The Technical Challenges of Seattle-Vancouver High Speed Rail: Part 1

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"^2 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"^2) than the max allowed for trucks (61"^2), 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"^2, 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: The Inland/Suburban Option