Sound Transit boardmember Dan Strauss is proposing three amendments to the ST3 system plan update on Thursday. The full list of amendments is expected Tuesday, and it sounds like there will be a lot of them. Strauss is a Seattle City Councilmember for northwest Seattle. His amendments are:
- Prioritize building the Ballard-Westlake Link segment now, and postpone the second downtown tunnel (DSTT2) until after it.
- Use the 70% of unused debt capacity ST has, and ask Olympia for permission to issue longer-term bonds. Strauss says there are some “limited, commonsense adjustments” to debt policy that ST can make.
- Set a target date for finishing Ballard Link, rather than leaving it indefinite.
To me, #1 is consistent with our call to build the best parts first. I’d still like to see automated trains and canceling DSTT2 (especially to eliminate the excessively passenger-hostile tunnel-to-tunnel transfers), but this is an important step. Let’s not make the perfect the enemy of the good. If somebody offers me half of what I want, I’d rather take it, and leave the rest as something to try for later.
Re #2, there may be room to optimize existing debt capacity, but I’m still not convinced of longer-term bonds. Re #3, setting a target date for Ballard makes sense, and would avoid leaving people in limbo for years.
Update: Seattle Subway endorses #1 and #3, has a petition to sign for it, and urges people to contact their ST boardmembers starting now.
They need a plan on how West Seattle as well as the Tacoma-Everett spine will be handled. These are also pressing issues.
It is also critical to send the 4 Line to S Bellevue station, and abandon the East Main station.
These changes must be implemented for the best outcomes in our regional transit system.
No, they do not. West Seattle and Everett north of OMF-North can be rescheduled to the future; they remain fully funded and in the plan, or the specifics on how the region will handle those corridors can be a decision made later. Same for Issaquah Link – ideally it evolves into a Stride line or commensurate bus investments, but if that project remains out in the 2040s, the specifics today are irrelevant.
In the short term, ST should move forward with OMF-S, which would include SFW station. On the north end, complete the EIS to at least Ash Way and build 1 or 2 station extension(s) for steady progress north. Otherwise, focus on Ballard-Westlake and all other decisions can be done by a future board.
West Seattle and Everett north of OMF-North can be rescheduled to the future
Exactly. By deferring West Seattle Link we also defer issues with the tunnel as well as which lines get paired. The rest of the system is exactly as it is today. Trains from Redmond and Federal Way go to Lynnwood. A new, independent line goes from Ballard to Westlake.
Extensions don’t change very much either. Tacoma Dome Link is supposed to be done long before the second tunnel. Thus it is clearly OK to run trains from Tacoma Dome to Lynnwood. The plan is to run trains from Redmond to Mariner. So again, if they extend trains to Mariner the Tacoma trains could turn back at Lynnwood while the other trains run to Mariner. It is only as the train goes further north (to Everett) that you need to worry about trains from Redmond going that far. I don’t think that is a problem. It is worth noting that they originally planned on running trains from Redmond to Everett (along with trains from West Seattle to Everett). So this is nothing new. But it really only effects Everett Link and shouldn’t be that hard of a problem to solve.
Tom Tertific wrote:
“Why abandon East Main? Yes, the big development plans for the space between it and I-405 seem to be “on hold” right now, but it’s in an excellent location. Or did you mean “abandon the junction just south of the East Main station”?
I agree, as do most ST posters, that a connection at South Bellevue would be more useful for Issaquah riders; it serves Issaquah – Seattle riders much better than the double-back at East Main. The most useful version is a junction south of South Bellevue so that Line 4 can continue to downtown Bellevue (and onward to RTS, not “South Kirkland”). But Al’s idea of a very frequent automated Line 4 “shuttle” would work also.”
[Ed. Moved the comment to the right thread position as best I could twice the pasting/clicking.]
Thank you very much, Mike. You aced it.
RE East Main Station:
ST has not released any station boarding data for any month after February. We still don’t know what East Main station boardings are (or any other East Link station boardings are) after the cross-lake opening.
And it has taken well over a year for Shoreline stations to report more than weak ridership. So even if the data was available and was disappointing for East Main, it’s premature to give up on it.
That said, it’s a disaster as a transfer station with its track crossings for the side platform. Not only it it a huge hassle to cross two light rail tracks just to make a transfer, the crossing gates will be down most of the time because there will be a train on one track every 120 seconds for the current peak 8-minute peak service (assuming 4 line also has that frequency) or every 90 seconds for the desired 6-minute service.
Stupid. SMH.
I can tell you with certainty that East Main has little ridership. This will be exceedingly obvious for anyone who tries to use it. There’s a huge parking lot across the street on one side, and a park on the other side. Past the park is a long, uphill walk to get to anything.
This (no riders and no way in) is what the surrey downs neighbors wanted.
It still has more potential than s Bellevue.
There is pedestrian access from 111th, so the station access from Surrey Downs is in fact quite good. Could be rezoned for midrise and the station access would be fine.
Ridership at East Main today is irrelevant, given the 3 megadevelopments along the east side of 112th have yet to break ground. The new apartment tower on the west side (Broadstone?) is now leasing though, which I’ll point out is taller than anything allow adjacent to half of Seattle’s Link stations.
The East Main subarea has been fully rezoned for high-rise north of SE 6th and midrise south of it. I know for a fact there is a major project in development on one of the parcels zoned for high-rise. As to the property (Hilton) with the large parking lot immediately across from the East Main station, like any property it needs to make sense to develop it but there are already requirements in place to provide for pedestrian and bicycle connections through this large parcel and to/from the station whenever it is developed. The location of the East Main station is a major driver in this rezone and projects need to show how the station is to be accessed as a focal point for transportation. My guess is that once the first project is permitted and underway it will be a catalyst for others.
I’m hoping against hope that some Board member will add a requirement that ST develop a long-term plan to eventually use automated trains.
I’m also hoping against hope that the Board requires ST to design SODO station for level transfers if they go ahead with West Seattle Link.
Finally, I’m hoping against hope that the Board mandates that any train tracks in SODO running in the same direction be adjacent to each other and can easily change tracks if needed.
ST cannot use automated trains on the existing 1, 2 and 4 lines. This is the push several of us have been wanting the 3 line to be automated if done properly and with different vehicles.
Why not? What would be required to run automated trains on the existing lines?
Extreme liability associated with at-grade crossings.
Why is that a liability? Do you think it’s impossible to automate at grade train driving?
Why would it be any less safe than today’s situation? Most of the accidents are caused by a pedestrian or car putting themselves ahead of the train. A human operator would not be able to perform any better than an automated drive in most scenarios… And even if it could in some scenarios, automation would be far safer in many more scenarios
The liability is obvious. It’s all about risk and insurance, at least the USA. If a human driver fails, it’s the driver’s fault, and the profits of a liability lawsuit extend only as far as their personal wealth. If an automated system fails and its system-maker’s fault, the entire company is at risk. No one will take that liability and also allow the unpredictable chaos of at-grade crossings with mixed traffic.
“Why is that a liability? Do you think it’s impossible to automate at grade train driving?”
It’s one thing for it to be possible. It’s another thing to convince the ST board to vote to do it. Government agencies are highly timid of potential lawsuits, and won’t do things even if the liability seems small. You have to wait until many other agencies have done it, and even then it may take thirty years after that.
If an automated system fails and its system-maker’s fault, the entire company is at risk.
Sure, and yet Tesla is trying very hard to build automated cars, despite all the accidents. In the long run I wouldn’t rule out trains in mixed traffic but like automated cars they aren’t quite there yet.
It also doesn’t mean the agency won’t get sued right now. They could sue Sound Transit for creating a system that is unsafe. You aren’t going to sue the train operator — you sue ST. That’s where the money is.
With a private car the driver is liable. With a train or bus the agency is liable because it’s their employee or contractor doing their job.
Yes, but self-driving cars are supposed to be safer than the average American driver.
Meanwhile, it will be very hard to demonstrate that a self-driving transit system with mixed traffic crossings is safer than a human-operated system under the US’ liability laws. I’m excited for the day they do, but it won’t be any time soon.
“self-driving cars are supposed to be safer than the average American driver.”
Yet their performance is mixed. Better sometimes and in some situations, worse other times and in other situations. Some of those situations are so major that it makes people more hesitant on allowing self-driving vehicles beyond the current small-scale deployments (or even those). In that environment, the most risk-adverse organizations (public agencies), will be the last to adopt self-driving vehicles in mixed-traffic areas. And they might be willing to do one short BRT line before the entire Link network.
If a human operator crashes, ST is still somewhat liable. Do they make the operator pay up? I don’t think the difference is huge. The same argument has existed for Tesla and Waymo, but it works out. The agency just has to deal with it.
Transit agencies are already testing automated trams in Europe, I bet the same will happen in the U.S. When the Amtrak Cascades train derailed over the I-5 bridge, the driver had failed to limit the speed through the curve. Automated train control would have saved the train and its passengers.
Nathan’s right, SKR. Even though a person probably cannot react more quickly than can a computer in “ordinary” conditions, in a snow storm the human is going to out-perform every time. And that’s a time that the tracks are extra slippery. Even “with AI” can automation avoid questionable collisions in poor conditions?
More to the point, will juries think so……..
In my opinion we should automate the entire system, and remove any at grade segments and replace with elevated segments
Even still, we can and should automate at grade segments. It might even be safer than a human operator, because we have install cameras and other safety sensors to prevent accidents that a human operator wouldn’t be able to use reliably. It should reduce the number of accidents.
And if a human can drive it, there should be no reason a computer can’t. A train just goes ahead and there is a fixed schedule… It’s way less complex than a vehicle on the road. The only complication is dwelling time, but that can be resolved using fare gates that do not let people enter after a certain point of time, security personnel stopping door holding, and a buffer for minor speed adjustment to catch up on lost time caused by large crowds.
In my opinion we should automate the entire system, and remove any at grade segments
I agree. To be clear, it doesn’t necessarily have to be on a separate grade. There is nothing wrong with surface running — as long as it is fenced off and completely separate from traffic. It is only when you have crossing streets that you need to be a different grade. I’m pretty sure there are parts of SkyTrain that run on the surface.
Link is different. It would cost about a billion to bury the train lines in Rainier Valley. In SoDo you might be able to just add more overpasses. I’m not sure the best option for Bellevue.
Without grade separation in Rainier Valley you have other issues. They don’t want to run trains that often, as it could mess up cross-traffic. Thus automation might save money but it wouldn’t necessarily lead to a lot more frequent service. A branch for Renton seems unrealistic, for example.
If you bury the line, it solves that problem. You can run trains a lot more often. Not only that, but it would be faster and more consistent. I think it adds more value — even for those outside the area — then Tacoma Dome Link. Trips from the north end to the airport are better. So are trips from Federal Way to downtown. Thus someone from Tacoma might decide to transfer there and take Link to Seattle after all (instead of taking an express bus to downtown). Of course it would also be safer and change the nature of the street. You would have extra space to work with, which means you could add bike lanes, bus lanes and/or widen the sidewalk. It is a far more worthy project than most of the things we want to build. But since it doesn’t involve anything longer, it isn’t being seriously considered.
Automating the entire Link system is very long range. It would cost oodles of money on top of the ST3 oodles ST is trying to address now. ST doesn’t have money or voter approval for a huge project to retrofit the existing Link segments to automation. The board is not inclined to even think about it now, or to hold up ST3 extensions for it. Everett wants Everett now! Pierce wants Tacoma Dome now!
Note that burying Rainier Valley or having 100% overpasses in Rainier Valley and SODO is just one part of automation, not all of it.
<>
In the document linked by Ross, in Appendix B, the cost estimate to bury the tracks in Rainier Valley between Mt. Baker Station and Rainier Beach Station is about $10.7 billion. The estimate to install an elevated guideway would be about $1.7 billion.
Let’s end all talk of burying the tracks.
The $10B option is for a deep-bore tunnel. The paper estimates $1.1B (2023$) for an open trench design, but that likely doesn’t include any relocated utilities or overpasses, so the total for a cut-and-partial-cover option would likely be a few times higher.
We should use trenches in a lot more situations. Even in downtowns. Can make cool pedestrian spaces and bridges over it. Easy access to stations, unlike the abomination we like to call UW station. Tunnels and subways are overrated unless you just have no way to fit a trench or elevated line (which unfortunately is the case in many cities that were built without proper transportation planning). Tunnels also make sense for higher speed bypasses.
All of the studied options assume that the replacement construction work would be in the middle of MLK. That would obviously curtail operation of the 1 Line for several years. But could an elevated guideway could be built on the eastern side of MLK that allowed the 1 Line to continue operations while the new guideway is being built?
The 2 northbound traffic lanes would become the base for the elevated guideway. During construction the 2 southbound lanes would be converted to one NB lane and one SB lane until construction was completed. Then the existing surface tracks would be ripped out to create 2 NB lanes and 2 SB lanes while the area under the guideway could be developed into bike lanes.
The paper estimates $1.1B (2023$) for an open trench design, but that likely doesn’t include any relocated utilities or overpasses
I think it includes all of it, given the detailed description of the work. For example, there is a “Utilities” section. There is also a “Structural” section:
There are 18 existing surface streets which cross the current alignment along with 10 pedestrian‑only crossings. The approximate average distance between these 28 crossings is 700 feet. Slab girder bridges would span over the open trench at each required crossing allowing vehicle and pedestrian traffic to cross at the current grade elevation. Figure 10 indicates the open trench section.
So yeah, about $1 billion for the construction. Of course that is an estimate (there is a 30% contingency) but it is a detailed estimate, not the wild guess that drove so much of ST3 planning (and why things are such a mess right now). The cost doesn’t include the extra bus service or mitigation during construction. But that likely wouldn’t add up to that much (a few hundred million, if that). As for the transit impact during construction:
The construction would be phased into four sections and be from station to station as shown in Table 1. Temporary bus bridge service would be provided to allow passengers to bypass the construction and continue service on the light rail line. Construction phasing alternatives for maintaining a single track in service during construction could be warranted. This would include studying constructing the trench in halves, or as a whole with a shifted alignment, however substantial challenges would be anticipated.
So at worse they have a bus bridge (something that is unfortunately quite common within our system now). But it is possible they could do some single tracking, although it would likely add to the cost (and they haven’t studied that option in detail).
> I think it includes all of it, given the detailed description of the work.
I do these sorts of rough-order-of-magnitude project cost estimates professionally (the table in Appendix B is a very common style), and engineers are careful to specifically list what’s included. It’s certainly unclear what exactly is included in their assumed cost per linear foot for trenching, but given that they explicitly state assumptions regarding sidewalls and sump, it’s not safe to assume costs for utility relocation and/or overpass construction are implicitly included.
My guess is that they wanted to include costs for overpasses and such, but it turned out to be too much effort to make them reliable. I think utility relocation is a much more complicated can of worms with a lot more risks than most people realize, especially on a major utility corridor like MLK. I’d bet the construction project managers for the MLK segment of Central Link have some wild horror stories.
I agree with Nathan’s analysis. At least five billion for the open trench by the time all the i’s are dotted and the t’s are crossed.
The Seattle Utilities map shows a sanitary sewer under EACH side of MLK in almost every block between Mt. Baker and Rainier Beach. So there’s no place to dig the trench without moving the sewer and re-connecting all the houses and tributary streets.
I don’t know why there are three or four blocks without one on the east side, but there are. Maybe there are no houses there.
I think this presents a really interesting “would you rather” question for transit advocates: with $4-5B, would you rather build WSLE with no Avalon station, no tail tracks at Alaska Junction (making future expansion extremely impracticable), and DSTT-style transfers at SODO, or would you would you put the MLK section of Central Link in a trench where it could run run at 55 mph with no traffic lights?
Of course, WSLE is a forgone conclusion and MLK won’t be grade-separated for decades (if ever), but it’s a fun hypothetical.
The Seattle Utilities map shows a sanitary sewer under EACH side of MLK in almost every block between Mt. Baker and Rainier Beach. So there’s no place to dig the trench without moving the sewer and re-connecting all the houses and tributary streets.
The report mentions that!
Look, I get it. We are used to half-ass estimates for things like Ballard and West Seattle Link. But this is a detailed plan that includes everything. There is an entire section dedicated to the sewer pipe that gives me more information than I every wanted to know about the sewage system there:
The major utilities within the corridor would need to be relocated before the open trench structure is constructed, including an existing 60-inch sanitary sewer line that runs from approximately Holly to Henderson. As-built of the central link light rail line were reviewed, and the following major utilities would need to be relocated:
• 60-inch Pipe Sewer (runs adjacent to tracks and may need to be relocated)
• 60-inch Sanitary Sewer (runs under the tracks and will need to be relocated)
• 42-inch Storm drain
• 2 – 66-inch Cedar River water transmission pipelines (72 inch under tracks)
• 51.5-inch Cedar River water transmission pipeline (60 inch under tracks)
That is all part of the estimate which is why it is a billion and not a lot cheaper. And look, there are questions. These are laid out quite clearly:
High water tables may influence the permanent solution, and whether vertical elements or tie‑downs are necessary to resist buoyancy. At the surface street crossings, the local geology would need to be evaluated to determine if deep foundations are needed to support the highway loadings from the slab girder crossings. Waterproofing the open trench would require evaluation with construction methodology and water table.
This is where the problems could occur. But it isn’t going to cost five times the original estimate. There just aren’t that many crossings. It also sounds like the next step is to determine the water table and local geology (at various places). This shouldn’t be that expensive which means they could get a better idea if this really will be a problem or not without jumping in head first.
If Ballard-Westlake is separate it can be automated now, and we can have the debate about retrofitting Rainier Valley, SODO, and Bel-Red later. That would be better than having no automation at all, because of the way it allows frequency to dramatically increase without increasing costs.
If Ballard-Westlake is separate it can be automated now, and we can have the debate about retrofitting Rainier Valley, SODO, and Bel-Red later.
Exactly. That is one of the big benefits of building Ballard-Westlake: You defer a lot of debates. This includes things that we have no plans for funding (like the burying of the line in Rainier Valley). But it also includes things that were part of ST3 that were never really resolved, like the CID Station. Or poor design decisions, like the SoDo station. Or budgetary problems that have led to draconian proposals (like removing South Lake Union and Avalon stations). All that mess is deferred. You focus on Ballard to Westlake, but with smaller platforms and trains coming more often. There are still unresolved issues in Ballard but at least you commit to going there.
ST has summarily denied that automated technology even exists. I’m simply hoping for a study to define how phasing and costs would or could work, along with what cost savings would be achieved.
Phasing may be that some sections of an existing lines remain operated by human drivers while others could be fully automated. With new cross-platform transfers designed into the system it could be possible to have an automated train waiting to quickly whisk riders away on what would otherwise be a segment with such low use that ST would find it very expensive to introduce or keep operating. Keep in mind that half of the travel time to Tacoma Dome from CID is KDM and half of the travel time from Westlake to Downtown Everett is Alderwood. Maintaining drivers and running long trains mostly empty for unbuilt, long stretches is a significant operating cost.
And automated trains are designed to have drivers when needed, so as long as a track connection is provided it could still access existing OMFs with a driver.
Someday all this automation denial mentality will look at short-sighted as automated check-out stations in groceries or automated elevators in buildings or self-driving cars. It’s going to evolve — probably in phases — over the next few decades.
Things look unlikely until they reach critical mass, or one little public statement or news item has an unexpected outsized effect, and then it turns around quickly.
It is bizarre to me. We already have working examples around the world including SkyTrain right nearby.
Automating a train is one of the easiest things to do. The complexity of the problem and variability of the automation is extremely low compared to the stuff we’re automating right now. The algorithms to automate a train has existed for decades now. Novice programmers could easily write it. It doesn’t even need AI if everything was grade separated… Because it’s literally following a fixed path and schedule.
It’s US myopia and Anglophone myopia. Pedestrian Observations (Alon Levy’s blog) blog has articles about this. American agencies/politicians are so stuck in the US way of doing things that they won’t look at what succeeds outside the US and how they’ve solved problems we have. Or if they do look at it, it’s just a cursory glance and dismissal. ST3’s and SDOT’s marketing materials early in a concept/study have photos of non-US solution/technology/policy alternatives, but then they’re immediately dismissed without further study or consideration, and instead we hear excuses that they don’t exist or can’t work here or would be too expensive.
At the larger Anglophone level, American/Canadian/British/Australian agencies are reluctant to consider solutions from non-English-majority countries. Partly because they can’t read the reports, partly because they don’t bother to look and see that English translations are available from the same agency, and partly because of cultural bias. Obviously Canada has less of that than the US does, but Levy argues it has more of it than continental European, Asian, or Latin American agencies have to not look worldwide.
This is more or less a “Not Invented Here” syndrome. This combines with American exceptionalism and car/highway/airport bias.
I’ve always found it odd that automation is understood by many as something done for an entire line at once. That’s not the case at all!
It is possible to phase it in with cross platform transfers. Yes riders have to change trains but the walk is fast and minimal. The transfer station only needs one automated track — because the trains would be automated! It could replace an existing track or be installed as an adjacent one.
And automating outer sections first would be better for the system operating and capital cost. Station boardings are lower. Stations are further apart so a faster moving speed could be achieved. Platforms would be smaller and safer with platform doors like an elevator. Automated locating would enable a single end station platform at the transfer point. Unlike urban sections like MLK, automation could even make a low demand extensions pencil out as operating costs should drop because drivers would not be needed.
Of course we are all speculating here. But most of the firms with existing contracts with ST have automated line experts within them. ST just needs to get real and explore what its advantages and costs are — and of course not produce a flippant report summarily trashing the idea without sincere analysis.
You don’t have to force a change of trains. Operators can board a train headed for non-automated territory a couple of stations before the end of automation, enter the cabin and prepare to assume control at the last automated station.
The opposite happens on a train entering automated territory except the operator lights at the same station at whichbthey board.
“Nathan’s right, SKR. Even though a person probably cannot react more quickly than can a computer in “ordinary” conditions, in a snow storm the human is going to out-perform every time. And that’s a time that the tracks are extra slippery. Even “with AI” can automation avoid questionable collisions in poor conditions?”
Computers / AI can be trained quite effectively even for these scenarios using reinforcement learning to handle complex conditions. Essentially, the AI can be trained on the human operator to mimic it. This may be less effective during extreme conditions, so ST can mandate human operators to take over or assist in operations.
Even after automating, we’d still keep attendants on board and they could even take over for manual operation if absolutely necessary. It’s like a plane’s autopilot. But the benefit of automation is that the trains will adhere more strictly to the schedule, and the trips can go longer without the attendants getting “tired” from driving.
How do we train humans on what to do? We give them *guidance* and rules. Those are quite easy to program into a computer. But for things that require judgement, which is quite rare, an AI system can generally replicate it quite well since the number of scenarios are not too high. Muscle memory related situation related to train physics and such are even easier to train onto an AI system. I’m sure Asia and Europe are training models on a variety of climates, environments, and weather conditions…things we could take a look into adapting…if we cared to go beyond the status quo, that is.
As for the liability issue, that’s going to be a problem with automation going forward. That shouldn’t be a blocker in adopting the technology.
“King County Metro and Sound Transit face direct liability when their employees cause accidents through negligence. ” Liability is already an issue even with human operators.
The current liability argument: “King County Metro or Sound Transit, by poor training or supervision, or negligent hiring”
> Even after automating, we’d still keep attendants on board and they could even take over for manual operation if absolutely necessary. It’s like a plane’s autopilot. But the benefit of automation is that the trains will adhere more strictly to the schedule, and the trips can go longer without the attendants getting “tired” from driving.
Spot the contradiction!
Also not sure what makes you think a human would perform well in snowy or poor visibility conditions 🤔 A computer can actually see even more, if equipped with the right sensors and cameras. There can also be checkpoint cameras at various intersections / crosswalks, which would signal to the computer far earlier than a human operator could see + much faster reaction time
The only “concern” is potential glitches or bugs with the system, that could cause it to fail… or potentially hacking. This is why I would think we need an attendant on board, who is trained and mandated to be fully attentive and prepared during segments like MLK without grade separation.
“SKR’s”. Apologies.
Ah the Seattle Process strikes again. You want gadgetbahn.
Also, extend the 1 line for now upto the Alderwood mall which is only one stop north, to increase footfall and ticket revenues.
Don’t change the 1-line that goes to the airport and connect it to West Seattle ! Does your survey support that ? Both Redmond & Lynnwood need direct connections to the airport.
The 2 line should do Redmond-Ballard.
Does your survey support that?
I don’t think we have a survey. But your point about the 1-line is a good one. The amendments proposed by Strauss would do that. The other lines would remain the same. By deferring the second tunnel we defer everything else that goes with it. That is another argument in support of this idea.
Extending up to Alderwood is really a different subject. Strauss doesn’t get into what is happening outside the city (for good reason). That being said, any extension is more likely if we defer the second tunnel (as mentioned in the other comment).
There’s no survey. There’s:
1. 1990s-2015: ST3’s early visions, which had Everett-Tacoma Dome and Ballard-West Seattle.
2. December 2015: DSTT2 and the split spine. ST said Everett-Tacoma Dome was too far for drivers to go without a break, so it suggested Everett-West Seattle and Ballard-Tacoma Dome. DSTT2 and SLU had gotten mixed into this by April 2016, and all this was in the ballot measure in November 2016.
3. 2022: A budget gap emerged (ST3 couldn’t be finished by 2041 under the debt ceiling), so the board reordered projects to finish its highest priorities by 2039. (The “realignment”.)
4. 2025: A larger budget gap has emerged.
5. December 2025: ST considers two scenarios that would eliminate DSTT2 to save costs, and upgrade DSTT1 to compensate. One has Ballard-Westlake like Strauss’s amendment. The other merges Ballard into the tunnel at Symphony (bypassing Westlake). 14/17 of the boardmembers are against both, while 3/17 have varying positions that support at least keeping the first option open. (Strauss; Ballard! Balducci: DSTT1 needs upgrades anyway, so let’s keep this option open and maybe do both sets of DSTT1 upgrades at the same time. Wilson: “Build the damn trains!” but [I guess] keep the option open.)
5. March 2026: ST has a board retreat on the budget gap, where three scenarios are outlined to illustrate potential changes. All keep the split spine and DSTT2, and all truncate Ballard in the first phase.
6. May 2026: ST presents a draft resolution to modify ST3, and two Executive Committee sessions debate it.
7. Saturday: Strauss announces his amendments.
8. Tuesday: All amendment proposals will be announced.
9. Thursday: The board will debate the resolution and amendments, and may vote on all of them. The board wants to decide it by the end of May, so that it can start construction on the West Seattle-SODO stub in June. (Which many STBers think is a bad project, but that’s another issue.)
So the split spine with no Lynnwood-airport one-seat ride has been part of the plan since step 2 in December 2015. There has never been a Redmond-airport proposal. There have been unofficial suggestions for Ballard-Redmond or West Seattle-Redmond, but no ST proposals.
The issue with Lynnwood-airport transfers and Redmond-airport transfers is the excessively long (10 minute, several level) transfer walks between the two tunnels. This was not anticipated in the ST3 vote, by either ST or voters. It was assumed DSTT2 would be the same depth as DSTT1, and CID2 station would be next to CID1 station at the same level. But later ST3 engineering studies claimed DSTT2 would have to be much deeper, thus the long transfers. And ST moved CID2 and Midtown stations to CID/N and CID/S, which exacerbated the transfer problem. So ST should say, “These excessively-long transfers are unacceptable”, and take a step back, and either try again with other tunnel options, or an automated line that can have a smaller tunnel maybe closer to the DSTT1 stations, or cancel DSTT2 and upgrade DSTT1 for three lines (as a candidate project in December 2015 would have done).
So there never has been an ST survey on these exact issues, because ST’s surveys are all “Do you like our proposal?” and “How many ways is our proposal wonderful?” So the only thing to do now is to try to influence your ST boardmembers, and influence city/county officials (because they have the most clout with the ST board), send feedback to the board, and get tens of thousands of your neighbors to do the same with a common message — all before Thursday.
To be fair, SODO is by far the easiest planned transfer point. And with a modest design change ST could still even design and build cross platform or level transferring there. It would significantly improve the impact of severing the spine.
It just doesn’t connect with the 2 Line. It’s why Eastside interests who able to understand station plans from a rider experience perspective (like Balducci) are more outraged about the terrible transfer designs Downtown.
I’m not addressing SODO transfers because I don’t have an intuitive understanding of what the passenger experience of each of the alternatives would be or how much it matters. And I hate transferring from Link to the 50 because that area is a depressing concrete jungle middle of nowhere. I fear a Link-to-Link transfer there would be just as depressing, and a lot of people wouldn’t do it or wouldn’t be happy about it for that reason. So I leave it to the SODO-transfer advocates to make their case, while I stay out of the way one way or the other.
A major benefit of cross platform transfers is that they can be timed! It happens worldwide. For example, it happens at BART’s MacArthur Station and has for over 40 years. Because it’s well under a 30 second walk people don’t even see the surroundings. MacArthur is in the middle of a giant freeway to freeway interchange approach (16 lanes of freeway traffic in total there) in a neighborhood that was considered pretty sketchy until recently.
If an area looks undesirable I’d rather walk across a platform quickly to the shiny next train in front of me than ride two escalators up and down inside a glass palace — and see the ugliness all around me for several minutes while making the trek.
And at least SODO transferring riders would get daylight. The tunnels at the other Downtown transfer stations are designed to be long and narrow shafts with only artificial light. And take several minutes to trek through.
Seattle Subway endorses #1 and #3, has a petition to sign (same link), and urges people to contact their ST boardmembers starting now.
Wow! This is huge. I’m going to sign it and encourage others to sign it. I don’t always agree with Seattle Subway (to put it mildly) but I think they are completely right in this case. There is a subtle suggestion there that is worth noting:
They could find further savings by building a service track connection to the existing spine instead of a new maintenance facility.
This is something we’ve discussed here, but I’m not sure if ST actually studied it. The report of the study is here. They looked at two ways of interlining. Either at-grade (Alternative 1A) or grade-separate (Alternative 1B). They also looked at a stub line with a new OMF facility. This would be a variation of Alternative 1A, but there would be a non-service connection between the two lines. This could lead to significant savings and less disruption. I have a feeling it would be similar (and a new OMF would prove to be the best option) but I applaud Seattle Subway for bringing up this issue. Once you commit to a stub line you can explore both options in more detail.
It also means that the line can be automated. In contrast, automation is problematic if you interline. The limiting factor becomes the tunnel (you can’t run trains that often through there) This defeats one of the key advantages of automation (much better headways). The Kubly-Reed plan falls apart. But this isn’t an issue with a non-service connection. The connection is only used when trains are out of service.
I signed it of course. It’s compatible with what I want, even if some other things Seattle Subway wants are the opposite of what I want. (And I’m ignoring the fact that “completing Ballard” is ambiguous on whether it includes DSTT2. Better to get one win than to get three losses.)
My public testimony is forming in my head.
1. Prioritize Ballard (amendment ####): Ballard and SLU is where the ridership is.
2. Automate the Ballard line, or at least give it real fair consideration. Automation is the international standard now (Hawaii, Vancouver, Paris), would serve passengers better and increase ridership and capacity by allowing 2-minute frequency without higher cost, and would probably lower capital costs substantially.
3. Cancel DSTT2, because those proposed excessively-long transfer walks are passenger-hostile. Good train-to-train transfers should be the first consideration in any multi-line subway, because half or more of the destinations require the transfer. If somebody must have bad transfers, it’s better to isolate that to Ballard/SLU riders who don’t have Link now than to impact everybody going between vast parts of the region, including those who currently have a one-seat ride (north-south) or a short transfer (Eastside-airport).
Eventually I think we need a new tunnel, but that would be for making a quality tunnel designed to actually operate a full subway system. The old trash tunnel can be repurposed as a bus tunnel, like it was before.
Yes, this is where I am landing as well!
I agree Mike. I would word it differently, though. I would use “defer” for DSTT2. I think a second tunnel — with a station in First Hill — would add value. Even if it went to West Seattle (instead of this — https://seattletransitblog.com/wp-content/uploads/2025/12/WSBLE-HackImage.png) it would still be a huge improvement. I think West Seattle Link is a big waste of money but I see no reason to rule it out. Simply defer it. It is possible it would be a better value in the future.
Thus I see one more thing I would add:
4) Design the new line so that it can be extended south from Westlake and east from Ballard (to the UW).
YES!!! Build the Snake! It breaks a lot of “rules”, but there’s a lake and a mountain right in the middle of Seattle north of downtown that make a “grid” not just “difficult”, but impossible. So stringing together a reverse “C” routing between Downtiwn Seattle and the U-District makes a LOT of tricky, three-seat rides on infrequent buses into slick, frequent rail-backbone journeys.
Fine, cancel, defer, anything that allows cooler heads in the future to do something better, and that doesn’t create expensive concrete facts on the ground now that would be hard to abandon/reverse.
I admire Strauss for trying, but everything I’ve read said that every single board member outside of Seattle is expected to vote against it.
If you try for something, you might get it. If you don’t try, you definitely won’t.
Even if it fails, it’s worth documenting what it could have been, and how far the actual network is from it.
I would like to see the Alderwood connection now. I personally would use Everett and tacoma link more than Ballard link.
The extensions outside the city are largely independent of what happens within Seattle. That being said, the other subareas are supposed to help pay for the second tunnel. If it is deferred then paying for it is deferred. This could result in those extensions being done sooner. So yeah, that is another argument in support of this idea. This argument is bound to have more weight with representatives outside Seattle. If you live outside the city I would emphasize that when contacting your local board member.
“This argument is bound to have more weight with representatives outside Seattle.”
Theoretically yes, but boardmember inclinations so far are the opposite.
This raises a potential article idea we’ve been kicking around: Who is DSTT2 for? Who is the primary beneficiary? Is there a mismatch between beneficiary and cost share? It’s hard to say who’s the primary beneficiary because there are multiple competing arguments:
A. Everybody: Downtown Seattle’s total north-south transit demand will exceed supply on all modes (Link, RapidRide, regular buses) for all trips (to downtown, through downtown, within downtown). A report ca. 2012 said that, and that and the need for more SLU capacity was the impetus for splitting RapidRide C/D, six more RapidRide lines (since reduced), and the second tunnel.
B. Ballard: Ballard Link wouldn’t be possible without DSTT2. (Or the DSTT1 retrofit the board deselected in 2016.) Ballard-Westlake started as a separate project, but the Ballard-DSTT2-West Seattle EIS strung them all together into one, and then West Seattle was split off so it could start earlier, so Ballard/DSTT2 are together, so therefore DSTT2 benefits Ballard (supposedly).
C. West Seattle: West Seattle Link wouldn’t be possible without diverting Rainier/FW/Tacoma to another tunnel (with Ballard). (Or the DSTT1 retrofit the board deselected in 2016.)
D. Everett and Tacoma: A second tunnel allows for the ridership increases these extensions will bring.
It seems logical that DSTT2 is in North King, therefore North King is the primary beneficiary (particularly Ballard). But paradoxically, it’s the North King boardmembers who are questioning it (to varying extents), and the Snohomish/Pierce boardmembers who are most for it. Does that mean they’re actually the primary beneficiaries? Or just that they’re deluded about their ridership prospects and how much their extensions will help their mobility and mode share?
The reason Somers/Snohomish is so against non-DSTT2 strategies is not so much that he thinks DSTT2 is wonderful and absolutely necessary, but he’s afraid that making changes like Ballard-Westlake or no DSTT2 will lead to unexpected cost overruns and delays that might jeopardize Everett/Paine from being finished on time or at all. And Pierce is worried about the same for Tacoma Dome, to a lesser extent, or can be convinced of it by Snohomish. (This is based on my interpretation of Somers’s statements in board meetings and newspapers.) Everett/Paine must be finished on time, whatever happens in Seattle. If he can be reassured that some Seattle change won’t impact Everett’s timeline, he’s willing to consider it, such as his endorsement of advancing Ballard’s design to 100% even if construction is on hold.
“Downtown Seattle’s total north-south transit demand will exceed supply on all modes (Link, RapidRide, regular buses) for all trips (to downtown, through downtown, within downtown). A report ca. 2012 said that, and that and the need for more SLU capacity was the impetus for splitting RapidRide C/D, six more RapidRide lines (since reduced), and the second tunnel.”
I’ve not seen this report. Where is it?
And it’s the planned Link segment between Westlake and Ballard that eases SLU overcrowding. The DSTT2 segment does not. Overcrowding south of Westlake is a separate topic.
“I’ve not seen this report. Where is it?”
It’s back in the early 2010s; I don’t remember exactly. We had articles mentioning it at the time, during the debates on splitting the C/D and adding DSTT2. I don’t remember if we had an article dedicated to it, and I’m not great at searching for such vague things without obvious keywords. The Downtown Seattle Association also weighed in with concepts to reduce 3rd Avenue’s lanes and beautify it; I don’t remember whether those articles mentioned the north-south capacity issue.
Can we all agree that the report is flat wrong? ST ridership on segments already open in 2016 isn’t yet back to pre-Pandemic levels. Total system ridership is up a bit, mostly from Lynnwood Link’s forced transfers.
Somebody needs to point out that the “estimates” are balderdash.
“And it’s the planned Link segment between Westlake and Ballard that eases SLU overcrowding. The DSTT2 segment does not. Overcrowding south of Westlake is a separate topic.”
The issue isn’t SLU Link crowding. That has never been suggested. The issue is Link crowding between Westlake-Symphony or CID-Pioneer Square (as ST maintains) or Westlake-Capitol Hill(-UW) (as many STBers believe). DSTT2 or changes to it could improve the Westlake-CID segment, but would do nothing for Westlake-Capitol Hill(-UW). So it depends on which crowding scenario you believe is most likely.
I wouldn’t call the general downtown-SLU capacity issue “crowding” because it’s not about a single route; it’s a more diffuse issue. The report (as I understood it) implied Westlake-CID area crowding; it didn’t address SLU. Or at least it didn’t explicity say so, and SLU wasn’t considered “downtown” as much then. Downtown was Weller Street to Stewart Street; everybody agreed on that in the ST/transit context. Views differ on whether downtown extends north to Denny Way, or stops at Yesler Way, or now extends to Valley Street, but those aren’t gemane to the ST/transit issue, and were not brought up during the north-south capacity/DSTT2/C/D debates.
Keep in mind that in 2012, DSTT was assumed to carry only 16 trains an hour in each direction (8 trains an hour for each line). The DSTT capacity was assumed to be markedly lower than what it could be. Even ST3 promised 20 trains per hour per direction.
Even the WSLE FEIS made this assumption.
It’s the little math trick that ST staff uses to demonstrate the overcrowding on paper.
Who is DSTT2 for?
It is intended to provide West Seattle with a direct line to Seattle without impacting the capacity of the other lines or requiring work on the existing tunnel(s). If there was no West Seattle Link then there would be no second tunnel. With three lines from the south going through downtown things get a bit messy. ST doesn’t want to run trains more than every three minutes. This works out nicely to trains every six minutes from the south and east. In such a scenario you might need service there more than through downtown. But if you start running trains from West Seattle then you have to run trains every nine minutes to the other locations or you “wouldn’t give our ridership as reliable a service”.
From a Ballard Link perspective, the second downtown tunnel is a “nice to have”. Westlake is basically the center of downtown right now. A lot of riders will have to transfer there anyway (e. g. UW to Seattle Center). The second tunnel certainly adds value, but given the stations are basically a subset of those in the existing tunnel, the benefit is minimal. Riders would get more out of the proposed South Lake Union station (a station they are ready to eliminate) or including Smith Cove, Interbay and Ballard (stations they are ready to defer).
While the second tunnel doesn’t negatively impact the capacity through downtown, it does impact riders. The existing One Line pairs riders from the north to the south. This is ideal. This will be broken, and riders will be forced to transfer. The SoDo Busway will go away. A lot of riders will see a significant degradation.
If he can be reassured that some Seattle change won’t impact Everett’s timeline, he’s willing to consider it, such as his endorsement of advancing Ballard’s design to 100% even if construction is on hold.
That is what needs to be done. We need to make sure that everyone outside Seattle understands how the Strauss amendments would actually speed up their projects. West Seattle Link (WSL) is supposed to start before Tacoma Dome Link. If WSL is deferred, then there is more money to spend on Tacoma Dome Link, allowing them to perhaps break ground sooner. You also have a fundamentally cheaper project. The same is true for Everett Link. Smaller projects in Seattle that start later benefit everyone else. It is a fundamentally cheaper project — this helps everyone in the short and long run. It also happens to be the most cost-effective section, which again helps everyone.
In the long run it makes sense to build the second tunnel with West Seattle Link. For now it makes sense to build the best part first.
What’s the actual impact of passing on West Seattle now, in terms of the inflation argument, and when is the soonest that Ballard can be in construction given the design and environmental permitting process?
I don’t have quantitative information on that, and we don’t know what inflation or other factors will do in the future. You’d need an engineer and a financial analysist who’s been deeply immersed in ST issues to answer this, but here are some factors:
A. The West Seattle-SODO EIS is approved, so ST just has some small intermediate issues and then it can sign construction contracts. It doesn’t have legal authority to sign the contracts until it resolves the ST3 budget gap, thus why it’s trying to rush through the ST3 plan changes by Sunday (May 31).
B. The Ballard/DSTT2 EIS needs major revisions because ST added alternatives after the alternatives analysis and initial work was done. ST moved Midtown station from Madison to James, CID2 station from Jackson to south of Dearborn, and added alternatives in the SLU area (affecting the alignment, Denny station, SLU station, and the approach to Seattle Center). A first draft of revisions is expected this summer. Then it would take a year or two to finalize, unknown time/certainty to get approved by the Trump-era FTA, a year or two to reach 100% design, and some months for construction to start. The assumed construction period was 2032-2037, so that’s four years, but it would likely be longer, maybe ten years or more.
C. If you split Ballard-Westlake from Ballard-DSTT2, then it inherits all the issues north of Westlake, and modifying the Westlake1-to-Westlake2 station interface. Automation would add more time to study. So it might need 2 years for the EIS, 2 years for 100% design, and 4-6 years for construction. That’s my guess. Others with more optimistic assumptions might guess less. You could subtract the cost and staff time for DSTT2 if it’s not pursued, but re-add costs to retrofit DSTT1 to add West Seattle, and it would affect the subarea contribution mix. Right now North King pays 51% of Ballard/DSTT2. If Ballard is split off, the other subareas’ contributions might disappear or shrink. Strauss is also arguing that all subareas should pay north to Seattle Center (not just to Westlake), because all subareas use and benefit from SLU and Seattle Center so much. The other boardmembers haven’t accepted this so far.
It is tough to say. It may actually be easier to build things later. That is because of the debt limit. We can’t spend too much too soon (https://seattletransitblog.com/2018/02/28/sound-transits-debt/). As you delay projects you get to raise more money. That is why it is common to see delays in projects as they get more expensive. But some projects get more expensive over time, negating that advantage. This can be caused by construction costs rising faster than we can raise money. But quite often it is simply different estimates. What we thought would cost a certain amount is just more expensive. Inflation has nothing to do with it. There is really no reason to assume that construction costs will rise faster than the money we are raising (which is also going up with inflation).
I don’t see much advantage in starting West Seattle Link. Not unless they plan on interlining it with the main line. If the board continues to oppose interlining, then West Seattle Link is highly depending on the new tunnel. A line from SoDo to West Seattle adds very little. With West Seattle Link dependent on the new tunnel, it really doesn’t matter that the West Seattle section is done first. You have to build the tunnel. This will likely take a long time. Not only because it is expensive, but costs are likely to rise as the design gets finalized. There is no consensus on the stations. Generally speaking, the more controversy there is, the more expensive a project becomes.
As for Ballard, they assume it will start two years from now. I don’t see that changing much if it ends at Westlake. They would have to modify the environmental evaluation, but that seems minimal. ST assumes it will take two years to modify or write a supplemental EIS. That seems way too high, given they are simply ending at a different place. The only significant difference is the tunneling work but even that change is fairly minimal.
So while it is difficult to tell for sure, it is highly likely that a delay for initial ground breaking would actually result in some parts of it (e. g. Seattle Center) getting build sooner. It would definitely result in Ballard getting a station much sooner.
“It may actually be easier to build things later. That is because of the debt limit.”
And a possible future Democratic administration and more pro-transit Congress. Congress could increase transit grants (with or without reducing highway grants), the FTA could disperse all the transit grant money Congress has allocated (rather than suppressing it, diverting it to other purposes, or arbitrarily firing the experts/staff who would administer it), congress/FTA could relax the “Buy America” provision by making it easier to get transit technology that’s not available in the US because the US transit market is too small/intermittent for American companies or foreign subsidiaries to set up factories for it in the US, congress/FTA could streamline regulations that make transit projects more expensive in the US than in other countries, etc.
It was fortuitous that ST2 passed in 2008, simultaneously with Obama’s election. FTA provided lots of financing that really helped in paying for the projects in those 8 years — both in grants and in TIFIA loans. At least 25% and as much as over 50% is traceable to FTA help.
(Does payback of TIFIA loans add to the current ST fund shortage? I’ve wondered whether this a limitation that’s not discussed.)
It seems to me that the Board may be tacitly dragging out ST3 until we inevitably get another pro-transit Federal administration. Getting projects shovel ready is quite strategic for that reason. That element of the current ST motion seems very wise.
By obsessing about funding and capital cost, it’s easy to overlook the two main structural problems I see remaining with ST3:
1. Awful transfers created for not only new rail riders but also existing ones.
2. Low productivity extensions and vehicle technology choices that will drain operating funds for decades .
Even if FTA gave away billions in capital funding to complete ST3, these two structural problems would remain.
So no matter what the Board dies this week, we should continue to encourage a wider consensus to resolve these two basic problems.
And of the two, getting transfers right is the most detrimental. Automated technology can be introduced on even existing lines and track profiles as it evolves. Service adjustments can be made to turn back trains that don’t get many riders. But bad transfers are a cruel and unnecessary 100+-year curse on the region’s transit riders.
I’m highly skeptical that ST will be the funds to build DSTT2 all the way to Seattle Center.
WSLE is listed as $6.2B for one deep subway station. Reaching Seattle Center would require at least five. Five stations in the middle of a congested Downtown area. There’s only $17.8B estimated in the resolution.
The boring is planned to begin at Holgate St through soft SODO soils too — almost a mile south of Jackson St.
And let’s not even discuss what happens to a TBM underneath Seattle Center.
I don’t know about needing five stations. West Seattle offers a precedent: drop stations to save money.
I’d say it’d be crazy to drop, for example, Westlake, but on the other hand, that’s probably the most expensive station in the entire system. I would have thought dropping Chinatown would be crazy, but then Constantine went and basically did that.
Actually there are six. I’m just anticipating that one in SLU will be eliminated.
West Seattle offers a precedent: drop stations to save money.
The precedent is First Hill. Generally speaking, dropping stations is stupid. Once in a blue moon it makes economic sense to defer a station. You build things out of order to save a little bit of money. That was the case with 130th (now Pinehurst Station). It should have been built at the same time as the other stations. But by building it later they saved money (at the time). The same thing is true with Graham Street Station. But at some point you need to build those stations. Graham Street Station is the most cost effective project left in ST3. Let that sink in. All of these projects involving bus lanes and miles of new track and yet a station that was not strong enough to warrant initial inclusion is a much better value. This shouldn’t be surprising, given the fundamentals of transit design, such as:
1) The longer the line, the more expensive it is to operate and maintain.
2) In most cases, adding a station is a lot cheaper than extending a line.
3) People take more short trips than long trips.
4) There is a network effect as you add stations.
5) Creating a complementary grid is a lot more difficult when you skip stations.
Of course there are exceptions. Sometimes an area is so undeveloped and lacking in potential that it doesn’t warrant a station. Some areas don’t have crossing streets so they can’t have crossing bus service. Stations deep in the suburbs tend to have very little in the way of travel between stations, thus negating the network effect. But none of that applies to First Hill or Graham Street.
Nor does it apply to the stations they are planning to omit. Avalon is similar to the Alaska Junction. It has plenty of apartments (although it doesn’t have as much retail). It is like Othello while Alaska Junction is like Columbia city. Omitting the station makes the transit grid worse. High Point is the most densely populated part of West Seattle. Riders take the 21 to downtown (via 35th). It is a straight shot that eventually goes on Avalon (right by a potential station). It is pretty easy to imagine the bus continuing north on either Admiral Way or Harbor Avenue. It is not clear what the bus will do instead but the options are more limited.
The same thing is true with South Lake Union Station. It is not perfect but it is still in one of the most urban, centralized areas in the state. If the station is omitted then ridership from every other station will be worse. It was also designed as an intercept for Aurora buses. Without a station there, riders from the RapidRide E, 5 or 28 will just ignore Ballard Link. They will take the Metro 8 to the Seattle Center/Uptown area. Or they will just drive. Skipping stations is penny wise and pound foolish. It is a way that the agency can brag about serving distant or well known locations while serving them poorly. It is symbolic transit.
Here’s a curious fact:
West Seattle is no longer a line in FTA’s CIG list. The latest list is here:
https://www.transit.dot.gov/sites/fta.dot.gov/files/2026-05/Public-CIG-Dashboard-05-01-2026.pdf
It was listed in 2025 here:
https://www.transit.dot.gov/sites/fta.dot.gov/files/2025-02/Public-CIG-Dashboard-02-07-2025.pdf
Note too that West Seattle Link Extension was listed at $3.2B.
Curious. The Wayback Machine shows WSLE listed on April 25, 2025: https://web.archive.org/web/20250425032624/https://www.transit.dot.gov/funding/grant-programs/capital-investments/current-capital-investment-grant-cig-projects
But not listed on April 27, 2025: https://web.archive.org/web/20250427214517/https://www.transit.dot.gov/funding/grant-programs/capital-investments/current-capital-investment-grant-cig-projects
The former CIG page for WSLE indicates it was adapted from the previous CIG listing for WSBLE. The FTA issued its ROD for WSLE on April 30, 2025, so they took WSLE off the list at the same time they issued the ROD.
Overall, the CIG list has gotten much shorter. I assume this is part of the Trump admin’s efforts to cut transit funding.
Hallelujah for #1. Its about time. Dstt2 as currently envisaged is the worst k8nd of boondoggle.
Examples of US-centric myopia in ST/SDOT/Metro:
In other countries, automated lines and platform screen doors are so long-established and standard now that agencies have to justify not using them. They usually don’t because they’d be laughed out of the room and people would say the alternatives aren’t as cost-effective or ridership-maximizing. Some cities are even retrofitting them into legacy lines. Yet ST is reluctant to pursue these because they’re not common in US networks, and they weren’t common in the 1990s when ST chose light rail technology, and ST can’t reverse a past decision (except when the board decides to anyway).
In the 2000s, SDOT asked the public whether the next generation of transit upgrades should be predominantly light rail, streetcars, or buses. I said light rail because it’s faster and buses because they’re low-cost (you can get more needed routes/frequency for the same cost). Streetcars are the worst of both worlds. In the ca. 2014 Seattle Transit Master Plan, SDOT had streecars where feasible (4 or so lines) and BRT where not (Madison is too steep). Later Mayor Murray converted the streetcar corridors (Westlake, Eastlake, Jackson-North Rainier) to RapidRide (except City Center Connector).
The streetcar and RapidRide plans missed what makes them so successful in other countries: transit exclusive/priority lanes for all/most of the route, robust signal priority, and don’t let car/parking/bike interests get in the way of that. And they have higher capacity than our implementations. So they give an almost light-rail-like experience rather than being watered down to little more than the status quo. Typical streetcars/trams are like Link on MLK, which wasn’t considered for non-Link corridors.
In 2016 before the vote, boardmembers made noises about considering automation for Ballard and future lines, and converting CID1 station to center-platform as part of the CID2/DSTT2 work. It didn’t pursue either of them.
We’ve begged ST for automation, center platforms, and open gangways for a decade. ST thought about automation for five seconds in 2016, thought about it for five seconds more in 2026, and dismissed it with spurious arguments that contradict Vancouver’s/Honolulu’s/Paris’s/other cities’ experience. ST is finally pursuing open gangways in the NEXT fleet of trains (or rather 2-car units, so Link would have 2 quad-long cars rather than the 4 double-long cars).
It would be beneficial now for central/north Seattle capacity but ST was unwilling to order them in the 2010s.
Center platforms, ST has taken an on-again, off-again approach. It build Mt Baker-TIB with side platforms, then Capitol Hill-Northgate with center platforms, and mixed beyond that. And now it proposes center platforms in the worst case: retrofitting a Graham station. (Article soon, but this drives up construction costs and is where center platforms are least needed — they’re most needed in underground stations where side-platform transfer walks are longer).
Transit-priority lanes on most of Aurora was something Shoreline did but Seattle didn’t, due to opposition from Aurora car-oriented businesses and landowners. Seattle finally got serious about it this year in 2026, a decade after RapidRide E opened. It could have done it in the 2010s, and then we would have had a more European-like BRT, ridership, and public satisfaction with transit.
Then there’s the whole land-use approach, which is more cities’ responsibility than ST/Metro. Human mobility options and ridership are maximized when you prioritize peds first, transit and bikes second and third, and cars/parking last. That’s what Paris has been doing for the past two decades. But here it gets watered down to reluctance to have full transit lanes on most of Aurora, letting cycletracks displace promised transit priority on Broadway and Eastlake, reluctance to take parking lanes on 45th, etc.
The businesses complain that they need all their street parking and GP lanes or customers won’t come, but that’s the same thing New York businesses said before congestion pricing, or other cities said about taking parking spaces/GP lanes for BRT, or pedestrianizing a street or adding protected bike lanes. But then you install good BRT or a pedestrianized street or or PBLs, and walk-up customers end up outnumbering lost car customers, and the businesses and people are glad it happened. You just have to get to the critical mass point where things turn around. It’s the same with ultra-frequent transit: you need it on the ground before you can see its benefits and how much better things would be. But countless cities that have done these things have shown they work very well.
The myopy isn’t helped by the fact that we have 3 separate agencies in Seattle (ST/SDOT/Metro) for roads (ST), bus service (Metro) and trains (ST). Translink in Vancouver is responsible for all 3 (except the Canada Line sort of)
Isn’t the monorail eventually going to reach end of life? It’s a fairly old and small so I would expect part shortages to become an issue. Would make much more sense to take the opportunity now to replace it with elevated link (preferably automated light metro)
I’ve been saying this for years, Ian. It’s just that it’s hard for many locals to accept that reality. There are still lots of people with fond childhood monorail memories who can’t face letting it close and going to a museum..
Alweg’s monorail technology is now owned by Hitachi, who does actually use it (mostly in Asia). Do you think Hitachi would be able to build new rolling stock for Seattle?
There are only two monorail trains. Even when they were built, you wouldn’t want to use a bunch of custom parts because it just isn’t economical to do so with only two trains.
The monorail is nearly 65 years old. Anything that would wear out has worn out decades ago and new sources found, and/or adaptations made for parts that are now available. Eg: if the contact shoes are difficult to find, then they should figure out how to adapt the equipment to use the same ones overhead traveling cranes use. The tires are probably a standard size, and the traction motors are probably a standard NEMA size. Door operators should have been something off the shelf from either the WABTEC parts made for subway cars or from the elevator industry. Control parts should be standard and off the shelf replacements available (eg, the same Eaton pushbutton made in the 1950s is still made today, with a few modifications, by Rockwell.
The only things that should need customization would be things like sheet metal and windows, and those aren’t that difficult. They had to be custom made for the trains anyway, since there are only 2 of them.
When they do decide to scrap the Monorail, I think they should copy Manhattan’s High Line and turn it into an elevated linear park. And I have the perfect name for it. The Monotrail.
Is that a trail for a Monotreme?
I like Strauss one and three as well.
I am curious what other amendments have been submitted.
I hope one amendment explicitly promises to improve DSTT1 in power, access, and signaling.
Most of the community discussion and the chair’s package has been about protecting X, Y, or Z line. All but Ballard are low ridership lines. Ridership maximization has not been a topic. It is we get Link because the 2016 plan promised Link. Of course, ST3 was not about ridership maximation.
To me, the DSTT2 is already broken and should be deleted. Its negatives have been listed by STB posters: high cost, deep stations requiring more minutes to access, fewer stations and none at the G Line, the split CID station denying a regional hub and increasing the transfer times between the east and south lines to about 10 minutes (unacceptably long). Deleting the DSTT2 would help the fiscal situation of all five subareas; 49 percent of the cost was to covered by the outer four subareas. The DSTT2 is to be served by the south and Ballard lines. But is not the south line limited to six-minute headway due to surface operation? Is six-minute headway a sufficient use of a tunnel costing billions? I think not. All south line riders wanting to access Capitol Hill or the U District would have the opportunity to transfer (e.g., walk and wait).
When CEO Constantine pitched the 75-year bonds before the Senate Transportation committee, he asserted the second tunnel was a regional asset that had that long a life span. But suppose the DSTT2 in not necessary at all?
I am very encouraged by Dan Strauss’ proposal to defer a downtown tunnel. Now Sound Transit should update their ridership estimates reevaluate their priorities based on the estimates. I bet that would lead to deferral of WSL stub. It may also lead to dropping the Boeing detour.
They should also study automation and a First Hill tunnel leading to Mt Baker as proposed in https://seattletransitblog.com/2025/02/25/automate-ballard-link/ and https://seattletransitblog.com/2024/01/10/focus-on-slu-and-ballard/. Even if we only build Ballard to Westlake for now, we will need to decide the location/orientation of the Westlake tunnel. The current ST3 plan calls for a very deep station. If the tunnel would extend further east, it could be shallower, potentially even above the existing tunnel which would make transfers and egress far easier.
I seriously doubt it can over-run the existing tunnel. But yes, absolutely, squeeze it in as shallowly as possible, under Sixth Avenue. Sixth rises at least three stories between Pine and Seneca, so a bored tunnel can too, at least somewhat, making a relatively shallow “Midtown” station between University and Seneca possible while still being deep enough to curve under the freeway and head on up into First Hill.
Obviously, any station(s) up there will, yes, have to be pretty deep, but not as deep as the original “First Hill Station” design, because it will be about three blocks as the train travels farther from Sixth and Pine than would the original one have been.
I do not understand the focus on trying to pin Sound Transit down on an estimated date for completing light rail to Ballard. Is the point to shock people into action based on how far in the future it would likely be? But isn’t “indefinite deferral” already serving this purpose…how would saying, e.g., “hopefully by 2059” help? Or, is the point to get Sound Transit on the record with a specific date, under the theory that once they’re committed, they’ll have to get it done by that date, somehow? But that theory seems obviously flawed—we’ve already been through numerous revisions to delivery date estimates and future board members years from now could certainly revise the date again to push it further out.
I’ll sign the petition because prioritizing Ballard to Westlake over the ill-conceived DSTT2 makes sense, but the “give us a date” push seems like a waste of an amendment on some kind of weird symbolic gesture.
I think the point is to force the rest of the Board to take ownership of the real impact of “indefinite deferral”, since they’re clearly hoping that the general public buys the rhetoric that by advancing BLE to full design, they’ll be able to find enough cost savings (and buy enough time to get more state/federal grant funding) to build it sooner rather than later. Simply “deferring” completion of BLE means they don’t have to take responsibility for the magnitude of the delay. The fact the Board hasn’t asked the Staff to present a feasible opening date for BLE means they feel the political cost isn’t worth the transparency. Strauss is taking a rare stand against the political winds by demanding a date for BLE, and it’s nice to see.
Edit: adding to this, I think the majority of the suburban members of the Board don’t think they’ll be on the Board in 2028 or 2030 or whenever the agency finally figures out how to finish BLE, but I think Dan Strauss is hoping to replicate some of Balducci’s success with East Link by advocating for BLE now and hoping still be there to claim the political win whenever it finally opens.
“I do not understand the focus on trying to pin Sound Transit down on an estimated date for completing light rail to Ballard.”
The point is to not keep area residents and people who would benefit from the line in long-term limbo. There’s been too much of that already.
The gap between U-Link and Northgate Link was initially going to be 4 long years (2016-2020) but it dragged on to 6. That’s 6 years where people had to bridge the gap between UW Station and the Ave area where most of the destinations are, and the gap between UW station and Stevens Way bus transfers, not knowing when it would end. But at least they knew it was under being constructed and would probably open in less than a decade.
East Link was planned in the 2010s and was going to open in 2021 but Crosslake service didn’t start until 2026, so it’s the same kind of thing.
Part of the reason ST3 was expanded to be a third larger than ST1 and ST2 was because cities wanted certainty on whether services were coming and when, so that they could plan their regional-center growth around it and not make incompatible changes. Kenmore and Bothell was keen on getting Stride S3 into ST3 for that reason. A large group of citizen-activists stood up to champion it, unusual in a suburban area. Kirkland wanted certainty on S2 at Totem Lake and 85th for the same reason. Issaquah wanted certainty on Issquah Link. Everett and Tacoma wanted to finally know whether and when Link would reach them, which they’ve been waiting for since the 1990s and which Pierce in particular has been saving up its ST1 and 2 tax revenue for.
In all those cases ST3 gave them reassurance that their Link and Stride projects were approved and an approximate time when they would open. (One decade? Two decade? Now they know.) The times turned out to be a false hope, but as of now the projects are still active.
It may not matter much whether Renton gets a P&R or when, no matter what the City of Renton thinks, but Ballard is Ballard, the highest-ridership project in ST3. Non-drivers need to know when they’ll be able to walk up to more frequent and faster service compared to the existing D, 40, and 44. If they really want to use that kind of transit in their neighborhood, they need to know whether it’s safe to move to Ballard, or locate a business there, or accept a job in Ballard, knowing Link will come in X-ish years. They knew it would be a minimum of 21 years (two decades), but then it slipped to 23 years (almost a quarter century), and now an unofficial estimate might put it at 25-30 years (almost three decades), and now ST won’t even say (if the resolution is approved unchanged). That raises the question of not only when it will open but if. That’s a lot of decades to be in limbo. Or for people to not move to Ballard or not shop in Ballard because the overhead of getting into and out of Ballard without a car is so high, and not knowing when it might improve.
Yes, I understanding why people want the “certainty” of a date, so they can make plans, not be in limbo, etc. But given the fact that the date would not be binding (as demonstrated by the history of passed delays) and that we are now talking about trying to predict what might happen decades from now (both reducing the ability to give an accurate estimate and negating any political accountability if the estimate is wrong, because all the politicians and staff will be retired by then), no one should bank on whatever date might be given, so we’ll end up being in limbo anyway. I do think it would be useful to know the best estimate for when a Ballard to Westlake stub could be built if Amendment 1 passed versus when light rail may get to Ballard if it doesn’t pass and priority stays with building the 2nd tunnel. That would better show the choice before the Board.
FYI horde
https://www.myballard.com/2026/05/26/dan-strauss-proposes-amendments-to-keep-ballard-light-rail-project-moving/
“The first would prioritize building a “Ballard Starter Line” from Westlake to Ballard”
That wording does sound like what Balducci used successfully to get parts of East Link open before the whole thing was finished.
The difference is Balducci was arguing for a subarea that didn’t have Link yet, while Seattle has the 1 Line. So it may be a harder sell.
If it works, so be it. At this point, I would not be surprised if the Ballard Alliance (some of whose members brought the Missing Link litigation/literally bullied Mike O’Brien out of running again for his council seat) use their powers of “persuasion” on the ST members.
What powers of persuasion? If they have his power, why hasn’t it worked at all so far? Why would bordmembers outside North King (the ones you’d have to convince in order to get a majority) care about the Ballard Alliance?
I thought the list of all ST3 amendments would be released today but I still don’t see it on the Thursday board meeting page or an email alert. I have an article started on it, but for an article on something, you need the something. Has anybody seen the list?
Probably a totally unrealistic idea, but here goes —
Could they break this thing into parts? Give the Sounder over to the state (under strict service requirements), keep bus service and core light rail with Sound Transit, and allow Seattle to expand its rail network — preferably automated, like SkyTrain — under its own agency or service so it can pursue its long-range rail plan (https://www.theurbanist.org/map-of-the-week-seattles-new-long-range-rail-plan-goes-big/)?
I saw this video about Montreal’s REM much as an organizational evolution. I found it to be a useful case study in approaches to expansion. Note too that the video is focused on expansion and not operations and maintenance.
https://m.youtube.com/watch?v=XlHqqA0onn0&ra=m
I’m not saying that ST should follow this model, but I am saying that it illustrates how different organizational structures often have a bigger impact on transit operation and expansion than local elected officials want to consider. Making no or few organizational changes is politically easier for an elected official.