A Duwamish Bypass: Lots of Bucks, Not Much Bang
Conceptual Duwamish Bypass
When all the construction is done and light rail spreads across greater Puget Sound sometime in the next decade, one segment will look distinctly different from the rest. The 7 miles of track through the Rainier Valley will be the only section one of few sections, along with parts of East Link, where the trains run at-grade. Though they have a dedicated lane, and pretty good signal priority, trains are limited in speed and still have to contend with the occasional bad driver shutting things down.
Nonetheless, the idea of bypassing the Rainier Valley comes up in conversation again and again, as a way to increase speed and reliability on a light rail line that may one day stretch nearly 70 miles from Everett Community College to the Tacoma Mall.
If you view Link as an airport express service or a long-haul commuter rail, then the Rainier Valley segment "feels" slow or wrong. But if you see Link as, well, light rail, then it's doing exactly what it's supposed to do.
All that said, the bypass is an idea that comes up often enough. UW Professor Cliff Maas once suggested that bypassing the Rainier Valley could help combat climate change. It seems worth considering. As an exercise, what would such a bypass look like? How much would it cost and how much money would it save? Estimates of 12 minutes travel time savings have been bandied about, though I have no official source for that. So let's see if we can do some quick math and figure out the pros and cons.
A reasonable bypass would run 5.6 miles from SODO to the South end of Boeing Field, where it would reconnect with the main line. You could extend the it farther south to Sea-Tac, but once you get south of Boeing Field the train picks up speed anyway, and it would be silly to skip Tukwila International Boulevard station. The Rainier Valley segment of Central link, from where it leaves SODO to the South end of Boeing Field, is about 1.4 miles longer, 7 miles of total track.
The bypass would have to make some stops. It would be unrealistic to build 5.6 miles of track without a single stop, especially through economically disadvantaged communities like Georgetown and South Park. Boeing Field would almost certainly want one for future commercial flights. So count on at least two stops in our hypothetical 5.6 mile route. That adds some dwell time at each station, plus acceleration/deceleration.
Service to the Rainier Valley would suffer. One can imagine that Rainier Valley riders might see reduced frequencies to accommodate trains entering and leaving downtown via the bypass. There may be ways to mitigate this, but it would certainly be an issue. Rainier Valley riders would rightly complain.
It would cost about a billion dollars. Using the numbers for Lynnwood link, which come to about $180m/mi, our bypass would cost roughly $1B. Possibly more with the river crossings. That would have to be evaluated against other possible extensions, to places like Burien, Crown Hill, Renton, Lake City, or the Central District. It's hard to see the bypass coming out on top in that battle.
It would save only a couple of minutes off of the travel time to the Airport. That sounds counterintuitive, but bear with me: In 2023 travel from SODO to the south end of Boeing Field via the Rainier Valley should take about 16 minutes*. To estimate how long the bypass would take, we need a straight, grade separated section of track to compare it to. Northgate-to-Mountlake-Terrace is perfect, about the same distance and also featuring two stops. Sound Transit says NGT-MLT will take 9 minutes, so the bypass saves 7 minutes in the best case. But remember that the RV line still exists, which means that only every other train (or so) will use the bypass. Now our time savings is down to at most a couple of minutes, as the average trip will involve more waiting time on the platform.
I hope this exercise makes it clear that a bypass through the Duwamish Valley would cost a lot of money and provide little benefit. To be clear, this is a separate question from whether or not the Duwamish Valley should be served with light rail at all. It might be a good idea, but it should be evaluated in terms of the potential benefits to the communities it might serve, not in terms of convenience for airport travelers.
Yes, it's annoying when some yahoo driver cuts off a Link train and service is suspended while they sort it out. It's doubly annoying when it happens at rush hour, and it will be triply annoying when it affects a 70-mile "spine" that covers the entire region. But instead of thinking of ways to mitigate that annoyance with another $1B in spending, we ought to re-think whether or not it's a good idea to build a single 70-mile light rail line to begin with. These questions should be top of mind as the board evaluates ST3 projects in the coming months.
* Once buses leave the tunnel, speed and reliability will increase noticeably. If you want the gory details: SODO -> Rainier Beach = 14 minutes in 2023. RB -> TIBS = 8 minutes. South end of the bypass is about 1/3 of the way from RB to TIBS, so roughly 16 minutes from SODO.