Governor Inslee and Premier Clark at the Cascadia Conference in September 2016 (Microsoft Photo)

Looking to foster greater ties with our Canadian neighbor, last fall Microsoft sponsored the Emerging Cascadia Innovation Corridor Conference in Vancouver. Many of the issues discussed were as you’d expect: the flow of skilled labor between the U.S. and Canada, easing of trade restrictions, and pre-emptive fear of the then-ridiculous prospect of a Trump presidency.

Seemingly out of nowhere, however, one of the most prominent topics was high speed rail between Portland, Seattle, and Vancouver. The mood was optimistic, with Governor Inslee, Premier Clark, and Executive Constantine joining regional tech leaders in a roundtable to lay out the vision. To say the least, the discussion was high-level; you couldn’t make out a single tree in their visionary forest, and the working paper Parsons Brinckerhoff prepared for the conference similarly lacked much technical substance. Many of us shrugged it off as loose boilerplate from our regional governments.

Then last month, Governor Inslee requested $1 million from the Legislature for an initial feasibility study, with the report due this December. The budget request is making its way through committee, with a lukewarm reception. When Microsoft testified their willingness to chip in for the study, Senate Transportation Chair Curtis King (R-Yakima) quipped, “We’d like your contribution to be $1 million.”

Vancouver in an hour would transform the region, but the prospect of building it faces enormous technical and political challenges. Besides the assent of Olympia, Victoria, Washington, Ottawa, and every other micro jurisdiction along the way, you’d need funding mechanisms for tens of billions of dollars that currently don’t exist and expertise that is weak on both sides of the border.

Current BNSF Alignment from Seattle to Vancouver

But the real challenges lie in designing and engineering the corridor itself. The current Amtrak Cascades service (157 miles, 4 hours, average speed 39mph) traverses BNSF’s legacy track, which hugs the shoreline wherever possible and never gets above 130′ in elevation (even Pike Place Market is higher). The result is predictable and unfixable: 14 additional miles compared to I-5, meandering curves, reduced speed, limited right of way to build parallel HSR track, and a corridor not at all future-proofed against rising seas.

So any HSR solution clearly lies inland, but that comes with its own challenges. Do you serve Seattle, Everett, Bellingham, and Vancouver? All of those lie at sea level along the coast. Do you weave back and forth? If not, which cities do you bypass and how? Even if you figure that out, do you access the central cities of Seattle and Vancouver or do you build peripheral stations and force transfers to Link or SkyTrain? Even California High Speed Rail is able to use existing right-of-way between Burbank and Anaheim, a luxury neither Seattle or Vancouver would have. As Mark Hallenbeck rightly quipped to the Globe and Mail, “The kicker in all of this is not the 100 miles in the middle; it’s the 30 miles on either end.”

So is it hopeless? No, but it’s damn hard. Here are a few ways it could happen. Let’s start in Seattle and move north. #long read

First things first: you wouldn’t use King Street Station. Built at sea-level on tidal fill, there would be no plausible path to access the likeliest right-of-way along I-5. Doing so would require a 4th tunnel that would cross underneath the Great Northern Tunnel, the Downtown Transit Tunnel, and ST3’s future Green Line tunnel, which would require a deal-breaking 7-10% grade. Instead, King Street would become the commuter terminal, much like Manchester – Victoria plays second fiddle to Manchester -Piccadilly in the UK. We’re gonna need a new station.

Stadium Station Option?
Midtown Station Option?

Where to put a central Seattle station? I see two possibilities: Sodo and 6th/Madison. If you assume the use of the I-5 express lanes, which seems by far the likeliest pathway, you either need a long descent to a sea-level station, or you need a high station. If you do the former, trains arriving in Seattle could conceivably punch under I-5 near Columbia Street and quickly reemerge to take a steady descent to a Sodo-area station. A Link transfer at “Stadium Intermodal Station” might work.

If you stay high, you could build a lidded station just south of Freeway Park. Call it “Seattle-Midtown Station”. Transfers would be provided to the Green Line and Madison BRT for quick access to King Street/Int’l District if continuing on a local train, as is done all over the world. The station box would likely be built atop I-5, with the passenger tracks below I-5, and central elevators/escalators in between. Certainly unusual, but not impossible.

The third option is an outlying terminal with a forced Link transfer. These could be in Redmond, Lynnwood, or a number of other places. This would exponentially reduce costs just as it radically increases total travel time.

North from Downtown, the next big obstacle is the Ship Canal Bridge. Do you retrofit it for trains, or do you build a new span? Retrofitting it may not be as impossible as it sounds. Free of freight traffic and the onerous weight requirements of the Federal Railroad Administration (FRA), HSR could use very light Spanish or Japanese rolling stock. Washington’s max truck limits are currently 99,000 lbs GVW (~50 tons). The latest Shinkansen N700 trains weigh 715 tons, distributed relatively equally over a 16 car set (only 44 tons per car). In addition, each railcar is longer (82′) than the max allowed for trucks (61′), distributing the weight even further. Though adding rails and catenary wire would add a great deal more still, the best hope for an easy Lake Union crossing would be to retrofit the bridge.

Shinkansen N700 Stock – Wikimedia

North of the Ship Canal, you could either use the existing I-5 express lane right-of-way at sub-HSR speeds, or you could make selective use of it and tunnel as necessary to avoid the curves at Green Lake. And of course, north of Northgate, Link will mostly hug I-5 as well, setting up numerous conflicts. Other napkin-planners, some of them relatively sophisticated, get around these problems with sleight of hand, by tunneling all the way to Everett. 

Once in the north Seattle suburbs, the primary enemy is elevation. South Everett lies at 500′, and Everett Station is at sea level. Assuming you want to serve Everett Station and maintain a <2% grade to get there, you’d need a 5-mile approach that likely involves significant tunneling.

At subway costs of $500m-$1B per mile, we would have already run up a $14-$28B tab. At more typical HSR costs of $100-$200m/mile, we’d be looking at $3B-6B for those first 28 miles. These eye-popping figures would make suburban termini very attractive.

Once in Everett, your wetland problems begin. But we’ll save that for Part 2.

Part 1: Seattle to Everett
Part 2: Everett to Bellingham
Part 3: Bellingham to Vancouver
Part 4: Terminal options

168 Replies to “The Technical Challenges of Seattle-Vancouver High Speed Rail: Part 1”

  1. There’s no reason to build a new station. Trains won’t be barreling into the city at 200+ mph. Plenty of examples of HSR that share very old stations with non-HSR lines.

    Anyway this entire study seems like a huge waste. Incremental upgrades are the only politically feasible option and BC refuses to even do that. Once trains pass the border north of Blaine it is a long, excruciating crawl into Vancouver.

    1. While that may be true, I think the bigger issue with using the existing track into Seattle is that it’s way out of the way and has issues with mudslides and the like. If you want to serve the metro area, then a stop at Lynnwood would be necessary. It makes the train substantially more accessible to the Eastside and to North Seattle.

      Of course, the real way to have solved this would’ve been to use regional trains for the trip out to Everett instead of Link and let HSR use those for the last few miles into Seattle downtown.

      1. So build another tunnel that links King Street Station to the I-5 ROW. We’re already talking about billions of dollars to get to Vancouver.

      1. They’ve had several decades to do so and have yet to budge. It took a lot of work just to get them to pay for border agents for the second daily run. As far as I know BC currently contributes nothing to Cascades.

      2. If so, this is a new attitude. Washington has been incrementally upgrading the track ever since Cascades started in the 1990s. BC has hardly done anything at all. If you take Cascades it’s only a couple hours to Bellingham and a bit more to the border, then it crawls for an hour to Pacific Central Station. Sometimes it waits on the east side of the Fraser River for a freight train to come first.

        I don’t expect this proposal to go anywhere because of the cost, international cooperation, and BC’s lack of investment in the existing service. But it may lead to a less ambitious measure, such as 110 mph Cascades from Bellingham to Eugene that’s the state’s long-term plan anyway. It has been humping along in fits and starts depending on the enthusiasm of the legislature, but a large cash infusion could accelerate it. And the state could buy the BNSF track and make it primarily passenger rail, and freight can shift to the UP track. BNSF and UP already work together on Washington’s freight load, the way Cingular/AT&T and T-Mobile built joint towers when they started GSM cell service in Washington.

      3. Yes move BNSF to the UP track. Why didn’t ST do this for south line sounder? It would get the freight traffic out of 4 downtown areas (Puyallup, Sumner, Auburn and Kent), and allow for an amazing commuter/all day service, and from what I’ve read about this idea it would of cost less than connecting Link from the Airport to Tacoma, Link could then have looped around from the Airport via Tukwila and back onto it self somewhere, or maybe even via Renton and South Center mall.

        Instead they have built a 3rd mainline, which could have been spent on double tracking the UP line for them to share.

      4. One cannot presume anything about the current BC government’s commitment to transit. They are marching ahead with Metro Vancouver’s second 10-lane bridge and freeway project (at one place with an illogical 21-lane interchange), and transit has always taken a distant second place to creating excessive road space like it was the 60s all over again.

        Moreover, the region was forced by Premier Christy Clark to hold a referendum on a very modest 1/2 of one percent additional tax to fund the Metro transport authority’s 10-year transportation plan with roads getting off scott free. There was a scant five months to prepare, and the vote lost on everything but transit (essentially a rage fest). In many people’s view, the referendum was an excuse for the government to abandon their duty to better the common good through the policies they are elected to decide on. Metro mayors must now contribute more from the region where senior governments take 90% of the tax revenue.

        Now there is an election campaign for a May vote and the BC Liberals (in fact, they are conservative and have been ruling for a long 15 years) have promised funding for two major Metro transit projects (a SkyTrain subway in Vancouver’s Broadway corridor and LRT in Surrey). But on closer examination the funding falls 7% short of the tax revenue stream going to the provincial government while the locals have to make up what the province pockets. This follows on the heels of a fresh commitment by a new federal government to focus on urban transit, and to the great relief of transit agencies and the mayors, they agreed to a 50% funding share, directly proportional to the level of revenue they take from the Metro.

        Given the current BC government’s penchant to favour mega-freeways, it is highly unlikely they will go near upgrading or creating a new rail line south to the border, and probably don’t have the vision to estimate what the world will be like by mid-century, and will likely say the private sector should build it all. Meanwhile, the BC debt is set to increase by $20+ billion in mega-highway projects built, under design, or on the books.

    2. The Vancouver region would like a whole new line from the border north. One reason that the current line has not been upgraded is that it isn’t really upgradable. It runs right along a busy waterfront, then through a swamp, and then over the oldest bridge in the region. That can’t be made faster without a considerable redo. Plans for such a thing have been kicking around for a while but contingent on money and integrating with other plans.

      1. I would note that the 1950s timetable for streamliners had Main Street to Sapperton in 20 minutes. They formerly did squeeze better performance out of the existing route,

      2. So they want it but they don’t want to pay for it. Got it. That’s exactly what I’m saying.

        Meanwhile, Washington and Oregon have been paying for new track for years.

  2. This concept is a political nonstarter without heavy subsidizing of cost from both the US and Canadian Federal Governments.

    The topographical challenges, coupled with property issues and the even more fun prospect of getting around other vital pieces of infrastructure to create a working line are pretty numerous.

    If Cali can’t get adequate funding for HSR from LA to SF, I just don’t see this as something happening in my lifetime.

    IF it does, however, it should be positioned in a way that allows for expansion southward, to Portland and Northern California at later dates.

    1. High speed rail at 200+mph isn’t likely and wouldn’t change things. Why does technology leave its adherents dazzled and hypnotized? Is it the money, the exclusivity of pricey tickets, a lack of understanding which issues most needs change? My own big project idea is to rebuild the RR lines south of Eugene, through Medford, Grants Pass, Ashland, Mt Shasta, Oregon small towns that have more population than Klamath Falls. I’d also like to see the Amtrak Pioneer returned to service with Talgo trainsets. My recommendation for Amtrak is 2-train daily service in each direction to allow passengers the choice of overnight hotel accommodations and catch the next train 12 hours later. Sleepers are too expensive and the recliners are cramped and uncomfortable. I’d like to see the Superliners replaced with Talgo trainsets along these lines. Talgo company has its XXI locomotive that should gain enough speed to reduce trip time. Give 200mph HSR a rest. It’s not gonna happen and those who don’t appreciate the Amtrak Cascades as it is should be ashamed.

      1. High speed for induced demand. Compete with air travel speed and make it easier and therefore more likely for people to travel between cities, tying the region closer together

      2. So you’d send the Coast Starlight over Siskiyou Pass and abandon the Natron Cutoff? Would that be faster, or just serve more cities?

        My biggest dissatisfaction with the Coast Starlight is timing. It’s never going to be anywhere near comparable to the speed of flying if you’re traveling between major cities (where more passengers actually are). One big differentiator is the view, and in the fall and winter you don’t get to see the area between Klamath Falls and Sacramento in either direction because it’s dark. Last time I took that leg was in early December and it was too close to the solstice to see the sun rising over Mt Shasta, which would’ve been lovely.

      3. Planes fly at 600 mph. HSR in the Eugene-Sacramento gap would require a massive federal investment that China would do but our government has other priorities. Even assuming it could run at 200 mph the entire way in spite of barriers like the Siskiyous, that’s still 4+ hours to San Francisco if I recall. Current planes take 1.5-2 hours, and they used to take 1 hour before they slowed down to save fuel during the post-9/11 year of plummeting ridership.

      4. @Mike Orr do you have a timetable showing 1hr flights between SEA and SFO, because pretty much impossible that it happened. Planes do fly at the most efficient speed, which isn’t always the top speed, but there isn’t a way the flight could be made in an hour.

        Top speed of a 737 is 564mph (airspeed, ground speed may vary, but prevailing winds usually aren’t North/South in this area, so they shouldn’t affect the speed notably.) SEA-SFO is 679 mi.
        (Just for comparison here are some top speeds of some other airplanes: B727 : 570 mph, DC-9: 558 mph, MD-80: 504 mph, B707: 600 mph) The speed of sound is ~767 mph.

        So even if a plane instantly reached cruising speed on the runway and you boarded it a second before it took off, there would be no way to get between SEA and SFO on a sub-speed airplane. The Concorde wasn’t usually operated on this route, and who knows how long it’d take after it went out over the water, sped up, then slowed down to land?

      5. When I used to fly between Seattle and San Francisco in the 70s and 80s it took slightly over an hour and LA took two hours. But for the past decade and a half San Francisco has taken two hours.

      6. Why, why, why do people use the speed of flying from one airport to another and consider it relevant? Do you live at an airport? Do you work at an airport inside of TSA security? This number is irrelevant!

        For business travelers it’s important to know how fast you can go from downtown to downtown. For non-business travelers it’s from their home to some other home or hotel.

        As such it takes about 40 minutes to take Link to Seatac, 10 minutes to walk to security, 30 minutes through security, 10 minutes to your gate and 30 to board. Then it’s 2 hrs to SFO in the air, 30 minutes to get off the plane, 10 to the BART and 30 downtown. That’s 5 hrs and 10 minutes!

        Granted, a train isn’t going to go from Seattle to San Francisco in 5 hrs and 10 minutes as it would have to average 156 mph. However, if it could average 136 like the TGV (probably still not possible) it could do it in 6 hrs and be a wonderful trip, much better than flying.

        Still this is not the best use of HSR. San Francisco to LA is good, LA to San Diego is good. Portland to Vancouver BC is good. All of these would decimate air travel between those cities just like it did between Paris, Lyon and Marseilles.

      7. I’ve done Seattle to San Jose business travel as a day trip. 45 minutes to get to Seatac (light rail or cab, the company is paying), 10 minutes to get through security (at 6am), 30 minutes to board, 1.5 hour flight, get off the plane at 8:15 and walk 10 minutes to the office park. Or take VTA rail to your destination elsewhere in Silicon Valley. Head back to the airport around 5pm for a 6:30 flight back. It’s a long day, but it’s doable.

        Anyway, I just don’t see rail as a viable option for that corridor on speed. Things that it could win on:
        1. Passenger experience. The current route of the Coast Starlight between Eugene and Sacramento doesn’t parallel the freeway and gives access to views you wouldn’t get any other way. North of Klamath Falls is a long stretch of remote forest, and the rolling valleys of northern California are very scenic as well if there’s enough light to see them. On the train, you can get up and walk around, you can bring your own food and beverages on board, you have real leg room. If we could get speeds between Seattle and San Francisco up to the point that it’s not an overnight trip, it wouldn’t need to be faster than flying.
        2. Less security. Especially for people who get profiled a lot or who have their names on a list. Not having to deal with the TSA is a big perk. This one isn’t as much of a factor for business travelers, but for people who need to get around and tend to get hassled a lot, the train is better. I haven’t tried to travel on Amtrak without any ID, and people seem to usually choose Greyhound for that, so Amtrak is the middle option for less but not zero security.
        3. Cost. This is also a reason people take intercity buses long distances even though it takes longer. Amtrak charges much less for the first tickets sold than the last, and it’s hard to know when you need to buy and when last-minute travel is going to be a reasonable option. I’ve paid $80 to go from Seattle to Eugene, one way, and I’ve also paid $120 to go all the way from Seattle to LA. HSR tends to be at least as expensive as flying, so we’d lose that advantage if we upgraded.

    2. Last sentence, Sean, I’m with you 100%. Only way, economically, that this project will ever make sense.


  3. And of course so called “High Speed Rail” (not all that fast) may soon be obsolete technology if the Hyperloop concept proves viable. In theory, Lighter, faster, and cheaper than centuries old technology. It is certainly worth consideration.

    1. The cost of “HYPERLOOP” goes up exponentially the longer the line is. Keeping a line in a vacuum-like state while factoring in heat expansion of material over a long distance and danger-proofing the ENTIRE LINE is unbelievably expensive.

      I’ll put it this way, a single failure in any part of the line would not only kill anyone in the tube, but the rapid re-pressurization of the tube itself to 1 atmosphere of pressure would destroy the entire thing.

      All it would take to theoretically kill hundred of people and destroy billions of dollars in infrastructure is for someone to shoot it at any point of the line with a high enough caliber bullet to pierce the hull.

      1. The other big and oft-unstated problem with hyperloop is an aesthetic and comfort issue.
        While all the artistic renderings show a transparent tube, there are no transparent materials strong enough to withstand the air pressure from the outside. This means, in reality, we’re talking about a steel tube. So, imagine, as a passenger, being in a car inside a tube that has zero natural light and no view of the outside. This doesn’t sound like the future of transport, but like a sci-fi distopia,

      2. Sean, engineers were experimenting with trains centuries before it first become usable for public use. Even once trains were running for commercial use, the rate of technological advancements in trains and rail led the first constructions of rail to become obsolete even with the same concepts.

        Let wealthier countries experiment with Hyperloop, like Dubai is, and let use contract countries that have perfected HSR (France and Japan) and do what needs to get done ASAP. We can’t screw around for decades trying new technology… we need to build the rail, it gives us a permanent right-of-way for future tech advancements.

      3. While I am suspect of the hyperbole loop, lets not blow things out of proportions.

        @Sean F.
        While bombing the tube could kill a whole trainload, something messing up the path of the train is a danger for current high speed rail too. But lets not exaggerate what a bullet will do to a metal tube at relatively minor pressure differentials.

        @Donde Groovily
        You should specify no economically reasonable for hundreds of miles of tube transparent materials strong enough to withstand the air pressure. There are definitely transparent materials able to withstand far more pressure, for example, the Ictineu 3 can descend to 1200m which is 120 atmospheres of pressure

      4. Sean F-Rapid depressurization also kills everyone in an airplane. System structure would be maintained by breaking the line into segments with individual vacuum doors. Only a few sections would be near vacuum all the time. Since these are modular tubes several could be kept on standby, like train tracks are, for,rapid rebuilding. Technical issues are significant, but not insurmountable. All forms of transport are vulnerable to something.

      5. To put this into context, betting on Hyperloop now would be like betting on transcontinental jet airplanes in the 1930s. Maybe they’ll be viable someday, but first we need to set up an airport with the technology that is in production and is well tested now. Let Dubai, California, and the northeast corridor be the first to try out Hyperloop, then if it lives up to its promises we can build it here.

        We have jet airplanes already. The rest of the world has high-speed rail, we’ve just neglected for fifty years. A hundred years ago conventional national passenger trains ran at up to 120 mph. We aren’t even using the technology our own country had then; we’ve let the tracks go to rot, let the robber barons put freight first, and got into a race-to-the-bottom in freight costs. Europe puts passenger trains first with high-quality tracks, while freight goes mostly by truck. The US has instead gone for cheap freight. The trains are slow but reliable and they cost less than trucking for cross-country container volumes. That’s good because trains use less energy and have less environmental impacts than trucks. But they won’t improve the tracks to speed up the freight trains because their customers don’t want to pay more for shipping. If they want something fast they fly it or truck it. But containerized goods are like rivers; it doesn’t matter how long it takes one doodad to cross the country as long as a stream of them keeps coming. But all that has negative implications for passenger service, namely slow speeds that aren’t competitive with driving.

    2. I’m not convinced hyperloop is physically possible for passenger travel. You need at lease 6 cfm of fresh air intake per passenger. The only way to do this in a vacuum is through space grade oxygen exchange, which have never been built on a scale to support large numbers of passengers.

      It could revolutionize freight traffic though.

      1. Glenn, just about done specking out the second LINK Tunnel with a giant fan pushing a barrel on wheels with a conductor in a tight collar (found the Albert Eli Beach patent!)

        So docket just about clear for next project:

        We can use cars in the form of tubes with rubber gaskets around each end. And I think it would give Elon’s image some class if mechanism could be given the classic-cash-register-and-slot-machine treatment.

        But Mike, would settle for following measures on the Sacramento Eugene Portland, Olympia corridor. Which is really my preferred route to the Bay area. Since flying now violates the Fourth and Tenth Amendments and the Geneva Convention.

        Or was, ’til Cascades schedule made it even more unusable than the ride on Greyhound already had. So while local wind power company is working on our hundred foot diameter turbine to push the barrels, I mean trains, let’s do this:

        Southbound: Cascades to highway bus to Eugene for lunch. Coast Starlight to Sacramento ok because sleep all the way. Caltrain to Richmond, BART to SF.

        Northbound changes: Cascades, put schedule back so I can head north an hour after my bus pulls in. Like before. But best improvement, technologically easy.

        Sell Greyhound to an intercity bus firm in Turkey or Argentina.


      2. The tube train didn’t run in a complete vacuum. Fresh air intake wasn’t a problem there as the low pressure side was basically atmospheric pressure.

        Hyperloop puts that tube train into a completely evacuated tube so there is no air resistance. The assumption is that it takes much less energy to go fast in a tube that has no atmosphere than it does to create the vacuum in the tube.

    3. Hyperloop is not real. It was Elon Musk’s (now Trump adviser) political attempt to derail a zero emissions transportation method that was faster than driving one of his cars. If it were a good idea (or even possible) he’d be building it. He wanted to offer a pie-in-the-sky option that made HSR look bad so people wouldn’t vote for it. Once he did that he hasn’t looked back.

  4. A very interesting article, and a good framework for thinking about the challenges of fitting HSR into a dense urban environment like Seattle. One note, though, particularly regarding the Everett approach – while a 2% grade is the maximum associated with traditional rail, HSR implementations can use grades up to 3.5% with no impact and up to 5% if necessary, at least according to the CAHSR technical standards here: The overall point, though, still remains: getting the road back down from a typical elevation to serve Everett will be a challenge.

    1. Seattle isn’t that dense. And… I have a feeling that Seattle can squeeze in HSR somewhere if cities like London, Paris, and Amsterdam are able to do it.

      1. The difference is cities like London, Paris, and Amsterdam use existing railway ROWs to access their cities and typically run in blended modes along these stretches, only opening the throttle when outside of the urban core. We don’t have an existing ROW suitable. The best option is along the I-5 (perhaps with some deviations), but we’re busy building a *light rail* line along it instead of a heavy rail line for northern access to/from Seattle.

      2. @erentz The HSR would likely run down the middle of I-5 anyway. Link runs on either side.

        Its not the best place to have light rail, but it does offer transfer opportunities.

  5. I actually don’t think a suburban station, with a forced Link transfer, would really be that big of a deal provided that the suburban station is directly adjacent to a Link station, with no bus connection required to get between the two trains.

    The reason is that:
    1) Link already runs very frequently, so not much wait time.
    2) Most people would end up needing to transfer to Link anyway, since King St. Station is no one’s real destination.

    That said, high speed rail of any sort does look like a political impossibility. Most likely, when the existing tracks along Puget Sound get flooded from rising seas, they’ll simply cancel the service altogether (and, in the case of freight, switch to trucks at Everett), rather than pay for the cost of entirely new track through lots of hills in the middle of the city. I am also not at all optimistic of WSDOT ever agreeing to transform the I-5 express lanes to any kind of non-auto use. Even if they could somehow be convinced to do it, the legislature would simply pass a bill to prohibit it.

    1. What about at the I-90 and I-5 interchange? Link will be passing through there anyway on the way to Judkins Park station.

    2. If the coastal route becomes endangered, BNSF will have plenty of time to rebuild the old Milwaukee between Ellensburg and Lind and dig a new “base tunnel” under Stampede and have direct access to the Green River Valley. The biggest problems won’t be north of KSS, but south of there, where SoDo will be flooded.

    3. “King St. Station is no one’s real destination”

      Good point. People are coming from all over Seattle, Lynnwood, the Eastside, south King County, etc. Link is like a giant DSTT; it allows destinations and homes and parking to be distributed so that people can go to events anywhere along the line. So just put the station somewhere and you’re done. At first I was worried that an outlying terminus would lose half the ridership. But where are airports? In outlying locations. That doesn’t stop people from using them. And they have massive shuttle/taxi fleets and sometimes frequent buses and trains, and people use them.

    4. A station in Bellevue is also an alternative, if the ROW and geography of 405 is better than I5. Having a “freeway” station adjacent to Bellevue TC or Wilburton plugs well into the Link network.

      Can a new HSR line also do double duty as a local commuter rail line? For example, if the line is going to be built from Everett to Bellevue, it could be triple track (in parts) to allow for local runs and a bunch of local-only stations? The express runs would of course skip everything between Bellevue and Everett.

      Building a HSR corridor and a new commuter line (like the baby bullet) at the same time expands the use-case and spreads the capital dollars over more constituents.

      If the HSR goes between Everett and Seattle, it is “duplicating” Link so there really isn’t value to local stations …. unless it runs elevated on the 99 alignment??

    5. There was that conveyor belt idea between a 405 Link station and the transit center. It could be dusted off for HSR.

      An HSR “line” can’t do double duty because it would be silly to buy expensive high-speed trains for a measly 30-mile peak surge, but the ROW could theoretically be triple-tracked for local trains as you say. That’s what Caltrain is doing, a mass of parallel tracks (which it has already). But who would pay for the commuter rail part? Not the HSR budget or the state, and the ST district is supposedly taxed out with ST3 projects up the gazoo. Also, commuter demand to downtown Bellevue is a fraction of what it is to downtown Seattle. It’s a bit hard to judge because the transit is skeletal so people can’t use what doesn’t exist. But the ST Express runs that do exist aren’t very full. Tacomans have been asking for STEX to Bellevue, so that’s something that’s not being met. All in all it looks like we should just do 405 BRT, complete the HOV/HOT lanes north of Kingsgate or wherever the end, and see how its ridership grows before doing something major. A bullet train from Everett to Bellevue plus a 20-30 minute Link ride to Seattle is not necessarily better than a 60-minute Link ride straight south from Everett.

      1. Yeah the double-duty is the ROW, and the associated tunnels, bridges, overpasses, etc. that need to be built. Hopefully the electronics could also do double-duty? But the actual trains running the slow routes don’t need to be high speed.

        It’s basically an upgrade to 405 BRT. If you think 405 BRT is a waste of transit dollars, you won’t support upgrading the corridor to rail. Given 405 is as jammed as I5, I’m optimistic, but that’s an entirely separate discussion.

    6. I think an urban station is quite important: passengers will generally need to get to and from the station without a car, and there is no place with as many people nearby or as good of transportation options as downtown.

      A downtown location would make a train useful to almost anybody in the puget sound region, and incredibly convenient for anyone coming from or going to Downtown. A suburban location would be impractical for a substantial portion of the region and walking distance for a vastly smaller group.

      1. As I said – what makes a surburban station is practical is that it would directly connect to the regional mass transit system, with trains every 10 minutes, all day long, to take you into the city.

        A downtown station may be faster for people going downtown, but it would force large numbers of people who aren’t going downtown to backtrack, while increasing the construction cost by tens of billions of dollars to build new high speed track in a congested urban area.

        It’s also worth mentioning that, if done on the cheap, rail lines through urban areas divide neighborhoods a lot like highways do, since every crossing point requires building an expensive bridge. On way to avoid this, of course, is by building the rail line adjacent to a highway that’s already there, but the existing highway has curves, so any rail line that followed the highway would not be high speed.

      2. Having stations that don’t allow trough traffic have been a problem for Paris. You take the TGV to Paris and then an RER to another station and the TGV on. It’s a pain. They’ve been building around it by going to the airport and then bypassing the city for through traffic. I don’t think Seattle really has that option.

        I don’t see why the tracks out of King Street Station have to tunnel under BNSF’s. Just have passengers load on the east side of the tracks (with either a pedestrian tunnel or overpass). That track would tunnel east and up to I-5 and across the ship canal bridge. Everett would need a long tunnel.

        Vancouver would need a new station probably at Bridgeport. I don’t see this as a deal breaker as Vancouver would be a terminus, not a station along the way to somewhere else.

      3. Fundamentally, it’s all about cost. Yes, if money were infinite, there would be some value in having a train system that allows for thru-riders. But, given that we can’t even afford a high speed rail line without building tracks into Seattle, itself, there’s no way we can afford such a line with tracks in Seattle, itself.

        While it would be tempting to try and use the I-5 express lane ROW to cut costs, just Northgate to Everett, alone, would already be tens of billions of dollars, and I find it very difficult to envision WSDOT giving the green light towards just closing the I-5 express lanes to cars in order to make room for rail – especially, when they have a multi-billion 520 project in the works that will dump even more cars onto the I-5 express lanes.

        Capacity-wise, each lane of freeway can carry about 20 cars per minute, assuming a 3-second following distance, which translates into 1,200 cars per hour per lane, or about 1,500 people per hour per lane. Multiply that by 3 lanes per direction, and that translates into a full train of 500 people running about every 7 minutes. At least in an approximate sense, this is about the load that Link is expected to carry post 2021, with all the local traffic of downtown, Capitol Hill, Northgate and the U-district. There is absolutely no way that a Seattle->Vancouver line, which bypasses all local Seattle destinations, would carry anywhere near this many passengers.

    7. I would consider Northgate a suburban location, and in that regard, just fine as a station. Not only is it easy to get to (via Link) but fast. Let’s say I’m close to University Station and want to get to King Street Station. I’m not walking, but taking Link. That’s a three minute ride, plus the extra walking. Now, instead of that, I go north, to Northgate, which is 13 minutes. So I’ve saved ten minutes. Except that the train — even if it is going 200 MPH — is still going to take 3 minutes to get from King Street to Northgate. Then you have the walking time, which according to Google, is 7 minutes. You’ve actually saved zero time! Nothing!

      That is for a trip at the south end of downtown. Given the fact that the business center of gravity has shifted to the north, and the biggest pockets of density are in Belltown and South Lake Union, in an ideal world you locate that station it makes sense to locate to the north. You also have the Central Area (which includes Capitol Hill and its station) along with (of course) the UW. All of those stops are not far at all from Northgate, as long as — as you put it really well — the train station is directly adjacent to a Link station. Not three blocks away. Not a “shuttle express” ride away, but right next to it. The transfer itself is far more important than the location, as long as you don’t get crazy with the location (e. g. Everett).

      You really have two issues here, competition and new customers. From a competition standpoint, you have flying and driving. Flying consists of two groups — folks from Seattle, and folks making a transfer. The latter group will just keep flying. You’ve already gone through security, you’ve already made it to your airport, you are just going to keep going. The second group of flyers, though, suddenly have a very nice alternative. If you are south of the city and getting to SeaTac is easy, then maybe you just fly. But if you are in the heart of the city, or north of it, then this train looks really good. Downtown to SeaTac is a half hour (or more) plus the oft mentioned long walk to the gate. Extra security, mandatory early boarding, waits at the terminal, and that train from Northgate looks really good. It is just faster (oh, and did I tell you the train station is right next to the Link station? How’s that for handy?).

      From a driving perspective, it is easy — take the train. By placing the station at the north end, you avoid the time advantage that comes from driving (not having to go north to go south). By adding a station or two along the way, you pick up those people anyway (and the number of people north of the city limits is actually very small).

      As far as new customers go, Northgate is fine. The speed advantage (if you get anywhere near an hour) is huge. It is so much faster that the ride to Northgate is really no big deal. At that point, the biggest question is really how often the train goes to Vancouver, not where the station is. If it goes every hour, then a day trip is really a piece of cake. Leave your house at 8:30, be in Vancouver by 10:00 (easy). If you aim for a the 7:00 PM return train but miss it, you still get home by 9:30 or so. Not bad at all (count me in).

  6. Just stopping by to say I love this. Freeways make terrible corridors for urban transit because of walkshed. But for inter-state transit? It’s excellent. Freeways tend to be very direct, usually have nice sloping curves, are safe and with well stabilized and beds.

    Ideally you just build it above the median staying high. But if we believe rail capacity would be on the same magnitude as car capacity (which it will be), we could go cheap and steal a lane each way outside of core urban areas (though obviously tough politically).

    I’m looking forward to the series.

    1. Agreed. Great stuff here — I’m going to renew my subscription :)

      I’ll get into the weeds and share my opinion later.

  7. Thorough and interesting article.

    But I have to step back and ask why?

    How many people are actually travelling to Vancouver on a regular basis? Is there evidence of a strong economic connection between the two cities? It’s fun to think about, but I can’t help but think it unnecessary and that the money and effort is best focused on improving mobility in the Puget Sound first. I would love to see an article that makes an economic and social argument that high speed rail is necessary between the two cities.

    1. Even before Trump immigration was an issue for Microsoft. It’s the main reason they have offices in Vancouver and Microsoft isn’t the only one. Canada will always have more favorable immigration policies towards fellow commonwealth countries like India.

    2. “How many people are actually travelling to Vancouver on a regular basis?”

      Every single rush hour on I-5 between Everett and Olympia, and I-5/I-90 interchange and Snoqualmie Pass points to good answer, Dave.

      Three years ago, Olympia was a reliable and pleasant two hour bus ride, or bus and Sounder ride, to Seattle. Now, since I-5 runs slower than average parking lot lane from 6AM on, the “Reliable” part went same place as the “Pleasant” one.

      At this rate, well before ST-3 is finished, car-bumpers will be touched end to end between Nanaimo and the Oregon-California border.

      ‘Til we can get the beneficiaries of present economy to pay its costs to the rest of us, we might be money ahead just digging and elevating fastest line we can get.


      1. Well, since Nanaimo’s on the island, I’m hoping the bumper-to-bumper traffic stops somewhere short of there. ;)

    3. Cruise ships out of Vancouver BC generate most of the summertime ridership gains.

      Roughly speaking, Vancouver BC trains are about 1/2 full in winter,
      and the Portland – Vancouver BC trains (513/516) are roughly 25% full of riders making the complete trip.

    4. Sea-Tac records about a half-million passengers traveling to and from Vancouver every year. That market is only slightly smaller than Seattle-Portland, but falls well short of SFO-LAX (1.8 million) and even SFO-SEA (1 million).

      1. Not really a fair comparison, when you factor in driving. It takes over five hours to drive between San Fransisco and L. A. if there is no traffic (and there is always traffic). Flying starts looking really good at that point. Even with the hassle — getting to the airport, getting back, security, the long boarding and possibly delays, it is still faster nine times out of ten. In other words, lots of people fly because driving just isn’t worth it. The same is true for SFO-SEA.

        That isn’t true of Vancouver to Seattle. Driving still probably takes longer, but not that much longer. From, say, the UW to UBC, you would be better off driving most of the time. The fact that our airport is a ways out of town makes driving a solid option.

        That is what makes this ideal for high speed rail. It is the ideal distance. A fast train trip like this — even if it started at Northgate — is much faster than either driving or flying. You really can’t say that with many trips (even San Fransisco to L. A.).

      2. @ Charles So… Portland – Seattle – Vancouver might have higher ridership than California HSR?

        No. California HSR is not just about the express (L. A. to San Fransisco) but all the places in between, that are actually a better distance for high speed rail (e. g. San Jose to Fresno). Given the huge number of people in California along that corridor, the numbers for that train will be much higher.

      3. California has six times the population of Washington. It has hourly flights from SF to LA on multiple airlines, and some parents put their children on flights every weekend to other parts of California to visit the noncustodial parent. Currently there’s no good alternative to flying or driving. There are regional trains from San Diego to Santa Barbara, and Sacramento to Gilroy. but when I tried to plan a regional trip from southern to northern California on regional trains I found little in the gap: it looked like it would require two more segments that each ran every couple hours midday, so it would turn into an all-day trip. Greyhound has one or two runs a day from LA to the Bay Area. So there’s a huge untapped market for something like hourly service running at freeway speed or higher. Washington has only a fraction of that level of demand.

      4. A quick Google search reveals that express bus service between San Francisco an LA exists – there are 5 daily departures on Megabus, with intermediate stops at just San Jose and Burbank. Not exactly hour service, but the service does exist.

    5. I’m not sure if there are huge numbers of people going back and forth. Nothing like, L. A. and San Fransisco, or the Northeast corridor (Boston, New York, D. C.). But what makes this corridor so intriguing is:

      1) The two coastal cities involved. Vancouver is the west coast city of Canada. It is the L. A. and San Fransisco rolled into one for a nation that is increasingly seen as different, and proudly independent of the U. S. As it, inevitably, moves from a resource dependent country, to one driven by tertiary development, Vancouver is poised to be another Toronto. Big, beautiful, important, full of everything you want, really, in a city. Seattle is, in comparison, bush league. But like a second division city that manages to be promoted, we have a lot to offer. Again, a lot of it is geography, as we are *the* Northwest U. S. city (no offense Portland).

      2) It is the right distance. So I guess I shouldn’t dis’ Portland, because the fact remains that Portland to Seattle transport is similar. Too far to drive, too close to fly. In that regard, it is way better than Southern California. While San Fransisco to L. A. benefits from an enormous number of people, it will still take a while to get from one city to the other. The promise here is one hour. But for the sake of argument, let’s say we have a 50% increase in the travel time. That is a huge loss, but just think about it — 90 minutes to Vancouver, BC? Sign me up. Let’s see, leave my house around 8:00 AM. Take the bus and the subway to the station. I’m on the train by 8:30. I’m in Vancouver BC by 10:00 AM. Holy cow, I’m in Vancouver BC, one of the greatest cities in the world, by 10:00 AM, and it was no big deal! I can work, play, hike all day long, and still get back at a decent hour. It completely changes the dynamic of travel between the two cities. There might not be a huge amount of travel between the two cities, but if it was that much faster, there would be.

      1. They need to get the immigration process faster though. Sitting on a train for 30-45 minutes waiting for people to look at passports is so 1990s. Agents could board in Blaine and get off at Surrey would speed up the process. They’d then catch the next train back and get off at Blaine.

    6. I’m inclined to agree. I’d love it, to be sure, and Portland even more so, but for ST3-like costs, this seems to offer far less value. The average person in the Seattle region goes somewhere that overlaps with eventual LINK coverage on most days, and if it isn’t useful for their commute, it will be useful at least now and then. But how often do people go to Vancouver?

      1. I would go a lot more often if it was this fast. This would be a dramatic improvement in travel time, which would lead to a lot more business and tourist travel (for the day or the weekend).

        You can’t say that about ST3. In the middle of the day, Everett to Lynnwood will be faster by car, and during rush hour an express would probably be faster. Same with Tacoma, Issaquah, and West Seattle. It is only Ballard where the time savings are going to be substantial, and even then, nothing like this. If this could be built for the cost of ST3, then it would probably be a better value.

  8. This is a silly, pie-in-the-sky distraction. Unless Donald Trump makes the economy so hostile to immigrants and foreign workers than Microsoft, Amazon and the other Tech Titans of Puget Sound® feel they must leave the US, relatively few people will want to make the journey.

    While it’s a nice excuse for a “transit” series, it will never take place. Rural freeways are exactly the place for “entrained” RoboCars. Getting to Vancouver in an hour and a half by car would certainly be preferable to getting their in an hour by train, since most travelers between the two cities are not “on business”, so they won’t be both starting and ending in the CBD’s.

    It’s a silly shiny thing to dangle in front of foamers, of which I am usually one.

    But not this time.

    1. Richard, since Inauguration day, find me one piece of evidence that the “Unless” in front of the “Donald” hasn’t long since turned into “As soon as…”

      History is still punishing the US for the genocide and slavery our Founders practiced and tolerated. But also shows what happens when a large country with a relatively small population decides it’s in its own interest to take in the millions of the world’s able and ambitious who would’ve been murdered back home.

      Easy to imagine Canada paying for a Beautiful bullet train line with stops at every private detention center up the West Coast, streaming two hundred feet under the ugly barbed wire fence protecting Freedom’s border from its own formerly-Great people’s escape attempts.


    2. Microsoft already uses Vancouver as a place to host workers that weren’t yet able to get an H-1B visa and start work in Seattle. There’s a pretty big presence there as it is.

      If Trump shuts down or curtails the Visa program significantly, the traffic between our two cities could increase dramatically.

      … assuming IT companies don’t leave the country entirely.

      1. Who would be travelling? The visa-less workers can’t come to the US. The American workers can continue working here until the companies lay them off. Managers may travel back and forth to meet the visa-less workers, but that’s not a flood of riders.

    3. >> Getting to Vancouver in an hour and a half by car would certainly be preferable to getting their in an hour by train

      Since when have you been able to get to Vancouver BC in an hour and a half? Holy cow, man, you are just passing by Bellingham in an hour and a half. You haven’t even hit the border crossing (which can take forever). It takes three hours, minimum, to get to Vancouver. That is why so many people (myself included) just avoid it. Fantastic city — love it — but no, I don’t want to spend most of my day getting there and back.

      A train, even at an hour and a half, would change all that. Go ahead, be pessimistic. 90 minute train ride, and a 30 minute transit trip to the nearest station (!?). It is still reasonable for a day trip, to say nothing of a weekender. Leave the house at 8:00 AM, in Vancouver (anywhere) by 10:00. Spend all freakin’ day in Vancouver, eat a fantastic meal, and leave at 7:00 PM. You are back at home at 9:00. Oh, and the time spent traveling was spent sipping IPAs and looking out at the Cascades. Sign me up.

      1. Ross,

        Because entrained RoboCars will be traveling 120 miles an hour in dedicated lanes. I am strongly skeptical about their ability to solve urban mobility problems, but they absolutely will revolutionize intercity travel speeds on freeways.

      2. I don’t think we’re going to have 120 mph robocars on freeways anytime soon, maybe never.

        For starters, as long as any vehicle on the highway is limited to 60 or 70 mph, you can’t have 120 mph traffic – it would be too dangerous. Even if you somehow assumed that all human-driven vehicles would be off the road, there’s still the fact that the highways are full of large trucks (and buses) which are physically incapable of driving 120 mph, so it still wouldn’t work.

        Energy-wise, air resistance of a moving vehicle is proportional to the square of the velocity, so, even if it were legal, operating a vehicle at 120 mph would be vastly more expensive than operating the same vehicle at 60 mph. In the case of large trucks, the costs of equipping them with more powerful engines and paying for all that extra gas would just be too much – especially if the trucks are driverless, so the faster speeds wouldn’t even save any labor.

        And, there’s still the issue that the interstate highways of today are full of curves with a design speed of around 70 mph, so 120 mph car traffic would require a massive rebuild, which, in a hilly state like Washington, would be massively expensive. As a rough estimate, you can use the cost of building high speed rail as a floor for what a highway capable of hosting cars traveling at high-speed rail speeds would cost. In reality, the highway would probably cost more, since it would be wider, with multiple lanes, plus entrance/exit ramps.

      3. Just yesterday we were going to Vancouver on the evening Cascades. It had brake problems near Tacoma and by the time we canceled the tickets hours later it still hadn’t moved. The interesting thing is we canceled our dinner reservations, brunch reservations and our hotel too.

        We have a car, we could have driven, we just had no interest in spending 6 hrs in a car RT and then needing to park it there over the weekend. We canceled an entire 3 day trip because we couldn’t take the train. True story.

      4. “as long as any vehicle on the highway is limited to 60 or 70 mph, you can’t have 120 mph traffic – it would be too dangerous.”

        And the 1930s infomercial cartoon about 1960s motorways said that adjacent lanes would be 50, 80, and 120 mph.

      5. Why did old speedometers go up to 120 mph anyway? Where did they think people would be able to drive that fast? I remember 120 mph speedometers as a kid, and then I think the 1970s energy reforms limited them to 85.

    4. I tend to think that robocars (to use your term) will use a subscription ownership model. People probably won’t own their own robocar; a company like Uber will own the car, and provide it to riders on-demand. Owning a robocar would be a waste of money because we have to store and maintain it even when we aren’t using it… we eat those costs now, but they’re an enormous expense if you have an alternative. We only own our own cars now because that alternative is (or was until recently) unavailable.

      So, assuming a subscription ownership model, why would you sit in (and pay for) a robocar for 3-5 hours roundtrip when you can take a faster and more comfortable train to Vancouver and grab a robocar when you get there? You can uber (or drive a rental) to another city now, but the costs–which are completely unnecessary if you can just grab another uber when you arrive–are prohibitive. Because of some of those costs (as well as some legal liabilities) I doubt the robocar’s owner would even allow you to drive the robocar to another locality in the first place. And taking a robocar to Vancouver would introduce another legal challenge for the robocar’s owner: it’s in a whole ‘nother country!

      If anything, robocars make suburban train stations more attractive than they are now, since the CBD and the suburbs would become equally accessible. I suspect that robocars will be a last-mile solution, not a 150-mile solution.

      1. I seriously doubt the subscription model will ever happen. The on-demand in three minutes things has never worked with taxis before, or now (using technology with instant communication), and for all the same reasons, won’t work for self-driving taxis.

  9. While we are considering alignments and just playing with lines on paper (or on a computer screen) why not take a page from the Chunnel design for the urban ends of this corridor and run it underwater? At least on this end that would solve problems from Everett south and maybe solve problems at the north end as well?

    1. The underwater topography in the Salish Sea is far more rugged than in the English Channel. Short of using the world’s first floating tunnels, it’s not constructible.


      Chamois, thanks bringing this up. Think the distance pretty much Seattle to Marysville. The book is a must-read for everybody transit-oriented in Seattle, because it covers the so much we need to know about all our tunneling.

      Still can’t decide my favorite part of the book. The whole population of 1885 England turning off the just-invented Tunnel Boring Machine because of panic over possible invasion by arrogant French waiters and Italians with mustaches, crank-music boxes, and monkeys?

      Good trial run for Brexit!

      Or the woman scientist reading the fossil samples collected by a core-drill through the hub of the cutter, only way to be sure they were still in the razor thin layer of chalk that prevented drowning in the Channel above or the mud underneath?

      Legend has it that DSTT Chief Engineer gave ironclad standing order to his TBM crews: “You don’t find ONE bone!” Maybe Bertha found one, but project lied and called it a pipe.

      Or the picture of English and French project directors screaming at each other in both languages in their offices far upstairs. While down below Irish laborers calmly cut cross-passages using hand-held jack-hammers with chisel blades like sculptors?

      But winner is when engineers turn on their surveying lasers and discover that heavy salt-content in the air of an underwater tunnel make it impossible to accurately locate France! Could be good side for us, though. Having missed Everett, might be our chance to go for Vancouver.


  10. Build the station at Northgate. Its the most logical as it has an easy Link transfer (only 5 stops to Westlake), doesn’t cross the ship canal, and has plenty of right of way and excess land in the form of parking. Its also already a major bus transfer point. Building it south of downtown makes no sense as it forces transfers anyway (no one is going to be a tourist in Sodo or going to business meetings there) and would require expensive engineering and routing challenges.

    1. If we’re talking about spending billions of dollars on HSR it seems ridiculous to stop at Northgate just to shave a few bucks on another ship canal crossing.

      1. Its also ridiculous to put a station in an arbitrary spot south of the stadiums. If the stop isn’t underneath Westlake, the Northgate option is way better. Going south of the city only makes sense in the context of continuing onto Portland. And this wouldn’t be a few bucks, easily 10 billion extra.

      2. I agree south of the stadiums is equally ridiculous. But if we’re talking about something as pie-in-the-sky as true HSR between Seattle and Vancouver we might as well do it right. Downtown is the only place where a station makes sense.

      3. Putting the station downtown gives it a strong walkshed. Much of the travel will be business travel, so downtown is clearly better.

        If we are going to stop is at Northgate, might as well stop it a bit farther north, like Lynwood, and save more money.

        Of course, if the HSR is built in stages, you might build an intermediate station at, say, Lynnwood and get that up & running before the downtown segment is built a few years later.

    2. A continuous line from Vancouver to Portland would have to have a continuous southern portion, not a gap between Northgate and Southcenter.

      1. moreover, a continuous line from Vancouver to Portland is probably the only way that you might have enough ridership for this to make sense, since you want overhead ridership and through ridership would be substantial at that distance, rather than some strange-dick around on link for through hpassengers

      2. That’s a fair enough argument if you believe that rail can actually be competitive with flying for thru-riders going all the way from Vancouver to Portland. Unless the train tops 200 mph, all the way, it probably wouldn’t be.

    3. Northgate to downtown is less than 15 minutes. Northgate to the UW is half that. Brando is right, Northgate is good enough. Much of the city lives north of the ship canal, or at the very least, can get to Northgate just about as fast as they can get downtown. Capitol Hill to Northgate takes as much time as Capitol Hill to IDS.

      Folks in the south end lose out, but folks in the south end are no worse off than today. There simply is no competition — if you trying to get from SeaTac to Northgate, you take Link. An express would be wonderful, but mostly for the first half (SeaTac to downtown) versus downtown to Northgate. If anything, you would want an express bus that skips over all the mess and simply connects SeaTac with Northgate. For the cost of saving a couple minutes to run the train into downtown, you could run express buses from SeaTac every five minutes for the next fifty years and everyone in SeaTac would come out ahead.

      So basically the main beneficiaries of direct service to downtown versus Northgate are folks downtown, in Rainier Valley, and West Seattle (and the only big hit is for the last two groups). It just isn’t worth the cost.

      1. What about using the Interurban ROW? It wouldn’t be true HSR, but you could probably get to 80 MPH from Everett to Seattle with minimal fuss. From Everett to Shoreline the trail ROW is pretty much intact (although runs through a lot of SFR), use Roosevelt Way to get from 99 to I-5, then use the express lanes to get into the city. This would also work well with the above ideas of terminating at Northgate until the express lanes can be pried away from WSDOT. As an added bonus, you could move Sounder North to this ROW with stops at Everett Station, Everett Mall, Mariner, Lynnwood TC, Aurora Village, downtown Shoreline, and Northgate.

    4. But the train doesn’t stop there. It keeps going south. So what are you saving? Put it at 6th and Madison.

  11. Why not go into Downtown Bellevue instead? A 405 alignment could have stops for Link in Snohomish, Bellevue and – if extended to a logical endpoint – Seatac with six minute frequencies. If the northbound roadway was elevated, there would be plenty of room for both high-speed tracks and light rail tracks. We could solve the glaring commute challenge of 405 as well as build HSR cheaper. It would be politically popular with many Eastside residents and tech employers. A nonstop train from Downtown Bellevue to Searac would be a game changer for example.

    I’m not necessarily recommending it, but I am noting that it is a viable alternative worthy of consideration.

    Much of Downtown north of Yesler will be on Link anyway. It would simply change the transfer point.

    1. I should clarify: Some parts of Downtown could be within walking distance of a Downtown Seattle station, but a good part of Downtown will still require a transfer to Link no matter where the station might be.

    2. I was thinking about a 405 alignment too. We’ll see what Zach says in part 4.

      One issue would be around Southcenter, where 405 makes a sharp turn to I-5. It would probably be better to go down 167 instead and join I-5 further south.

      There are also the turns in Renton. It’s not as bad as the S-curves were but it still may be too sharp for a train.

      1. Agreed – the geography is much tougher south of Bellevue. Again – the train doesn’t need to be high speed in the urban corridors – even if it’s just going 60 in Renton & Southcenter, as long as it’s fast elsewhere it really doesn’t matter.

        OTOH, if you do it right you can add a local station in Renton.

      2. Yes the area through Renton and Tukwila could be tricky to design. In some places, it may require that the high speed rail be elevated above the roadway. Given the extrodinary cost of the entire project, an elevated gentler curve at I-405/I-5 through the Southcenter parking lot would probably be minor compared to some other challenges.

        If SeaTac is the terminus, I would suggest using SR 518 and the Airport Expressway, ending at about where the Airport Link station is today. There would be no need to curve for I-5

        If Tacoma or further south is the terminus, then either the 167 corridor (with a direct connector to SeaTac) or one that goes back to Seatac then jogs back to I-5 (405 to 518 to Airport Expressway to a tunnel at or south of 188th to get back to I-5) would possibly work.

        A final crazy option would be to turn the train north from 405 in Renton up the current track corridor into King Street. Boeing Access Road could become a transfer point as the area’s Union Station, with light rail to SeaTac, Sounder, giant multi-day paid parking garages like SeaTac has, and maybe a higher-speed train between Downtown Seattle and Tacoma/Oregon.

      3. 167 is another option between Renton and Tacoma, if the route is going through Bellevue. The ROW is comparable to an interstate.

    3. Can someone more knowledgeable speak to if this hypothetical HSR line can do double-duty as a local track?

      To repeat what i said in another comment, if we can triple track the line and run local commuter trains between Everett and Tacoma through Bellevue, then I think that it a much more compelling project than HSR to downtown Seattle that (unfortunately) runs parallel to Link.

      Conceivably, ST and WSDOT could share the cost of building the HSL corridor within the ST territory. WSDOT would operate the ‘express’ trains that would skip most of the local stops and try to achieve HSR performance, while ST would use the tracks to run Sounder-esque trains serving local stop.

      1. I believe that shared use is what they are considering between San Francisco and San Jose using Caltrain tracks. Some issues about radii, electric power supply, station stops and platforms lengths and other things would have to be considered.

        For this corridor, it would probably require that local and long-distance trains use the same equipment. That equipment would likely be incompatible with Link’s light rail vehicles. After all, if it was compatible, we wouldn’t be looking at a new corridor between Everett and Seattle in the first place! Anyway, a train operation with identical vehicles could have half the routes end in Marysville or Burlington or Bellingham, and the other half could go all the way to Vancouver.

    4. Not a bad idea, but I like Northgate better. As said above, the key with Northgate is that it is very fast from there to the two primary destinations in the greater Seattle area, the UW and downtown. Bellevue is not. Here are some times:

      Bellevue to ID — 19 minutes.
      Bellevue to Westlake — 24 minutes.
      Bellevue to UW — 30 minutes.

      Northgate to ID — 16 minutes
      Northgate to Westlake — 13 minutes
      Northgate to UW — 5 minutes

      Meanwhile, it is a shorter distance from Northgate to the border. By my estimates it is about 15 miles, which is a large distance from a cost, if not speed perspective.

      1. “Northgate to ID — 16 minutes
        Northgate to Westlake — 13 minutes”

        Is Westlake to ID really only going to take 3 minutes when the buses are no longer in the tunnel?

      2. Good catch. It should read:

        Northgate to ID — 18 minutes

        So five minutes from Westlake to ID (like it used to be, back in the old “ride free” days). The numbers I’m using are from these charts (which came from STB posts):

      3. “Is Westlake to ID really only going to take 3 minutes when the buses are no longer in the tunnel”

        It takes two minutes now after 8:30pm. I’ve timed it on my stopwatch a few times. 2 minutes northbound evening. 12 minutes southbound daytime. (I travel mostly northbound evenings because that’s where my home is.)

  12. Thanks for taking this one on, Zach. Thing I like best about it is my take on the reason these two officials have gone public. Encouraging response to some crystal clear and by every day’s non-fake news, intensively present dangers.

    Cost and technical complexity? Let’s see total balance sheet for our every war since the Second World one. Call this project National Defense, which it is, and numbers should clinch it.

    Same reason the Army should make I-5 transit, not HOV, lanes from SR101 to Everett happen rush hour Monday. Authorizing legislation uses the “D-word”. Not “JDP”, for Joe Diamond Parking.

    For transportation category, think airline, not railroad, for both speed and number of miles at grade level. Will postings deal with geology, Zach? Because the deeper the tunnel, the less ‘quake damage.

    So put LINK’s every elevator in same defense contract as for Seattle station. Which can fit into one jet fighter contract without denting the fuselage. Tell me there’s no Congressional precedent. But if local geology says no…

    For cost and travel time, regular-rail connection at Yakima still beats other alternatives of time and money. We’ll be able to pay Yakima’s share for Beautifying a certain Wall. And showing the Legislator we’ll spare some expense to keep Yakima in the same State as Seattle.

    Leading back to my first paragraph. These last four decades have seen a very large number of our country’s out-State economies go from manufacturing, logging, and mining to private prisons. The one current promised infrastructure measure with delivery guaranteed.

    Westward view from Angle Lake Station probably already on the cover of the regionwide corporate prospectus.

    Personally, I think a couple thousand miles of international ocean ports and a lot of technical advancement have earned us a different industrial base. Which we”ve got the wherewithal to build for ourselves if we have to.

    I’m not the one redrawing the borders of the United States to put a Vancover BC high speed rail terminal in the same country as Seattle. And Portland. And San Diego. But glad to see our State’s chief exec meeting with a foreign leader who isn’t Vladimir Putin.

    Secession is treason. So let’s just say that if somebody else has already decided to use the Cascades for a wall, we’ll be glad to pay for it. And make our side really, really Beautiful.

    Mark Dublin

  13. 7% grade really isn’t that bad next to a station as it is a short distance at low speed.

    It seems to me you want the main station where the best transit infrastructure is. So, maybe putting it near Westlake would work?

  14. Two things I’d like to add to the study.

    1) What travel time do we need? We don’t need an hour to Vancouver or Portland. We don’t need to facilitate daily commutes to there. Why would we do that? For cheaper housing? Vancouver and Portland have the same housing-affordability problem we do. If Mircosoft steps up recruting in Canada because of strict US immigration, they won’t be coming to the US all the time, they’ll be living and working in Canada. If the tech companies continue to employ a large number of US citizens, they’ll leave the US offices open. The only people traveling back and forth are those going to business meetings, and those aren’t every day. And only a few people go to the ones that are every week. 2-3 hours is fine as a target for Seattle-Vancouver and Seattle-Portland. And don’t those tech people have fancy video screens and VR prototypes that are supposed to put you “virtually there” for long-distance meetings? Maybe they should work on that technology instead.

    2) They should study not only a new HSR line but an incremental Cascades alternative. There’s already a plan to upgrade Cascades. Update it and analyze what some additional money and ideas could do to improve it. If we’re going to improve north-south rail travel, we should compare the best of both approaches. And even a third approach of “early deliverables” now and the whole enchilada later. It’s dangerous to put all your eggs into a shoot-for-the-moon basket. Washington has incrementally upgraded Cascades for two decades and has something to show for it. Florida put all its effort into high-speed rail and got nothing, not even somewhat better travel between Florida’s cities.

    1. The maximally useful line not just have Portland or Vancouver in reasonable HSR time from Seattle, but also have Portland and Vancouver be within reasonable travel time of each other. In that case if you’re going to be making any substantial new build rights of way, 2:15 is about the most you’d want to allow, ideally lower if it can be done at incrementally reasonable cost.

    2. 1) Faster is always better, and it is always a judgement call as to when something is “fast enough”. The value of an hour trip to Vancouver, though, is huge. It isn’t about commuting, it is about every other reason you go from one city to another. Business trips and tourist trips — visits to friend and family. At an hour, you now have the possibility of just going up for the day. There are plenty of people who do that right now, but they fly (and spend even more time getting there and back). It is why the east coast rail lines, and the European rail lines are fairly popular. The cities are fairly close together, so travel is fast and day trips make sense for a lot of destinations. New York to Philadelphia or Baltimore to DC.

      At two and a half hours (plus transfer time) I’m just not interested. I might as well drive. Get the numbers down to an hour and a half, though, and it probably good enough. Even with transfers on either end, taking the train is still a lot faster (and a lot more pleasant).

      2) Good point. I think that sort of planning (looking at alternatives, both short and long term) is exactly what is needed. There is no need to spend a huge amount on a very small improvement in travel time — you want to focus first on getting the best bang for your buck (which is always true for mass transportation).

      1. Ever tried to drive into or out of YVR near rush hour time? If you’re trying to get anywhere near downtown, it’s not easy. The transfer time for the train (which drops you off right at a skytrain station, right on the bike network, etc.) has to be compared to that. I do a day round trip via train and boltbus and it’s brutal, but so would a car trip be.

        I also think Zach’s (very interesting) analysis could stand comparison to what it would take to bring the current 4-5 hour trip on the current tracks down to 2 hours. This would only be an average speed of about 70mph. Could we spend a billion or two and get that (along with much increased frequency)?

        That strikes me as a far bigger bang for the buck than 30B (or whatever) for a one hour trip. There’s rapidly diminishing returns under 2 hours.

  15. I like the prospect of such a line, though it will likely never happen in my lifetime. I think a midtown station at 6th and Madison is an excellent terminus location, turning the old federal courthouse there into a multi-modal station for both it and the proposed Link 5th/Madison station with Madison BRT extending this all to ferry users. Given the expense of tunneling it isn’t likely it were to ever occur but I think it should be a tunnel underneath I-5 namely for all the existing surface impacts of noise from freeway traffic and Link service, substantial mitigation for improvements to manage noise and vibration impacts and unsuitable freeway geometry. Not sure what the delta for that would be but I think the property costs would be fairly more substantial as a result. A deep bore tunnel would certainly mitigate both, additional design costs and tunneling notwithstanding. Additionally given the width of the freeway ROW envelope it should allow for good design and track speeds without having to acquire sub-surface rights from hundreds of private property owners.

    I would cite stations at Mariner so that Lynnwood and Everett each have shared and equidistant access and one other in Bellingham. I would surface the track somewhere from Everett north where the built environment begins to taper off.

  16. Did I misread, or did this completely ignore the elephant in the room. That *instead* of Link along the I-5 north of Northgate we should be building a blended HSR and regional rail line north to Everett. It would’ve saved a lot of money, space, and also provided superior transit service to Link along this portion of the corridor. Now instead we have to build yet another line in the same approximate ROW as Link, with another set of stations. It’s frankly absurd.

    As for tunneling into downtown, yes, I support using the express lanes if that can be made to work, I’m not sure it’s a fight worth having in the long run. But tunneling to King Street is clearly feasible. And if we make this part of a system – and stop being afraid of mixing transit modes – this continues through King Street station to extend along a blended HSR/regional rail line south towards Tacoma. Bringing regional service from Olympia/Tacoma (and in between) right through to King Street station, a new North Downtown station, and beyond to a Northgate station where connections can all be made to Link for transfers to local service.

    1. That’s an ideal that’s off the table. You can’t build something unless the government and taxpayers and business leaders agree to it. (Although the business leaders can sometimes be overcome if there’s overwhelming public support). Many of us advocated for a two-level system: a smaller Link network in Seattle, and heavy rail on new ROW to the suburbs. We pushed for Link on 99 and more stations. But the city and county leaders, the majority of the public, and the business leaders were not interested in that. High-speed rail, or commuter rail compatible with HSR, was not within the region’s power or budget even if it wanted to. That requires state leadership and choosing an HSR technology and design; the tail can’t wag the dog and predict what the HSR design or alignment will be, or even what technology will be available when/if it’s decided. King County did an admirable job of prebuilding the DSTT for Link, but they designed it tor high-floor trains and a few years later the industry switched to low-floor trains.

      So that “elephant” was not a possibility. Link and Sounder were, and are running and are being extended. This HSR proposal has the support of business leaders and the governor, so it might go somewhere. (Although I think the business leaders are delusional and the governor won’t follow through when a concrete design and cost estimate comes out. But their efforts could catalyze a second, more realistic proposal later, which is more than we’ve got up till now.

      Given that we have Link and Sounder, the only path forward is to continue improving the transit system, and that might possibly mean heavy rail next to Link on I-5. C’est la vie. It’s better than not having it.

      Ideally the state would have prioritized transit in the 1970s and laid the groundwork for statewide passenger rail at a much higher and more frequent level than Cascades, with enough ROW to adapt to HSR later, and led a process for the cities to integrate regional and local subways and buses and commuter rail with it. We could imitate Germany’s transit network for instance. But our state has only a mild interest in statewide transit, and no interest at all in regional or local transit.

  17. As a blog that (supposedly) supports mass transit, the tone of this article is a disappointment. The framing of this issue feeds into the “Seattle Process” stereotype of finding every way to criticize a potential project, instead of realistically pitching challenges as obstacles to be overcome. High Speed Rail exists across the world; it is not impossible. Seattle Transit Blog can do better.

    1. Give us some credit, Mike. Today we’ve been both hitting back at limited vision and suggesting improvements, haven’t we? Zach’s also given us some necessary technical considerations. And this is only the first piece in a series.

      Over my working years there, Seattle’s process, including intelligence and lack of bribery, got the city a subway through a crowded central business district without having to condemn one square inch of property. In three years, one of which for utility relocation only.

      And suburbs who wouldn’t get actual trains for thirty years paid for our dual-power bus service. Buses themselves stunk, but gotta admit, the Walls of every station really are Beautiful. How’s that for meeting and matching a challenge?

      Public or private, no “Process” runs itself. Delays attributed to it are really caused by the people who support the process being able to decide what to use it for. Like any other tool. Including leadership itself.

      Can’t be seized and still be leadership. Followers have to literally seize their chosen leader, who doesn’t want the job, by the collar and demand they lead. So be careful of the opposite approach. Still can’t find out who TED is, but all his TALKERS need to click this link.

      The higher the tech, the more basic the questions it has to have answers for.


  18. Just for reference, the State has produced a Long-Range Plan in 2006 for the Cascades Corridor, which suggests 4 daily trains between Seattle and Vancouver with a travel time of 2:37, by 2023, with incremental improvements to the existing tracks. The report identifies $6.5 billion in capital improvements between Portland and Vancouver, but doesn’t break out the cost for improvements north of Seattle.

    Long-term, an entirely new rail corridor between Vancouver and Seattle will be needed in order to see significant reduction in travel time, and I-5 is a logical right of way. I’m not sure that anyone currently has the stomach for the actual costs. The time to bring this up is whenever I-5 maintenance enters a crisis mode, and the State is discussing spending $10s of billions rebuilding it.

    1. Perhaps I am the first to break this news, but unfortunately the long-range plan is dead.

      All of the work recently (or soon-to-be) completed on the stimulus projects, work that was supposed to advance the long-range plan, was completed using track geometry that cannot ever support the speeds necessary to achieve a 2.5 hour trip time to PDX from SEA.

      I am currently researching this and delving more into the story—and there certainly is a story here.

  19. For a different reason, I have come to the same conclusion that a transfer to local transit outside the CBD makes sense. Any HSR plan would require an entirely new grade separated path through the city. And any such path would garner more local users than users of HSR. Thus it would be wasteful to keep such a path open only for HSR every 15 minutes or so. It makes more sense to transfer to local transit services. And because most HSR riders will need a connection to get to their final destinations, this makes sense in any case. A station at Northgate in would be the sensible way to handle this in Seattle, and I have thought about similar ideas for a Surrey station in Vancouver. However, the complication would be traffic traveling through Seattle to points further south. Obviously in that case, expansion would need to keep the possibility of through train traffic under Seattle.

    As it has been mentioned, HSR trains have powered wheels all along the train and are better able to handle higher grades. Five percent is not a problem. The only wrinkle is that the vertical curve, like any horizontal curve, must be gradual to avoid the roller coaster effect. (Curves may be sharper when the train is moving more slowly like near the stations.

    As to other technological choices, they are less realistic. Hyperloop has the pitfall of other evacuated tube concepts. Horrendous cost. I suspect that some day does concepts will come into use, but only when construction is done by robots and is much cheaper. (Some of the original cost estimates were just silly, knocking two zeros off any realistic cost.)

    Using automatic cars at higher speed and possibly entrained has a more geometrical problem. People in cars take up more space than people without them, so building a new ROW for these vehicles into Vancouver and Seattle is going to be more expensive than building a train. The stations are essentially underground interchanges and are much more expensive to build.

    1. You’re right that Robocars aren’t an answer for in-city travel, but the volumes between Seattle and Vancouver will never be huge. There just aren’t enough people in the region and probably never will be with shrinking populations by the time such a line would be complete.

  20. Anybody know where the Ship Canal Bridge sits in the line for replacement or major seismic retrofit? It was built in 1962, so it’s not that old, but it does predate modern seismic building codes (or any understand of the magnitude of earthquake we’re subject to from the Cascadia subduction zone).

    It would make a lot of sense to push for HSR accommodations when work is already being done. (I think our shot to get light rail on the 99 corridor is probably when the Aurora bridge gets replaced and we can push for rail to be included in the design.)

    1. We’d want to build rail on the Aurora Bridge much lower than the highway level, though, so we can hit Lower Fremont too. Though I guess we could always include a huge elevator instead.

  21. So from what I’ve seen I offer this

    1) Downtown to downtown is the whole point of intercity rail. As soon as you add in two transfers to get anywhere, many will simply default to taking a car.
    a) Yes this is more expensive but likely worth it given the density within downtown but unless someone can point me to rail outside downtown that gets decent ridership for intercity rail, I’m inclined to push for a downtown Seattle staiton.

    2) If you paralleled I-5, it would act as a potential ST Sounder express from Everett to downtown using an RER like service. Imagine getting from Lynnwood to downtown Seattle in 20 minutes or from Everett in 45 minutes or less? I would aim for making Marysville within an hour of downtown Seattle by rail.

    3) This doesn’t overcome the political issue of the border. As has been seen with Eurostar, even with travel times, the fares are astronomically expensive. Even booking in advance, one way is about $60. If needed in a week or less, it can be upward of $180. Paris to Lyon domestically is around $40 when booked in advance (3 months) or $120 if tomorrow. Typically the border will kill part of the ridership even with frequency simply due to the immigration/customs check. Eurostar hasn’t had the ridership of other HSLs. Mind you, it takes about 30 years to develop into a mature line and we are in year 8 of the true high-speed trip from St. Pancras-Paris and Brussels but ridership is still underwhelming.

    1. Brexit will likely only make things worse in regards to customs and travel time between London and Paris. With the current administration or at least the president’s enthusiasm for Brexit, customs checks and border times will likely only become worse in the short term here.

    2. 1) Downtown to downtown is ideal, but for the distances involved, a station a bit outside (Northgate) is just about as good. Seattle is in the process of building a suburban express system. This makes lemonade out of lemons, really. There should be half a dozen or more stops between downtown and Northgate, but there are only four. This means travel between the two areas is very fast. When you consider that the UW — a major destination — sits roughly half way in between, there really is no great loss. What are your customers, really? Tourists? No problem — very few of them would be able to walk to the station if it was downtown anyway. Business workers? The transfer is annoying, but there is no way they would drive; and flying is much worse from a transfer standpoint.

      3) If done right (and I see no reason why they can’t do it right) then the border situation is actually a bonus. Right now crossing the border is a big pain. But with a train, you can actually do the screening while on board. There are three things that border people are concerned about:

      A) Terrorism. A terrorist is likely to blow up a bomb in the station or on a train headed anywhere. It is unlikely that someone in Canada will load up the train with bombs ready to explode an hour later. That is also something that is relatively easy to check (unlike bombs that go off right away).

      B) People sneaking in. This is easy to check, as you walk through the train. Suspicious people get pulled out, and questioned in more detail in a separate care. If things don’t go well, then they are sent back on the next train (an unlikely and rare occurrence). That isn’t that different than what happens at the airport, except that it happens on the train itself.

      C) Contraband. This again can happen while the train is in motion.

      In contrast, you can’t really make the border crossing much faster. It simply doesn’t scale, because you have only so many lanes at the border.

      1. Isn’t “Northgate” Seattlish for Gare du Nord?

        Access to the HSR station won’t be mostly walking. So while there’s some value to having it in the core, there are lots of precedents for well-connected intercity rail stations nearer the edge.

      2. Dan

        It would be one thing if you had access to major metro lines and many bus routes but it is another matter when you don’t. Paris Gare du Nord I would consider is a part of the core and given the separation of Nord to de Lyon is around 3-4 miles. that would place Northgate at least twice the distance of the separation of those two stations to Seattle’s downtown core. It would be one thing starting from SLU with a connection to a subway but Nord has two metro lines, 3 if you count M7 at the next door Gare d’Est.

        I see this as the opportunity to give Sounder North a proper burial and make a true higher speed corridor with a downtown station come into the area. Marysville-Seattle RER.

        The main thing is the potential transfer options. From Jackson Street there are many from Link to many buses. Northgate that can depend. You change the advantage to North Seattle rather than downtown and if you are coming from the south you lose out. There isn’t a strong desire to make Tukwila’s transfer environment much better and the location of the station doesn’t support it.

        I would argue a 2nd trunk across the Ship Canal would be beneficial not just for HSR but for fast frequent rail from Snohomish Link will be frequent but from Lynnwood north not fast.

      3. “The main thing is the potential transfer options. From Jackson Street there are many from Link to many buses. Northgate that can depend. ”

        Presumably there would be a bus restructure after the HSR station opens….

  22. I think you are overthinking the right of way. any high speed rail will require a new subway downtown and an underground station. Subway lines won’t collide… you just put them at different depths.

    I don’t think you can use I-5 to run high speed rail. It’s too curvy for a train moving at 200mph.

    If you were going to go north, you could run an elevated line along 99. 99 is almost completely straight. Once you are past Everett you can just carve new right of way out of farm land.

    That said, I think a Seattle Portland route would get more ridership than a Seattle Vancouver route. It would also allow us to connect Tacoma to Seattle via high speed rail, which would out a huge amount of housing within a quick commute to Seattle.

    1. Any route plan within a city has to consider noise. HSR at speed is noisy, almost entirely due to air resistance. Elevating HSR along the 99 route would still not allow for higher speed unless it were put into a tube.

      Elsewhere HSR slows down through urban areas, partly due to noise and partly due to tighter curves and points. But that means that 150kph trains would be ok with I-5 curves.

      1. 99 is already plenty noisy. It’s a highway.

        What makes you think a train can go even 150mph down I-5? Light rail will only move down I-5 at 55mph. It takes a solid hour to get from Seattle to Everett on I-5 via rail in ST3.

        And where would you even fit that track, with the light rail track taking up the spare right of way?

        This discussion is very airy, but I think I-5 as a high speed rail right of way doesn’t pass the sniff test.

  23. Although a fascinating conversation, this is as likely in the near future as a real Sasquatch or Unicorn sighting.

  24. Serious question: Is stopping in Everett a serious goal of this line? Should it be?

    Sounds like an attempt to serve Everett in would be costly. Perhaps a stop in “Everett” (as in, wherever the cheapest version of the line can pass by everett) would make more sense. I know regional planners think Everett is about to grow exponentially but it’s worth noting that Snohomish county is largely sprawl and Everett population is currently just over 100k – there are not substantial hints that that trend will actually change soon. Not really the scale (or distance fro Seattle) to be an ideal HSR use case.

    Similar thoughts to the south – the stations in between the major cities don’t have to be urban – though if we went “Tacoma” instead of Tacoma – integration with Sounder (or better yet, Link) would be critical.

    For Olympia the desire for a station feels more psychological than a reflection of a market for this service.

    1. I feel the same way (and said that below). By all means you should have a station in Everett, but you shouldn’t spend a huge amount of money on it. There is no reason to deviate the line to serve it (or Bellingham). If the train follows the freeway right of way (which seems likely) then it isn’t much of an issue, since the freeway is very close to the heart of town. If it has to go inland, then so be it. Either way you don’t serve the existing Everett station (it just isn’t worth it).

      1. If it doesn’t serve Everett city center, then surely CT & Everett transit would build a new SWIFT line connecting the station to Everett.

        As Zach sketches out, I don’t think you serve either Everett Station or King Street station – you end up building a new station a few blocks away. Seems close enough for me.

  25. The third option is an outlying terminal with a forced Link transfer. This would exponentially reduce costs just as it radically increases total travel time.

    Which is why Northgate is a great compromise. It is about 15 minutes from there to downtown. It is about half that to the U-District. From the only other major pocket of density in the state — the Central Area — it is just about as fast to Northgate as it is downtown. From Capitol Hill Station to Northgate is 7 minutes, which is how long it takes to get to the International District (i. e. the other train station). All in all, for the bulk of potential passengers, Northgate is almost as good as downtown.

    Do you serve Seattle, Everett, Bellingham, and Vancouver? All of those lie at sea level along the coast. Do you weave back and forth? If not, which cities do you bypass and how?

    Vancouver should be served in the same manner as Seattle. Ideally you go right into town, but if it is a lot cheaper to connect to the (frequent and fast) SkyTrain system, then do that.

    There is no need to serve the heart of either Everett or Bellingham. Everett is a sprawling city, and it will soon have a light rail connecting to it. With the Seattle terminal being in Northgate, it is questionable whether you even need an Everett station, but given the proximity of the freeway to the city, you might as well. A station close to the intersection of I-5 and Highway 2 would be very good. That would be about as far away from the heart of town as the current train station (and terminus of Link). With a little bit of connecting bus service, you would have great service to Everett, which would not only make for a fast trip to Vancouver, but an excellent shortcut for a trip to Seattle.

    Similarly, Bellingham would be fine with a freeway station. A station close to a downtown freeway exit would actually be closer than the existing train station, which is a couple miles away from downtown. In general, it doesn’t matter much, since the city is fairly sprawling as well, and with decent shuttle bus service (possibly connecting with the other train station, downtown and the university) you would have something more attractive than driving, even for those at the south end of Vancouver.

    The main reason folks are so excited about this idea is that the distance is right. Much longer and flying makes more sense. Much shorter, and driving is easier. If you land at SeaTac, you’ve already been through security, so you will likely still take a connecting flight to Vancouver. But for lots of other people, this would represent a huge time savings versus the alternative, even if it only went from Northgate to the south end of Vancouver, and barely skirted the cities in between.

    1. A lot of HSR lines in Europe go between major cities so travelling on would actually require connecting from station to station. For example Paris has six major rail stations including Gare du Nord, Gare l’Est, Gare du Lyon, etc. In part this is because a lot of the rail lines are operated by different country groups in Europe with Paris as a major destination. So, Thalys might primarily serve Belgium, but extends a high speed line south to Paris.

      The question for HSR from Seattle would be whether direct service from Vancouver, BC to Portland is important. Or, whether it makes more sense to consider the service to Vancouver, BC separate from that to Portland. In that case the HSR from the North could connect to Link at Northgate and the HSR from the South could connect to Link at some point South of the city with Link providing the final connection to downtown.

    2. Ross is basically advocating for treating the HSR terminals like airports – stick them farther out and connect them to urban areas with secondary transit.

      The recent Overheard Wire podcasts on HSR station planning have some great discussions on this choice with respect to French HSR. Basically, the french experts argued both/and … you build a suburb “airport” station for people who drive (drop-off, paid parking, etc) with some transit service, AND you build a station in the urban center with urban access (rail, bike, pedestrian)

      Of course geography is the big question here in Cascadia, and there are cost savings to not serving Everett & Bellingham city centers … but shouldn’t a big public works project like this strive to make places, not just serve trips?

      Basically I think the French experts would argue for a suburban station (Northgate, Lynnwood-ish?) in addition to an urban station in Seattle’s core because they would service distinct markets.

      Similarly, HSR between Portland and Seattle should have a station that serves SeaTac, to both facilitate airport-HSR transfers and to leverage the vast parking infrastructure already there for people who want to drive & leave their cars for HSR trips, just like they do for flying.

      1. I think, when they say suburban station, they me one usable for car travelers. This basically rules out Northgate, which does not have reliable parking nearby (since the mall is quite possessive about theirs).

  26. The Vancouver side is silly that they’re not putting any effort into it. It should look something like this, and then it’s also available as rail from the burbs into town.

    Canadian routing should be:
    – Start at roughly the truck crossing
    – Past the new shopping/housing area at approx 176@16 (not looping via WhiteRock as the existing line does)
    – New Station w/ park & ride at 176/16
    – replace/upgrade most of the existing line roughly paralleling Hwy 99, Hwy 91, Hwy 17
    – New station at Scott Road (interchange w/ Expo Skytrain)
    – New bridge over fraser (they’re building already a new Patullo, dual use bridge)
    – Non-stop grade separated all the way into Pacific Station (interchange Main St Skytrain)

    25-30 minutes from South Surrey into downtown is a win win win just on the BC transport side, without even considering the Vancouver- Seattle benefits.

    now you have a new upgraded rail line
    – ends downtown (or very close)
    – two interchange stations with existing light rail (Main St, Scott Rd)
    – takes pressure off the existing Fraser Bridge (91) and Massey Tunnel (99)

    1. This is pretty much the lower cost plan currently on the table. It would get rid of all the very slow speeds, but still would not be very fast. The higher cost, higher speed plan is continuing north from the truck crossing along Pacific Highway #15 to connect to the CN tracks along the Fraser then over a new rail bridge to connect to the BNSF line at Sapperton and into town. This would be more direct and faster, but would require much more work to be able to maintain speed on what are already busy rail lines.

      Combing a new rail bridge with a new Pattullo was looked at by engineers and found not to offer any savings because the lift bridge and high lever bridge were very different. Also not sure that the road and rail bridges are ideally located at the same location. Rail bridge needs to be double tracked but there isn’t much room for that at the current location or the future location of the Pattullo.

      Some local news reports about the White Rock section of the line:

    2. I don’t think paralleling 99 is feasible as the descent from White Rock down to Delta is way too steep for rail.

  27. Going on the Northgate Station idea a bit further, even if high speed rail is a pipe dream, bus service between Seattle and Vancouver exists today, and it would be wonderful if, in 2021, at least one bus company would offer service just between Vancouver and Northgate that didn’t go all the way into downtown Seattle. For anyone who lives north of the ship canal, you waste so much time crawling in I-5 traffic the last few miles into downtown, only to turn around and go back the other way on the local transit system. Even with Link available for the northbound trip, the extra backtracking to/from downtown can add close to an hour to the trip, and at present, the only real way to avoid it is to drive the entire way. Even for people headed downtown, a Link connection at Northgate can still be much faster if it’s anytime near rush hour.

    One may also ask – why not route the bus to stop at Northgate, then continue on to downtown? On the surface, it’s the best of both worlds, as it provides both the Link option and the one-seat ride to downtown. But, in reality, this option is not so good. From a passenger’s standpoint, anyone headed would would still be delayed by I-5 traffic near downtown, even those getting on at Northgate, and, from an operational perspective, the bus would still need to run between downtown and Northgate (albeit, half-empty), which means higher fares and less frequency possible with a given fleet size.

    On the Vancouver side, one could make a similar argument, that a Seattle->Vancouver bus should exist that just drops everybody at a SkyTrain station (perhaps Surrey station, to avoid the bottlenecks trying to cross the Faser River). If a bus service existed that could just go from Northgate to Surrey, non-stop, while charging cheaper fares than the downtown->downtown competition, I would absolutely ride it, and I’m sure others would too.

    1. I could see a station a bit farther north like Lynnwood transit center leasing out some bus bays to private operations. Lynnwood TC seems the best option b/c it also pulls in ST & CT buses from east Snohomish & King.

      In a similar vein, South Bellevue P&R is a perfect location for private (or public) buses going over the pass, along with north or south using 405 rather than 5.

      Once Link is built out more, I could see SDOT try to move the Greyhound station out of downtown to 1) get those buses off of downtown streets, and 2) re-purpose that space for another use.

      A good example on how to not move an inter-city hub

      1. Agreed – once Bellevue Link opens, South Bellevue P&R would be a great place for Greyhound service to eastern Washington, and once the train goes to Lynnwood, Lynnwood would be a logical terminus for inter-city buses to Vancouver. I suggested Northgate only because it opens two years before Lynnwood does, but, post 2023, Lynnwood would be better because it’s closer to the freeway and has less traffic to deal with, getting in and out.

        FWIW, Greyhound already uses Everett Station, as does Amtrak Thurway buses. Amtrak is pretty much stuck at Everett, forever, because that’s where the train tracks go. Greyhound is too slow going to Vancouver because it has too many stops along the way. A bus that simply went nonstop from Lynnwood TC to Surrey Station, with Link/Skytrain connections available to bypass all of the traffic, on both sides, would be totally awesome. (Of course, for it to work, the bus would actually have to turn around at Lynnwood and not continue into downtown Seattle; otherwise, it would still be subject to the whims of I-5 traffic around downtown, and, even if you use Link to bypass it, it just means you get to wait at the station longer).

  28. I love that the STB is now engaged in the discussion of long-range rail planning, which is my bread and butter.

    This is a very fine post, but I would like to add my perspective, which is one educated on this topic.

    First, I would argue that the distances and populations of PNW cities do not necessitate a full build-out of 220mph high-speed rail. Our cities are not large and dynamic enough to warrant such an extraordinary investment into this type of infrastructure. HSR should be reserved for primate and primary cities that can deliver loaded trains with great frequency, two categories for which Seattle and Vancouver cannot qualify. Furthermore, the cities are rather close-by, which would neutralize the benefit of 220mph operations, especially if intermediate service is expected. Compare ours to the distance between that of Paris and Lyon, one of the finest high-speed lines in the world. Japanese high-speed systems do have shorter distances between city pairs, but their individual cities are enormous and densely populated.

    For the Cascades corridor, our sweet spot is between 125mph to 186mph. Fortunately, that eases the complication of building a HSL dramatically compared to 220 mph.

    Next, King Street Station is certainly in central Seattle and is perfectly suitable as a station/terminal. High-speed terminals have dual roles as train stations and as the rail equivalent of regional airports, so the fact that KSS is but blocks removed from Seattle’s definite center is not problematic at all. In fact, it’s location is beautiful. And on top of that, it already exists and is well connected by the existing transit network (though improvements can be made there). None of Paris or London’s terminals, either, are flawlessly central. Local transit connects all other points to the regional rail terminals.

    Finally, new passemger rail lines are not needed into Seattle. The existing corridors, improved with higher superelevation (or cant), featuring widened curves wherever possible, double-tracked, and operated over with modern electric trainsets that tilt, would be transformational for trip times. What is needed, rather, is the removal of freight trains from Seattle, which is surprisingly possible.

    Critically, freight trains into Seattle are either trying to reach the port of Tacoma or Seattle, or get to points east of the Cascades. Both of those objectives can be achieved by using a heavily upgraded Stampede Pass, with funding support provided by the State (it would require new signalling systems and either a heavily upgraded, or a modern bored, Stampede Tunnel). All other trips to points north of Everett can be completed using the Stevens Pass line, avoiding the Seattle metropolitan area entirely. The Interbay Yard near Magnolia, which the BNSF has already considered closing for real estate development purposes, would have to be decommissioned.

    That effectively removes freight trains from the Seattle-Everett line, and whatever cargo traffic remains could either operate outside of passenger hours, or we could eliminate it outright for the benefit of the traveling regional public.

    Now, of course, some things would be a bit messy, but the transition is feasible and possible with a little long-range thinking, as this blog is doing now.

    With freight trains off the line, or outside of passenger hours for local work, the corridor can be heavily improved to eventually accommodate speeds that, while not high-speed, get the job done and reliably and quickly connect legacy track trains to any higher speed line north of Everett, which is gorgeously flat.

    That is how you deliver express rail service to Seattle in a most responsible and spend-thrift manner.

    1. I don’t understand how Stevens Pass can reasonably replace service north of Everett. I have no clue, seriously. How?
      Adding the distance and unreliability involved with crossing the Cascades twice would effectively eliminate freight rail between Seattle and Vancouver, forever.

      1. Just a miscommunication here.

        Stevens Pass would handle all cross-Cascades traffic to points north of Everett. All other trains would access Seattle and Tacoma, or head east, using Stampede Pass and avoid the Everett to Seattle coastal trek.

        The Seattle to Everett line currently only exists to shuttle trains to Stevens Pass or to Cherry Point, and alternative routings can eliminate that need.

        Following an upgrade and reorganization, Stevens Pass would be used to exclusively access Cherry Point and Vancouver, B.C.

        With Stampede Pass fully operational, which it currently is not, it would handle all the traffic, and more, that currently travels over Stevens to get to the ports of Tacoma and Seattle.

        Hopefully that explanation is clearer.

      2. I don’t think you truly understand how important freight rail is to every element of the state’s economy.
        The refineries currently ship a lot of their crude oil by rail from North Dakota. Crossing the international border twice would increase travel times, costs and reduce reliability (due to border delays).
        Besides, the improvements you make to the line for passenger rail benefit will also greatly benefit freight, as long as you let them use it.
        We can choose economic development, or economic obliteration.

      3. I would also add here that, eventually, you do encounter a point where passenger and freight services become mutually exclusive, where investments into one form of traffic is detrimental to the other.

        On the corridor between Seattle and Everett, with tight curves and poor track geometry, upgrades to the alignment are critical for quality passenger service, but those same upgrades would render existing freight service impossible.

      4. How would upgrading curves for higher speeds render the track unusable for freight? Is it the cant? You could always use tilting railcars for the passenger service.
        And most of the investment for increasing speeds would be grade separation, which is unquestionably a big benefit for both passenger and freight, not to mention non-rail travel.

      5. Not only do the higher cants of the track pose an enormous problem for running huge freight trains, huge freight trains would destroy the high-speed trackage and right-of-way.

        Additionally, if you still have operational freight trains traversing the line while passenger trains are doing the same, passenger operators are legally required to use trains that are forcibly heavier and which do not fully tilt, compromising any new investment into the SEA>EVR line, or compromising the superior operation of modern electric trains on the existing infrastructure.

  29. Just to inform, I am a former railroader whose current employment revolves almost solely around the planning and construction of railroad infrastructure.

    I am not suggesting trains travel across the Cascades or through the border twice, no.

    What I am suggesting is that all coal trains sourced from Midwest coal fields or points farther east, traveling to Cherry Point, bypass Seattle and go through Everett via Stevens Pass.

    All other trains, bound for the Ports of Seattle or Tacoma, would use either the Gorge or Stampede Pass, again avoiding central Seattle. That, along with the SPIRE plan, would have transformational impacts on both our passenger and freight services.

    There is very little operational need for freight train service through and underneath central Seattle, and that would certainly be the case after the completion of the upgrades I have specified. The service most impacted would be loaded coal trains being redirected from the Gorge to Stevens Pass headed for Cherry Point, but that is only a minor business headache. Coal trains take mountain grades perfectly fine everyday over on the East Coast. That is not a radical proposition.

    1. I understand better now.

      Currently the coal trains use Columbia Gorge to deliver, and the empty trains return on Stevens, or at least that was the plan for the proposed massive coal terminal at Cherry Point. I presume this is because a full train is not easy to get over a mountain pass, and I imagine that coal and crude are the heaviest trains there is. Sure, they can do it, but I’m guessing the railroads would rather not.

      1. That’s right: it is definitely feasible, but it currently is not in the railroad’s immediate financial interest to operate oil and coal trains over Stevens. That is why you make it in their interest and provide financial assistance to the upgrade of Stampede Pass, and perhaps also provide political support in some manner that facilitates the reorganization.

        Certainly, this should be the option to explore before spending multiple billions of dollars on an unnecessary high-speed line through established neighborhoods and homes into Seattle.

Comments are closed.