1. Train/Bus station

Last week, I covered part 3 in the series about Seattle-Vancouver high-speed rail, covering the Bellingham-Vancouver segment; the first two parts, by Zach Shaner, covered Seattle-Everett and Everett-Bellingham. This piece, the last part, covers the possibilities for suburban stations. Is it better to build stations in Downtown Vancouver and Seattle, or in less constrained suburban locations? Are there in-between compromise options, urban but less central? The answer to the last question turns out to be yes in the intermediate cities, but in Seattle and Vancouver the answer is no: downtown stations are required.

Endpoints and Intermediate Cities

In Seattle and Vancouver, the main tradeoff is that a station in the central business district is more accessible but more expensive. Between the two cities, an additional drawback of central stations, beside the cost, is the noise a 220 mph express train produces. If the location is too constrained for 220 mph, then the speed restriction becomes another reason to build outlying stations.

It is not common internationally to build central stations for in-between cities. In France, all stations located on high-speed lines are outside built-up areas. South Korea and Japan do have some central stations, often with speed restrictions or extensive noise abatement; but in Japan, the trend on the newer lines is toward outlying stations.

In comments on part 3, Mark Dublin zooms in on another factor: the construction of an HSR station in a city may cause its CBD to shift toward the station. The best example I know of is Lyon Part-Dieu, an urban neighborhood that is not the historic center of the city. Thanks to both local investment and TGV service, it became France’s second-largest business district, after La Defense, just outside Paris.

Metro Lyon has 2.1 million people, and the TGV connects it to Paris, metro population 12 million. The ratio between the population of Seattle or Vancouver and that of the intermediate cities is even larger than that between the populations of Paris and Lyon. This suggests that building near-downtown stations in Bellingham and Everett, next to I-5, would cause those areas to become more central, helping build up ridership. In both areas, car ownership is high, and the good freeway access is a boon for ridership. But there is no hope for the same CBD shift in Vancouver and Seattle themselves—they are too big.


Vancouver station options: Waterfront, Pacific Central
Vancouver HSR station options (Map by Oran)

In last week’s post, I explained why the only reasonable station locations in Vancouver are in or near city center: there is relatively easy right-of-way from the border right to the existing station, Canadian National’s Pacific Central, two SkyTrain stations outside Downtown Vancouver. The alternative location is Canadian Pacific’s Waterfront, right in the CBD. It is relatively easy to build a suburban approach that feeds the CN track, but not the CP track.

There is a CN branch going from CN’s mainline to the CP mainline, just outside the CBD, the Burrard Inlet Line. It is single-track, with several grade crossings; CN runs six daily trains on it. While HSR can cross roads at grade at lower speed, the neighborhood is fairly dense and has criticized the disruption to surface traffic coming from just six trains. Grade separation and double-tracking are required, with minor takings; this would involve about two-thirds of a mile of trench. Vancouver should be able to build this for $100-150 million.

The Waterfront location is superior to Pacific Central. Employment in Vancouver is concentrated in the CBD, and in a number of secondary centers, none close to Pacific Central. Tourist attractions in the area (such as Stanley Park, the steam clock, and the historic Chinatown) cluster in the CBD as well, which is why the hotels are all in that area: according to the Five-Star Alliance, there are 18 hotels in Vancouver, two near the airport and the rest in the CBD. It is almost certainly worth it to build a trench to get HSR to Waterfront.


Seattle high-speed rail station options: King Street, Midtown, Redmond
Seattle HSR station options (Map by Oran)

In Seattle, the only feasible alignment into the CBD is I-5, taking over road lanes near city center, where land acquisition for an alignment running alongside the Interstate is too expensive.

The main option for where to go from I-5 is a short tunnel, perhaps a mile long, to King Street Station. The tunnel would have to cross under the Link tunnel (as the legacy tunnel carrying the BNSF tracks under the CBD does), but should not be any harder to construct than existing tunnels. The only real difficulty is ducking under the legacy tunnel, but even that is (barely) doable with 4% grades descending from the station.

Other options involve staying on I-5. King Street is at the southern periphery of the CBD; an elevated multi-track station capping I-5 could provide a more centrally-located station. The cost would be losing easy access to Link. With no room for parking, the station would not even have good highway access in lieu of public transit access.

Instead of continuing on I-5 and stopping in the CBD, it is possible to stop south of the CBD. In the first part of the series, Zach identified the Stadium area, one stop on Link south of King Street, as an option. This spares Seattle the need for the tunnel, while maintaining Link access. The problem is that it lengthens egress time for arriving passengers.

The more radical option is to avoid Seattle entirely. There is ample land to the east, the land of I-405 and 520 and low density. Microsoft has spent a little money on the Pacific Northwest HSR studies; is it possible to make Redmond the Seattle terminal? In short, no.

The two square miles of Redmond around Microsoft’s headquarters have 60,000 jobs. The single square mile of the Seattle CBD west of I-5 has 150,000. As in Vancouver, the CBD job concentration means business travelers are overwhelmingly headed downtown: the Five-Star Alliance recognizes 22 four- and five-star hotels in Metro Seattle, three well outside the built-up area, three east of Lake Washington, and the rest in the Seattle CBD. The Space Needle and the university are just outside the CBD, far from Redmond. Nor is a Redmond route a feasible first step, to be followed by a connection to Seattle: Redmond is east of Seattle rather than north of it, and building HSR from Redmond to Seattle is both circuitous and expensive, since tunnels and a long bridge over the lake are needed.

Downtown-to-Downtown HSR

It is tempting to save money by avoiding constrained CBDs. In outlying cities, this is usually the correct decision, including in particular anything between Seattle and Vancouver. But in Vancouver and Seattle itself, it is the wrong approach. In Vancouver, missing the CBD saves a little bit of money and reduces the line’s utility by more. In Seattle, it can save a lot of money using the Redmond route, at the cost of making the line useless to people not visiting Microsoft or nearby tech firms, who form the majority of Seattle-bound travelers.

The good news is that it is not that hard to get to both CBDs. Vancouver is easier, but even in Seattle, what is needed is a small project by the standards of the ST3 tunnels. The real difficulty is construction alongside I-5 in the Seattle built-up area, but even that requires more political will to take freeway lanes just outside the CBD than intricate civil engineering. Project costs under $10 billion seem achievable.

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

79 Replies to “Seattle-Vancouver High Speed Rail Part 4: Terminals”

  1. If Microsoft wanted to help fund a diversion, one could also split the line in Lynnwood and send 1/4 of the daily trains to Bellevue. That might have enough draw from both north or south to fill the trains.

    Completely skipping Seattle would not.

    1. The suburban station mentioned was Redmond, but I actually think Bellevue would be an even better choice. We could probably build Link across 520 with the money savings. If you put the station in the right place and make the train -> Link connection simple, you’d be 10 minutes away from UW, 10 minutes away from Redmond, and 15 minutes away from the CBD. And I’m guessing the time savings may balance out. First, a lot of people won’t be right around the stop anyway. Second, I’m guessing you can maintain faster speeds for longer if you don’t have to tunnel around everything.

      You could probably even put regional trains on the same line running from Everett to Bellevue. I’m guessing synchronizing trains should be pretty easy – just have the regionals stop at a station side track while the high speed train comes through.

      1. Like I said though, Bellevue instead of Seattle is not going to happen. Not enough people would ride it. They’d fly instead.

        The train needs to go to the bigger urban center or it won’t get used enough to justify the cost.

        You could envision both Seattle and Bellevue, but not Bellevue to the exclusion of Seattle. HSR needs to have the right terminals for folks to use it.

      2. Why would you fly? It’s 40 minutes to the airport + an hour waiting for the flight + an hour flight + 30 minutes from the airport to the center of Vancouver. That’s over 3 hours. Instead of adding maybe 20 minutes for the train in exchange for substantial cost savings and letting many other people save time.

        Practically, you’re not going to get just people from around King St. Station. You’ll also get people from UW, from SLU, from Redmond and Bellevue. It won’t be 90% from walking distance to the station and 10% from outside. And, as I mentioned, I’m betting running along 405 will let the train run faster longer.

      3. @David — You’ll also get people from UW, from SLU, from Redmond and Bellevue.

        Which is why the east side is not a good choice. From Westlake to Bellevue takes 24 minutes, which means that from Belltown or South Lake Union it takes well over a half hour. From UW to Bellevue takes a half hour, and that is if you happen to be standing right by the station. There are way more people living and working in Seattle than there are in Bellevue. If you wanted to save money and serve a suburb, you would choose the closest one, which would be to the north. Northgate, for example, might be substantially cheaper, yet close enough to the city core to be palatable.

    2. Downtown Bellevue could work – it’s only 20 minutes by Link from downtown Seattle, or half the time of Link to the airport. Seattle is still much, much better, but Bellevue might pencil out as the next best option if there’s some technical or political reasons why I5 won’t work but 405 works just fine.

      1. The competition isn’t just the airplane, it is the car and simply not doing it. If you put the station well outside where most of the people work or live, then you will lose a lot of people. Twenty minutes doesn’t sound like much, but that is on top of the time it takes to connect to Link. For example, from Belltown it would take at least a half hour. From the UW, Ballard, First Hill, Central Area, Lake City or any other high density area, it is much longer.

  2. Would it possibly be a less expensive option for Seattle, to go around the east side, and into Seattle along the BNSF tracks from the south?

    As a bonus, when we at some future point go to HSR to Portland,, it would mean that the egress from Seattle is already started

    1. That’s what I was wondering too. With Link, riders could get to Downtown from the north or from the east. Riders will need to use Link anyway for most of Downtown Seattle if the destination is King Street; the station area is off-limits to buildings over 20-25 floors.

      ST says it will be 28 minutes from Lynnwood to Downtown Seattle. HSR would probably still be 15 minutes given the slower speeds needed for urban geometries — and reaching much of Downtown would require some backtracking if the stop was King Street, taking more travel time.

      A 405 alignment could do double-duty for faster local service too, and have a profound change on accessibility to jobs and residents.

      Ideally, a third stop with a Link station to the south would also be great. Either one at Boeing Access Road or at SeaTac Airport, depending on cost or constructibility.

      1. TIBS might also work nicely as a Link transfer spot if the HSR station was sited between the rental car garage and the Link station — and there was an added tram to the airport check-in locations (connecting the rental car facility too, btw) from there. The aerial alignment could follow 405, 518, 509 to South Park and next to 99 as an aerial track almost to King Street Station. That would enable a faster connection to SeaTac airport from both Downtown Seattle and Bellevue, as well as a HSR connection northward.

  3. The only way into the CBD is going to be via a tunnel. WashDOT isn’t going to give up the express lanes or GP lanes and you’ll have an absolute revolt and alienate commuters. Let’s be realistic if you want to get this built. I-5 north of Everett is basically the last useable bit. The corridor between Everett and UW has zero room for expansion to accommodate HSR without a massive eminent domain process.

    It does make sense to do a branch to serve the Eastside corridor but that shouldn’t be the main core/consideration.

    1. Hell will freeze over or the zittenial generation will not have cars before we can reduce the I-5 lanes near downtown. It backs up to a standstill every weekday and game day, so subtracting two lanes or three would cause third-world level traffic jams. (On the other hand, fears that the 99 tunnel will be underused would vaporize.) This HSR line would exist in the context of deciding to spend tens of billions of dollars to build it, and if we’ve decided to do that, then a short tunnel downtown will be a small detail. And as Alon said, downtown stations in large cities ensure the maximum ridership. One of the advantages of trains over planes is that they go downtown-to-downtown, and Vancouver-Seattle-Portland is within the distance that getting to the airport is a significant part of the trip. Microsofties can take the shiny new light rail from King Street Station to Microsoft Station, and back to the downtown Bellevue for their hotel. If they try to do that from Sea-Tac airport…, well, they’ll only do it once. If SDOT were in control of I-5, then there might be a chance of taking lanes, but this is WSDOT, which is so adverse to inconveniencing SOV drivers that it won’t even guarantee 45 mph in the HOT/HOV lanes for 405 BRT or I-5 express buses.

    2. Once Link Light Rail reaches Lynnwood (2023), it will be much more feasible, practically and politically, to convert the I-5 express lanes to rail.

      1. What Mike said. I just can’t imagine that Seattle would replace the express lanes with high speed rail to another city. I could see if the city had replaced the express lanes (or the HOV lanes) with Link, but it is way too late for that. The only change that I see being possible is that the express lanes be made bidirectional. The best thing about Link is that it will make the trip that doesn’t involve the express lanes (i. e. Northgate to downtown in the evening) much, much better. Right now that is much worse than the traditional commute, and that is where Northgate Link will really be a huge improvement (well, that and trips to Roosevelt, the UW and Capitol Hill will be a lot faster).

    3. Using the express lanes can be a slam dunk if the ROW does double duty as commuter rail. The ROW is plenty wide for 4 tracks downtown, right? So you repurpose the express lanes with 2 HRS tracks and 2 commuter rail tracks.

      The commuter rail can be DMUs, and there can be multiple commuter stations in places like Marysville that would never be served by HSR. The commuter rail can also deviate from the HSR alignment outside of Seattle.

      1. We are building a multi-billion dollar commuter rail system. It is called Link. I really don’t see us building another one on top of that from scratch, while reducing the capacity of our freeway system. Just not gonna happen.

      2. I disagree – they would serve completely different populations. This commuter rail would operating similar to a Sounder line – I would envision it bring in people from north & east Snohomish (Marysville, Monroe, etc) into Seattle CBD, and from there they can walk to work or transfer to Link to get to jobs in UW, Bellevue, etc.

        If the commuter rail was duplicating anything, it would be North Sounder, not Link. Imagine how much better Sounder would be if someone could get from Everett station to Seattle in 25 minutes! That could completely change commuter patterns in & around Everett in ways Link never will. We have built Link with an extra track so we could run express trains from Snohomish into Seattle, then sure commuter rail would be unnecessary, but no one with any real power has suggested that’s a good use of Link.

        Arguing we don’t need commuter rail because we have Link is like arguing Chicago shouldn’t bother investing in Metra because it has the L. It’s a both-and solution.

    4. This is the tragedy of NorthLink. Given the stop spacing is really at commuter rail distances northlink should have been built as a heavy rail alignment that could also carry HSR. HSR trains would express through stops while the commuter and regional local trains from Marysville and Bellingham would stop at all stops. A more frequent stop spacing alignment for local travel would in future have run up 99 via Paine Field and connected with the main heavy rail network east of PAE or at Everette.

      1. Edit: to add. Yes it would cost more. But the outcomes would be far superior and it would have been cheaper overall compared to building duplicate incompatible rail lines (Link and heavy rail) along the same route.

  4. Something I do not understand… Why have all those other stops? Vancouver, and Seattle. That’s it! If the train has to stop all these other places, why build it at all. It is a High Speed Train after all!

      1. There’s a point at which the additional passengers outweigh the time increase of deceleration and dwelling. A stop every 30 or 60 miles is reasonable. A stop in suburban Pugetopolis and the Lower Mainland is also reasonable; Surrey or Everett could do for that.

        We’re not talking about trains to California or Chicago, but within the 400-mile corridor that is considered a good fit for regional rail. So it’s not like a 10-hour trip stopping too many damn times, or like a Metro bus stopping every other block. It’s just stopping every ten or fifteen minutes for cities larger than 20,000 and for the outer part of large metropoli, more or less.

      2. Even the trains to California and Chicago tend to stop roughly every 100 miles, except in areas where there are very few people. The wonderful thing about trains is that this doesn’t cost much: if it’s done well, maybe three minutes. For that, you can get a dozen people per train at a tiny city, or several dozen at a larger place like Bellingham.

      1. Part 1 & part 2 of this podcast is an excellent start on “WWFD.”

        Basically, they build both urban station and suburban stations along TGV lines, because both urban & suburban residents deserve to be served. “Suburban” stations resemble airports with lots of long term parking, etc.

        With this model, there would be some urban stations in CBDs (Seattle) and some suburban, non-CBD stations (Bellingham). Both can work along the same HSR line.

      2. I don’t actually know what France would do. France isn’t building HSR in such constrained environments. In less constrained environments it would build a line avoiding the built-up area and site intermediate stations tangent to urban areas (e.g. Avignon) or sometimes well outside them (Champagne-Ardenne, Haute-Picardie). But neither France nor other countries I know of would have every train serve those intermediate station. Perhaps 1 hourly train would run local and 1 would run express.

    1. The other thing I don’t understand is why are we assuming one train and one line pair? In Italy I took the HSR and the rail line went through stations with multiple trains and multiple lines.

    2. What you described is basically what an airplane can do today, so why spend tens of billions on duplicating that? The nice thing about trains is they can stop along the way without a significant time penalty (with reasonable spacing, of course). That means better use of expensive infrastructure and opening more trip possibilities for people along the line.

      It’s a political thing too. You can bet that Everett and Bellingham will want a station on a line being built through their cities. Not every train has to make every stop. You can have express and local trains as well.

      1. Come on, Oran. Last time in Europe, all I could think of every time I saw a Finnish jet airliner take off for a fifteen minute flight to Madrid was:

        “Damn! Why couldn’t they see their mistake and quit imitating the United States when they put the first McDonald’s in Vadstena, or intercity traffic jam under the wind turbines in Skane (Sko-nah) Province?

        One thing a jet can’t do? “Stay On Trolleywire!” Or even have pilot get out and re-wire.


      2. “What you described is basically what an airplane can do today, so why spend tens of billions on duplicating that?”

        And another angle- while the laptop ban didn’t happen this time, it hasn’t been taken completely off the table. If it becomes an issue again, those little planes doing the Seattle-Vancouver run are no longer going to be a thing, international out of Sea-Tac is going to become more difficult, and more buses aren’t necessarily going to be the best alternative. HSR from Seattle to Vancouver (then possibly international flights out of YVR) will be the most practical option to allow people to keep their laptops on them.

      3. I wasn’t arguing against HSR. I was arguing against the original commenter’s “why build it at all [if it makes stops between Seattle and Vancouver]” statement.

        How long would the laptop ban last? By the time this line gets built? There are better, more fundamental reasons to build HSR than getting around a policy that can change on a whim. There’s no guarantee that HSR would be exempt from security theater.

    3. It is always a judgement call. Neither Everett nor Bellingham is that big, so you can make the argument that it simply isn’t worth it. However, both have advantages from a geographic standpoint. Everett would draw from all of Snohomish County, and perhaps part of north King County. If it is a pain to take public transit to downtown (too many transfers to Link) then someone will want to drive to Everett instead. For example, if I’m in Bothell I could drive to the park and ride, take the bus, then the train into Seattle, then take the high speed train back north. Or I could drive to somewhere in Everett — possibly even the station itself (especially if the station is in the outskirts of town).

      Bellingham is a bit different. The argument there is that it is a growing city, and far enough away to have no logical alternative. This would be twice as fast as driving, even when you count the time spent waiting and getting to the station. It is also a college town, with frequent travelers heading to Seattle (and over the border to Vancouver). This means that you would have higher ridership than the city size would suggest. It also means the city itself would boom (which might get folks in Bellingham to wonder whether they even want the thing). It isn’t that I see a huge number of commuters, but that day trips to Bellingham would become common.

      But more than anything, the cost of adding a couple stations is pretty small. It does slow down the train, but not that much (and you could easily have an express that skips the stops). The key point is that we shouldn’t spend a huge amount of money making sure this serves downtown Bellingham and Everett. We should simply add a stop along the way, and if it turns out very few people use it, then we just skip it.

      1. … and Canadians shopping in Bellingham when the exchange rate is favorable, or picking up mail-order deliveries from vendors that won’t ship outside the US.

    4. For a few years, the Lyndon LaRouche people were demanding a whole planet bored through for either HSR or vacuum tube trains, I forget which. Bet the Republicans wish they’d never put him in charge of the Republican Party. Nobody ever knew his make-up people ever knew he had a wig like that. Explains a lot.


    5. There would presumably be express trips that don’t stop, but not all trips would be express.

  5. While this series has been an excellent analysis of the technical difficulties such a project would face, it does not make the case why it should be undertaken.

    That’s understandable, because the case simply can’t be made. A high speed rail line between moderately large secondary cities in two countries joined by a historically excellent but rapidly decaying relationship is doomed to fiscal abandonment.

    Di you really think that the howling Baboons of the nativist Right will allow what they view as the subversive, increasingly Chinese dwellers of Vancouver free entry at the levels necessary to support high speed rail? It’s ridiculous.

    The only way such an integration might come about is after the disintegration of the United States. That will not happen without an economic catastrophe that would moot the need for the project.

    The discussion has been exciting, but in a way reminiscent of a visit to the Midway at “The Puyallup”.

    1. Oh come on. The relationship is not decaying, it is only suffering from the stupid rants of a demented man who temporarily occupies the White House. Only 37% of the country actually approves of the guy, which means that it is quite likely he (and his thinking) is just a passing phase. These things come and go. By the time we actually start funding this, it is quite possible we will have a completely different administration, as well as Congress, ready to do whatever it takes to improve the finances and well being of the country. In that case, a high speed rail line connecting the countries makes a lot of sense, given the distance. It is simply too far for cars, and too short for planes. This is where high speed rail works really well, and eventually is very popular. Neither city is huge, but both are growing, and likely to increase in importance as time goes on.

    2. https://www.amazon.com/When-Mississippi-Ran-Backwards-Earthquakes/dp/0743242793

      Well, Richard, I imagine right now the Canadians are having a “Next Time We’ll Do Different” moment.

      1811. The Southeastern US First Nations people finally had a leader who got them to quit fighting each other and take a week or two getting serious about deporting undesirables aliens.

      And Chief Tecumseh his British military counterpart in Canada really liked each other and were ready to finish the deferred problem of a rogue nation next door. With prize-winning special effects courtesy of the New Madrid Earthquake Fault, which makes the San Andreas look like a pencil scratch.

      (Yawn) too bad England didn’t feel like it. Well, won’t take the EU long to realize how little they miss them, except for the Indian (from INDIA this time!) food. Proving the English don’t have the monopoly on Understatement.


  6. For station arrangement, I prefer “suburban option” in the article. However, appropriate transit connection between suburban stations and downtown should be done at the same time, which needs transit agencies such as TriMet, Sound Transit and Translink to cooperate. As I know, many Chinese high speed railway stations are built far away from city center, but there are subway or bus lines carrying passengers between downtown and stations, many cabs are waiting there too, so it won’t make people feel too inconvenient to use those stations.
    PS: I think bridges and tunnels could be widely used in order to make HSR lines tracks flat and straight. In fact, more than 80% length of Beijing-Shanghai high speed railway is built on bridges, not on land.

  7. Brian, are you our locomotive engineer? Because we’re into some operating questions that need the view from the cab. Also, from the seat of a Tunnel Boring Machine.

    We certainly are going to have to tunnel. But not over land-use. True, 220 mph is a very low altitude thunderstorm. But question that needs asking: Southbound into Seattle at over 200, where exactly do you put the controller to “regenerative?”

    What’s minimum curve radius, and exactly where? Leading to next question: What are the rocks, soils, and underground rivers like anywhere in the Downtown area? Northbound under Third, roughly at Century Square, Big Bertha’s little pet Mighty Mole found our how Spring Street got its name.

    But major compensating factor is the amount of freedom we’ll really have down there- which should add a lot of black problem-solving to counterbalance the volume of red ink coming out in tank-cars. And excellent political funding justification, from little neutral countries Finland.

    Whose defense forces would likely voluntarily get obliterated in battle in a few days if Russia decided, from its own history, that it was soon going to be mortally vulnerable from the west. In Finland, and Norway, and Sweden, every subway station is an air raid shelter.

    Reason I’m not kidding about undiscussed purpose of currently-unusable I-5. Whatever crap Congress has been demanding in the name of national defense to create jobs in its districts since the end of World War Two, major land corridors, freight and passenger, and their facilities deserve every dime of Defense money they can get.

    Like the freeways, underground transportation complexes can provide benefits that really do make our country healthier and stronger. If we get into it with somebody that can really hurt us on the ground, the enemy’s balance sheet will be generous with money to take out these exact targets.

    Mark Dublin

  8. This is an enjoyable discussion (if a bit fantastical), but as long as we’re dreaming…. in the course of this study why not look at a floating tunnel from Vancouver to Olympia (with stops along the way of course), exiting the water at Olympia and heading south to Portland from there.

    Sure the engineering challenges would be enormous, but by the time you figure in costs of right of way acquisition, tunneling, and bridging…. it might end up being cheaper.

  9. I’m noticing that you didn’t examine the possibility of a Northgate terminus, as several past comment threads here have discussed?

    1. I agree, good point. An east side terminus is not a great idea. It would take almost as much work, and put the station in an awkward place. Even downtown Bellevue is over 20 minutes away from downtown Seattle, which puts it way outside the bulk of people in the city as well as the businesses.

      Northgate, on the other hand, is 13 minutes from Westlake, and 7 minutes from the UW. You can make a pretty strong case that the “Central Business District” of Seattle stretches from Madison to the UW, which means that a Northgate terminus, while obviously less than ideal, is still very good. If the savings from a Northgate terminus are big enough, it could easily work.

      For that matter, what about a UW terminus? It seems like that again is less than ideal, but still pretty damn good. You are less than ten minutes from every downtown station, and closer to South Lake Union than King Street Station is. Our downtown is very spread out, which means that you really can’t find a perfect spot that would allow someone to have a short walk to everywhere. Even if cost wasn’t an issue there are arguments to be made for places like Madison (where the biggest buildings are) or Westlake (which is closer to the center of downtown). Yet those stops are half a mile away, meaning you can’t please everyone, and like every city, there is no perfect center. The great majority of travelers will either transfer to transit or take a cab, and if have the terminus be at a Link station only a few stops away from both downtown and the UW, I think it would work really well.

      1. I disagree that Seattle’s downtown is spread out. The job density in the CBD is higher than anywhere in Paris proper. American cities generally tend to be spread out in that there is a lot of employment in the suburbs, but the cities proper have dominant CBDs, more so than in Paris; their neighborhood centers, the equivalents of Montparnasse, Gare de Lyon, etc., are weaker.

        The reason I don’t like a north-of-CBD terminal is that it’s an 1840s solution – avoid the CBD by building outside it. It means you can forget about connecting the line to a future HSR line to Portland; it also requires a great number of arriving passengers to connect to Link, whereas King Street lets people walk a kilometer to their destination. The compromises involved in this design are literally why London had to invent the Underground – there were huge traffic jams (of horse-pulled cabs and buses) between the train stations and the CBD.

      2. He’s talking about viewing the entire area from downtown to the U-District as a quasi-downtown. At that point you might as well extend it to Northgate because all of that area will be denser in couple decades.

      3. UW or a future Northgate are not part of the downtown central business district, not with them being surrounded by miles of low density residential areas. It doesn’t pass the sniff (or map) test.

        Midtown Manhattan and Lower Manhattan are regarded as NYC’s #1 and #2 CBDs, not as a single CBD, even though they are separated by a ten minute ride on the subway with the area between them completely built up.

      4. OK, it probably is a stretch to consider the UW part of the “Central Business District”, but it could easily be in the future (as they allow more tall buildings). Yes, Midtown and Lower Manhattan are considered two different CBDs, but from a practical standpoint, it makes no difference. They are both part of the central core of the city, which is why favoring one over the other is rather arbitrary.

        Furthermore, I think we are putting too much emphasis on the businesses. The core of the city (from a population and destination standpoint) is really between the UW and the southern end of downtown, which is really no farther than the two Manhattan CBDs. From the UW to Westlake is only three miles, and while it includes some relatively low density areas along Eastlake, not that many, really. The biggest obstacle to growth is simply geography (water).

        Which is not to say that I think Northgate will soon be part of all that. The point in including Northgate is that it should be considered, simply because it isn’t that far to that central core. This makes it substantially different than Bellevue, which is well outside the central core (20 minutes just to the edge of it). Again, I’m not saying Northgate is ideal. The ideal location is probably Westlake. But I’m saying that adding a station at the outskirts. is not the end of the world if the savings could be large enough.

      5. @Alon — Good point about connecting to places south (like Tacoma and Portland). It really comes down to cost, and we have no idea. Ideally you get into the heart of the city, but if not, I would be OK with getting close, with future extensions to the heart.

        But I disagree with the idea that King Street lets people walk a kilometer to their destination. From King Street it is over a kilometer to Belltown, the state’s most densely populated census block. It is over a kilometer to Pike Place and Westlake. You won’t have too many people walking to First Hill, either. King Street is really no longer the center of downtown, it is on the southern edge, with very little to the south of it. The city is still growing, and it is growing to the north, towards the UW. You would still want to locate in the heart of downtown (i. e. Westlake) but no matter where you locate, the bulk of people will connect to their destination via public transit and cabs. I just don’t see locating close to the city’s core (at say, the UW) being that much worse than location at King Street.

    2. A Northgate terminus only makes sense if you don’t plan to go any further south than Seattle, and I would suggest that the case for a high speed rail project would be greatly strengthened by the inclusion of a Seattle-Portland line.

      1. Maybe it’s too much detail to get into, but we also haven’t talked about phasing the construction (and the funding) so we can get the line up and running sooner. What if we terminate at Northgate with our initial operating segment, and then build the new tunnel from there to downtown Seattle later as a second phase?

      2. Thanks ColumbiaChris – such a project will unlikely be successful with only a line north of Seattle. Portland and Eugene must be included.

      3. @Alan — Exactly my thinking (for both ends of the system, really). Eventually you want to go right into the heart of the city, but if you get this thing built by going to the outskirts, it really isn’t the end of the world. If anything, it is turning the lemons that is our line to Northgate into lemonade. Ideally that line should have twice as many stops. But the trade-off would be fast service from downtown, and thus not much of a time penalty for the bulk of riders (who would likely come from somewhere north of King Street).

  10. In Paris, 4 out of 8 train stations serve as TGV terminals – Gare du Nord. Gare de Lyon, Gare de ‘Est, and Gare Montparnasse. All of those stations are within the city of Paris, in very tight quarters. Gare de Lyon and Gare Montparnasse are as centrally located as it gets. IDTGV and OUIGO, two low-cost TGV services operate in suburban stations, but actual TGV lines serving Paris, serve Paris. Once the trains enter city limits they slow down, but they’re coming to a stop anyway so it doesn’t really matter. It just gives passengers time to disembark.

    City options are feasible (though probably expensive like you noted). Could the rail company (assuming it is not Amtrak?) buy Union Station from Sound Transit and bore another tunnel headed there? This would function like Lille, where they have Gare de Lille Flandres for normal trains and Gare de l’Europe for Eurostar and TGV right next to one another.

      1. Interesting. Though it sounds like Paul Allen still owns the land the building sits on, which would certainly factor in to any attempt to put the station back in use for rail.

      2. According to King County property records, Union Station is organized as a condominium. ST owns the Union Station unit. Paul Allen’s Vulcan has one building.

        The larger issue with the Union Station site is all the stuff already underground. You have the existing DSTT station, the bus layover area, a thousand car underground parking garage, and several office buildings sitting atop it all. ST plans to build a new Link station under 5th Avenue adjacent to IDS for Link to Ballard in ST3. So any new rail platform would have to be deep under the complex which could be more expensive compared to a station lidding I-5.

    1. 1. The only station in the Paris CBD is Gare Saint-Lazare, which has no TGVs, only commuter trains and some short-range legacy intercity trains. Gare de Lyon is in a secondary CBD, though.

      2. iDTGVs leave from the same stations as regular TGVs, they just don’t have conductors collecting tickets – instead tickets are checked at the station. OuiGo indeed leaves from middle-of-nowhere stations.

      3. The Lille-Flandres vs. Lille-Europe separation isn’t an ideal situation. Lille-Flandres probably would’ve been the Eurostar station if it had been a through-station rather than a terminal. In the US, very few cities have this problem – the only one I can think of is Los Angeles, which is building run-through tracks at Union Station. (Boston South Station is a terminal as well, but it’s also a natural northern terminus for intercity services.)

      1. And the Boston North-South tunnel connection has been repeatedly proposed.

        For that matter, a southbound exit from Grand Central to Penn in New York City has been repeatedly proposed.

        Through stations are the way to go, period, end of story. Nobody builds terminals any more, and there are good reasons for this. (Grand Rapids just built a terminal but it’s an unusual situation.)

        You’re kind of doing everything wrong north of Seattle. The real problem section is Everett to Seattle and Seattle keeps foreclosing on the possible routes. Tunnelling all the way from Everett to Seattle may be the only viable option after a while. Which is not going to happen any time soon.

  11. The Downtown Seattle stop goal is just another reason why losing Convention Place is so tragic. If the site was still available, that would be a great Downtown location for a number of reasons.

    1. I was thinking the same thing. There it is, centrally located, right next to the freeway. I have no idea if it is feasible to run a train there, but if it is, then it is a shame to lose that spot.

      1. For God’s sake, no matter how low their bid is, Breda is NOT going to get the contract for the buses!


    2. All that underground space is becoming loading docks & parking. There’s still an option in the future to replace all that space with a station. It’ll certainly be more expensive than if a station could be constructed without, you know, an building in the way. But with Crossrail and other modern projects, entire stations are being built underneath existing buildings. So can see a solution where we replace a giant underground parking garage with a station.

  12. I still don’t buy the argument that extending the line into the heart of Vancouver is worth it. Every argument you can make applies today, for the existing railway. Yet they haven’t seriously considered it, because doing so just isn’t worth it.

    The existing station is in the middle of nowhere, but is only a very short distance to the rest of downtown. No matter where you put the station, the vast majority of people will not be right by their ultimate destination. They will have to transfer to transit, or take a cab. Who cares if the cab ride is ten blocks or fifteen.

    It is true that Pacific Central does not connect directly with all of the Vancouver SkyTrain lines. But it connects to the two most popular (the Millenium and Expo). Getting to the Expo Line is really not that difficult, either (just a couple stops and then a nice, easy transfer). I really don’t see that as being a big deal, compared to the enormous amount of time you save versus getting to the airport. Extending this to waterfront is ideal, but sounds like one of those “nice to have” things that can be put on the back burner for a long time.

    1. First, Pacific Central doesn’t actually connect to the Millennium Line now that it’s turned into a separate line rather than an Expo Line branch. I did look briefly into a Commercial/Broadway option, at the intersection of the Expo and Millennium Lines, but rejected it because that site is too constrained.

      Second, the reason Vancouver isn’t building this trench is that there are two Cascades trains per day. Without HSR, there’s not nearly enough intercity rail traffic to justify this. Nor is there any point in commuter rail on the CN line, which is directly parallel to the Millennium Line.

      Third, you can literally walk from Waterfront to many of the downtown hotels and office buildings. A train doing Seattle-Vancouver in 1:15 would induce a lot of day-tripping: maybe I live in Seattle, take the train to Vancouver in the morning, have meetings all day there, and then return home in the evening. Or vice versa. In both cases, direct CBD service easily saves you 10-15 minutes in the city you don’t live in.

      Now, the main counter is that Main Street, as you note, isn’t that far off. 10-15 minutes isn’t a lot of time. But then, $150 million isn’t a lot of money, especially compared with the cost of saving those 10-15 minutes via higher speeds through Everett and points south.

      1. >> First, Pacific Central doesn’t actually connect to the Millennium Line now that it’s turned into a separate line rather than an Expo Line branch.

        Fair enough, but that means, of course, that the Millennium Line doesn’t connect to Waterfront Station either. It also means that Pacific Central could connect you to the Millennium Line faster than Waterfront Station. The only line that Pacific Central doesn’t connect to very well is the least important one (Canada).

        I don’t buy the 10-15 minute thing. It is a very short and frequent ride to the rest of downtown. Simply taking the next train and riding to the end is ten minutes at most, and there is a good chance you will get off before the last stop. There is a reason they added all those stops north of there — they are worthy destinations in their own right. If you end up taking a cab, then it probably costs you an extra five minutes or so, depending on your destination.

        Without a doubt the ideal station is Waterfront, and maybe $150 million isn’t much, but compared to what Vancouver still needs to build, I just don’t see it as a priority. Like King Street Station, Pacific Central isn’t the ideal location, but it isn’t that bad, either.

      1. Which makes sense given the rapid pace of development in Richmond and near the airport.

  13. I’m surprised that the importance of not-precluding Portland thru-running hasn’t been mentioned much since I would imagine that any HSR scheme that put Portland in HSR time of Vancouver would be much more useful

  14. Can you clarify what “project costs under $10 billion seem achievable” covers, exactly? Is that for the urban approach into Seattle from Everett along I-5, or is that the whole HSR from Vancouver to Seattle? I’m trying to compare it to Zach’s cost estimates in part 2 of this series.

  15. The obvious terminus in Vancouver, given both cost and travel time (the train would have to be moving quite slowly west of the Grandview Cut due to tight turns, etc., and the cost to move the train westward form there would be enormous), would be a station in/around where the existing rail line in the Grandview Cut intersects the Expo and Millennium lines – either at the existing Commercial/Broadway station or immediately west at VCC/Clark, which is soon to be extended westward and intersect with the Canada Line.

  16. I’m not sure that it would be worth it to bring the train all the way to Waterfront (in Vancouver) – that would be extremely expensive. The transfer to the SkyTrain via Main Street – Science World station directly across from Pacific Central is convenient and runs on a frequency of minutes, not tens of minutes, thanks to the automated system, and you can get from there to Waterfront Station in just a few minutes with delays being extremely rare.

    Downtown Vancouver also has high density over a larger area, so plenty of people could walk quite easily to work from Pacific Central (< 10 minutes)

  17. There is always the possibility of elevating the two tracks before entering Everett and keeping them elevated along one side of I-5 and above the upper deck of the Ship Canal bridge and into that Midtown station over I-5 in downtown Seattle. As long as there is an easy connection to the 5th, Madison (Midtown) station in the second Light Rail tunnel to be built for ST3, maybe long escalators from mezzanine to mezzanine. The elevated tracks can work their way south out of downtown Seattle and eventually down to the ground level around Boeing Access Rd and on to Portland. What do you think?

  18. What about a Westlake terminus? It would be pretty easy to bore from I-5 under Howell and put the platforms under Olive. Take a left to go to Westlake Link platforms, a right to go to the terminal. Just a quick turn right past the platform down 2nd Ave to join with the BNSF tunnel would let it extend to Tacoma and Portland as well.

Comments are closed.