When Toronto faced escalating cost on their relief line, instead of extending their existing subway network, they decided to build a separate Ontario line and use lighter and shorter automated trains rather than their prior subway trains. Lighter trains mean faster acceleration, allowing steeper climbs including from within a tunnel to shallow stations and even elevated portions and smaller tunnel diameters. Shorter trains mean smaller stations which can fit more easily into the existing urban landscape. A separate line minimized the impact on the rest of the transit system.
As Sound Transit ponders escalating construction cost, it may want to consider the same. It may not only save them money during construction and operation, it may also provide a better ridership experience.
As subway ridership in the downtown core grew, a new “relief line” was discussed to ease congestion at the intersections of Line 1 (yellow) and 2 (green). Soon its cost estimates escalated. They finally hired Michael Schabas as a consultant. He had worked on automated lines in Vancouver, Honolulu and London. Within 3 months he came up with a plan for the Ontario line. By building a separate line with the latest light metro train and signal technology, it would not need to integrate with their older network and could be procured more independently. As lighter trains allow for steeper climbs and tighter turns, they provide more flexibility in using tunnel, at-grade, and elevated sections. That included using a bridge over the Don River rather than a deep tunnel underneath.
With new leadership in place in Seattle, a similar approach could be used in Seattle to accelerate the planning efforts and use the technical advantages of automated light metro trains to build a better line for Ballard:
- Shorter stations mean better stations. They can be closer to the surface or in better locations (e. g. squeezing in between building foundations). The Denny station may fit on the SLU Discovery Center park block. This would reduce construction impact.
Stations and their approaches need to be straight and level as the driver needs clear sight lines. Shorter automated trains are far more flexible: trains can turn or go up or down right out of the station. That may allow for pushing both the SLU and Seattle Center stations further east. - Automated trains allow for higher frequency and longer operating hours. Vancouver has trains running every 90 seconds. Copenhagen runs their trains 24 hours on weekends.
- Modern light metro trains also allow for smaller tunnel diameters which reduces construction costs.
- Better for current riders. With a new tunnel, riders from the south (e. g. Rainier Valley) are much worse off. They lose their good downtown stations. They lose their direct connection to the UW and other north end destinations. Their transfer for trips to the East Side will likely be much worse. Riders from the north (Capitol Hill, the UW, Northgate) will lose their one seat ride to places like Rainier Valley and the airport. This can all be avoided by building a second, independent, automated line.
With the recent ST3 cost increases and Sound Transit rethinking their plans to serve the region, several board members have suggested that a second downtown tunnel (DSTT2) may not be affordable. On their behalf Sound Transit has been evaluating alternatives such as interlining or a separate stub line (the Stub-End alternative). The blog had four articles on this; see the archive December 15-21, plus the open thread December 17 starting here.) The Stub-End alternative is similar to an alternative I outline last year and that many on this blog contributed to.
While the Spine has a lot of political momentum, a separate Ballard/SLU line would provide the highest ridership opportunity for our regional network and an opportunity for our region to catch up with the latest urban transit technology our neighbors in Canada, Hawaii, and around the world have been using for many years.
Thanks, Martin. This is very well organized and written, complete, and not too long for an otherwise interested but not technical person to read.
At this point doesn anyone doubt that the promised Federal Funds for ST3 projects will not be forthcoming? We’re one of the “Bluest of the Blue” states, and Seattle is “like Portland”, so Trump probably thinks that Katie is an Antifa.
The financial short-fall will get even worse when that happens. South King, Pierce and Snohomish Sub-Areas can’t afford to meet their own rising costs, much less increase their contributions to DSTT2. In the case of Pierce especially “sub-area equity” demands that their project be completed as a priority, with the extension to Everett second. East King could certainly delay Line 4, but it’s already at the tail end of the projects list, so they can’t really help either, except maybe with a loan.
Seattle has to accept that North King can only afford one of the three projects comprising WSBLE — it’s arbitrary to include the portion between SoDo and Westlake in either WSLE or BLE, because it’s categorically different than any other ST3 project: it duplicates an existing facility and service. On its face it is the least essential segment.
Oh, maybe in 2050 it will really be needed, but the existence of the United States in 2050 is in severe question right now, so there may be much bigger fish for the region to fry than overcrowding in DSTT1.
Actually, ST3 was predicated on less than 9% Federal funding in total even back in 2016 at the end of Obama. And that’s for everything including Link, Stride and Sounder.
The funds also would be received many years down the line. There are several administrations to come that could change policy or future transportation bills in Congress that could simply dole out the money. .
Federal funds are at this point almost a side issue. The cost overruns are by far more significant.
I said “the financial short-fall will get even worse.” I did not blame it for the short-fall.
And to Mike’s admonition, the budgetary problems are the only thing that will get ST to consider automation.. I funds were plentiful, there would be no chance that they would revisit the favored design.
That’s why we’re talking about it now. The board has started to be open to some things it wasn’t before (no DSTT2) due to the difficult cost of the preferrered alignments. We want to support those and nudge it a bit further (automated Ballard). The time to do it is when the board is debating alternatives and making a decision, which is now.
Let’s keep this article’s comments on the the issue of locally choosing to switch to automated, and the technical/logistical issues that would arise if so. We don’t know whether the feds will cancel Washington state’s transit grants or not, or when, or by how much or which parts. So let’s wait and see. When you think you know about the future, you often think wrong, because many different factors are going into it and influencing each other.
It’s pretty funny, though, that the U.S. Department of Transportation’s “Build America” webpage uses a photo of people working on Line 2 across Lake Washington: https://www.transportation.gov/buildamerica/
.
Now, back to your regularly scheduled programming.
Confused about the shorter station thing, I’ve seen people complain the trains aren’t long enough to handle capacity and yet also saying they need to be shorter when Skytrain runs similar length trains to Link on Expo and Millennium lines and manages capacity well for a light metro.
Canada Line being the outlier and on some level was woefully underbuilt for the ridership it gets, especially during peak where crush loading is not uncommon.
The platforms on the Expo Line — the original system — are 260 meters long, and fit five-car “Mark II”-length trains. Right now, when not every train is a five-car Mark II, capacity is about 18,000 passengers per hour in the peak direction. There are estimates that it can reach 23,000 pphpd with 75 second headways. I guess that would require a new communications system or they would already have done it.
Does 20,000 pphpd not sound like enough capacity for the Denny Way to Westlake portion of BLE? I expect that will be the most crowded segment, though the difference between peak and reverse-peak ridership will be less than farther north.
260 feet, not meters.
Yes, thank you Nathan.
I think the call for shorter stations is short-sighted. It would leave a very busy section of Link with permanently shorter stations than the rest of the system. And by permanent I mean literally permanent.
Don’t consider it the “rest of the system.” T-Link has shorter stations than C-Link and nobody cares. Call this B-Link: it will operate independently of C-Link forever, even if it shares rolling stock with C-Link for some procurement and O&M savings.
What matters is VPH, not TPH. With single-line operations, automation, and different station designs (e.g. Spanish solution island platforms, platform edge doors, etc.), it’s plausible to run 2 car trains at 90 second headways, creating *more* passenger capacity than the DSTT with 4 car trains at 4 minute headways with lower capital investment upfront by shrinking the station footprints
As Troy Serad has painstakingly documented, T-Link was supposed to become the tail of the 1 Line but at some point that vision was abandoned. Probably because its operations are too slow and its platforms are too short to accommodate the 400-foot “light metro” trains Sound Transit ended up building.
See Brent’s comment, below, for more points to consider.
I think there’s general consensus among pragmatic transit advocates (including, apparently, the Seattle contingent of Sound Transit’s Board) that the Ballard Stub is the way to go. What I think is unclear is whether or not this means the Ballard Line will never become the tail of the 1 Line.
I think a permanent Ballard Stub Line is a bad place to end up, but what seems to be missing is consensus on where the line should go instead.
Nathan Dickey,
Choice #1. A smaller, stand alone, above ground train line to Ballard.
Choice #2. No train to Ballard in your lifetime.
There’s no third choice because the money just isn’t there. However bad the current ST budget projections are, I think anybody with any common sense knows they’ll be worse, maybe much worse, by the time ST gets around to the Ballard project.
And Sound Transit is just the tip of the iceberg for Seattle Progressives. There’s been so much talk, so much hype, so much planning for progressive ideas in Seattle in the last 25 years, yet there’s zero to show for much of it.
If Katie Wilson wants to be remembered as one of Seattle’s greatest mayors, she needs to get the light rail lines to West Seattle and Ballard sorted out and under construction on the budget Sound Transit has to work with. Compromise is key word here.
If you’re a Lefty in Seattle, it’s time to “move it or lose it!” Politically, the planets have aligned for progressives to get something (anything?) done. Light rail is a heck of a great place to start.
No more on how bad the feds might be in this article, or it’ll be moderated. There are open threads to discuss the likelihoods of what the feds might do in the future and how the region could/should respond to such contingencies.
Yes, T-Link was going to be the same system (Troy does great work), but it evolved such that now it is both technically incompatible and no longer intended to integrate (ST3 assumes C-Link heads west to Tacoma Mall rather than turns north in Tacoma CBD). I’m suggesting the same think can happen with Ballard Link – it was proposed to be within the C-Link system and now may evolve into a distinct line, and that’s OK because it can have a different operating framework, similar to how T-Link operates very differently than C-Link. C-Link is metro rail, which is why the trains are long, frequency is good but not great, and it has to compromise between stop spacing and end-to-end speed.
B-Link and T-Link are urban lines, which is why frequency is of utmost importance and train/station size should be sacrificed to facilitate frequency. (for example, no one advocates for 2-car T-Link, but adding a few extra streetcars to boost frequency would be great; OTOH with Sounder ST invests in longer trains rather than better frequency to address crowding pre-Covid)
Mike, can we start moderating tacomee’s near-incessant off-topic proclamations that leftists and liberals should move to dead midwestern towns, too?
Mike Orr,
This isn’t really anything about what the Feds may or may not do, or what the economy may or may not do. It’s about Sound Transit delivering a workable light rail system in a reasonable amount of time for budget they have to work with. Right now it looks like interest rates and Federal funding are working against Sound Transit, but that’s fluid as you well know. Things change over time.
The author of this blog post, Martin Pagel, understands this. Looking at what other systems did in times of funding difficulties is the right answer. He’s a smart guy. Sound Transit should listen to him.
One of the worst things a person can be is a “disappointed expert”. It’s easy to sit back as some sort of self appointed arm chair urban planner or transit expert and point out all the flaws in any system. As far as Sound Transit goes, it’s rally time. Look at the fiscal realities and new timeline formulate a new way forward. I congratulate STB for being leaders.
What I often see is progressives acting like a person who believed they would inherit a million dollars when their great aunt dies and ends up depressed and angry about “only” getting $500,000. So Sound Transit can’t afford a second tunnel! And the Feds might pull some funding! And interest rates and inflation are going up! There’s still money coming in and there’s still a transit system to improve.
AJ,
If the T-link had some sort of automated 2 car trains…. and maybe 50-75% more of them, it would be a much better system. Let’s say to systems have the same capacity per hour….. the one with the most frequency is better in real transit world.
Let’s say a system surrenders capacity for more frequency…. that’s a big question! What’s better? And I know there’s a lot of other factors that go into it, but my guess is frequency is the most important factor.
Frequency is key. We currently have 8min headways in Rainier Valley. Even if the light metro trains are half the length (but no driver), if they run every 90 seconds, we would have still have 3 times the capacity. That would provide enough capacity for Ballard for the next century.
Martin Pagel,
I’m not a train expert, but technology usually makes things smaller, (and better) not bigger. Many subways around the world were built before the DOS operating system. I’m not why we’d want to clone them.
“It’s about Sound Transit delivering a workable light rail system in a reasonable amount of time for budget they have to work with”
Switching to automated and ditching DSTT2 is a way to do that. That would lessen the need for federal funding and finding so much local money to fill the gap.
“It’s easy to sit back as some sort of self appointed arm chair urban planner or transit expert and point out all the flaws in any system.”
We’re trying to get ST to listen to professional transit planners and network designers who are aware of international best-practice standards and other ways to deliver more for less. We want ST to hire somebody who can do that. But ST and the politicians have a blind spot about that, so we have to try to fill the gap as amateurs as best we can. Our message is to study these alternatives and give them a fair comparison, not to just blindly do what we amateurs say.
Mike Orr,
A big thanks to STB for being a voice in the wilderness on transit. There’s a huge difference between being an amateur exploring various solutions for transit and the sort of negative nellie who points out why any new solution isn’t going to work. Martin Pagel certainly is a positive vision guy! I mean he’s looked into a West Seattle gondola… I was deeply surprised how much negative comments that idea got. I mean of course a gondola would make transfers to other transit trickier, but I can’t think of a cooler rider experience than riding a gondola!
The thing that STB really is aware of that other pro-transit outfits miss are the subjects of rider experience and financial reality. Even if Sound Transit had the money for that second tunnel, (it does not) would that second tunnel make the overall transit experience better? That’s an honest question STB is wrestling with. Good luck to you all!
“Martin Pagel certainly is a positive vision guy! I mean he’s looked into a West Seattle gondola… I was deeply surprised how much negative comments that idea got. I mean of course a gondola would make transfers to other transit trickier, but I can’t think of a cooler rider experience than riding a gondola.”
I say this cautiously about gondolas: Their limited speeds make them mostly practical only when there are elevation differences and for a distance of under two miles.
Here are some hypotheticals:
1. Stop West Seattle link at Delridge and save $4B. Some of that money could go to a gondola hub above the station not only going to Alaska Junction (G1) but also Admiral (G2) and Morgan Junction or High Point (G3). Even Westwood or Alki may be reachable via gondola.
The same is possible for ending Link at Smith Cove with gondolas to Upper Queen Anne, Fremont and Old Ballard.
Or maybe a gondola to Univerrity Village and Children’s Hospital from U District or Factoria and Bellevue College from South Bellevue.
And don’t forget one connecting Harborview to Link.
Theees no impetus to study any of these things among the general public unfortunately. If some developer billionaire wanted it considered, elected officials would however rush to support it.
I think the call for shorter stations is short-sighted.
OK, so you are basically calling for more capacity than the existing stations. I’m not sure why, but OK. What then? Make the stations in Ballard bigger than the ones in the rest of they system so they can handle five car trains? How is that supposed to work if the trains are interlined?
The only way to realistically get more capacity is to run more often. But just assume that we do that with the current setup. The train is running from Tacoma to Ballard every 8 minutes. So now we run an additional train from SoDo to Ballard every 8 minutes as well. That means a train every four minutes. That is a lot of capacity. That is as much capacity as we expect from the UW to downtown — by far the most popular and busy section of our system. Yes, theoretically you could run the light rail trains that often from Ballard to SoDo but you run into problems because the other train (coming from Tacoma) is on the same line. So just consider light rail trains running every four minutes as the baseline.
But now assume the line ends at Ballard. Instead of running every four minutes we run every two minutes. There is no interlining (and it is all automated) so this is easy. Instead of light rail trains (that are by their very nature not very capacious) we run light metro trains. Instead of trains with cabins we run automated trains. If the trains were half as big but running twice as often they would still have considerably more capacity.
Except they could also be, say, 60% of the size. They could run every 90 seconds. At some point you have to ask yourself why this line has so much extra capacity while the rest of the system doesn’t.
So again, smaller trains but more capacity. It is like sports cars. When I was growing up, every muscle car had a V-8. A 1972 Camaro V-8 (performance edition) had a whopping 255 horsepower. It was the last of the big block engines. But now a Ford GT V-6 gets over 600 horse power! You can get a Mercedes V-4 that over 400 horsepower. Then there are the hybrids of course which can get over 800 horsepower out of a V-6. The point being you can still get a V-8 (like I remember from my youth) but there really is no point. It is not about the size of the engine anymore. Smaller can be faster when it comes to cars. Likewise, smaller can carry more people when it comes to metros.
Many automated metros come with platform screen doors. It’s like the ones at SeaTac. Those shuttle trains have run at high frequencies for decades and move lots of people. The platform screen doors help lots.
I don’t get how this would be much different than what’s been operating at SeaTac for decades. We’ve had the local example running for decades.
As far as capacity issues go, I do not see West Seattle nor Ballard lines having the crowds that are anywhere close to the crowds on the 1 Line today. The 1 Line serves much more population and much bigger sports venues.
There is this odd myth in the air that the current 1 Line is somehow carrying less riders than these projects will. That’s just not true! These segments will carry less!
We’ve already built rail transit in our most productive corridor. It’s why the 1 Line was built first.
And automated metros around the world carry many more riders with these shorter stations.
As Ross demonstrated, this is a non-issue. Sure it may get a temporary surge after a game ends but that will quickly dissipate.
>OK, so you are basically calling for more capacity than the existing stations. I’m not sure why, but OK. What then?
To be clear, I didn’t explicitly called for that at all in my comment, so claiming such is some level of fallacy. That’s not to say I disagree – who wouldn’t want more capacity?
I strongly support an automated Ballard Link but, again, I don’t think it’s wise to stub the line and its platforms. A stub line can be extended and connected for improved network effects. Stubbed stations (especially grade-separated ones) cannot be extended without extreme cost.
Maybe you’re right that there will never be a day that a Ballard Link station will ever need to move more than 18,000 people in an hour. Maybe you’re right that it will never be the tail of a line with 400-foot platforms. I prefer to be more optimistic.
Nathan Dickey,
The problem here I see is asking today’s tax payers to pay for their grandkids transit system. Maybe their great grand kids transit even. I honestly don’t think that “ask” is even possible with the current finances. It’s don’t really believe there’s much political will to raise taxes for Sound Transit either.
My guess is the ST board will make zero changes and just let these projects get pushed back… for decades?
Tacomee, you really need to read ST’s annual financial plan updates if you’re going to continue bloviating about long-term finances.
Nathan Dickey,
I’m not reading Sound Transit budget projections because I don’t have the background to tell how bad things really are. Are you a civil engineer? Large construction project manager? At least an accountant with a background in…. in don’t know, building large road projects? If you have some sort of magic insight to how to corral the West Seattle costs, now at 8 billion and climbing, please feel free to share it with the rest of the class.
I do understand there’s 4 moving parts to Sound Transit. There’s the tax revenue coming in, there’s a debt ceiling ST has to stay under, there’s the construction costs of building projects, there’s the cost of running and maintaining finished transit projects.
So when the price of the West Seattle project went from 4 billion to 8 billion (and climbing) what’s your solution? The incoming Sound Transit taxes didn’t double, did they?
And let’s get one thing completely clear here. I’m far from some GOP hack. I care deeply about these issues. I care deeply enough to call out the bogus math that isn’t going to work. Sound Transit needs to build a system it can afford that actually works.
>> so you are basically calling for more capacity than the existing stations.
To be clear, I didn’t explicitly called for that at all in my comment
You implied it. If the stations are the same size as the existing stations they would have more capacity. Quite a bit more capacity depending on the headways of the trains.
To put it another way, you can have the same capacity as the existing trains with much smaller stations. That is the point.
Ross, you’re implying more capacity is somehow bad or unnecessary. You’re not addressing my actual argument, which is that designing a new line with shorter platforms than the rest of the system seems problematic from a planning perspective.
Tacomee, you constantly question ST’s finances, yet you refuse to even attempt to read (much less understand) the very legible reports.
> I care deeply about these issues.
Sure, Jan.
tacommee, you simply cannot have an automated “at-grade” line, just as you cannot have third-rail power distribution at-grade. At the very least, not without rigorous fencing and NO at-grade public crossings.
It’s just too dangerous, and no government would embark on an automated railroad without strict grade separation and exclusionary fencing.
Since such a standard is completely at odds with an urban cityscape, it is off the table for Ballard Link.
It’s possible to have a stretch of such at-grade exclusionary right-of-way alongside the BNSF tracks through Interbay on an otherwise elevated and tunneled BLE, but a surface LRT line would require human operators and certainly have to bypass South Lake Union and Lower Queen Anne, which are the parts of the line which have the greatest ridership potential.
Without service in South Lake Union and Uptown, the likely ridership is inadequate for any rail line.
Overall, I like the stub plan with a well-thought-out allowance for a dedicated base if necessary.
That said, shortening the stations seems like a major cheap-out we will live to regret, since we’ve seen Vancouver’s experience with having to lengthen stations, at great expense, with lengthy station shutdowns. The Ballard line will have more tunnel stations, which will be nearly impossible to lengthen.
Various videos shown here have pointed out the pound foolishness of spending gobs of money to tunnel under a body of water, only to have short stations stunt future capacity.
Automation will indeed deliver much better headway control, with occasional need for humans to override, but that is no substitute for building for future capacity.
Absolutely. Build the stations to Link standard now, at the very least. My preference would have been to build everything like a premetro, but that ship has, I fear, sailed.
Atlanta has built stations for longer trains, and 40 years later the trains are still only about half the platform length.
Two Link car train equivalents operating at 2 minute headways is 12,000 people per hour.
Seattle just doesn’t have the dense development patterns that Vancouver has.
Good on Atlanta! They built stations that will not run out of room this century.
Here, we do things like tear down a 4-story library to build an 8-story library, and build one-story libraries with no TOD above in the walkshed of stations under construction. Those single-story libraries do not generate tax revenue. But I digress.
Also, Ballard is building more TOD than most of the MLK stations. Lower Queen Anne is already full of apartments. South Lake Union is now the northern half of downtown, but with residential towers. The capacity is already needed.
Again, Vancouver lengthened its Expo/Millenium line platforms to 80m, which is still shorter than Link’s 120m platforms. The Expo and Millenium lines also can carry far more (nearly double) people per hour than what the currently-planned Ballard Link could. I agree we should not shrink station sizes too much in an automated BLE (50m Canada line platforms are too short!), but we can still make the stations significantly smaller than the existing Link stations without sacrificing capacity.
Overseas in Taiwan, they’re building new automated lines with 85m platforms on the Taipei MRT Circular line. I’d say the engineers in Taipei’s Dept. of Rapid Transit Systems know what they’re doing as this is their sixth rapid transit line (they are working on several more).
Really though, someone needs to run some numbers to estimate how ridership demand we can expect from a Ballard line over the next few decades. Then we’ll see if we need platforms longer than 80m.
“ Good on Atlanta! They built stations that will not run out of room this century”
MARTA built station to hold 8-car heavy rail. That might be overbuilt to some people. It also has to do with the region has done very little to deserve its heavy rail since its operation.
Meanwhile we are talking about 4-car LRV platform being too big for Seattle? Seattle could put the 8-car heavy rail system to good use if it had MARTA’s money.
> Seattle could put the 8-car heavy rail system to good use if it had MARTA’s money.
Hah, funny you mention that. MARTA was built from the money Seattle refused.
As for whether or not 4-car 120m LRV is too big or not, I think the more important point is that we could move twice as many people as such an LRV using automated trains, 80m platforms, and a lot less money.
@littlefish
As for capacity, the key issue with LRV is its low-floor configuration as the indoor space has to make room for certain equipment. High-floor train with open gangways can expand a lot of capacity without increasing platform length. This can be done with or without having an automated train.
Until recent years, US heavy rail systems have a history of making every train car a control car so they can operate in minimum 2-car unit although operating in such small capacity is missing the point of building a heavy rail. Having control cabin in every car is just not how many other countries build their subway trains. Elsewhere, a typical 6-car unit is consist of 2 control cars, 2 motor cars with pantograph, and 2 trailer car without power and control. This makes it possible to open gangway no matter the train is automated or not.
The benefit of automation is more related to operation rather than capacity.
The stations would be shorter but have the same capacity (or more). There are a couple reasons for this. First is because they would run more often. Second because the station cabs themselves take up room. So again, shorter but more capacity.
Also keep in mind, Martin specifically cited a line in Toronto that was specifically designed to deal with capacity issues (on the old, big subway). It was literally called the “Relief Line” because it was intended to relieve overcrowding. Obviously they didn’t want to shortchange capacity for a line specifically designed to deal with capacity issues. That is why they had smaller trains but more capacity.
Beyond the fact higher frequency + shorter trains > standard link capacity, we also need to look at demographic trends.
Standard models are showing we reach peak-human (the population of the earth starts shrinking) around 2070, and has been being revised downward. The little I’ve dug into this, I suspect it will be closer to 2050-55, or about the time they fix the Westlake escalators.
Ballard’s currently planned station area is so built out with sparkling new townhouses that won’t be ready to be replaced for at least 50 years (long after peak-human), as well as the geographic isolation of Ballard, I don’t foresee Ballard turning into “New San Jose”.
Sure, we might buck global trends, and Ballard might become the massive recipient of a hoard of climate refuges, even as populations are severely declining globally, but I suspect a capacity of 18,000 folks an hour will be vastly more capacity than it will ever, ever need.
Great comment Cam Solomon!
I think the City of Seattle and Washington State put out these population growth projections that are not going to happen. It’s not like Seattle is going to morph into New York City any time soon. I think the State population is a little behind the curve of Oregon…. Washington State will be losing population within 5 years.
How does the City right size transit projects? I don’t know.
Cam, two things:
1. I agree that there won’t capacity problem, even if the line isn’t automated, except during events at Climate Pledge Arena, during which the capacity from ab automated line might be cheaper. In any case, the first consideration should definitely be cost.
2. You keep bringing up how Ballard is filled with townhomes that can’t be replaced, but people pretty thoroughly debunked this in the previous thread. In any case, this line isn’t just about serving Ballard, it also serves SLU, and these are the number two and number one densest locations due for service, respectively. If ST prioritized ridership, cost per rider, farebox recovery, or any other standard performance metric, the BLE would be first in line.
Where are Ballard townhouses? East of 14th?
Go to page 205:
https://www.soundtransit.org/sites/default/files/documents/ballard-link-extension-station-planning-progress-report.pdf
Now go to google maps and go over to 14th and market.
https://maps.app.goo.gl/fZy3MWaXVdCP55Dh6
Now get your walking shoes on and realize that even the few lots where google maps is showing single family homes exist West of 9th, are in the process of being razed for more townhomes and apartments.
What, exactly, was debunked?
Yes, old town is building higher on the few buildable lots left there, but that’s not where the station is.
There is some commercial and retail I suppose they could raise, and there may be some of that.
But the cheap lots, aka the small single family homes there were there 10 years ago are all replaced by townhouses, and some apartments on the arterials.
The issue is the tons of multifamily and businesses between 24th and 15th. That’s the station’s primary purpose. And add the industrial area to the south, because the new “Ballard Industrial Center” was one of the justifications for Ballard Link.
The townhouses east of 14th are just a flaw in Ballard’s zoning. 7-story apartments should be allowed, as befits a regional center and station area. We already knew Ballard’s station location has “half a walkshed” because of this.
You’re suggesting that townhouses are the only thing or predominant thing on both sides of 15th, and that’s patently false. The growth of 7-story buildings and “15-minute city” businesses was already underway when I started working in Ballard in 1999 and lived there in 2003 — a quarter-century ago.
I’m not saying the walkshed is zero. I saying half a walkshed isn’t worth 22 billion.
I feel like Seattle is just so desperate to find any nugget of density to serve, that they have lost all rationality with regard to how much they are willing to spend to serve anything. There is no going “beyond Ballard.” There is little hope of a future Ballard, even if wildly optimistic projections come true and you can ignore the massive zoning the mistakes that have crippled any sort of serious growth that can justify 22 billion. There just isn’t.
Yes, it would be nice to serve uptown and Seattle Center a little better, but this isn’t really the way. And you could serve it really well for a tiny fraction of the money if you didn’t insist on going to Ballard and crossing the Rubicon, er.. ship canal.
We can’t just throw away northwest Seattle as part of the housing crunch solution and non-car solution. We need the Ballard regional center to have good transit access to the rest of the region. That means shrinking the current 30-45 minute overhead to get to regional transfer points (Westlake and U-District) on the current D, 40, and 44. We must shrink it substantially, down to 10-15 minutes or so. We can do it with ST’s preferred Link alignment or some other way, but we must do it.
Cam, the whole “22 billion dollars” isn’t worth it for Ballard is kind of a false dichotomy—of course that’s too much for ST’s tax base, but there are other options than just not serving Ballard, like automation and elevated sections, not to mention that a huge amount of those 22 billion dollars are for a worse-than-useless new tunnel. It’s really deceptive to attack Ballard as a destination for cost reasons while leaving out that the biggest part of the project cost is not in Ballard. If anything is the rubicon here, it’s the new tunnel, not the ship canal or any of the new destinations.
In any case, I think a more reasonable trichotomy would involve evaluating against the projects that are on the table, since at least one of them will have to be modified or delayed and they are unfortunately what voters approved. The Ballard extension is obviously the best project by most standard transit performance metrics. Another great metric for this very expensive mode is “this is faster than an equivalent bus,” and indeed, Ballard is the best, er, only project that meets this standard, due to that pesky ship canal that you’re trying to use to disqualify the project. By your own standard of “possible growth” and existing density, Ballard, LQA, and SLU have much more of all of these things than any of the other destinations in the other projects. If Ballard is some backwater devoid of jobs and transit potential to you, what’s the Tacoma Dome, for example? The moon?? It’s boxed in by freeways, the port, and heavy industrial land use, you know, things that are even worse to have in your promising TOD zone than a few townhouses and more than half of a real 15-minute city.
The Congressional Budget Office release a report last week saying this too:
https://www.cbo.gov/publication/61879
It says that our national growth rate will drop from 0.3% annually to 0.1% annually to 2056 when it will start to shrink.
In the 1990’s, the US added 33 million people. It dropped to 27 million in the 2000 decade. Then 23 million in the 2010’s. And that’s all before our national leadership decided to make immigration and foreign workers so expensive, difficult and even violent.
It’s not discussed much, but there is a linguistic difference between a projection and a forecast. A projection should simply assume that past trends continue (think savings account) and a forecast should look at the gamut of contributing factors (think stock market). To often, the media and citizens and elected officials and even report authors don’t grasp the difference.
“In any case, this line isn’t just about serving Ballard, it also serves SLU, and these are the number two and number one densest locations due for service, respectively.”
It’s unfortunate sometimes that ST calls extensions only by the end station. The major exception was East Link ( or perhaps the original Central Link).
I’ve tried calling it Northwest Link Extension or SLU/ Ballard Link Extension at different times. It doesn’t seem to catch on.
Finally, I will remind folks that ST forecasts routinely estimate Ballard Link Station all by itself —- even in its current location — has more boardings that all the West Seattle Link stations combined. However, we still don’t know what the Smith Cove to Ballard segment will cost. I can’t imagine that it’s as high as $8B like West Seattle is though. The soon-to-be-released EIS should tell us.
> [I’m] saying half a walkshed isn’t worth 22 billion.
> And you could serve it really well for a tiny fraction of the money if you didn’t insist on going to Ballard
Cam, this a gross misunderstanding of the estimated costs to build the planned Ballard Link Extension. The majority of the costs are associated with drilling a tunnel from Elliott Ave to SODO, and building deep stations in Uptown, under Aurora, in SLU, under Westlake, under the King County Civic Campus, and south of CID. We don’t know the full cost of the preferred tunnel to Ballard and underground station at 15th and Market, but it’s nowhere near the majority you make it out to be.
Look, no one is realistically suggesting deferring any part of the Tacoma Dome Link Extension to pay for the Ballard Link Extension. In fact, it’s the only project that’s even remotely close to still being affordable at its full scope under the current financial plan, so it’s probably the only extension that will proceed without significant delays. Perhaps you should celebrate that more than denigrate neighborhoods in Seattle that are denser than the densest parts of Pierce County. Assuming the plan holds, Tacoma will have three versions of rail-based mass transit before Uptown or Ballard have any.
Hah. Okay.
Except I’m not in support of completing TDLE either. Ross has brainwashed me into hating pretty much all the remaining ST3.
I feel like we should spend the money on things that will actually be worth the cost, including potentially the purchase, improvement and (I can dream) electrification of the heavy rail tracks for Amtrak and Sounder.
@Cam Solomon
The “preferred alternative” is tunnel to 15th/Market, page 216, not 14th/Market
I am tentatively in support of this electrified stubline proposal, however.
The only reason I started commenting here at all was to express skepticism of the concerns around capacity.
And you are right, I am too harsh on Ballard. I spent a lot of time there and really liked it in the previous century. It’s not Ballard, it’s me.
“The “preferred alternative” is tunnel to 15th/Market, page 216, not 14th/Market”
There’s still a risk ST might change it to 14th, like it did moving Midtown and CID2 stations. The Ballard Draft EIS isn’t finished yet, and who knows what might be in it.
> There’s still a risk ST might change it to 14th
They might, but the only motivation to do so would if somehow the shift would make or break the affordability of the project. They’re more likely to truncate the line at Smith Cove and wait until they can afford to build the station at 15th than change the preferred alternative to the 14th Ave location. Remember, the preferred alternative is also being advanced to 10% design with preliminary engineering, whereas the other alternatives are parked at the “conceptual” stage.
They’re just as locked into the 15th Avenue station location as they are into the N/S CID station locations, which is why the potential deferral of the tunnel south of Westlake is such a critical opportunity to go back to the drawing board for that segment.
“STB… this is the 7th week in a row you’ve shown Automated Ballard Link in class.”
What are we building again? Oh right a light rail extension. I’ll change my tune the moment North/South CID gets chosen for DSTT2, for now it’s either 5th Diagonal or bust.
North/South CID is already the preferred alignment. The argument is that ST can’t afford any version of a second downtown tunnel without cutting some other major project, and we think it should build the Ballard to Westlake section first.
yes. The preferred alignment for DSTT2 is deeply flawed (e.g., too deep, leading to longer access time; split station, leading to long transfers between the east and south lines; and too costly, leading to the ST3 fiscal crisis. A task of the Enterprise Initiative could be to reorder the ST3 projects. Some could be deleted. Some could be improved.
Nathan Dickey and Jack Whisner,
I think cutting costs by building some sort of cheaper version of projects is possibly on the table, but it would be a hard sell to the board. Smaller automated lines, at grade lines or elevated lines *could* happen, but honestly changes like that aren’t likely.
Changing the order of the projects or canceling any of them? Not politically possible. There’s no way the ST board is going to bow to Seattle. There’s no way to even have a richer, Whiter Ballard jump a poorer, more diverse West Seattle in the light rail race. Dow would let that happen.
Did the Ontario Line go with shorter trains and stations because it delivered a better experience, or because that was the value-engineering forced by an anti-anything-but-cars state government?
Oops. I did not mean this as an embedded comment in a thread.
Definitely not because it
.
That’s because Ontario isn’t a “state” (at least, not yet…)
They went with automated trains because it was a better value. They could have the same capacity with smaller, more frequent trains. Capacity was the main reason they built the line in the first place. So they go that, while also saving money both in construction costs and operations. It also means the trains run more often which is better for riders. So basically cheaper and better. Like a modern smartphone compared to when they first came out.
“Ontario isn’t a “state” (at least, not yet…)”
I can’t resist veering off-topic for a second. If Canada were to join the US, it would make the most sense for the provinces to come in as states. Maybe consolidate the Maritimes because the small New England states would object to them having as many senators as they do. But the Orange One is saying Canada would be the “51st state”, implying the whole country in one. Of course, I don’t think he’s thought about how large or populous that would be, or how large and powerful the “state government” would be.
Or about the French language issue.
Well, we aren’t building anything currently, given the project is still in the planning stage.
““STB… this is the 7th week in a row you’ve shown Automated Ballard Link in class.””
This article is because we thought the articles since December 15th didn’t sufficiently lay out the case for automated Ballard. The “Build the Best Parts First” article advocates for the Stub-End alternative but mentions automation only briefly in one sentence.
This is THE issue we’re trying to convince the ST Board of, since it’s already considering Stub-End but hasn’t acknowledged automation as a potential alternative or commissioned a study for it. That’s what we’re trying to get ST to do. This is our attempt: an article that makes the case and lists the reasons as best we can, that can be a reference when people bring up the issue with boardmembers, politicians, or neighbors.
So I don’t think there will be another Ballard automation article unless we’ve missed something, six or twelve months have passed, or if the board does something interesting when it debates the Enterprise Initiative further over the next few months.
I call for a line from IDC (King Street) to Ballard rather than the automated redundant stub line, it’s Sounder North, except truncated at Ballard and not going up to Everett (due to low ridership). The line would run on two to three passenger cars, and have two tunnel stations at Symphony and Pioneer Square (running in the existing BNSF tunnel), and at-grade stations at Belltown (Broad Street), Smith Cove, Interbay, Magnolia, and of course it’s terminus Ballard. Station locations and layouts TBD, and we’ll see how long it takes for ST and BNSF to get into an agreement. I know it runs on BNSF-owned tracks, but I would like to see cooperation between BNSF and ST. This would probably not save millions, but billions, and has a similar path to Ballard Link and already provides service to a second tunnel. It could be an extension of the existing south line (which is gonna get capacity fixes, though increased frequencies and shorter trains would provide a better rider experience).
What would be “redundant” about the Ballard stub line? Aren’t you calling for a redundant tunnel within downtown? (for which BNSF will charge an arm and a leg)
It’s not really redundant as there’s not many competitive routes. Also I bet BNSF will charge less than 22 billion.
BNSF would never (and I mean never) allow Sound Transit to run any kind of metro-level service through their tunnel. It’s too critical of a connection, and there are no remaining redundancies within the freight rail network. Unless Sound Transit is willing to provide a roughly equivalent parallel line (which doesn’t exist), this proposal would involve buying BNSF out of all their Western Canada operations, and likely all of their Seattle area operations as well. And Stevens Pass.
I think this is arguably in the public interest, but only in the context of nationalizing the rail network in general. ST cannot bear that cost, which would be well north of $10B for trackage (with net profits of ~$6B a year, even giving the WA/BC trackage a yearly valuation of $500M is conservative, and 20 years of profits is a low ball price anyways), plus billions on the infill stations in the downtown tunnel.
A more interesting downtown tunnel to consider a ST takeover for is the SR99 one, but I imagine infill stations there would be ruinously expensive too.
“Unless Sound Transit is willing to provide a roughly equivalent parallel line (which doesn’t exist)…”
I have pondered whether it would make sense to move the BNSF tracks into a long, bored tunnel under Seattle. There would be no need for station vaults (the aspect that makes Link expensive to build), and the Great Northern Tunnel may or may not be salvageable. The thing is though that the tunnel is quite old and will someday need upgrading anyway. I’m not sure how messy and costly it would be to breach the tunnel walls or ceilings for stations.
The Great Northern Tunnel will never be voluntarily reconstructed. Like many pieces of legacy infrastructure, it will only be replaced if/when it’s destroyed when The Big One hits.
We need the northern track to continue carrying freight, both for normal goods distribution, the family-wage blue-collar local jobs it provides, and to keep freight on rail rather than switching to more energy-intensive, space-consuming trucks. The track can’t support passenger trains every ten minutes or every hour without cutting significantly into freight capacity. Additional tracks can’t be added because of the fixed-width tunnel downtown and the narrow hillside in north Seattle. A Ballard station on the track would be off the western edge of the village, the same problem we’re trying to address with a 15th/14th station but worse. A Belltown station would be a steep hill below Belltown.
One of the proposals for an HSR line was a deep tunnel under Beacon and Capitol Hill to circumvent downtown. With unlimited funding we could build two tunnels: an alternative for freight and one for HSR and regional passenger (Sounder) rail. No tunnel station, just a station by the Link OMF. Then we could repurpose the existing BNSF tunnel for an urban rail line. But this would take decades.
It is extremely expensive to run Sounder trains that only run a handful of times each day. Running every six minutes all day long (like a normal metro) would be way more expensive. Even running as frequently as Link (every ten minutes midday) would be extremely expensive.
From a ridership standpoint you end up skipping Uptown and South Lake Union. At best, Belltown makes up for losing one of those, but not both. But the Belltown stop would be by the water, which is not good at all (fish don’t take transit). It means you are skipping the middle of downtown. Assume you want to take a trip from Ballard to Westlake Center. What do you do? Take the train to CID and then a bus back? Get off at the Belltown stop and walk fifteen minutes? The existing, slow bus sounds better. From a network standpoint it is poor. The only transfer point is CID. Someone trying to from Capitol Hill to Ballard would be out of luck. They would just ignore the train and take whatever bus options are available. Same goes for First Hill.
Then there is the cost. You still need to cross the ship canal. The existing BNSF bridge is fairly low. Unlike the planned movable bridge (that voters approved) it would have to open and close a lot. This would add to the cost and could lead to some delays. Then you have to serve Ballard itself. The line would have to branch from the BNSF line and then what? Maybe it can curve onto Market but then you going elevated (or underground) just to serve the main part of Ballard. Obviously this would save quite a bit of money but it would still cost a fair amount.
The main savings — if there are any — come from skipping South Lake Union, Uptown and the middle of downtown (Westlake). Hard to see how it would be worth it.
Then run it when crowding is at it’s peak, at least every 15 or 20 minutes, though you would also upgrade the C and D Line in return of no West Seattle or Ballard light rail expansions.
This is certainly more expensive than a greenfield line, given the need to buy out BNSF. Even if BNSF gave the tunnel up for free, I’m not sure this would be cheaper than a new line, unless you are proposing zero underground stations? The Great Northern Tunnel opened in 1905 – adding underground stations would involve shutting down the tunnel for years and building entirely new station boxes. Saving the cost of tunneling between the station boxes is probably negligible, given the offsetting cost of retrofitting a century old asset.
No, No and No. The BNSF tunnel is a diesel-soaked nightmare of a place for humans not in a locomotive cab.
The tunnel is BNSF’s premier line to Chicago from the Ports of Seattle and Tacoma. Running passenger trains at a trunk-line frequency with all those stops would leave NO slots for those twenty-four trains per day which pay the bills for the Northern Transcon.
And finally, Shilshole Marina is a terrible place for a bus intercept.
No. No. And No.
The shorter station concept is concerning. I’m always thinking about event crowds. If the stub line were built (which I fully support), I dread the image of hundreds of people exiting the 1-Line at Westlake and trying to cram onto a shorter train and smaller platform. Additionally, in 2075, the public will have the same gripes about lack of capacity due to overcrowding.
Keep in mind the current plant is Ballard will have the same capacity as the Rainier Valley. It’s very likely an automated stub will have higher peak capacity than the current proposal.
There may be queuing during major events, but that’s true for any rail network, and higher frequencies is often a better solution for irregular peak crowding than longer trains.
The stations would be shorter but the trains would have more capacity. That is because the headways would be lower and the trains carry more people per linear foot.
Thanks for a succinct article on the benefits of an automated train line, Martin.
I would caution that the Ontario Line cost estimate has ballooned from $11B CDN to $28B CDN. That’s a costing mistake as wide as the West Seattle and Ballard Link extensions’ costing mistake.
Automation is inevitable in my opinion. It has so many practical advantages beyond this. These include:
1. Automated reversal of trains at ends of lines.
2. Ability to run longer lines — like from Tacoma to Everett or Redmond to Everett.
3. Lower cost overhead to run trains more hours or more frequently.
With this in mind, I think ST should quit planning on building any new track and spend a year to recast all extension planning including OMFs to switch to automated trains. Once the specs are developed, all of the extensions can be reassessed and planned as automated trains.
10 or 20 years ago, automated trains were a bit novel. But now across the world they are the de facto standard. Going forward, everything from vehicle specs to station specs to power specs to signal specs will default to being automated. And that ultimately affects what future vendor prices will be.
While a Ballard stub is important to automate, the topic is much more systemic. ST seems to have summarily ignored this basic technological reality.
I would propose a slightly different lesson to learn from Toronto, amid lessons *not* to learn from the Ontario line and the newer generation of light-metro infrastructure being built in the GTA region now. What makes Toronto’s existing rapid transit system unbelievably successful and productive?
1. No branching lines: every line runs end to end with no branches to achieve maximum frequencies. No branching means no delays or difficulties managing mixed schedules.
2. Fast transfers: the busiest transfer stations in Toronto have a short, one-story set of stairs between connecting lines, such that a transfer between platforms takes 30-60 seconds by foot at most. This means that riders don’t see transfers as inconvenient.
3. Maximum frequency all day every day: why are transfers convenient? Because every line runs every 5 minutes or less all day every day, so the maximum transfer time is about 5-6 minutes. You can make two transfers and still wait less than ten minutes on average. Before the pandemic, line 1 ran trains every 3 minutes or less until 10pm to accommodate events and sports games downtown. New investment improved the min headway on Line 1 to 100-110 seconds, and the Ontario line can support 90 second headways.
4. Short dwell times: Toronto does not have express trains, even though the subway lines themselves are getting quite long. Trains only dwell a few seconds at each stop, in part due to high frequencies and no branches. If you miss your train while entering the station, the next one will arrive in 2-3 minutes most of the time. Imagine cutting 20 seconds of dwell time across 20 stations on a line– that’s over 5 minutes of travel time savings. It adds up.
And here’s the thing- all of this can be achieved without full automation. It’s up to you. Legacy subways in Toronto can run every 2m20seconds with manned operation. ATC decreases headways from 150 seconds to 110 seconds. Seattle is nowhere close to these headways, and isn’t even building infrastructure that can accommodate them. So I don’t know if automation is the biggest lesson here.
And the Ontario line is getting things wrong in important ways, similar to Link, mainly in its cavernous, overbuilt transfer stations that make transfer times longer: look at the drawings for Ontario line’s downtown transfer stations. To reduce complexity, the tunnel is bored deep under Queen street, necessitating a multiple story descent for passengers wanting to transfer, whereas legacy lines were built cut-and-cover under the surface street. Toronto’s legacy subway can be used for short trips because it takes just a couple minutes to get to the platform. The Ontario line will not be practical for short trips, meaning that the surface streetcar will still be required in parallel.
I don’t really see no branches as much of a positive. If there’s distributed demand, then a branch is a pretty logical choice. Tons of systems world wide use branching, and it works fine. I’d say the reason Toronto hasn’t is more about a history of lower investment levels than good planning. I’d consider lack of branches to be essentially neutral. It’s possible to run good service with branching lines, and many places do. And high frequencies is probably because there’s high demand. Not saying we don’t have that, or that demand and frequency aren’t related, but Canada in general is more transit oriented (and dependent) than the US.
In terms of the depth of stations and poor system integration for transfers, I’d agree. But the decision to forgo convenient transfers to minimize construction impacts is an issue basically everywhere transit lines are being built.
1. As blumdrew mentioned, branching has its place. It really depends on the geography of the city and how much coverage you want.
2. I agree, transfers are very important.
3. Same thing with frequency. Part of the reason Toronto can afford to run their trains so often is that the system is very efficient. It has very high ridership per mile (higher than Mexico City). Seattle is the opposite. We focuses more on quantity (lots of track) and not on quality.
4. Link has always had poor dwell times. I’m not sure why. It may be due to the design of the trains. Light rail with two levels makes it tougher to move about the train (and thus tougher to get on and off).
It is important to learn from places like Toronto but we shouldn’t try and mimic them. Toronto is a very different city. It is just a lot bigger. They have arguably the only sensible streetcars in North America (although they should be faster). Everyone else could just replace each streetcar with a bus and handle the loads just fine. Toronto would need a lot more buses.
The only reason Toronto is relevant is that they changed their plans and are building a line that will use automated trains. They had the same issues as us. The change was quite late in the process. It means they have to buy different trains. But it is clear that it was worth it just like it would be worth it in our system.
Our model should really be Vancouver. To begin with, the cities are similar. Both are fairly young, West Coast cities. In terms of mountains and sea they are about as similar as two separate cities could be. They both have dense areas but they were built after the automobile so they sprawl. This makes building a good transit system challenging.
Vancouver has risen to the challenge. The city does extremely well when it comes to transit. It is one of the best per capita in North America (if not the best). One of the keys to this success is very good integration between the buses and trains. SkyTrain carries a lot of people. The buses carry more. One of the keys to their success is very frequent service on the train. Yes, that could be achieved without automation but I doubt it would be. The O-Train is way less frequent despite having good ridership per mile (and being relatively short). If SkyTrain had drivers it is quite possible they never would have created the virtuous cycle like they have. The trains would be infrequent so not that many people would take the bus to them. The buses wouldn’t have that much ridership either so they wouldn’t run them that often. People just drive. Instead they’ve built a system that is arguably the best on the West Coast and the best for a city its size in North America.
It’s obvious to me that light rail is absolutely the wrong technology for Seattle (and I used to be a big cheerleader). Having now experienced automated lines like Vancouver Skytrain and Copenhagen metro, I think that smaller, more frequent, fully automated solutions like this are the way to go. It should more than offset any savings that we might realize from the very little surface/at-grade running that we might consider in the future.
The costs for the Ballard to West Seattle line are so astronomical that we need to look at new options. A second tunnel through downtown is going to kill the entire project. Ballard is not dense enough (nor will it ever be) that it requires 4 car light rail trains. I strongly support an independent automated line from Westlake to Ballard. The smaller stations and automation aspect are critical to freeing up the necessary budget.
The governments’/agencies’ vision is shortsighted. They should start with the ultimate goal: getting car mode share down to 50%. What kind of transit do you need for that? Far more than the current Link plans. 2-6 minute frequency full time throughout the inner core of the network (Northgate, Rainier Beach, maybe Bellevue and Redmond and SeaTac). More Seattle lines, not just to Ballard and West Seattle. West Seattle multi-line BRT instead of Link for the reasons Ross has outlined. Regional-type rail or frequent express buses in enhanced right of way to Everett and Tacoma, faster and separate from Link and not limiting how many Seattle stops Link can have. Comprehensive frequent local buses everywhere, every 10 minutes full-time on all of Metro’s core routes. That’s what cities that are serious about transit have.
But our politicians are nowhere near ready to consider this, so we have to make incremental steps that at least gets us towards it. Frequency is king, and good train-to-train transfers. And maybe don’t extend Link past Federal Way and Lynnwood or in West Seattle, both for cost effectiveness and to serve passengers better throughout those areas (most don’t live near a proposed Link station, and with the current and proposed bus networks, would have trouble getting to it).
Yeah, I agree Ryan. Light rail has its place, although it is kind of a niche mode. It makes sense if you already have a comprehensive Metro (like Paris). It also makes sense if you already have the tracks and the demand to match. But otherwise it is a compromise between cost and quality. A metro is just a lot better when it comes to moving a huge number of riders from place to place. A bus is better (and cheaper) if you don’t need the capacity. So it becomes a niche mode for when you don’t want to spend the money on a metro but you still want that capacity — it rarely makes sense.
Seattle kind of stumbled into it. Part of the reason we embraced light rail is because so many other U. S. cities did. We saw that as the model (not a traditional metro). This was a mistake. It was also political. Traditional heavy rail was seen as too big and dirty. But light rail was seen as new and clean. In this way it appealed to smaller cities. Heavy rail was seen as overkill while light rail was seen as much cheaper, easier and appropriate. But it really wasn’t. Very few cities in North America make sense for light rail. Either build a real metro or run buses.
But the other big reason we had light rail was because of the bus tunnel. We had one foot in both worlds for a while. In this case there really was a very good reason to have low-floor light rail cars. Otherwise bus riders would have to step up to the platform to get off the bus. We also started with a line to the south. Rainier Valley was controversial from the start. It made sense to run elevated or underground (and on Rainier, not MLK). But given the way the city (and region) treated the area, the neighborhood objected. They didn’t want to go through the disruption caused by digging up the street. They didn’t want an elevated line. So it went on the surface and we’ve been stuck with an inappropriate mode ever since.
This goes back to the root of the problem and the politics surrounding it. You want to build the most important transit piece first. The part that will get you the most riders per mile. The part that will replace the most buses while saving those riders a lot of time. In the case of Seattle that was downtown to the U-District. At that point you can kick out all the buses and still come out ahead. You probably want to take it a step further and go to Northgate (which would have allowed us to kick out even more buses). With the buses gone you can be high floor. Since all of that is grade separated it would be automated. Once you have that line (from Northgate to downtown) you think about going south and east.
But ST was designed as a regional rail system. They wanted to go as far as Everett and Tacoma (which is really crazy for any kind of metro, let alone light rail). So they ended up going south as far as possible even though it meant picking up fewer riders. This was a signal to everyone — the train can take you to places far away. Places that you always hate driving to (like the airport). It was much cheaper to run on the surface in SoDo and Rainier Valley. At the same time they didn’t kick out the buses. They ran through the tunnel too.
It really goes down to quality versus quantity. We have (or at least had) this idea that if we make a few compromises (like using light rail or running close to the freeway) it can be built fairly cheaply but be just about as good as a regular metro. Then we can use those savings to go far and wide. But it really doesn’t work that way. We are spending a fortune on this thing (more than just about anyone in the world) and it really isn’t that comprehensive. Major urban areas (Belltown, First Hill) won’t have coverage. Nor will secondary areas (like Fremont and the Central Area). Nor will we address the biggest network weaknesses (lack of fast east-west transit inside the city north of the ship canal). But if things go as planned we will end up serving places that don’t benefit very much from light rail (like Ash Way or Fife).
The inappropriate choice of light rail should be not be surprising given the inappropriate nature of the main line itself. Tacoma to Everett is a ridiculous idea for a metro. In comparison the choice of light rail is a minor issue.
Sorry for the rant and thread drift but I agree with you. We should use every opportunity to build what makes sense to build in the city. In the case of Ballard Link that is an automated light metro to Westlake.
The truly nutty cost problem with all parts of what was originally called WSBLE is the deep stations created by using tunnel boring machines. That’s because each station vault is as big as a football field that must also be very deep. Does anyone remember how massive the Capitol Hill station hole was and how long it was fenced off? This will be repeated for each and every deep station constructed!
And deep stations are really awful for riders to use when making short trips. It’s great for long spacing — but that’s not what this corridor needs. The mindset is a vertical version of expecting everyone to use a freeway a mile away that will require going through two huge, closely spaced interchanges when the travel that people want is at street level. There’s a reason why the SF Central Subway is not getting the riders they expected — and it’s driven by the extreme station depths.
The biased truth is that too many people with power here have bought into the aversion to aerial tracks and stations. The horrific costs are a result of this aversion. As a rider, aerial stations would be closer to the sidewalk than deep stations are — and would be half the cost.
Even cut and cover would be better. Sure we would have to close street blocks sequentially — but all those trucks hauling all that soil is going to de facto have the effect of a street closure anyway.
Consider the budgeting mistakes from ST3 on a project basis. While every project has become more expensive, the cost increases of every other one is less than double — usually less than 30 percent in fact. The bored tunnel segment costs are by contrast triple what was estimated!
So let’s talk about the psychology of this choice. Are aerial tracks ugly? Isn’t there an element there that looks at tracks as a visual or noise nuisance? And isn’t this somewhat based in an aversion to visual change and perhaps wanting to hide transit and its riders in the urban fabric? So riders get to look at subway walls so drivers can keep their surface views?
Who advocates for closing the monorail because of it having overhead tracks? No one. Did aerial tracks create blight in Chicago, or did it enable a super-tall building market that otherwise would have not evolved?
Overhead tracks and stations do not have to look brutalist. Riders shouldn’t have to be relegated to becoming moles. The mere fact that there aren’t at least partial if not full alternatives above ground for West Seattle (the 2026 adopted plan), DSTT2 and SLU demonstrates how everyone conveniently ignores internal biases about aerial trains.
I may have been a little over-dramatic with this comment, but sometimes I think we need to take a hard look as to why we have assumed underground trains as a given when we have this truly massive shortfall.
I agree – elevated tracks really should be back on the table if the Ballard Line doesn’t need to tie into the main line. It will certainly be billions cheaper than tunneling and likely will result in better passenger access than the proposed deep stations.
Link likely cannot make the SLU turn, which then suggests we should pivot to another mode or simply shorter trains, allowing for the ST3 alignment to be met with an elevated line. For example, a rubber-tyred metro or guided bus could make the incline up from Smith Cove to LQA easily (as proven by the buses that already run that route).
Could even switch to a Belltown alignment and straight-up replace the Monorail between the Seattle Center and Westlake (which is perhaps at the end of its useful life anyways). Closing the monorail for several years during construction would be a bummer, but much much better than closing the DSTT for a year(s) or the need to close various avenues for extended periods of time during excavation.
Also, I’ll note that with rubber tires and with concrete supports, the train (or bus) is much quieter than the legacy elevated lines in Chicago, NYC, Philly, etc.
TSB guideways allow for tight turns, steep inclines and quiet operation. Such features allow construction in urban spaces. There is no additional sound proofing required as maglev trains don’t touch the track. Prefabrication reduces the disruption of an elevated line to a minimum.
https://seattletransitblog.com/2023/12/03/evolution-of-urban-guideways/
Great post Al S.
I’ve worked in construction all of my life and one thing I know is… a house is only “new” for about 1 hour. After the owners are given the keys, the movers and kids will undoubtedly start scratching up the paint and staining the carpet. A house is place to live and it just gets dinged up over time. Sometimes repairs are needed and a home never stays the same. Things get funky. Floors get a slant or run to them. Nothing is truly square anymore. But we don’t tear down 100 year old houses because they’re out of square and have “funky” out of style remodeled kitchens. We just live with these imperfections for the most part.
Cities are just like houses. Seattle has a lot of funky details added over the years. We just live with these because there’s no changing them. We can’t afford to even if we wanted to. Let’s think of light rail as a remodel…. we need to somehow shoehorn it in with what’s already there AND we have a tight budget. All of the pie-in-the-sky plans Sound Transit had with tunnels from Hell to breakfast need to go into the recycling bin. Build something we can actually afford. Sure it’s going to be a little funky! That’s just the economic and political reality of the project.
Over the next 4 years, will “the prefect” kill off the “the good” in a progressive Seattle? Kudos to STB for exploring realistic and workable solutions.
One of the objections to aerial trains is aesthetic. But there are practical limitations. In a dense urban environment (with big buildings) you have to follow the street grid. The same thing is true with cut and cover. In contrast, deep bore tunnels can avoid two-dimensional issues by taking advantage of the third. If you go deep enough you avoid the whole mess and can curve any way you want. But of course that comes with a cost (both literally and figuratively). Deep bore tunneling has become less expensive but the deeper the stations the more expensive they are. As you mentioned, it also takes a long time to get to the platform. You can mitigate the latter problem by getting a bunch of high-speed elevators. This might be cheaper but most likely it won’t be (unless, like Beacon Hill, you don’t expect that many riders).
But smaller stations served by more frequent trains can help in various ways. The smaller the platform the cost per station is bound to be cheaper just because it is shorter. It doesn’t matter if it is elevated, underground or even on the surface. But there are other savings. By running more frequently you have smaller surges. If you do decide to just go with elevators (and they are fast enough) you need fewer of them. The same is true of escalators. Instead of a huge group of people trying to crowd onto an escalator every five minutes you have a smaller group getting on there every three. But it allows for more flexibility with the stations. A station has to be flat and straight even though the line itself might be curving. The shorter the station the more options you have. Sometimes it might enable a station closer to the surface.
Of course one issue in Seattle (and San Fransisco) is the hills. The reason the station is so deep at Capitol Hill is because it is so much higher than downtown. The UW Station is extremely deep because we went under the ship canal. Folks seem to prefer that for Ballard which will inevitably push up the cost. Elevated through the Seattle Center is challenging. Maybe you could weave your way around using Harrison in South Lake Union and then Republican the rest of the way but it wouldn’t be easy. I agree, I think it could be done and would probably save a huge amount of money but I doubt they would consider it now.
The combination is not out of the question either. From Elliott it could go into a tunnel to serve the Seattle Center (as planned). But the tunnel could get close to the surface as it leaves the Seattle Center. It would pop out of the ground on Harrison between 5th and 6th. It would then run elevated past to Westlake.
Automated light metro trains can also go up and down more steeply that current Link trains can. They don’t need line of sight at the station approaches either. Rubber based trains or maglev trains can do that even steeper. That allows for shallower stations.
I honestly don’t get the aesthetic objection. Sure the Alaska viaduct was ugly, but the monorail is a tourist destination, so the specifics matter.
Also, any objection of views from the ground is counterbalanced by views from the train. In a tunnel there is nothing to look at, but elevated riders have a view of beautiful Seattle. The proposed bridges over the Duwamish and Ship Canal will both be spectacular; no reason the train winding through LQA and SLU cannot also be highly enjoyable.
“I honestly don’t get the aesthetic objection.”
Look at a half century of people pushing back against overhead wires, streetcar tracks, cut-and-cover construction impacts. density, “ugly” new buildings, changing the “character” of a neighborhood, etc. Objection to overhead guideways and noise is part of that. In order to do anything in these regards, ST or SDOT/Seattle has to overcome inevitable political opposition.
We’ve made headway in some of it: urban villages are expanding, midrise and missing-middle housing are being allowed in more areas, parking minimums are decreasing, Link is being built, etc. But elevated or surface trains in residential/retail areas is one of the hardest things to get approved, along with cut-and-cover. There’s too much politicians’ and public opposition to them, and politicians favoring nimbys over transit riders’ needs.
Ballard and West Seattle were elevated in the representative alignment in the ballot measure, and the ST3 budget was scaled to that. Then after the vote, some neighborhood groups insisted they must be underground to minimize impacts on residential and business properties and visual impacts. So ST is pursing that.
Federal Way Link had two alignment alternatives: one on 99 and one on I-5. The 99 one would have served walk-up passengers better. The City of Kent wanted 99. But the Cities of Federal Way and Des Moines wanted I-5 both to lower construction costs and guarantee it would be built (Federal Way) and to avoid changing an auto-oriented strip-mall area with “inextensive immigrant startup storefront opportunities” (Des Moines). In other words, “keep your elevated trains away from me”.
The Monorail Project plan was moved from 2nd Avenue to 5th Avenue between Seattle Center and Intl Dist because 2nd Avenue businesses didn’t want trains in front of their 3rd-floor windows and passengers looking in, or losing surface parking for the stanchions. I guess 2nd Avenue businesses had/have more clout than 5th Avenue businesses somehow.
The one east-west street that could easily accommodate an aerial alignment is Mercer Street between Seattle Center and Westlake Ave or Fairview Ave. Getting under the 99 tunnel is part of why SLU can’t be easily cut and cover.
To curve southward is where the challenge would be. It becomes an expensive 3D jigsaw puzzle to configure the routing (with stations) between there and south of Westlake where a cut and cover segment could possibly begin.
The transfer hassle problem is real! Too many of the ST3 cooks don’t fully understand that it’s serving their tasty dishes cold because the added time required to use the stations for short distances (an attempted analogy) is like food sitting out for 10 minutes after it is prepared.
As I’ve said before, there are other much easier and cheaper ways to partially address this. The easiest: The SODO Station area needs to be redesigned for cross-platform, same-direction transfers or same-platform transfers by having two lines stop at the same platform. Making people have to use multiple escalators and transfers just to transfer between train lines creates bottlenecks at the escalators and elevators as well as adds several minutes to a transit trip (and worse when a device is out of service).
This design change would benefit Link riders regardless if the trains are automated or not. Even the West Seattle stub would greatly benefit from this.
I’ve been saying this for years. Does anyone else get its basic importance enough to actively advocate for it? Every time I read about how awful the transfers will be at Westlake or CID or how useless the sweat Seattle stub period will be I want to make the point that this is a partial remedy.
Sure it does not directly benefit 2 Line riders or SLU riders headed north, but it would benefit many riders making the trips mentioned in the post.
Another lesson to learn from Toronto, including the Ontario line, is that you really should max out bus frequency *before* you invest in rail.
How frequently did surface transit run before each line was built in Toronto?
Line 1: two-streetcar trains every 60 seconds (peak)
Line 2: two-streetcar trains every 60 seconds (peak)
Line 4: local and express buses every 90-100 seconds (peak, 5-10 mins off peak)
Line 5: local buses every 1-2 minutes (peak, 5 mins off peak) west of Line 1, east of Line 1, replaces 3 separate local buses each running every 10 minutes or less (all day)
Line 6: local buses every 3-5 minutes (peak, 5-10 mins off peak)
Ontario line: North of line 2, replaces 3 separate local buses each running every 10 minutes or less, and an express bus running every 8-10 minutes (all day)
In each case, surface transit built a ridership base of at least 30-40,000 daily riders well before rail construction started. And there are numerous surface transit routes in Toronto where buses or streetcars still run every 5 minutes or less, all day every day, even post-pandemic and post-service cuts, often including overnight (2-5am) service every 15-20 minutes. So in all of this debate, why do buses to Ballard run every 30-60 minutes at night?
And thinking of lessons from Toronto, would those riders feel differently about the need for rail if those buses ran every few minutes 18 hours a day?
Then we can never have rail because of the politicians’/agencies’ adversion to that high bus frequency? UW-downtown was melting down with bus bunching, traffic congestion, and overcrowding even at 7-minute daytime frequency. We should have just lived with that forever because the agencies are unwilling to run buses every 2-5 minutes and they’d get caught in I-5/Eastlake congestion anway and there’s no room to widen the streets or create a busway?
Look at the Eastside: the 2 Line will run every 8 minutes peak, 10 minutes off-peak. The 550 runs every 15 minutes weekdays and Saturdays, 30 minutes Sundays and evenings. We’ve begged and begged and begged ST for over a decade to increase the 550’s frequency in the interim, but ST’s attitude is “wait for Link” no matter how long it takes. (ST did try to increase frequency to 15 minutes ca. 2022 but the driver shortage swallowed it. The governments didn’t lift a finger to improve the situation.)
Or look north: Link will run every 4 minutes peak, 5 minutes off-peak, but the best ST could do with the 512 is 15 minutes daytime, 20-30 minutes evenings.
That’s not because there’s no demand for higher bus frequency. It’s because of lack of political will or vision.
Jonathon and Mike Orr,
I wonder of the threat of tearing up West Seattle for 10+ years for construction of a light rail line might make adding “bus only” lanes a little more politically possible?
I really believe Sound Transit wants to build tunnels to avoid political backlash. But if tunneling isn’t a reality now because of the budget…. maybe West Seattle would rather have a word class bus system rather than light rail?
“I wonder of the threat of tearing up West Seattle for 10+ years for construction of a light rail line might make adding “bus only” lanes a little more politically possible?”
It hasn’t for twelve years. Why should it suddenly do so now?
“I wonder of the threat of tearing up West Seattle for 10+ years for construction of a light rail line might make adding “bus only” lanes a little more politically possible?”
It probably depends on where in West Seattle someone lives or works or owns or shops. Generally, I think most are unaware of what the construction impacts will be.
Here is what I see happening:
>The Delridge station surround disruption will be modest as ST is buying out properties in an area including the limited commercial activity near the station. There’s also not much pedestrian activity there with RapidRide H being the main pedestrian driver.
>The Avalon station if built will affect local traffic circulation but there isn’t much heavy pedestrian traffic there either. Traffic management is probably the main challenge.
>The Alaska Junction area is the one that will create disruptions that will be long-lasting and dramatic. The station site will be a donut hole in the commercial district. The excavation trucks will be constant for months at a time. The on-street businesses will lose customers as lose on-street spaces after encountering construction vehicles around the station. The long construction period of several years (expecting at least 5 years) will impact nearby businesses by discouraging customers. The neighborhood character will be forever changed.
As the possible construction impacts get closer on the calendar, I expect that those interests around Alaska Junction will shift from enthusiasm to concern and maybe opposition.
Unlike other places like Capitol Hill or U District with lots of renters within the walkshed, patrons in West Seattle are often area homeowners who walk or drive to the commercial district. Link will fundamentally change the district if the station is built and then opened (after local merchants have long disappeared because of losing their core local customers no longer wanting to go there).
Al S.
I’m wondering about what if the ST board decided it can’t afford tunneling? Given the population and density of West Seattle. I don’t think a tunnel is really the right answer. I’m saying the board sticks with that 4 billion number than builds the best it can with that. That’s still a whole lot of money.
Sound Transit has used tunneling machines to quell popular opposition to projects…. I don’t really believe West Seattle or Ballard are any different than Portland neighborhoods served by just good old surface light rail. The only reason for the tunnels are political in nature. Seattle isn’t NYC after all.
So if West Seattle was offered an elevated or surface train…. I’m guessing the local population would just say “fuck no!”
“Sound Transit has used tunneling machines to quell popular opposition to projects…. I don’t really believe West Seattle or Ballard are any different than Portland neighborhoods served by just good old surface light rail. The only reason for the tunnels are political in nature.”
Unless there’s a topographical reason, I’ve long said that subway stations should be earned by a place creating enough tripmaking density to make the station productive. It shouldn’t be a mere bargaining chip to ease neighborhood opposition.
Subways take many years to build and nowadays cost billions to build. It’s not like adding a RapidRide stop or streetcar stop. That kind of investment should be made only when it’s strongly wanted to the point of supporting high density development — and not just tolerated because other alternative designs are considered too much of a visual nuisance.
It’s a political decision. Sound Transit is structurally designed to build what the subarea counties and cities say they want. Boardmembers make their decisions based on how it affects their reelection chances in their county/city position. More voters care about not having a surface, elevated, or cut-and-cover alignment in their neighborhood than about whether ridership numbers justify a station, or where the line goes or how difficult the transfers are. They don’t know about those things and have unrealistic ideas about what transit can and can’t do and what it costs.
If it weren’t that way it wouldn’t be such an uphill battle to get ST to do the right things.
Mike Orr,
I understand the political nature of Sound Transit all too well…….
Currently there is this big ground swell of support for “Build the Damn Trains Already” . The idea the reason Sound Transit is floundering because of the board is just dithering and everything can (and needs!) to go just like the voter’s pamphlet is popular. But is it true?
“Currently there is this big ground swell of support for “Build the Damn Trains Already” .”
We don’t know what that means. Does the group that published it and all its endorsers mean “No changes to ST’s current preferred alignment and projects” or “Some sensible changes are OK or necessary, but don’t dawdle unnecessarily”.
“It’s a political decision. Sound Transit is structurally designed to build what the subarea counties and cities say they want.”
That pretty much screams out the core cultural problem with ST3. It’s about what officials want. It’s not about what’s the best benefit to the public or the riders based on performance metrics.
One would hope that elected officials would be responsible stewards of taxpayer money enough to ask for metrics before deciding things. But they don’t. And staff offers them no metrics either other than costs (and even then those can be way off).
That pretty much summarizes the last decade of ST planning.
Go read any other Board agenda where they’re spending transit money in the billions in the US. They openly discuss multiple technologies, rider benefits and objective tradeoffs. ST leadership carefully orchestrates what they want to happen beforehand — and makes it public only when they have already crafted the outcome that they want.
Even the “Build the Damn Trains” presser appears orchestrated. Why was Dow there? Why was the presser announced and conducted so quickly, rather than be first stated in the press releases of the participating advocacy groups weeks beforehand? This has all the fingerprints of a PR effort organized by some people who feel threatened about not getting their way.
“he core cultural problem with ST3. It’s about what officials want.”
It’s how the state legally structured Sound Transit. The counties told the state they needed regional transit because the local agencies couldn’t provide all-day inter-county express buses because of state-imposed tax caps and the agencies prioritizing it last. The state didn’t want to get involved in deciding regional transit corridors and operating it itself, so it allowed Sound Transit to be created to do it. The counties, cities, and parochial voters wanted to keep control of what was built and who paid for it, so they got the state to embed that into ST’s governance structure. So ST boardmembers are like a committee of county/city officials rather than independently elected.
Since the name is actually ID/C can we switch off of CID?
CID is easier to write and read. I still write “Intl Dist” sometimes based on its former name. ST can reconcile its wording with the city’s someday. In the meantime I consider they’re all interchangeable. I don’t say “Lynnwood City Center” or “LCC” just because ST puts arbitrary unnecessary words into the name. I just say “Lynnwood”, unless I’m speaking with somebody who might not recognize the station if I don’t use the exact words on the sign.
The name of the neighborhood is not the “ID/C.” Certainly nobody calls it that; in my experience, it’s either “the CID,” or if you’re older, “the ID.” It’s not called the “ID/C” in any city comprehensive plan or by the city clerk, and indeed the International District name ST puts first is rather recent (it dates to the 1950s) and has always been somewhat controversial, to the point that the city passed Ordinance 119297 in 1998 to change the name to “Chinatown/International District” in the comp plan. The “ID/C” name is entirely a Sound Transit idiosyncrasy, as far as I can tell. It’s a clunky malapropism and kind of makes Sound Transit look ignorant of the history of the neighborhood—I’m not saying it’s super important, but if the name of a neighborhood is highly contested, maybe go with the agreed-upon name instead of something made up?
Apparently it was just the International District Station when Metro first opened it. It appears that the name change was requested and made about 20 years ago. I’ll let others chime in on details.
I personally think it should have just stayed the International District Station. There’s just too many letters in today’s name. But that was the era where systems around the country started giving out stations longer names with slashes or hyphens. Nowadays, seasoned riders know the shorter names are better — as they can fit on signs on trains and at stations easier, and make announcement’s quicker.
> The “ID/C” name is entirely a Sound Transit idiosyncrasy
It’s King County’s misnomer, carried over from the original Downtown Seattle Transit Tunnel station naming scheme. As far as I can tell, the push to append “Chinatown” to the neighborhood’s previous name of “International District” was ongoing when King County established the name of the station. It appears the planners simply guessed wrong as to which end of the name “Chinatown” would be added to. Today and for the foreseeable future, the hard-won name of neighborhood is “Chinatown International District”, so the station name ought to be corrected. It seems like a correction that has to come from the Board, but the Board simply has bigger fish to fry.
June Foamer,
All names are used and correct. It’s Chinatown/International District or just Chinatown if you’re Chinese and International District/Chinatown (or just International District) if you’re a non-Chinese Asian.
As a White guy I know better and just stay out of it. The Asian leaders have had many pissing matches about the name over the decades. All sides sorta make up and agree to disagree in end. Call it whatever name you like. In Tacoma, it’s always going to called the I.D. by the SE Asian families with connections there.
Ah you’re right Al. I posted hastily and didn’t consider that, of course, the station predates the 1998 neighborhood renaming ordinance. My bad! The name was definitely given by the county. If you read King County Ordinance 15074, which renamed the station in 2004, they point out that both the community center and the library were named International District/Chinatown by 2004, even though the comp plan was changed years earlier (in 1998) to call the neighborhood the Chinatown/International District. I’ve never noticed that the library and community center invert the name even though I’ve been to both a fair amount. Oops. In any case, I still insist that I’ve never heard anybody call it the “ID/C”!
The station was called International District and the neighborhood was renamed because of 1980s-era misgivings that “Chinatown” left out the other ethnic groups. Then they asked the other groups and found out they were fine with the name Chinatown, so the change was for no reason. Although I’ve always felt “International” is inaccurate because it’s really only Asian.
By the 2010s sentiments had changed and there was a wish to bring back the “Chinatown” name, both for historic reasons and because visitors would recognize it more easily (several American cities have a Chinatown next to downtown with a walkable street-level market like ours does). But they didn’t want to eliminate the “International District” name they had promoted thirty years before, so the city came up with “Chinatown/International District” (CID), and ST somehow distorted it to “International District/Chinatown” (ID/C), perhaps to keep the “International District” name first as the station had been called, or because international is more international.
To me all these names are interchangeable and it’s not worth a whoop to insist on one and suppress the others. Ideally it should probably just have remained Chinatown throughout, but it’s too late for that, and I’ve already gotten used to International District.
Based on the history of Sound Transit station renaming processes, I’d be 100% sure that they avoided changing the name to Chinatown/International District because they would have had to rename the internal station code. Which is a shame. Getting the correct name here actually matters, given the history, and especially considering that their headquarters are literally in the neighborhood too.
Fortunately “University Street” and “Symphony” have the same letters in the same order…. Oh wait, there’s no U in Symphony, and S is first.
So let’s talk about branching again. It applies to what seems to have happened in Seattle. As a Torontonian who recently moved to Seattle, I don’t quite understand American cities’ obsession with every bus from every single neighborhood running downtown. It sounds great in theory, until the thing you just described (bunching and erratic service due to overlaying mildly-frequent routes on top of one another) starts to happen.
Ottawa (Canada) managed to get 180 buses/hour on a single lane couplet of bus lanes on at-grade streets downtown. When congestion became a significant issue, the transit agency used ridership to prioritize which lines could go direct to downtown, and which riders should transfer at major bus stations (to an express bus running every 3-5 minutes).
When lines don’t branch, and you maximize frequency above all else, even including above transfer-less rides, something magical happens. If the bus runs every 60 seconds and all the buses are going to the same place, then you can get on the first bus coming and transfer to another bus at another station.
I would think that Seattle’s experience with bus congestion on downtown (and other) transit malls would be instructive here. When “high frequency” means every 7 minutes, then yeah riders might object to a forced transfer.
We tend to run buses to Link stations. We may want to consider the same for express bus lines at some point as Ross and I suggested for West Seattle’s RapidRide lines.
https://seattletransitblog.com/2024/06/07/west-seattle-by-bus-instead-of-light-rail/
“As a Torontonian who recently moved to Seattle, I don’t quite understand American cities’ obsession with every bus from every single neighborhood running downtown. “
The history traces back to how comprehensive and popular Metro was before Link arrived. Keep in mind that Link opened in 2010 with it only running between SeaTac and Downtown until 2016, and only to Northgate after 2021.
I don’t think there is another large urban area in the US or Canada that has overlapping agencies like we do. And Metro staffs the ST operations! It leads to disjointed systems planning — even though both agencies claim that they’re coordinating well. For some reason, each agency decides what it wants and when they finally tell the other one they call it “coordination”. Few point out the ludicrousness of it all.
And layer on SDOT staff who want to force their pet projects that may or may not align with the system that Metro or ST wants to run. That’s how we got the FHSC that didn’t replace any bus route!
As Metro tried to more radically restructure things in 2015 for U Link opening and got heavy local pushback AND Metro has not faced the systemic pressure to make big changes in recent years (aka “structural budget crisis”), routing hasn’t really changed.
Heck, Metro hadn’t even walked away from serving park and ride lots even though ST has built several large parking garages just a few blocks away.
My observation is that many needed transit routing changes don’t happen until they are forced — usually forced by budget issues.
A final issue is that Link station spacing beyond Downtown (original Metro tunnel) has been further than other metros have built, so that results in more buses still running on prior routes.
Only a couple routes run every 7 minutes. We’re lucky to get 10 minutes on a handful of routes. The rest run every 15-20 minutes, and not even that on evenings or weekends or Sundays.
The 43 was a full-time route for decades between downtown and the U-District going east of John and north on 23rd. That overlapped with the 48 on 23rd and the 8 and now 11 on John. In 2016 Metro tried to delete the 43 and compromised by making it peak-only. The argument was that people could ride the 48 and transfer to an east-west route, and soon RapidRide G would start. That breaks down when each route runs every 15 or 30 minutes, the transfer stop is 2 1/2 blocks away or up a hill, congestion makes the buses unreliable. People don’t want to transfer in that environment. Oh, and the G opened years after expected.
What people are afraid of is that Metro will eliminate the downtown routes and promise to increase frequency on the feeders and trunks, but then the frequency increase doesn’t happen due to a recession or driver shortage, so you’re left with crappy service and a two-seat ride.
Metro did do a bunch of trunk-and-feeder restructures between 2012 and 2018. And there were earlier ones going back to 1990 when the downtown tunnel opened. There are 1984-era Metro maps around online showing how far we’ve come. (I lived through that period from 1979 in junior high to 2012 when the biggest restructures started occurring.)
But after 2020 Metro got timid again about restructuring. Not as bad as it had been before 2012, but we’re missing a bunch of opportunities for route consolidation that could lead to more frequent service for the same cost (service hours). It seems that the bold 2010s-era planners retired and the new ones aren’t as bold, and management isn’t encouraging them or pushing back on opposition as much.
ST’s mandate is “regional transit”. Metro’s mandate is “local transit”, and filling in express needs left over beyond the ST Express network. So most Metro expresses are gradually going away. A few hang on for historic reasons or because ST Express left the corridor out.
(E.g., the 255 continued to downtown because there was no Kirkland-downtown route like the 545, 550, or 554. It was truncated at U-District station in January 2024 I think. But then the pandemic hit, the Link trunk was reduced to 30 minutes weekends, the 255 lost most of the feeder-frequency increase it had gotten, Husky Stadium events prevented the 255 from getting to the U-District in a timely manner, and many highway 520 construction closures forced the 255 to go all the way around via I-90. So Kirkland residents have gotten the worst end of the stick as far as bus restructures and subsequent calamaties.
We’re trying to raise the word that 15 minutes is the absolute minimum for a core route, and 7-10 minutes is better, even on suburban routes as Reece Martin says, but it’s hard to convince the politicians to recognize this or put together a plan to fund it.
It is unfortunately how people are using it most of the time. It is the small-town planning mindset that you have to have a business district that runs all the business. Therefore all the bus has to go there. On one hand, I think it is stupid to run all the buses to the same destination, but more often bus network is just a result of the city form.
The reason other places don’t have buses all going to downtown is because many places don’t really have US style downtown.
If someone ask different people from my hometown where is the downtown of my homecity, I am pretty sure they won’t have the same answer and some might find it hard to narrow down to a specific area. As many cities outside NA has a much bigger urban core with all the mid/high density mixed use district. Many cities may only have one or two CBD that features all the landmark supertall buildings like the US downtown, but on the street level, there is a bigger area that looks just like “downtown”.
So it is not the bus network, but the city form leads to all this.
When I was in London I was really puzzled to figure out what “downtown” or was. Victoria Station? Charing Cross? But they aren’t denser than other inner-city areas. I finally came to the conclusion that London doesn’t have a downtown as I’d think of it.
Toronto certainly has a lot of destinations all along its subway lines, which is why so many people ride it all directions all day. Ditto Chicago. San Francisco and Vancouver have streets with destinations going on and on, convenient for a single bus route.
Seattle in contrast is more islands of high or medium density. Broadway and University Way are short. Rainier may be the closest to a long destination-packed street. Aurora has a lot of businesses but most of them are spread out with large parking lots in front and it’s a stroad. Ballard is a 15-minute city but a rather small one with a limited range of things. All that is because people have been historically allergic to density or having businesses in residential areas, or a wide range of businesses outside downtown. So we get kind of blah neighborhoods, and people have to go to a few places for things they don’t have.
London’s Central Business District is Canary Warf. It’s political center is Westminster, with road distances to London traditionally measured from a statue of King Charles at Charing Cross, and the retail center is traditionally Oxford street. Looks like UK has a “Central activities zone,” which is probably gets to the spirit of your question “where is downtown”?
https://en.wikipedia.org/wiki/Central_London
“London’s Central Business District is Canary Warf.”
I think it is probably a typical perception Americans would have about cities outside North America. Canary Warf is probably closest thing London has to a North America style downtown because big part of it was developed by a Canadian developer. This part of London doesn’t feel like London but more like Chicago. Three are less foot traffic on the sidewalk compared to other part of the city and there are fewer curbside stores. Most retails are along the indoor concourse. They even have a shopping mall there. The kind of design is not exactly ped-unfriendly but it does keep foot traffic within the buildings whether they arrive by car or transit.
But it is not THE Central Business District for London. I think east part of City of London is more likely to be the considered “the CBD”. This financial district has most London’s famous landmark high rises like The Gherkin. Calling Canary Warf the CBD of London is like calling SLU the CBD of Seattle, which is not exactly wrong but it is certainly not “The One”.
Both Canary Wharf and SLU are outside (adjacent to) the area that could be a downtown candidate (somewhere between the City and Oxford Street), and weren’t even built up until the 1990s and 2000s. Unless maybe you argue that London didn’t have a downtown until the 1990s but now does. But there was a lot of business and activities and services centralized in part of London for decades before that, and that would be London’s downtown.
My doubts sprang about London not having as much centralization or a density difference as I’d expect for a downtown.
https://www.london.gov.uk/sites/default/files/gla_migrate_files_destination/londons-cbd-jan08.pdf
“Central London” is too large an area to be downtown. A downtown needs to be neighborhood-sized: around 1-4 square miles so that you can walk from the middle to the edges. Otherwise you can’t “go downtown” and “do downtown things” and come back: you’re just going to a very large area where you need a tube line to get from one part to the other. That’s a large urban area, not a downtown.
So I don’t see how Canary Wharf, the West End, and Camden Town can all be part of downtown. I’m looking for something like New York around Wall Street, Chicago’s Loop, downtown Los Angeles, San Francisco’s Market Street area. Not Manhattan all the way north to Midtown and Central Park. And ideally it would be were jobs, government, retail, services, and culture are all concentrated together — although that’s not possible with some cities that have government or retail somewhere else. But not just an area of 9-5 office towers and a midday concourse between them and nothing else. That only addresses a narrow range of people and activities.
Look at the PDF, page 8 – in London the retail, culture, government, and business districts are distinct. This is common in European cities (Paris: La Defense, Lyon: Part Dieux), where the office districts are brownfield developments because the historic urban neighborhoods are not amenable to high rise development.
No, branching is a powerful tool when it comes to serving lower density neighborhoods and secondary destinations.
https://pedestrianobservations.com/2012/10/24/branching/
Even with a bus network, there should be a trunk and branch operating pattern unless your city lacks urban nodes : https://pedestrianobservations.com/2025/03/20/open-brt/
Seattle’s urban core clearly has a southern endpoint at the ID, so that is the appropriate branching point to the south. To the north, rather than branching the rail service, ST, CT, and KCM plan to terminate many routes at Link stations, so key routes like Stride 522 and Swift Blue are essentially branches off of Link.
The urban core of Seattle will be served by the trunk line with no branching: U District to the ID. Otherwise, Link + Sounder create a radial network centered on King Station, which is the best practice for major cities.
https://pedestrianobservations.com/2019/05/13/the-importance-of-radial-urban-rail/
Also, the Seattle isthmus creates a narrowing point at the CBD, which makes it natural to converge North/South routes. If Elliot Bay did not exist, the RR D & C would probably be a single north/south corridor, but since both of those routes need to curve east to follow land, might as well serve the primary activity center directly via 3rd.
Finally, not every bus goes downtown. The 8 is a key circumferential line within urban Seattle, the 44 and the 48 are key circumferential lines for Seattle isthmus*, and 405 Stride routes provide a circumferential service around Lake Washington. There are also important routes that don’t go near any CBD, such as the A and F.
*the 50 kinda sorta completes the 44/48 circumference, and this blog has heavily covered the issues with the 50
With the urbanization of SLU, Cascade, and the Denny Triangle, the 8 really isn’t a circumfrental route anymore. It isn’t going around the CBD, but connecting it’s Northern parts to the areas to the East. And this is really why Denny needs Bus lanes, it’s not the edge of the core anymore.
I agree FDW, but the point remains the same. It makes sense for geographic reasons to send a lot of buses north-south, through downtown. That is why the biggest bus spine is north-south even though Link also runs north-south.
But at the same time, having a few east-west lines adds a lot of value. Several of them serve downtown as well, including the smaller spine on Pike/Pine as well as the 8.
I don’t quite understand American cities’ obsession with every bus from every single neighborhood running downtown.
That is no longer the case with Seattle. It used to be. If you look at transit map from the 1980s you can see a lot of routes went downtown. It may be hard to see that from the map but look at the route listing on the left. Bus after bus has the word “downtown” in it. There were a few reasons for it. First of all, downtown was a major destination. Not only that but there were plenty of people who drove everywhere but hated driving downtown. That is still the case but it isn’t as extreme. The other reason is Seattle’s geography. For example consider a bus serving the top of Queen Anne. It would be nice to go directly east, towards Capitol Hill. You can’t. The hill is too steep and you run into Lake Union. Maybe you could run directly west, towards Magnolia. Again, the hill is too steep. The only direction that really makes sense is south, directly towards downtown. It is very difficult to make a good grid system in much of the city which is why they gravitated towards a hub-and-spoke system with downtown as the hub.
At the same time, even back then you had buses that didn’t go downtown. Notice the 48 running from the UW to Rainier Valley via the Central Area. Or the 44 running from Ballard to UW. The 8 was another east-west bus that at the time didn’t really serve downtown (it skirted it). Now it would be considered downtown but also an obvious route that is doing something that the other routes aren’t.
But the big changes occurred with Link. Look at this map from right before U-Link (when the trains got to the UW). There are a bunch of buses that go downtown. Not just by route but by the number of buses each day. In contrast, look at a modern map. North of the ship canal you have a lot fewer buses going downtown and a lot more that don’t. Buses like the 31/32, 44, 45, 61, 65, 67, 75, 372 and 522 don’t go downtown. As you get further north it is pretty much *only* buses that connect to Link. There are a handful of routes that run express to downtown but they run so infrequently they don’t really matter.
There are buses from the East Side that go downtown but certainly not most of them. There will be fewer once East Link is done. To the south you probably have the highest proportion of buses heading downtown but there are several reasons for that. One is that the freeways connect very well to downtown from the south. West Seattle to downtown or Renton to downtown is very fast. Quite often the freeways are skipping over areas that aren’t bound to get much ridership (like the industrial area between West Seattle and downtown). But in the south end of Seattle you still have buses like the 50, 107 and 128 that don’t go downtown. In the case of West Seattle it is just natural to continue downtown. If a bus is running north from the south end of West Seattle it is really easy (in most cases) to keep running downtown. The idea with the 50 and 128 is that riders transfer to a bus (often RapidRide) to get downtown. Other than the 125 I don’t see any other bus I would have skip downtown. Meanwhile, I would run the 56 all day (which means roughly the same number of buses — if not more — running downtown). As you get outside the city you have more and more buses serving the local area but not downtown. You still have plenty of express buses but that is because there just isn’t that much demand for destinations on Link between Tacoma and downtown (other than SeaTac).
There really are only a handful of routes that I would see changing (to not go downtown). The 106 for example. Maybe it could be truncated. On the other hand, maybe it could go downtown via a different pathway or go to First Hill (which is essentially downtown now).
How do we learn from Toronto where agency can procure whatever that’s best for their operation all over the world?
Are there automated trains the size and length of our current light rail fleet?
Of course. Any multiple-unit passenger equipment, even standard LR vehicles, can be automated, but if one used standard LRV’s one would end up wasting the cab space in every car even more severely.
Automation is an electronic technology that doesn’t take much space, and does not need cabs at all, though somecsystems include “fold-out” stand-up manual control stations for moving individual cars around Maintenance Facilities.
Think of the Sea-Tac People Mover; it has very primitive automation because there are few stations, the lines are hermetically sealed by platform screens and being in tunnels, and there are no junctions. But it has been fully automated for decades and runs constantly with no fatalities.
Any number of cars can be linked together in a train under a single designated controller. A controller will typically be present in every car, but only one which issues commands it receives from the off-board computers is active in any given train. It uses the standard Multiple Unit controls throughout the train to control the other cars.
Glenn may know od fully distributed systems, but I have not heard of any.
Tom,
The problem is what’s qualified for federal funding not what are available out there.
Airport landside/airside shuttle system are built in very similar use case and environment. So Bombardier was able to sell one product to 10-20 different clients without much change.
Meanwhile mass transit system requires more specialization to meet all the local need and USDOT requirement. Every time a foreign company build some new transit models here to meet By American Act requirement, it comes with all sorts of problem and cost overrun. If ST will be able to use the same model as Hawaii Skyline, that would be ideal.
I agree with Tom. But more to the point if we are going to build a new line anyway we might as well make it automated so that we can have shorter platforms but more capacity. The only drawback is that it wouldn’t be integrated with trains going south. I would consider that a feature, not a bug. It makes way more sense to go with the original pairing (Ballard/West Seattle) than Ballard/Federal Way. Way more riders go to the existing south end locations (like SeaTac) than West Seattle. Way more riders are heading towards the existing north end locations (like the UW) than Ballard Link locations. It is better to match up the most riders as possible and the existing line does that.
Not only that, but it means West Seattle Link — if it is ever built — would have smaller, automated trains as well. West Seattle Link is extremely dependent on transfers which means that a high frequency route is more important. But this would also reduce the cost of building it. Thus just like an automated line to Ballard it is cheaper and better for riders to design it with automation in mind from the beginning.
I get why people focus on the entire line but we can’t possibly build the entire thing at once. We can only build it step by step. The next step should be Ballard to Westlake. It adds the most value. Eventually we have to consider whether it makes sense to build West Seattle Link or whether it is too expensive and we simply can’t afford it (without additional funding). Then, with the next levy package (if there is one) we should consider whether it makes sense to build West Seattle Link (and a second line through downtown) or to serve First Hill (or both).
Yes exactly this, Ross. A switch to automation for both Ballard and West Seattle — along with a long-term vision to connect them when the funds are there — would lead to lower station costs as well as other things that could save billions that ST staff has not yet demonstrated to the Board. That could be a huge savings for just West Seattle if the staff would study it and the Board would consider it!
The critical next big irreversible decision the Board faces will be to take further action to build West Seattle Link stub. The Board doesn’t seem willing to pause its construction commencement. Maybe the staff and Board would reconsider things if the issue could be made as a simple either/or: Use the current technology and stop at Delridge or change the technology and get further to Alaska Junction.
Kubly weighs in on ST’s cost solutions:
https://www.theurbanist.org/2026/01/13/op-ed-how-to-close-sound-transits-35-billion-dollar-gap-without-breaking-the-system/
My read:
Overall, he proposes a 30 percent farebox recovery — which is never something the Board talks about and should. As I’ve said many times recently, it’s terribly irresponsible how the Board does not discuss performance measures when deciding Link extension details — and the staff won’t talk about them either. Of course, so much of ST3 Link would go away if the Board was so serious about farebox recovery at it was based on Kubly’s proposed 30 percent. But it would be a useful “what if” exercise for ST staff to study then tell the Board how to get to 30 percent, or even 25 percent by cutting the less productive elements of ST3 — and at least daring to not offer frequent service to the extremities if the ridership isn’t there.
This alone puts Kubly’s general thoughts here a bit into the “unrealistic” category. Farebox recovery cannot be achieved with a magic wand.
That said, he does make a strong plea for automation as a solution.
He also makes a strong plea for revisiting excessive design standards, and also references looking at cut and cover rather than deep bored tunnels. He also talks about scaling back to 8 minute headways for each line as a huge way to reduce OMF and vehicle purchase costs (yet does not question whether the expected passenger loads would be acceptable).
He unfortunately doesn’t discuss endorse an automated Ballard stub line other than to indirectly say that if it can reduce the costs by up to $4.5B it should not be summarily disregarded. (Note that many of us think the savings could be much more if the scenario revisited the design standards assumed in the estimate too.) Yet his big picture points are curiously similar to what many of us have been saying here — and the Board never discusses (and the executive staff won’t dare bring up to them).
Where is Kubly getting deleting BAR or or Graham or abandoning Aurora/Harrison from? Have boardmembers indicated they might consider those, or is this just his own guesses? So far I haven’t seen any indication the board is ready to debate these.
And what does abandoning Aurora/Harrison mean? Deleting the station? Moving it?
Where is Kubly getting 6-minute headways on the north and east lines from? The south line in DSTT2 is the only one I’ve seen ST say will have 6-minute headways. ST considered 6 minutes for ST2 but went with 8 minutes, and I haven’t heard that will be increased in ST3.
ST should prioritize 6-minute all-day minimum on all segments, because that’s a major factor in why cities with real subway networks have such high ridership and low car mode share. When a station is just steps from your location and a train will come in 2-6 minutes both day and evening every day, that makes transit very competitive with driving or taking a taxi, and encourages people to do trips that would otherwise be inconvenient (like popping to a neighborhood for lunch in a short timeframe).
But frequency has to be weighed against fiscal constraints. I’d rather ST give frequency a somewhat higher priority than insist it must be 6 minutes now, everything else be dammed. The fact that a 15-mile segment will be 6 minutes where the 1/2 Lines overlap (including downtown, Capitol Hill, U-District, Roosevelt, and Northgate) is a major step forward. We can gradually improve the East and South corridors toward that as some of these other issues are resolved. The important thing is that ST commits to ultra-frequency, has a long-term plan toward it, and is taking visible steps toward it. If it has that, we can wait a bit for it to be fully implemented in all segments.
Kubly’s suggestions do seem off, both in the sense that they’re not the things we should cut if we want to retain a productive system, and in the sense that many are not realistic or things the board even seems to be considering. I wish that The Urbanist’s reporting on ST3 was better, or that it at least had a comment section. People go after STB for being particularly armchair, but The Urbanist seems to consistently miss many of the things that have been discussed for years on STB, like automation.
ST cleverly removes vehicle purchases and OMF construction out of adopted extension budgets in ST3. It lowers the published price tag.
Consider that OMF South is $2.5-$3.0B not included in TDLE. Each train car is now up to $7.4M, up from $4.9M in 2019.
Of course, trains today can be shortened from 4 cars to 3 to avoid buying and housing more vehicles. But we’d still be paying for more drivers. I don’t see ST reducing train lengths as it expands.
In total, going from 8 minute frequencies at the peak to 6 is not just a driver cost. It’s an expensive capital cost too in both OMFs and vehicles if we expect to have four-car trains. If the capacity is not an issue, I’d leave it at 8 (or 7.5) minutes.
I would even be willing to go along with putting three lines in the DSTT and go to longer 9 minute frequencies. Those extra three minutes would otherwise be taken up by the terribly cumbersome transfers with well over 3 minutes required to make the transfer for a large share of riders anyway.
The OMF-S is needed to support the WS Link extension, so it would be incorrect to include within the TDLE project cost. ST appropriately manages the OMFs and vehicle procurement as distinct capital projects as they support multiple expansions.
Combined response to a bunch of comments above, but:
One additional advantage of short, automated trains is speed.
Eg: if you have to make a street radius turn from 15th onto Market, either at street level, elevated, or in tunnel, it takes you half as long to make that curve with a 200 foot long train as with a 400 long foot train, and you can get back up to speed faster.
Great article. I fully agree. This is the only solution Sound Transit should consider.
West Seattle can be automated in a similar fashion, and I think even starting to think automating the entire 1-2 Line spine will do great wonders.
The main concern with the Tacoma to Everett Line is the length, which operators may not enjoy. Thus automation solves this constraint, as well as allowing us to algorithmically maximize frequency and timeliness for near perfect transfers.
The perfect rail spine in my opinion is what is planned, as well as an eventual replacement of Stride with higher capacity rail, and a new line connecting Renton along with Rainier Valley segment to the UW and Lake City. Then smaller automated *light* light rails that act as a highly dense subway network within downtown.
ST could automate the 1 and 2 Lines and grade-separate Rainier Valley, SODO, and Bel-Red as part of the project. They don’t have to be a permanent blocker to automation.
I’d say that is much better value than building out the 4 Line, or West Seattle Link. Unfortunately they have to carry through with the voter approved projects.
Unfortunately people have been brainwashed to dislike buses, even though they can be faster and more pleasant than light rail in many situations.
I think they may be more afraid than averse to the idea of automation. It’s the issue with people who don’t have an understanding of technology. Even if you aren’t an expert/specialist in tech, it helps to have literacy in it as it’s a quickly evolving and ever so useful field. In the public sector, they won’t know this. A private company would be more confident in investing in automation.
It’s not difficult to automate trains all. Contract the right people to build the software. Even go to the universities. I’m sure there are brilliant students in our state universities who are willing to join internships and do this research. Plus, there are working examples around the world.
Anything a human vehicle operator can do can be automated in today’s age. With a train, it’s even easier. There are almost zero “intelligent” non-deterministic decisions that a train operator has to make, in comparison to driving a bus on busy roadways. All you have to do is program the train to follow the script, develop a monitoring/resiliency/disaster recovery plan, and let the software do its magic. We can simply leverage the existing tracks and infrastructure too. The only changes needed are to allow the train controls to be operated by the software directly, without compromising security and making them vulnerable to external manipulation.
Hell, they could just spend an hour at SeaTac riding the automated people mover there.
From my time working at ST (now >5 years ago), the leadership with the Operations Department considers automation impossible. Not hard, not impracticable, but simply impossible. I asked a few times and was laughed at as if I was asking why the trains can’t fly. They knew how to run light rail the same way they were trained in the 70s and 80s and were uninterested in learning anything else. If the ST Planning staff tried to schedule a meeting to review the idea of automating Line 1, they would just decline the meeting invite. I would bet good money they know nothing about the automated bus routes current operating in America, let along the myriad of automated rail lines running around the world. So it’s not fear but ignorance and incompetence.
South King Resident is correct. The automation of a rail line, even with at-grade crossing like in the RV and Bel-Red, is a solved problem. If there is an obstacle on the line (pedestrian, car, whatever), the train makes an emergency stop and then a person sitting at a desk the OMF can intervene and determine if it is safe to continue.
@AJ, thank you for that telling account.
This convinces me that the Sound Transit planning and operations departments are actively worsening transit in the region, in part by taking up space that prevents actually-good planning and operations people from getting involved.
Who hires for these departments? Which political leaders need to be lobbied to get new people in there?
Operationally, the long spine will at some point in the future will have to deal with another rather obvious outcome: crowded trains in Seattle (about 45 minutes of the trip) contrasting with few riders on the trains for the last 30-40 minutes at either end except for maybe 2 hours a day in each direction inbound or outbound.
The only operational cost solution that will make sense is to truncate some of the trains way before the line ends.
The irony about automation is that the additional costs of running a full line with long segments carrying few riders can be more easily absorbed with it. Automation will help productivity (segment cost per rider) of these long, low ridership segments. If Board members outside of Seattle really wanted to make sure that they will always have great light rail frequencies, they should sign onto the notion of automation ASAP.
Maybe they just don’t understand ongoing transit operating costs and productivity metrics and automated technology.
ST staff has already fully considered the crowding issue and has recommended Mariner as the truncation point.
Additionally, every station between Northgate and Mariner are bus-rail transfer nodes, where the higher frequency from interlining both lines is impactful.
The main concern with the Tacoma to Everett Line is the length, which operators may not enjoy.
Most likely you would have a turnback to the north. Not only to save operators from really long trips but also for cost concerns. Ridership goes down and is less influenced by frequency the farther you get away from the city. It is common all over the world to run trains less frequently in the outskirts of town. I just don’t see them running trains to Everett every five minutes.
Lynnwood seems like the logical choice for turning back. You would run trains from Everett to Redmond as well as Lynnwood to Tacoma. Lynnwood to Tacoma seems like a long way but they plan on completing Tacoma Dome Link long before they build Ballard Link. In other words, they are planning on running that far anyway (at least initially). Redmond to Lynnwood could also be automated at some point.
Katie Wilson seems to oppose the idea of dropping DSTT2, per Urbanist interview: “We need a reliable system, and I think it’s pretty clear that with the single tunnel, we don’t have a reliable system. And so as we’re building it out, we need that capacity, that a second tunnel seems to be the only way that we’re going to get that. I’m happy to have things looked at, but to me, it’s pretty clear that that’s the direction we need to go.”
https://www.theurbanist.org/2026/01/13/one-week-in-katie-wilson-charts-the-path-ahead/