would run from Bellevue to Totem Lake.

        Does anyone notice how the LRT alternatives proposed here don’t actually go to Downtown Bellevue (the densest development spot on the Eastside) and they don’t use the existing Link corridor between Downtown Bellevue and South Bellevue?

      4. “Why would we possibly want to engage in the sort of rudimentary future-proofing that the Brits figured out 150 years ago?”
        — geniuses at Sound Transit

      5. There’s no reason for a line to terminate at Bellevue TC. Just like how all trains that go into the DSTT will end up in Northgate, all trains that goes to South Bellevue can end up at Hospital. You can put a branch stub at either of these if necessary, because in the future they’d become extensions to Kirkland, Issaquah, or Renton.

      6. Mike, what do you think about the side platforms at Northgate? Given the operations proposal to turn half of the trains around there, I think it’s going to be an operational disaster once half of the non-peak trains turn around there and the other half go to Lynnwood.

      7. There really is nothing to prevent putting in a junction anywhere in Rainier Valley. It is a surface line with grade crossings so flat crossings wouldn’t further limit headways.

        Due to those limited headways a line to Renton is going to involve a transfer anyway, at least during peak headways. So you’d probably want a station adjacent to an existing one. Most likely near RBS.

        I don’t really see the point of providing one-seat rides between East Link and South Link or to a hypothetical Renton spur. A track connection between I-90 and Mt. Baker isn’t going to provide enough travel time savings to gain much in the way of ridership for the cost.

        I suspect any station for an E/W line between Balllard and the UW will be under 45th. The U-District station box ends south of the Neptune so there is no conflict other than passing over/under the North link tunnels. A short pedestrian tunnel can connect the two mezzanines. Sure it isn’t as nice as a shared or cross-platform transfer, but it isn’t the end of the world.

        BTW what. Is with the obsession with providing one seat rides between any two possible stations on a rail based transit network? It really isn’t necessary. Rail to rail transfers are some of the easiest to get riders to make. Real rail networks assume people will transfer to get where they are going. Endlessly branching and running every possible combination of endpoints leads to joke headways like in Dallas or Denver.

      8. Center platforms are always better because they allow the widest variety of transfers or turning around without leaving the patform. But anywhere trains turn back, there would have to be turnback tracks somewhere so I’m not worried about that. Is Northgate really side platforms?

      9. My mistake. I could have sworn that I saw a drawing with side platforms at Northgate, but I checked the latest design on line. It turns out that it will have center platforms after all. Whew!

      10. Re Northgate: The Everett extension report said East Link trains are going to Lynnwood, implying full time. I don’t know if the speaker simply forgot about the off-peak turnbacks ot ST has changed its mind on that. In the latter case there may be no Northgate turnbacks after all.

        Re Renton: the missing piece in your Renton – Mt Baker – Bellevue proposal is that Renton is (A) not very large (compared to Bellevue, Kirkland, and Redmond), (B) not particularly transit-riding (compared to Kent, Bellevue, Redmond), and (C) not particularly important to North King (for the Mt Baker – Rainier bypass). It’s just too far from Bellevue when East King has higher priorities, and the 101, 106, and 560 show Renton won’t even use off-peak routes when they’re available (unlike the 150, south half of the 169, 550, B, etc). The Renton-Burien line has a greater chance because Southcenter is much closer to Renton than Bellevue is, and the 140 is more heavily used.

  17. I was depressed on reading this, but on a careful read of the study, I think we’re missing half the picture.

    We’re not really looking at the optimal alignment yet. There’s a combination of ERC and I-405 that works much better than either the ERC or I-405 alone. Everything in this study is paying minimal attention to Kirkland proper. Of course it’s hard to justify transit on the ERC if you exclude the biggest population center north of Bellevue. I think we’ll have a better picture when the Kirkland-Bellevue-Issaquah study is rolled out.

    If that one is as negative as this, then maybe it is time to write off rail on the ERC, but right now it’s premature.

    One other thought. Getting rail to Woodinville is really hard to justify on its own terms (ridership will never be there), but it does get you within a half-mile of the Snohomish County line. If I understand the financing, I think that means Snohomish could then get rail into Bothell/Canyon Park at a rather low additional cost? Expect to see some interesting politics around that.

    1. You could bring it close to Bothell’s downtown. Right at Bothell Landing…
      On the other side of the Sammamish River. The ROW is somewhere in the weeds there.

  18. Perhaps the highest and best use of this Eastside corridor is for coal and oil trains bypassing Seattle due to the highly explosive nature of these cars. As a benefit, several coal slurry and gas pipelines could also be built along the ground and new electric lines overhead.
    This could go a long way to launching the next Renton PSA campaign of being Waaaayyy ahead of the Curve.

  19. The Eastside rail line remains the only freight rail bypass of the unstable bluffs between Seattle and Everett. Its abandonment as freight rail is extremely short-sighted and it should be restored as a through-freight route.

      1. With the track ripped up and a missing crossing of I-405 at Wilburton, not to mention the 100+ year old trestle, it will be an expensive proposition to run a freight train on it.

Comments are closed.