The next 18 months may be the most important period of Sound Transit history as it tries to close a $30-40 billion program gap (The Urbanist). Additional coverage by the Seattle Times ($). Mayor Harrell thinks a new technical advisory group and regulatory reforms will get the job done with no program cuts or new taxes (The Seattle Times, $)
Sound Transit is starting a mid-life overhaul of the Sounder railcar fleet (The Urbanist)
The King County Water Taxi will run extra service from Vashon to Downtown Seattle on Saturday to get islanders to and from the Grand Opening Celebration of the rebuilt Waterfront (Metro Matters). Earlier this year, the State Legislature approved funding to maintain the Water Taxi’s midday service through June 2027.
SDOT thinks putting bus lanes on Denny would paralyze general traffic (The Urbanist). More on the SDOT blog. Here’s what the Fix the L8 campaign proposed on STB earlier this year.
Join the “Light Rail Relay” race on September 27 (The Urbanist)
Kitsap Transit’s summer service change begins on Sunday, September 14 (KT Headways)
Last week, Amtrak began operating its new high-speed Acela train fleet along the Northeast Corridor (Citylab)
Adaptive, automated bus lane enforcement is speeding up buses across the country (Fast Company)
Parking is the invisible “dark matter” of the urban planning universe (Planetizen)
The White River Bridge in Buckley is closed for emergency repairs after a truck crashed into the bridge structure. In the meantime, King County Metro is increasing weekday service on DART Route 915 to get riders to the nearby pedestrian-only Foothill Trails Bridge (Metro Matters)
This is an Open Thread.
Balducci in the Urbanist article: “The current plan has been based on assumed constraints on headways and throughput and reliability through the tunnel [DSTT1].”
In the March 2023 board meeting, several people in public testimony asked ST to study a single-tunnel solution (the ST3 candidate project to upgrade DSTT1 from 3 to 1.5 minute frequency instead of building DSTT2), as part of ST’s “reviewing past assumptions”. In 2016 ST had told us the tunnel would need a “signal upgrade’. In 2023 a boardmember asked the staff rep to clarify what the issue had been in 2016. The rep, speaking from memory and probably not having thought of it for years, gave a different reason: the platforms becoming overcrowded and not having enough escalators/elevators/stairs for egress. The board said it would not even look into questioning the previous board’s decision for DSTT2, and rejected an DSTT1-upgrading study.
This was at the same time Balducci and WSDOT boardmember Millar proposed a “Restored Spine” alternative (amendment 2 in the agenda). I don’t remember what that was. Maybe returning to a Lynnwood-Tacoma Dome line and a separate Ballard-West Seattle tunnel? It was to mitigate the long-transfer issues with Eastside-airport and Rainier-UW trips. The rest of the board rejected it.
So that was Balducci’s innovation then; and single-tunnel is her innovation now.
Still, this budget shortfall may be forcing ST to study what we asked it to study in 2023.
I don’t understand why they are so reluctant to look at the single downtown tunnel option. Seems like a no brainer especially for WSLE which otherwise will be truncated at SODO for a currently estimated 7 years. Those signal upgrades would be worth it.
On the other hand with how many issues the 1 line has been having with reliability it could be a catastrophe
Main objections are that a single shared tunnel for three services would seriously constrain the long term potential of the system, and it presents operational challenges due to the required three-way merge.
On the operational side, a good example is the Washington DC Metro, which has suffered ever since the opening of the Silver line from combining three services at Rosslyn. Even at peak headways of < 2.5 minutes, there simply aren't enough trains to adequately serve all three routes. In the DC case that meant the Blue Line got dropped to 5 trains/hour at most times times of day, which leaves an entire line with 12 minute rush hour headways. So you spend the billions on Ballard Link for an every ~12 minutes train, or let's say 10 minutes if you can get highly optimized 2 min headway performance from Link (unlikely). That might be acceptable for some period of time but it's really not paying off the investment.
And then in terms of long term network design, you've just put off the need to build an entire new tunnel for any new service until a later, likely even more expensive date. For example, an Aurora Link line is far off, but it's the obvious next investment and with a second tunnel in place it could be built out someday in phases. If Ballard goes into the existing tunnel, then any new service needs an entire new dedicated tunnel (and the new tunnel couldn't serve to address the existing capacity problem at that point, since you've already sent Ballard into DSTT1.)
All that said, I agree with most of the objections to the existing proposed DSTT2 design raised on this blog. But I'd rather stub Ballard link at a (well-designed!) Westlake interchange so that the second downtown tunnel can still be built later. Ballard passengers get more frequent trains and the system retains its growth potential at the cost of transfer times.
(Yes, WSLE would also need to be stubbed in this scenario… I don't know if there's a chance of building the new SODO station and tracks such that you could divert some 1 + 2 line trains out to West Seattle until a second tunnel is available.)
based on the terrible reliability on the current tunnel, seattle would basically build its entire transit system on a single point of failure that is constantly failing. But in doing so, it would continue the long tradition of America not having a clue how to build reliable transit not putting all your trains in one tunnel
The second tunnel could enhance reliability if there were crossover tracks between the tunnels like in Manhattan, but it’s refusing to. Just like it neglected proper turnbacks at Northgate and Lynnwood, and a crossover inside DSTT1. Or how it refused to order open-gangway trains for Northgate Link and Lynnwood Link to get 20% more capacity “for free” — that would have helped in recent crowding moments like the delay of the full 2 Line. Lynnwood and West Seattle trains will not be able to switch to DSTT2 if necessary, and Ballard and Tacoma Dome trains will not be able to switch to DSTT1. So if one of the tunnels stops functioning for some reason, all the people on those branches will have to transfer to a shuttle bus and maybe have lower frequency — the same as if the other tunnel didn’t exist — because the other tunnel won’t go anywhere near their destination.
I think a lot of the supposed need for a capacity of a second tunnel was predicated on pre-COVID assumptions about office commuters going into and out of downtown during rush hour. Travel patterns have since changed into less peak travel, and more consistent demand throughout the day. That, alone seems like good justification to revisit the second tunnel. The argument that the existing tunnel can’t be upgraded to more frequent trains at less cost than building an entirely new tunnel is ridiculous.
Jackson,, Ballard can NOT “go into the current tunnel”. We have hashed over this for ten years and determined that there is no way to make the northbound diversion without tearing down the new mid-rise (25 stories?) building in the northeast quadrant of 3rd and Union.
You CANNOT have a level diversion at Third and Pine northbound with the sharp curve there and the long trains on the Spine. It would be an operational catastrophe.
You also can’t engineer a northbound diversion under Pine Street, because the structural box narrows to two-and-a-half lanes just east of Sixth. If it had been built three lanes wide, you could put a burrowing junction in the middle,but it is not wide enough.
You cannot puncture the south bearing wall without bringing the street above down on your head, not to mention a block’s worth of the lovely buildings on the south side of Pine.
It might be possible to divert to the south inside the TBM vault just east of The Paramount, but then you are heading 180 degrees the wrong way.
Read Ross’s synopsis of the consensus solution and my discussion of preserving Symphony’s center lane for a non-revenue holding track, ponder them, surrender your allegiance to ST’s over-engineered, bloated and impossible-to-access DSTT2, and join us in advocating for a short-train, small station, automated Ballard-Downtown with its potential extensions into First Hill / North Rainier and east to UW / UVillage via Fremont.
> You CANNOT have a level diversion at Third and Pine northbound with the sharp curve there and the long trains on the Spine. It would be an operational catastrophe.
I mean it wouldn’t be easy but Chicago el and some nyc subway lines do it fine
Three lines in the tunnel means West Seattle-Everett, Redmond-Mariner, and Tacoma-Lynnwood-or-Northgate. Ballard-Westlake would be a shuttle line with everybody transferring. This is to avoid breaking into DSTT1’s watertight shell, which would be expensive and risky. Alternatively, a Ballard-West Seattle line could be automated (smaller tunnel, less expensive), or elevated or surface. Or Ballard-Westlake could have an option for a future extension to First Hill (instead of through downtown).
It’s important not to break the one-seat ride on Seattle’s primary axis (airport-Rainier-UW), nor the 1/2 Line transfer in DSTT1 (Eastside-airport). That would be taking away a great achievement Seattle has made, and affect more riders than Ballard would. Link just attracted riders away from the 48, and breaking the eastern north-south axis would force them back onto the 48. Ballard doesn’t have light rail yet so it wouldn’t be losing anything. And Ballardites will still have the option of taking Link to U-District and the 44 to Ballard.
WL, the “corner” crossings of the El generally run considerably shorter trains than does Link, so they block the diamond for a shorter period. The long trains are in the Red Line subway.
If you had full automation north of the junction at IDS it might be barely possible with six-minute headways on each line, but ST has publicly ruled them out on that portion of The Spine. Maybe they’d change their minds given the lack of funding to complete the system, but a shallower Westlake with better connections makes more sense, because it gives the potential for an extension into First Hill.
With decent connections at Westlake you’d have the excellent situation that many people riding south on Spine trains would deboard in order to take Ballard-Westlake north a couple of stations into SLU, freeing space for folks coming south on Ballard-Westlake desiring to go to the south part of downtown or the airport. It’s likely to be a pretty close to neutral balance.
WL (and others), actually there might be a way to add a northbound diversion to The Spine without a level crossing of the northbound Ballard line and the southbound Spine Main.
About three years ago I spent about six months advocating the addition of a burrowing junction between the north- and south-bound tracks in the Pine Street tunnel. Since the trackway rises about one story from Westlake to the flat section just before the curve into the TBM vault, the diverging track wouldn’t have to dive much at all.
Alas, as I detailed in another reply, the Pine Street Tunnel is too narrow.
But the center roadway in Symphony is not.
Now what you’re about to read may or may not be feasible depending on the placement of the station box supports at the north end of Symphony. But if there is a twelve foot gap between them halfway between the tunnel portals — or if an alternative support with an equivalent span can replace the current middle supports — then I believe that Symphony is long enough that a burrowing junction can be dug in the center roadway and get deep enough to enter a newly bored tube at the north wall.
Here, then, is the layout. From the north wall centerline the bore would continue north, at first centered below the mid-line between the Spine bores. About Pike Street it would move west such that it is directly underneath the southbound Spine tube where it would resume the heading of the street and Spine Tunnels only it would not make the turn into Pine.
Immediately north of the north wall of the Westlake Station box it would enter the lower level of a “stacked” station with the tracks in the west half of the Third Avenue ROW and the platforms in the east half. The upper level, which would be at the same level as the platforms of Westlake, would be for southbound trains. Leaving the station southbound it would pass through a new hole devised in the north wall of the Westlake Station box and merge via a trailing-point turn-out with the southbound Spine Main.
Immediately north of the newstation the stacked tracks would turn obliquely into the Stewart Street right of way and continue northeast to Westlake. There the upper southbound track would transition to become the southbound Ballard-Westlake line and the lower northbound would underrun a projected future extension southbound into First Hill. The NB track would curve northward under the streetcar plaza and Sixth Avenue and become / merge with the northbound Westlake-Ballard line.
This solution has spectacularly good transfers!. The most important one, Spine from the North to SLU requires a walk of one block from the west end of Westlake Center’s northbound platforn down a gentle slope to the south end of the Downtown-Ballard northbound platform (the lowest level) in the new station. Can any transfer that requires two platforms be much simpler?
A similar connection from the mezzanine level of the new station, which would extend the entire width of Third Avenue, to the mezzanine of Westlake above the walkway described above would allow a two-level-change transfer between SLU southbound to the northbound Westlake platform for Spine northbound passengers. That transfer will more typically be the “going home” journey when time is less critical or a going-to-UW transfer by a young person.
This would be the “fast way” to transfer SLU to Spine North, BUT less energetic people could simply ride two stations farther the south and reverse using a new center platform meant only for reversing at Pioneer Square. This is badly needed for Eastside to Airport transfers, and collecting all such “out-of-direction'” transfers in one facility makes the expense of providing ADA access that will almost never be used more cost effective.
One platform would solve two out-of-direction teansfer problems.
People from Ballard, Magnolia, Lower Queen Anne to anywhere south or east would simply ride to Pioneer Square or International District get off and wait in place for their train.
To support the high-frequency of operations in the tunnel full automation north of IDS would be required.
“But the center roadway in Symphony is not.”…..
It seems to be an elegant solution, but I don’t fully understand this paragraph.
“Immediately north of the north wall of the Westlake Station box it would enter the lower level of a “stacked” station with the tracks in the west half of the Third Avenue ROW and the platforms in the east half. The upper level, which would be at the same level as the platforms of Westlake, would be for southbound trains. Leaving the station southbound it would pass through a new hole devised in the north wall of the Westlake Station box and merge via a trailing-point turn-out with the southbound Spine Main.”
Does this idea have two connections to the spine then? It sounds like going north the train would dive at Symphony, (thereby skipping that station which is fine because of the super easy transfer at Pioneer where they could hop off a Ballard destined train and get on an Everett one to get to Symphony!); and coming south the train would merge north of westlake and not miss any stations.
Do I have it right?
Either way, we’re talking about some pretty expensive changes to the tunnel.
My first question would be what can be done to what exists now?
There is an existing junction segment where the old Convention Place station used to be. What can be done with that? Is there enough space between whatever the convention center built and the Link tracks to allow a sharp curve to get Link directly under 7th? As I remember it, the tunnel was built to be fairly straight to get from Westlake to Convention Place, so coming off at a curve might be enough to get a branch under the street.
This branch would only be used as a maintenance connection, or to provide SeaTac-Ballard service in the event the UW tunnel gets closed.
jas, yes, you have it exactly right. I have been advocating for punching a hole in the north wall of the Westlake Box at Third Avenue in order to create a non-revenue connection to Ballard-Downtown for cars needing “heavy maintenance” at Forrest Street for three years. So I sort of short-cutted on describing the upper-level southbound tube. The tubes would be stacked all the way to Westlake Avenue and Stewart Street. My apologies; most people don’t read my posts because they’re too detailed, so I tried to be brief.
Glenn, I simply do not believe that ST will put a level crossing in DSTT1. Now maybe eventually in desperation to serve Ballard inexpensively they will. But I wanted to give them a method to avoid it
I think there is insufficient room in the short block between Sixth and Seventh to dig a burrowing diversion. The platforms in Symphony are nearly two of those blocks long, and I came up with over a six degree grade to get under the box at the north wall. Those pans always seem to be a clearance problem.
Anyway, the tunnel narrows right at Sixth Street so you’d have the “bring the street down on your head” problem from widening the box.
Since I really respect your engineering, I have to consider that I may just not understand how you intend to get from the northbound Westlake platform to a track northbound under Seventh Ave due without rebuilding the Pine Street Tunnel or having a level crossing of northbound Downtown-Ballard and the southbound Spine track. Would you expand on your idea? Thank you.
Glenn, I think I understand now. The snag is that you called “Ninth Avenue” “7th Avenue”.
The Convention Center expansion is between Ninth Avenue and the freeway, and there is indeed a widening of the tunnel box there and a pair of tail tracks that stretch some way to the east of the curve into the TBM removal vault just northeast of the Paramount. These tail tracks were used for reversing during combined operations. They got the trains out of the way of the buses which started diagonaling into Convention Center Station just west of Ninth Avenue. That bit of tunnel is underneath the addition to The Camlin at Ninth and Pine and may indeed still be there and open.
It would help make the curve into Ninth Avenue. However it would have to include a level crossing of the southbound Spine Main, because, again, the tunnel box to the west is too narrow to host a burrowing junction.
If it is just for non-revenue moves to and from Downtown-Ballard from some heavy MF, then it’s probably fine. Those can run late at night.
However, the elevation at Ninth and Pine is quite a bit higher than that at Third and Pine because the Pine Street Tunnel rises a full story between Sixth and Ninth, so the connection tunnel would have to dive that extra story to get down to the new tunnel.
It would have a couple of extra blocks to do iy,but all the corners it would have to turn would be right angles. The Third and Pine to Stewart and Westlake has two oblique intersections to leverage for less sharp curves.
I do think that for a strictly non-revenue connection it could work, and it would free the center roadway at Symphony to host a standard scissors cross-over.. This proposal for the burrowing junction in Symphony is for full revenue service between Ballard-Downtown and points south, probably West Seattle.
Again, it would probably require full automation of all three lines north of IDS, all the way to Lynnwood (and eventually perhaps to Paine Field or Everett). It would make seamless transfers at Pioneer Square possible simply by adding back the reversing center platform, but with ADA compliant egress included, but unused except in evacuations. Normally nobody could access the center platform from the Mezzanine.
Innovative, forward thinking, and prepared for the future. I like it. Surely by not having to dig another tunnel to Symphony, and then only a single tunnel from symphony to westlake – a few billion could be saved.
More billions could be saved by buses to West Seattle – an idea that I also like.
Alternatively, if West Seattle has to get a train, I think they could look at changing the scope and doing a low bridge, and taking a slight right turn after the West Seattle Bridge to a station on Harbor Ave, and then punching a short tunnel through the bluff to a station at Admiral Junction with a contingency to turn the line south at California in the future. This would be one fewer station than is planned now.
The promise would be that in the next expansion the line would go south down California – perhaps in a cut and cover tunnel to save cost over a bored tunnel.
Sure it’s a slight out of the way jog if you are sitting at the West Seattle Junction, but having a station at Admiral adds coverage to the north end of West Seattle, and it wouldn’t be that much of a delay I think.
I need to add a further refinement. DSTT1 cannot handle all the trains for a short-train, small-tunnel automated Ballard-Downtown line. There are not enough “slots” in the tunnel for all of them. So, most Ballard-Downtown trains would have to continue as planned in the “consensus” model down Sixth Avenue to an initial stub station around Pine Street. That line would be constructed with the ability to be extended south and then east into First Hill, but it would be as shallow as possible.
Since it would be designed to turn uphill to the hospitals on First Hill instead of downhill to IDS2, it could be as shallow as the supports of the Westlake Station box around Sixth Street allow.
I would estimate that only one out of four trains on Ballard-Downtown could be accommodated in DSTT1, or one every eight minutes. That would mean that there would be a 2-2-4 minute headway pattern at the Westlake stub for northbound trains. That’s obviously not ideal, but the longest wait is four minutes, not a miserable one.
The once-every-eight minutes trains through DSTT1 could auto-reverse in the center track south of Stadium. Running all three lines through DSTT1 would require overpassing Holgate as well as Lander and closing Lower Royal Brougham anyway, so reversing there would be relatively safe.
But the money saved by not four-tracking the busway and building that Skycastle at Lander would certainly pay for the modifications to Symphony and the connecting tunnel.
STB commenters launched the single tunnel concept. Reece Martin exposed the fallacies of the split CID station advanced by ST Board Chair Constantine, now CEO. The reset mandated by the fiscal and feasibility crisis and the Covid ridership changes allows an opportunity for study. Trump 2.0 is not a time to apply for FTA grants.
Hasn’t existing operating condition says enough about Feasibility of single DSTT. What do they need to finally shut up about single DSTT scenario?
Basically at this point it isn’t about building a reliable system. It is about building the cheapest system that will underperform and underserve. It will be a case study on how not do do it. But it allows the symbolic goal of building something even if it is ultimately more expensive to fix eventually with a second tunnel.
The operating issues are largely due to the system being too big. They are basically spending too much money expanding, and not enough money on the existing system. Thus a second tunnel would make the system less reliable. Oh sure, there would be times when trains are running through one tunnel but not the other. But during those times, transit sucks. If you can’t get from downtown to the UW it really doesn’t matter that the other train is running just fine. You are still out of luck.
The only place where there would be redundancy is where it is least needed (downtown). I can get from CID to Westlake just fine on a bus. But getting from Northgate to downtown via a bus takes a really long time (with the normal buses). So you are still dependent on those shuttles. But if you spend a bunch of money building a second tunnel and more money maintaining that tunnel you not only have less to spend maintaining the existing tunnel, you have less money to spend on those shuttles. If you want reliability, you don’t want a second tunnel.
The Link reliability issues have not been focused in the DSTT; they have been everywhere.
https://www.seattletimes.com/seattle-news/transportation/link-light-rail-service-down-between-capitol-hill-and-northgate-stations/
DSTT2 would be somewhere around 9 floors under Westlake, requiring some 11 minutes of escalator rides to get from one to the other. With that amount of time added, it’d be faster to just go to a bus on the surface.
You could build a lot of backup Link lines on the surface for the price of this thing that would provide service where would serve the need better.
A second tunnel might be nice for various reasons, but this particular proposal really doesn’t have anything positive that I can see.
“The Link reliability issues have not been focused in the DSTT; they have been everywhere.”
I think that’s true. A couple recent signal outages and others were not in Downtown Seattle.
But Single DSTT isn’t really a reliability issue anyway.
When we talk about reliability, we are not talking about train being too slow or infrequent constantly. The issue is there is unacceptable level of chance that service gets disrupted. For example, a typical 20-minute trip becomes 40-minute trip 3 times a month is a reliable issue. I think that’s what happened this summer.
Single DSTT on the other hand is more of a capacity issue that will suck every day. Single DSTT means the service will never operate at a frequency level that can fulfill the demand of each corridor because single tunnel cannot handle that many trains, but the system can still be super reliable and your waiting time is constantly 10 minutes every day.
Single DSTT on the other hand is more of a capacity issue that will suck every day. Single DSTT means the service will never operate at a frequency level that can fulfill the demand of each corridor because single tunnel cannot handle that many trains
Says who? It is quite possible that with very little effort it can handle the loads just fine.
You have it backwards. Two tunnels means it will suck every day, all the time for a lot of riders. People heading to the airport from the north end will have to transfer. This will suck. People from Renton and Kent who are used to a fast trip to downtown (and a fast connection to Link at SoDo) will suddenly find that their trips sucks. Riders from Beacon Hill, Rainier Valley, Tukwila and Federal Way will have to transfer to get to the UW. That will suck. Their stops downtown will be a lot worse. That will suck. Chances are, life will suck a lot worse if they build the second tunnel than if they don’t.
In addition to what Ross says, I’d add overcrowding to be a significant problem if the second tunnel is built.
Nobody is going to want to transfer at Westlake if they are able to due to the vertical distance, so most everyone will want to transfer from the south line to the UW/Lynnwood line as far south as they can. So, you probably wind up with 3/4 or so of the passengers in the curent tunnel.
“People heading to the airport from the north end will have to transfer. This will suck.”
“People from Renton and Kent who are used to a fast trip to downtown (and a fast connection to Link at SoDo) will suddenly find that their trips sucks.”
“Riders from Beacon Hill, Rainier Valley, Tukwila and Federal Way will have to transfer to get to the UW. ”
Have you thought about the possibility that some of those people won’t transfer but use the stations along second tunnel because their destination is between 3rd and 5th Ave anyway?
Yes all these will be problems if there is a second tunnel, but at what scale? Do you have data that shows how many riders will be on the Origin-Destination pattern that requires painful transfer in 20XX?
Same thing about transfer to Airport. I wonder if there is any survey done to estimate what are the top Link stations riders going to Airport are from.
Single tunnel in a city this size is reckless. 3 tunnels is a luxury. You need redundancy. If you’ve lived in a 2 tunnel city it’s a lifesaver to go to the other line when one is blocked.
@George To Ross’ point though, the only redundancy between the two tunnels will be between Westlake to the International District – where there is already TONS of redundancy in the form of surface buses.
As it is planned right now, the second tunnel both isn’t far enough away from the original to provide more coverage *and* doesn’t provide redundant service to an area lacking in that regard.
That seems shortsighted. Ballard Link could be connected to Northgate or U district in ST4, giving redundancy for a much larger section.
Mixed feelings about this. While it is true that the people relying on the one tunnel that is blocked are SOL, it is also true that the other half of the system is still operating. On the other hand again, disruptions *in* the downtown tunnel have been rare, and will continue to be rare. There was the instance where part of the ceiling fell in at Westlake, requiring extensive delays to fix, but the tunnel has otherwise proven very reliable, despite its age! Whether it’s worth it to spend the $$$ on a second tunnel for these edge cases, AND to introduce some long, inconvenient transfer situations in the process, that is more of a concern to me. I would argue the second tunnel is not practical unless it can be done more cost effectively as cut-and-cover, with better transfers and better access to the surface. If not, scrap it and upgrade DSTT.
Delta, part of the vision for an automated, short-train, small station Ballard-Downtown is that it can be extended both directions fairly inexpensively. From Westlake it could continue southeast through First Hill and terminate at either Judkins Park or Mt. Baker, providing the wanted redundancy to the south of town. The portion south of Yesler would be elevated. The area there is ideal for high rises because they wouldn’t block anybody’s view.
To the north we mostly are of the opinion that turning more sharply east to UW and UVillage would serve more trip pairs. Most folks would like it to pass through the north edge of Fremont along the way with a tunnel station by The Troll and covered escalators up to Aurora and down to 34th. Since two-car, frequent, automated trains can carry the same number of riders as half-as-frequent four-car trains, five stations would not be ruinously expensive.
This would be that redundancy you mentioned, but also vastly improve north end Metro bus connections to UW.
Delta, my apologies. Upon rereading your original reply, I see you did say “or the U District”.
To connect west Seattle and Ballard I would like to see something similar in concept to what Tacoma has.
In other words take the south lake Union trolley line that is already there, and follow through with the plans to take that line down First Ave. Tie into the First Hill line if you want, but also extend it all the way to west Seattle. At Mercer the tracks would have to turn west to get to Seattle center and Ballard.
ST could keep it grade separated from west Seattle to sodo, and from Ballard to Seattle center, but through the heart of the city maintain typical urban stop spacing.
Don’t build the second tunnel. Don’t interline through the tunnel. ST takes trolley operation from the city. The city would love it!
A major problem with running on the surface downtown is Seattle’s short blocks versus the needed train capacity. Link’s four car trains are just too long for a single block.
There are also the crossing challenges of the Duwamish and the Ship Canal.
If Portland could figure it out I bet Seattle can.
If Portland could figure it out I bet Seattle can.
Sure, but Portland does not have a very good transit system. Vancouver has by far the best transit system in the Northwest if not the West Coast. The trains are grade-separated (and automated) which is basically the standard now, world-wide. It is especially cost effective when you have small trains (i. e. a light metro) but even in areas where they have gigantic trains sets and relatively short lines (i. e. a lot of riders per driver) they have moved towards smaller, more frequent, automated trains. For their new Ontario Line, Toronto switched to using automated trains (late in the planning stages).
The point being if Portland did it all over again, the would probably be more like Vancouver. Or maybe they would be more like Brisbane, and build a series of busways to accommodate the sprawling suburbs. Very few cities should build a system based on surface trains and Portland isn’t one of them.
“Sure, but Portland does not have a very good transit system.”
Ross, you’re letting your fanboying over SkyTrain blind you to the fact that Trimet’s MAX is pretty solid as a transit system. It’s a system that while yes the downtown and steel bridge sections is slow and should be faster with Stadtbahn like downtown tunnel. MAX still punches above it’s weight for a city of its size, alongside a solid bus and streetcar network. There’s a reason why Portland is considered a model for peer cities who want to improve their transit systems.
MAX is an example of how not to build a transit system. Designed to suck forever which is the American transit way: serve as a warning to others. If you’re going to model a system, make it be better than Link, not worse. Skytrain is objectively better. If you can’t do that, just BRT it.
Paul, MAX for all intents and purposes is a good transit system. It’s basically a mix of an interurban and statbahn system. You can keep saying that “it’s not good” all you want till your blue in the face, but the reality is different on the ground with how many ride it everyday for work, school, errands, and events.
“If Portland could figure it out I bet Seattle can.”
Portland’s solution is to have up to 2-car train running on every route so light rail is almost as long as one block. With its light rail system more built out and relative smaller metro area population, Portland’s light rail demand probably doesn’t hit its capacity too hard, but it won’t be that way in Seattle if a second downtown light rail corridor can only accommodate 2-car Light Rail. It’s not like Portland didn’t struggle with this capacity constraint. MAX’s Siemens S70 orders are single-ended with cab only on one end of car to make more space for passenger.
“the reality is different on the ground with how many ride it everyday for work, school, errands, and events.”
How satisfied are they with it though? A lot of people are exacerbated with its 20-minute slog downtown, and it loses potential ridership because of that. Or they ride it anyway because there’s nothing better, but it really isn’t ideal, and not like what other cities have achieved.
Zach B, Ross is right to fanboy over Skytrain considering it gets 6x the ridership of MAX (roughly same population metro area).
Portland has low modal share for transit. Ridership on light rail is not that high. While the buses get more riders it still doesn’t get that many. The one good thing you can say about MAX is that it didn’t cost that much. But it isn’t clear that it was a good value. Jarrett Walker (who knows a lot about transit and lives in Portland) even suggested they might have been better off with a BRT system like Brisbane (see the bottom of this post: https://humantransit.org/2009/11/brisbane-bus-rapid-transit-soars.html). To be clear, he didn’t say that would have been a better choice, but it is would have been a reasonable one.
In contrast, no one in their right mind would say that about Vancouver. The trains carry huge numbers of riders. They form the backbone of a hugely successful system. Ridership on the trains is about half a million — dwarfing cities like Seattle and Portland. But since the trains complement the buses so well, the buses do even better. Third best in English speaking North America!
Light rail has its place. They are great if you can leverage existing rail lines. That is the idea behind S-Bahns. They make sense if you have already built a full-fledged metro and still have enough demand to justify a surface train (e. g. Paris). They also make sense if you need the extra capacity but just don’t have the money to build a subway line (like some of the extremely dense, but smaller European cities). But most North American systems aren’t like that. They are built often built as a cheap substitute for what they should build: a light metro or busways.
Light rail is often used as a compromise between systems. For all intents and purposes, our system if mostly a subway. But in a handful of places, it operates like light rail. This is fairly common. You run underground (or less commonly, above ground) inside the city and on the surface in the suburbs. Running on the surface on Rainier Valley is quite reasonable. But it comes with a cost that just wasn’t an issue fifty years ago. Link can not be automated. Not only that, but the city won’t allow increased frequency in Rainier Valley anyway. So rather than running every 6 minutes most of the day (like SkyTrain) the trains run every 8 minutes at best, and every 10 minutes midday. With a grade-separated line the trains could be running faster and more frequently (and more reliably and safer). It could also open up the possibility of branching. If trains are running every six minutes down Rainier Valley then running every 12 to Renton and SeaTac is quite reasonable. In contrast there just isn’t much point in running trains every 20 minutes to Renton (not when the bus runs every 15).
In our case, light rail was really the wrong choice. To a certain extent, we stumbled into it. We wanted to run the trains and buses in the same tunnel. This meant low-floor light rail trains. Then they wanted to run really far outside the city. This led to compromises, like running on the surface in Rainier Valley.
It is easy to see in hindsight, but we made a big mistake. We should have started with Northgate to Downtown. That not only gets the most riders per mile, but it replaces a lot of the buses. So many that you can kick the remaining buses out of the tunnel. At that point, you can forget low-floor vehicles. You rebuild the downtown tunnel for a high-floor, automated light metro system. This means higher capacity trains and shorter dwell times. Then it can be expanded as time goes on. Again, that is hindsight, but light rail in our system was a mistake and it probably was for Portland as well.
Mike, they have deleted stops in Downtown Portland to speed up the train as part of a plan by the Trimet board in 2018. Skidmore Fountain was the last station to close as part of their Downtown stop consolidation on the 24th of August, they also closed Mall/SW 4th & 5th Ave (next to Pioneer Place) and Kings Hill/SW Salmon (South end of Providence Park) in 2020. The only reason Skidmore Fountain station didn’t close in 2020 was because of push by the Portland Saturday Market who wanted it to stay around for convenience of people visiting the weekend market.
Realistically, the only issues Trimet have left in the downtown section that can improve speed and times before considering a downtown tunnel for MAX is allowing for the signal priority of MAX at stoplights downtown and doing something about the Steel Bridge and some other slow spots to allow for faster travel speeds.
There’s a reason why Portland is considered a model for peer cities who want to improve their transit systems.
Since when? Transit modal share is tiny. Portland has an excellent bike network and an outstanding bike share system (unlike Seattle). But their transit system is second rate. The fact that biking is so good and transit ridership is so low is an especially bad sign. The two generally go together and yet it hasn’t translated into the kind of transit ridership you would expect in Portland.
To be clear, many cities started building light rail lines after Portland did. But that is only because the US generally doesn’t know what they are doing (our transit system is often considered the worst for the money we spend on it). We fell in love with the “light rail” concept without ever considering the advantages and disadvantages of it. Again, it has its place, but I think it is pretty clear that cities across America have spent way too much on light rail and way too little on buses and metros (light or otherwise).
Link functions as well as it does mainly because it so rarely operates like light rail. Put it this way. Imagine if it ran on the surface between the south end of downtown and the UW. It would go on Eastlake, skipping Capitol Hill. It would be slower than the old express buses that ran between the UW and downtown (using the express lanes and bus tunnel). How is it a better system if it is slower than the buses it replaces? The short answer is, it isn’t.
Light rail is a compromise of sorts. There are places where it makes sense to run on the surface. These are typically in locations where there are very few crossings and not that many people. Might as well save a few bucks and run on the surface. But that clearly isn’t the case through downtown. If anything it is the opposite. Running on the surface in West Seattle or in Ballard (and certainly south of Ballard) is quite reasonable. But downtown is about the worst place to run on the surface.
“the city won’t allow increased frequency in Rainier Valley anyway. So rather than running every 6 minutes most of the day (like SkyTrain) the trains run every 8 minutes at best”
The ST3 plan was 8-10 minutes but it got changed to 6 minutes all day sometime this year or last year. ST wouldn’t have changed it without SDOT’s consent. I assume the reason for the change is capacity need in SLU and Ballard, and to avoid the spectacle of an expensive second tunnel with only 6-8 trains per hour. Rainier Valley did have 6-minute peaks between 2012 and sometime in the late 2020s, so it’s clearly possible to do without screwing up cross-traffic.
As to why ST won’t run 6-minute service all day now, ST has never given a reason; it just ignores our pleas without responding. Maybe ST doesn’t want the operational expense or wear and tear on the trains. Maybe it thinks ridership is not sufficiently high. (That again comes down to thresholds: what’s the threshold for increasing frequency? How many riders or how full trains? ST won’t say.) Or it may be because SDOT doesn’t want it. But we can’t just assume that. If ST is OK with 6-minute trains on MLK in fifteen years, how can it object to them now? If ST changed the plan without telling SDOT, that sounds impossible: it would need SDOT’s consent and traffic configuration. If SDOT still doesn’t know about the change, that sounds impossible too.
While we dither, our neighbors to the north are building a new “SkyTrain” under Broadway in Vancouver BC, opening 2027:
https://www.broadwaysubway.ca/about/stations/
Click on the images. Look at the scale. This is what the Ballard line or any other new line in Seattle should look like.
Vancouver: “All stations will feature: Up and down escalators and elevators”
That Great Northern Way – Emily Carr station looks like our Roosevelt station.
@mike
I guess to be fair in Japan a lot of stations only have up escalators not down
“Portland has low modal share for transit.”
“Since when? Transit modal share is tiny. Portland has an excellent bike network and an outstanding bike share system (unlike Seattle). But their transit system is second rate. The fact that biking is so good and transit ridership is so low is an especially bad sign. The two generally go together and yet it hasn’t translated into the kind of transit ridership you would expect in Portland.”
You may have downplayed Portland’s transit a little to much.
There aren’t many metro area in the US with 5+% population commuting by transit and Portland is one of them, but I agree that transit in Portland is not good enough for Seattle to look up to.
Seattle should look up to place like Vancouver, BC. Portland is not the same size of metro area as Seattle and Seattle certainly has higher share of commuters taking transit than Portland does. Nationwide Portland is definitely an exemplar for places that are falling around average.
We actually have many similar problems as Seattle.
The busiest of the three new lines is Yellow, which is median running on Interstate. It’s great for enticing local traffic and is a huge improvement over the previous bus route 5.
It’s not going to be great if they connect it to Vancouver, because it’s too slow for downtown to downtown service.
Green and orange lines are faster, and run next to highways, but really don’t attract that many riders because they don’t get too close to where people want to go. Green has been converted to single car trains as its ridership is reliably poor. Shaving a few minutes off the travel time from downtown to mostly nowhere by installing a downtown tunnel isn’t going to solve the basic problems that not many people have their transit time shortened by having fast light rail to mostly nowhere.
They could have made the Interstate line a bit faster by putting it in a tunnel, but it wouldn’t make it that much faster because you really want the frequent station spacing for all the locations along Interstate.
Vancouver (and Portland, at least in the core) have a major advantage over Seattle in that they are located in river valleys, making for relatively flat terrain – particularly in Vancouver – and allowing for a continuous grid system of streets and hence bus routes throughout the city. This allows for an optimal transit network which in turn leads to increased ridership. It stands to reason that locating rail within the grid (Canada Line, Broadway) only aids in strengthening transit ridership. Seattle definitively does not have this advantage; as we all know traveling any distance east-west by any mode of transportation is a PITA (hell, there’s only one single street in the entire city – Madison – that extends all the way from sound to lake).
Now, if you want to talk about lousy transfers, Vancouver’s Expo/Canada Line transfer between Granville and Vancouver City Center stations is terrible. You have to leave the system, go up to street level, walk a block and a half, then go back down again despite the fact the lines literally cross there. While I’m certain for engineering reasons the Canada Line station had to be placed away from the existing Expo Line, not finding a way to connect the two without going up to the street would have been optimal (yes, I’m aware that you can make a direct transfer at Waterfront, but that involves traveling a station out of direction, both ways).
Scott, soon a rider will also be able to transfer at Broadway-City Hall directly to the Millennium Line. That will be a great shortcut for folks along it, and a decent one for those who are headed out the Expo Line.
That’s great news, Tom. I fly out of YVR occasionally and often take the train up, meaning a transfer between the Expo and Millennium Lines when traveling between the train station and the airport. It’s not the worst thing ever, but it’s a long way from “world class.” Any improvement will be welcome.
Vancouver (and Portland, at least in the core) have a major advantage over Seattle in that they have relatively flat terrain allowing for a continuous grid system of streets and hence bus routes throughout the city.
Sure, but that doesn’t entirely account for the fact that Vancouver is doing so well and we aren’t (despite both cities investing a huge amount in transit). Just to back up here, most of the city has good north-south streets. That is because the glaciers flowed that way. Going east-west is often challenging. So we basically have a “half-grid” if you will. We have invested quite a bit in subway systems and subway systems don’t have to follow the grid. When they don’t they provide a service that the street network can’t. To a certain extent we do that. The trip from Capitol Hill to the UW doesn’t follow the street grid. The trip is faster than driving at noon. But we’ve failed numerous times to take advantage of our geography. There are several ways in which we’ve failed:
1) Lack of stations. For example, between Capitol Hill and the UW there are no stations. The obvious station location would be 23rd & Madison. It would get decent walk-up ridership (likely a lot better than some of our suburban stations) but more importantly it would complement the street grid. It would connect to the 48, making trips from the north end to the Central Area much faster. It would connect to the RapidRide G, giving it an excellent anchor and providing another way to get to the heart of the city. That is just one example. We only have two stations in the U-District. Just from a walk-up standpoint this is a failure. But it also screws up the (potential) grid. The 31/32 (trying valiantly to follow an east-west grid pattern despite the geographic obstacles) struggles just to connect to Link. A station at Campus Parkway and The Ave (justified merely for walk-up riders) would greatly improve the 31/32 as well buses that go across the University Bridge. Even when there is a grid of sorts in the city, Sound Transit seems to ignore it. Instead of putting a station at Northgate Way, they put it at what is essentially a dead-end. You can’t go east-west from Northgate Station on a bus. It requires a huge effort just to get them to build Pinehurst Station — something that should have been obvious to anyone who can read a map. Keep in mind, this is not for a brand new (multi-billion dollar) line. This is simple adding stations on that line.
2) Duplicates fast travel. Some of this is inevitable. The city is an hourglass and we built freeways to downtown. But too often we are building rail lines that do the same thing that the express buses do. The result is a minimal improvement in travel time. Running next to the freeway is a classic example of this and it is about the worst thing you can do (https://media4.manhattan-institute.org/sites/default/files/economics-of-urban-light-rail-CH.pdf). ST3 is full of projects like this. We’ve come to expect this in the suburbs but this is even happening in Seattle. West Seattle to downtown has some of the fastest transit travel times in the city. There is an expressway connecting the two areas. Once a bus gets to the north end of Delridge it doesn’t make any stops until it is downtown. That is over three miles — in the city — without any stops. This isn’t a special “express” either (that only runs during peak). It is just a normal bus.
3) Disinterest in improving the grid. While it is easy to criticize West Seattle Link, Ballard Link has similar flaws. It is almost four miles from Ballard to Uptown. For that stretch, the train will essentially run north-south. During that stretch, there is only one east-west pathway (Magnolia and the north end of Queen Anne). That’s it. Again, this is in the city*. Contrast this with other proposals. A line from Ballard to the UW would intersect seven north-south transit routes. Many of these routes are extremely fast and carry a lot of people. The same is true with a “Metro 8 Subway”. When running east-west it would cross dozens of north-south routes. When running north-south it would cross the east-west routes in the heart of the city.
4) A focus on regions, not networks. The Spine is a great example of this. The neighborhoods served by the future stations have very few people. They are very far from large destinations. This means they won’t get many riders. Yet serving these stations isn’t cheap. So why is the Spine a priority and not, say, First Hill? Regionalism. Sound Transit treats light rail stations like they are community centers. They spread them out, far and wide. There is this assumption that a station anywhere in Tacoma helps the entire city. Or that it really doesn’t matter where the station is — it is “Tacoma’s Station”.
The decision to build West Seattle Link instead of Ballard-UW epitomizes all of these errors. West Seattle Link will only have three new stations despite being almost five miles long. It will duplicate the fast transit that exists already. It won’t enhance the grid. Yet it gives the impression that it serves all of West Seattle. The assumption is that the entire peninsula benefits even though the vast majority of people live nowhere near the stations. In contrast, consider Ballard to UW. There would be plenty of stations per mile. It would replace buses that struggle to run 10 mph in the middle of the day. It would dramatically improve the grid. It would be focused on the network, not serving a region (although it would serve Ballard better than Ballard Link).
*I’ve used phrases like “in the city” and “the heart of the city”. For those not from Seattle, I should give a little more background. Seattle proper is not a particularly large city when it comes to area (https://en.wikipedia.org/wiki/List_of_United_States_cities_by_area). Of course that is the actual legal city limits. Greater Seattle is much bigger. But Greater Seattle is not like L. A. While Seattle definitely sprawls, almost all of the density and destinations are in the city itself (a relatively small area). This matters when designing a transit system. Density and proximity are two of the key elements when it comes to transit ridership. They are also where a subway does especially well. Dense urban areas tend to have slow surface streets. They also have popular destinations along the entire route. Thus a subway — making stops every half mile or so — can do really well there. In contrast a subway between distant cities (like Seattle and Tacoma) will struggle because of the lack of density and destinations along the way. You have little to no network effect and riders get delayed with all the stops. That is why most cities (L. A. being the obvious exception) just run express buses or commuter/regional rail (leveraging old railways) to serve those riders.
So you basically want the train to be slower than a bus. What is the point in that?
err… no. the point is they have no money, and if they can repurpose the existing track and plans for the connector it would be hugely less expensive than digging a new tunnel, or figuring out how to run another line through the existing tunnel and then somehow sending it off to Ballard.
I also said that it would, and should, run grade separated except for the segment through downtown, where I presume signal priority would be required. Having urban stop spacing in the middle of the dense urban environment that is downtown seattle seems appropriate to me, and my guess is would be very useful to a lot of people.
Perhaps an option for a train from West Seattle to First Hill, or Ballard to First Hill could be considered too, instead of all the trains being Ballard to West Seattle.
To account for ‘short blocks’ run two car trains, or two car trains with open gangways more often if more capacity is needed.
Just trying to think outside the box here is all.
Ross,
Portland doesn’t have a downtown bus tunnel (and honestly will not have one in the next 50 years). and the trains work reasonably well there.
All of the ST budget woes can be fixed with one sentence… “No more tunnels” . It really is that easy.
the point is they have no money, and if they can repurpose the existing track and plans for the connector it would be hugely less expensive than digging a new tunnel
So again, it would be slower than a bus but it would be a symbolic victory. I really don’t see the point in that. If we want to run on the surface then the best option is to just run buses. You can run fast while saving a huge amount of money.
Just look at West Seattle Link. The plan is to spend about 7 billion dollars to get from Alaska Junction to SoDo. Instead they could add ramps from the Spokane Street Viaduct to the SoDo Busway. Which is faster? The short answer is, it depends. If you are trying to get from one of the three stations to downtown, the train is faster. But if you are somewhere else in West Seattle heading to downtown (or various other locations) the bus is faster. By avoiding a transfer you save time. This is especially true of riders on the RapidRide H Line. The station will be right next to the freeway ramp. Even if a rider magically timed the transfer perfectly it will take time to get to the platform. By then the bus would be half way to SoDo. Thus the vast majority of riders would be better off with the bus.
Or consider Ballard Link. Yes, there are certainly places where the train could run on the surface (and probably should). But it still needs to cross the bridge. That itself costs a bundle. Riders still need to get to the station. Last time I checked, there will only be one train stop in Ballard and it will be nowhere near the most popular bus stops. But a new bridge just for transit would help a lot, even if it cost a fortune.
But the tricky part is Uptown and South Lake Union. More can be done, but there are still intersections where the buses run slowly. But a streetcar would run slower. That is the problem. So now you’ve spent a fortune building a rail line across the bridge only to have it go slower than buses through Uptown and South Lake Union. What is the point of that? You might as well take some lanes on the Ballard Bridge and just run buses on their current path. It would end up being just as fast if not faster.
Surface rail is a niche mode. It only makes sense if you need the extra capacity but can’t afford a real metro (or you’ve already built one). But we don’t need the capacity. Keep in mind, the whole idea with the increased capacity is so you can run the trains *less* often. So instead of running buses every three minutes, we run a train every six. Does that sound like West Seattle to you? Or even Ballard? No, of course not. Hell, Ballard isn’t even running their express routes anymore *during peak*.
In contrast, an automated Metro is *not* a niche mode. It is a standard way to both increase capacity *and* reduce travel time *and* have very good headways. By being a grade-separated metro, the trains are faster than driving, at noon. By being automated, you are able to run frequently, all day long, even if ridership isn’t that high in the middle of the day. An automated metro system from Ballard would certainly be a big improvement compared to improving the bus system. It would cost a lot more, but get you a lot more. In contrast that isn’t the case with West Seattle. West Seattle is just very will suited for buses. There is already a huge bridge over the Duwamish. There already is a major viaduct over Spokane Street. There already is a busway that runs to downtown and connects to Link along the way. All you need to do is connect the pieces and run the buses more often.
All of the ST budget woes can be fixed with one sentence… “No more tunnels” . It really is that easy.
Current estimates for West Seattle to SoDo is $7 billion. The problem isn’t *just* the tunnels.
“So again, it would be slower than a bus…”
Would it be? I believe they would be able to design it so that it is faster, and it would absolutely be faster north and south of downtown.
It must be nice to be so 100% positive about your opinion …
I believe they would be able to design it so that it is faster, and it would absolutely be faster north and south of downtown.
But you are saying the trains would run on the surface. The streetcar is slower than the buses. Why do you assume a train would be faster? Put it another way — if we can make the train faster why not just make the buses faster? There is no fundamental reason why a train would be faster than a bus. If anything it is the opposite. A bus can avoid temporary obstacles. If some idiot “blocks the box” and can’t back up then a bus will go into the other lane. A streetcar can’t do that.
it would absolutely be faster north and south of downtown
Yes, but most riders would be worse off. Consider the H Line by segments:
1) Westwood Village to the freeway — Slow.
2) The freeway to downtown — Very fast.
3) Through downtown — Slow again.
Now imagine you have built your train as you suggested. The only place where it is really fast happens to be the place where the bus is fast already. More to the point, imagine if they ran both. The only people who would use the train are those that can walk to the station. Everyone else would just stay on the bus. It would never be worth the transfer. You have to walk to the platform and wait for the train. The train would never catch up.
“the point is they have no money, and if they can repurpose the existing track and plans for the connector it would be hugely less expensive than digging a new tunnel”
What’s the point of building it if it’s slower than a bus? Not building it would save 100%. The purpose of a metro is effective mobility, and that’s what Portland’s MAX is poor at downtown.
I could maybe see a surface alternative as a way of getting out of the bad transfers and budget-breaking cost, but it would have to have exclusive lanes and be significantly faster than the exiting streetcars. Unlike Portland, Seattle has the primary Link north-south trunk, so you have the option of taking it to U-District and transferring to the 44, or taking it to SODO and transferring to West Seattle (someday). Portland doesn’t have that option.
fyi the times were discussed in https://seattletransitblog.com/2025/02/28/ballard-light-rail-at-grade/
Route 40 takes around
* 40 min in traffic from ballard to downtown seattle
* 28 min without traffic
* 15~19 min would be the at-grade light rail
* 11 min would be elevated/tunnel light rail
It would take around 4~6 minutes to exit to the surface though https://www.theurbanist.org/2022/01/14/deep-seattle-light-rail-stations-other-st3-details-emerging-ahead-of-january-28-draft-plan-release/
So if just heading from ballard to west lake the at-grade light rail has around the same transit travel time.
Continuing further south or transferring is where the tunnel alternative has much larger time savings
I thought we had a surface alternative article somewhere. Thanks for finding it.
What I have been told is that the philosophy among transit planners in Europe is that you never build a new rail line unless it can be made faster than driving.
For new tram lines, that means excellent signal priority, dedicated lanes at minimum, and various other improvements that make it thus.
“ Yes, but most riders would be worse off. Consider the H Line by segments:
“1) Westwood Village to the freeway — Slow.
2) The freeway to downtown — Very fast.
3) Through downtown — Slow again.”
Ross’ point here is very important. Why is the current intention to build the light rail segment (SODO to West Seattle) that has the fastest bus speeds first — at a cost 3 times higher than estimated already? Let that sink in.
It’s pretty similar to the original st3 candidate plans. They had the west Seattle/ballard link travel on 1st avenue at grade
ah slow trolleys…more costly than buses and way less capacity. worse of both worlds.
Route 8 and Denny Way. My short term suggestion is for SDOT to change Yale Avenue; either close it at Denny Way or convert it to northbound only. The traffic bound for the southbound I-5 ramp could use Howell Street; Howell Street is carrying far less transit after ST2 Link. SDOT blog: https://sdotblog.seattle.gov/2025/08/28/route-8-updates-supporting-reliable-travel/
They may hint at my suggestion here: “Assess and develop options to distribute some general-purpose traffic off Denny Way with improved wayfinding, signal timing, turn restrictions, and changes to lane markings.”
Both decision making and Route 8 move slowly.
The late SDOT planner Jemae Hoffman described the issue with the shifts in the street grid 20 years ago.
My short term suggestion is for SDOT to change Yale Avenue
I agree. I think the simplest is to get rid of the eastbound slip lane from Denny to Yale/Stewart. Once you get rid of that, you can move the bus lane to the curbside. There is no reason to be in that curbside lane, as you can’t turn right at the next intersection. I would also ban right turns from eastbound Denny to Fairview. There is no reason to turn right there — the only people turning right there missed the right turn on Boren. So now you can run a bus lane from the bus stop on Boren all the way to Stewart, thus avoiding the weave. I think this would help alleviate the problem but it wouldn’t solve it. You would likely move the backup to Boren.
One of the big things it would do is set the stage for more BAT lanes. Once you move the bus lane to the curb, you only have one general purpose lane heading east. This gives you space to add a BAT lane the other direction (from Stewart to Fairview) without impacting traffic. Eventually the BAT could be extended both directions. When this happens you eventually end up with a lot fewer cars on Denny. The roadway still has backups (with cars turning right) but the bus can more easily avoid them since the other lane would have fewer cars.
https://www.seattletimes.com/seattle-news/transportation/mayor-harrell-dont-backtrack-on-ballard-west-seattle-light-rail/
Per Yogi Berra, this is deja vu all over again; Sound Transit was reset during Sound Move, ST2, and ST3; now, it is really needed.
The Seattle elected officials lined up and said: our Link lines are important. The SDOT changes can help with millions and months; the ST3 Link issues are billions and years. They may be addressing their electorate who voted yes in 2016 and the boardmembers from the outer four subareas. But the fiscal and feasibility issues of ST3 remain after the posturing.
They hint at willingness to consider the single DSTT. Bravo. But the ST3 deal that Dow cut included financial contributions from the four outer subareas to help pay for the DSTT2, supposedly needed for capacity. The deal will have be negotiated again.
Note that Trump 2.0 is a good time for reconsideration. That administration is very unlikely to approve FTA grants to ST.
There are also other issues with a single tunnel. To connect West Seattle Link to the existing tunnel is trivial. To connect Ballard Link is not. That is why I’ve changed my mind on Ballard Link. It makes more sense as a stand-alone line. This saves even more money. Not only do you not build a second tunnel but at most you run a single, non-service connection to the main line. This is much cheaper than branching the main line. Since the new Ballard Line is independent it might as well be automated and have smaller stations. This saves even more money.
Oh, and you might as well have a stub heading to First Hill. This means that if you do eventually build a second tunnel downtown it won’t be crap. It won’t be a second-rate copy of the original tunnel. You will do what every city in the world does when building a second downtown tunnel — they maximize coverage. Send the trains to First Hill (eventually) and you have a project worth the digging. But again, that can come later.
Now you’re spending spending 7 billion dollars running West Seattle Link through the existing tunnel. Remember, that is without the work needed to handle the extra train traffic. To be clear, it is definitely a lot better and a lot cheaper than the ST3 plans but I don’t think it is the best value. Replace West Seattle Link with bus service and you don’t have to mess with the existing tunnel at all. It can handle trains every three minutes. Speaking of which, spend a billion and bury the trains in Rainier Valley. Now you can run trains every six minutes all day from the East Side and South End through the tunnel. A little more work and you can automate the trains. The capacity of the existing system goes way up. Not only are the trains capable of running more often, but each train can carry a lot more people (no driver cabs). Hard to see why it makes sense to spend 7 billion dollars on a new line to West Seattle but we can’t spend 1 billion making trains in Rainier Valley safer, faster, more reliable and more frequent.
Kudos for the board for rethinking some of the poor assumptions that went into ST3. Take it is a step further. Rethink all of those assumptions. Is West Seattle Link the best value for transit riders? No. Does Ballard Link have to be light rail? No. Should trains from the South End be more reliable, safer, faster and more frequent? Yes.
This would be a political battle. A lot of the Ballard Link fans want a train that goes to the south end of downtown (without a transfer). You have to convince them that smaller, automated trains make up for that. Perhaps a field trip to Vancouver is in order.
You also have to convince West Seattle riders that investing in buses would actually be better. This could be a tough sell. Much of the region now assumes that the closer they are to Link, the better their transit will be. ST has perpetuated this myth. Convincing people that not only is this a better value but simply a better system will be difficult.
I thought that is where many of us were already. The Ballard line to a transfer at Westlake.
Is there any way to bury trains in the Rainier Valley without significant service disruptions? I can’t think of any unless you want to do a buried reroute along Rainier Ave, which seems difficult due to street disruptions and even more fiscally problematic than the elevated Rainer Ave reroute proposal under the current funding environment :(
The one option I can think of that preserves the existing (obviously worse than Rainier Ave) alignment with possibly shorter service disruptions involves box-jacking in vehicle/pedestrian underpasses, which the LIRR was able to do in a day per underpass. I suspect that on its own this would turn MLK into more of a highway in practice, and this would really need to be combined with the traffic calming and multimodal improvements that MLK is sorely missing anyway.
Closing some intersections instead of adding an underpass everywhere would decrease costs and service disruptions, but it creates a tricky pedestrian crossing problem―a 2-4 block detour is negligible for cars but not for many pedestrians, and ignoring this or even adding unwieldy pedestrian overpasses will result in people crossing at grade anyway (and then potentially being brutalized by cops for it, as we have seen…) Keeping at-grade pedestrian crossings at closed intersections with the automated pedestrian crossing gates that they’re planning to install anyway would probably still remove line-of-sight rules and the headway bottleneck, but at that point just adding vehicle and pedestrian gates with no grade separation at all would achieve a similar reduction in collisions and consequent service disruptions, and would likely also remove the headway bottleneck, at least if you believe the HTNB report.
The bigger question is whether or not gating every grade crossing would allow increased speeds, since if it does, it seems like that would give us nearly every advantage of grade separation at a much lower cost. The trains seem to run relatively slowly from S Lander St to S Royal Brougham Way in spite of the vehicle gates there, but the stop spacing is a bit lower and the pedestrian gates have not been installed yet so maybe that’s why.
In spite of the name, the “At-Grade Crossing Master Plan” drafts only really talks about the planned band-aid safety improvements which have somewhat improved pedestrian safety but have only reduced vehicle collisions by 3% so far (see section 2.1) and does not evaluate vehicle gates at all, let alone the tradeoffs between multiple comprehensive changes like vehicle gates vs full grade separation and their impacts on signaling, headways, and travel times. The draft master plan does mention a number of studies in table 6-3 that I cannot find anywhere that might go over this? This indicates that vehicle underpasses were not studied as part of the grade separation study, so my box-jacking thing is probably a pipe dream :)
@Jack — Yes, a lot of us favor that approach. I’m just saying most on the board aren’t there yet. They are only now talking about sharing the downtown tunnel with all of the different lines. They need to progress from there to the idea of just ending the Ballard Line at Westlake. They also need to think about replacing West Seattle Link with better bus service. It is a good sign that they are rethinking some assumptions but they have a long ways to go.
@June — There is this report: https://www.documentcloud.org/documents/24785840-wsp-mlk-separation-concept-estimates-st_task_order_7001_final_memo/. It suggests that elevating would be very disruptive. Building a trench less so. But it still might require closures (and bus bridges) or single tracking. It is an initial report and they make it clear that it wouldn’t be easy to maintain service.
In any event the first thing they should do is add gates. That will improve safety. I don’t think it solves the problem with speed and frequency though. The problem is that the cross streets are too busy. For something like Elliot it would be really easy to run trains on the surface. There are only a handful of places where cars have to cross and they can wait. The traffic lights always favor the trains. It isn’t as simple with MLK. But it makes sense to install the gates and then see if the trains can run faster (and more frequently).
That Long Island RR article provided by June indicates eliminating cross traffic could be done pretty quickly and without many disruptions. It’s a pretty neat technique.
In that report they looked at building pedestrian bridges and pedestrian tunnels. For the tunnels they assume a “Jack & Dig” approach. As they point out, these pedestrian tunnels and bridges have drawbacks. They are only effective if people use them. As June pointed out, there is a history of problems with them on that very street. Pedestrian overpasses are unpopular — pedestrian underpasses even more so. There are safety concerns, especially at night.
For whatever reason they never looked at underpasses for cars. My guess is this would be quite complicated. An automobile underpass is fairly simple. But it is all the turns that get complicated. For example consider Alaska Street (https://maps.app.goo.gl/acadzNoUTf64CcNz7). The simplest thing is to add a two lane tunnel underneath (one lane each direction). The far right lane is (on the surface) is for those turning right. Those turning left have to go under the street and then make a U-Turn at the next chance. Realistically this means taking three rights. This means pushing traffic onto side streets. That works, but it might prove to be unpopular. If we are willing to do that then we should ban left turns along MLK. A lot of the accidents and a lot of the delays are due to people turning left. Ban left turns, get rid of the left turn cycle and things are smoother.
Either way you still have pedestrians on the surface. You’ve improved things, but it is less than ideal. In contrast if the train runs in a trench then riders never have to deal with it. They only deal with the cars and the road is narrower. This in turn makes the street more pedestrian friendly which leads to further safety improvements.
[ad hominem]
Explicitly refusing to put bus lanes somewhere because there’s too much car traffic? I hope the next mayor takes a blowtorch to SDOT. Fire everyone and start over.
SDOT has actually done a lot of good things under Harrell. They have taken lanes in Westlake and Rainier Avenue. This is something that no other mayor (even McGinn) was willing to do. I never thought it was going to happen. It is quite likely there are people within SDOT who are willing to take a lane. I’m really not sure why Harrell isn’t listening to them. I wouldn’t fire everyone at SDOT but I would definitely tell them that they need to start taking lanes on Denny just like they did on those other streets. There is simply no reason for two general-purpose lanes both directions on Denny.
Greg Spotts did a lot of good work for SDOT, and was a surprisingly progressive pick for Harrell. SDOT’s current leadership does not have the same motivation to prioritize transit & pedestrians over cars.
Nathan: cars v. transit v. bikes v. peds. It is more complex. In several cases, the placement of PBL on transit arterials has left buses behind in congestion (e.g., Roosevelt Way NE, Broadway, Pike-Pine streets, 4th Avenue, NE 65th Street). It may be okay to slow cars down for safety, but it is strategically weak to slow transit even more than it was.
Nathan, Spotts became a streetcar foamer, rebranding the Center City Connector Streetcar as the Culture Connector Streetcar. It died of its own fiscal weight. See: https://www.seattle.gov/transportation/projects-and-programs/programs/transit-program/culture-connector
It may be okay to slow cars down for safety, but it is strategically weak to slow transit even more than it was.
Not only that, but slowing transit means the transit system isn’t as good (the buses are slower and run less often). A weaker transit system means you have made the city less safe (https://www.bloomberg.com/news/articles/2018-09-11/cities-with-good-public-transit-have-fewer-road-fatalities).
I’m disappointed that Spotts favored the streetcar. To be fair, this may very well have come from Harrell. It seems as if no mayor has the guts to cancel the project for fear of being seen as “anti-transit”. It doesn’t help that the Seattle Subway crowd is so obsessed with it. In any event I talked to Spotts but didn’t get a chance to talk to him about the streetcar. It was clear that we shared the same goals for the city. It was refreshing to here an SDOT head so disinterested in traffic congestion. Normally there is talk of “striking a balance” or “we don’t want to screw up car traffic too much”. There was none of that — his model was Amsterdam.
Which may have been the problem. Amsterdam has an outstanding bike network. It really is a model for the world. It also has a good transit system. But from what I can tell, more ride the trams than ride the metro. The trams are certainly more prominent if you travel through the heart of the city. It is a very beautiful city (Rick Steves called it the most beautiful in Europe) and so there is a lot to be said for traveling on the surface. Buses have a smaller impact (from what I can tell). Given all that, I can easily see how Spotts would be overly enthusiastic about trams. They make sense for a city like Amsterdam. For Seattle, not so much.
If people think the new fid alignment is a mess; just wait until st tries to figure out dstt2. Its time to kill it. Im a little surprised at all the shake thrown on MAX. it beat link into service by 23 years. It has a far superior airport station. It covers a huge part of the metro area . Hillsboro to Gresham east west. Columbia River 5o mil2aukie/claxkasnas north south. It makes like look like a joke.
too slow. not enough capacity. getting it done wrong first isn’t good long term. see Rainier valley segment.
MAX capacity really isn’t particularly constrained. When they need to they have done 3 minute headways. It’s a bit crowded after events but so would anything.
The biggest problem is the highway parallel running, which drastically reduces the demand over what it could be if it went where people actually want to go. Eg: Tacoma Park and Ride vs center of Sellwood retail and residential area. Clackamas Town Center station isn’t really that close to either the shopping center or the hospital and retail area on the other side of the freeway, and walking those several blocks is a time consuming undertaking thanks to the freeway ramps. One of the biggest transit destinations on Route 33 is the Fred Meyer at Oak Grove, but MAX falls 3 miles short of it.
So, not too much different than some of the problems with the outer end of Link.
A tunnel downtown might be nice, but for the money required an awful lot of other speed improvements could be made.
Paul Ventresca,
You seem to have a lot of opinions on transit and I guess that’s your right. I’ve actually ridden a lot of transit in a lot of different cities (and I’m not alone in this here on this blog by any means). Here’s what life in a big city looks like…
There’s always a lot of traffic. Bumper to bumper traffic is pretty much the norm. Adding transit does nothing to solve this, traffic is just part of life in the city.
Transit works the best from downtown to the airport, around universities and in few high dollar urban neighborhoods. Get off the beaten path to lower rent ‘hoods and transit doesn’t work near as well.
It doesn’t take a genus to point out the transit failures of any city just like it doesn’t take a genus to trash expensive restaurants on Yelp!. Doing so doesn’t make a chef however… nor transit expert.
Big City politicians lie…. Whatever bullshit the pols were shoveling before the ST3 vote is now turned into compost. Most of the good projects, the ones with high ridership and the ones that made sense are already built.
Now that all the high ridership rail lines are near finished…. Sound Transit has run out of money… imagine that!
“There’s always a lot of traffic. Bumper to bumper traffic is pretty much the norm. Adding transit does nothing to solve this, traffic is just part of life in the city.”
I’ve spent a bit of time in London, Berlin, São Paulo, Rio de Janeiro, Copenhagen, and a couple others.
I only once saw congested traffic in Berlin, but mostly didn’t seem to be a problem. City center streets didn’t seem to have problems in Copenhagen or London either, but there were a few problem areas on the outskirts.
Såo Paulo has infamous traffic, but that’s mostly because their transit system is way behind where their population is. They’ve had to do a lot of catching up in the last few years, after attempting to do everything with highways and buses.
Rio’s metro seems to do a decent job of keeping huge numbers of people from clogging many of the roads at peak periods.
Glen in Portland,
I have family who live 50 miles outside of London and it’s a 2 hour trip to get there. Driving straight though, driving to a transit hub then taking a train or bus… doesn’t matter, it’s 2 hours one way. Transit is popular not because it’s faster or cheaper, but there’s the problem of what to do with the car in London…. if you have the money, cab is the easiest.
And it’s the same deal in Frankfurt and Hamburg when I’ve stayed on the outskirts of both cities…. it’s just not easy to get around with millions of people around you. Building transit (and roads) is expensive and cities that are doing it are growing…. so as a city adds transit and people at the same time, will getting around really become faster? Or maybe the best case scenario is not making transportation worse?
Portland certainly is a recession currently I’d guess? Seattle is still growing a little, but as the National economy cools, I’m sure a recession is on the horizon. Both cities have education systems that are broke and a nasty housing crisis. Oregon has a big problem with the State pension system being underfunded. The chances the State of Federal government are going to rescue PNW cities is slim to none.
And yet there’s still endless talk about Sound Transit’s second tunnel in Seattle. I’m not saying it won’t happen, but it shouldn’t.
upgrading the sounder even as it is not even at 50% precovid levels is a travesty. Cancel it. We’re in a budget crisis and the idea no new taxes and no cancellations or delays is illogical.
Screw over South King County again why don’t we?
It’s okay, we’re kinda used to it.
Sounder is actually a good service (at least the south one, haven’t taken the northern one). It’s faster to Tacoma Dome than TDLE will be from Chinatown
Cancel Dupont.
Cancel longer platforms.
Cancel any parking structures that haven’t yet been built.
Cancel TDLE.
Implement every 10 minute expresses from Commerce to Federal Way, with half going to Seattle with HOV 3 and new on and off ramps.
Cut a deal with State, Amtrak and BNSF to help buy the line between Tacoma Dome and King Street outright and improve the track to support faster, 45 minute or quicker runs. Lease slots back to freight, but develop a real utilization plan for that incredibly valuable, but mismanaged resource.
Maybe even get Metro to pitch in, because with a S Sounder with increased frequency and span, they can basically just run feeder buses in the Kent Valley and cancel everything else.
Cancel N Sounder, but extend S Sounder to Ballard with a few stations in between. Downtown. Seattle Center. Interbay.
Good ideas Cam!
And all of these could be done in reasonable time frame I believe?
Sound transit has big problems with budget AND time.
“extend S Sounder to Ballard with a few stations in between” would be a multi-decade, multi-billion dollar project on the scale of a Link extension.
ST3 already has a billion in funding for higher frequency on the Tacoma to Seattle segment. When I was at ST, the rule of thumb was that $1B could get you 15 minute frequency (currently 20 minutes) on roughly the same span of service. Buying that ROW outright would be >$10B … at that point, it will clearly be cheaper to finish Link between Tacoma and Seattle.
Cancelling projects can always be “quick,” but there is no political support to delay TDLE or the ST2 South Sounder garages. The only suggestion that is helpful and could be done quickly is HOV3 on I5.
AJ – Why would extending Sounder and adding stations, mostly (or completely, if you want to save time and money) in the areas where it runs on the surface, be so time consuming and expensive?
10 Billion is about what you would save by just cancelling TDLE and the other things. If you canceled Ballard Link Extension in the process, where the last price tag I’ve seen is 11 billion and rising, it’s substantially cheaper, and benefits a vastly larger population that has far more room to build densely around stations than Ballard does.
Ballard’s a backwater that is built-out within walking distance of the proposed station with townhomes and low-slung apartments almost completely. The same cannot be said of downtown Kent, Auburn, Sumner and Puyallup. Not to mention Sunset Hill. The train crosses the ship canal almost as far west of central “historic” Ballard (24th) as the proposed light rail station is east. Upzone Sunset hill to 10 stories, and see a huge boom in dense building with incredible views.
You can eat a lot of lutefisk for 11 billion.
They are going to have to replace the Ballard Bridge eventually. When the do, upgrade the RR D to complete dedicated lanes and eliminate or mitigate with queue jumps the few lights and crossings it needs to deal with. Ban mega-yachts from the locks, and your cost overruns are pretty much taken care of (assuming they do the obvious, and West Seattle is cancelled as well).
extend S Sounder to Ballard with a few stations in between.
OK, consider those stations:
1) Belltown. This definitely has merit. I could see how this would save a lot of riders a substantial amount of time. It fits in with the schedule as well (riders from say, Auburn would commute to their job in Belltown).
2) Interbay. I just don’t see it. There aren’t many jobs around there. There are some apartments but not enough to justify the cost of service (even if construction was free). Ridership would be minimal unless you ran the trains often (like a metro). That is unrealistic.
3) Ballard. The problem is it doesn’t really serve Ballard. You thought the plans for Ballard Link were bad — this is worse. Keep in mind, the area doesn’t even have Metro service (despite the rest of the 44 being very strong). That is how desolated it is. To be fair, there are a handful of buildings on Seaview but only a handful. There is very little room to grow as well.
I could definitely see an extension to Belltown, but otherwise it just doesn’t make sense for commuter rail. Neither does North Sounder, for that matter. North Sounder should be replaced by bus service. To a large extent it has. Both Everett and Mukilteo have very good express bus service to Lynnwood right now. The only city that doesn’t is Edmonds. Replace North Sounder with an express bus from Edmonds to 185th Station (https://maps.app.goo.gl/UGdFm9mRububq5UF8). Make the usual stops along the way in Snohomish County along with a stop on Aurora. Run it every fifteen minutes midday. Cancel the 909.
Ballard’s a backwater that is built-out within walking distance of the proposed station with townhomes and low-slung apartments almost completely. The same cannot be said of downtown Kent, Auburn, Sumner and Puyallup.
Oh come on. That is ridiculous. Ballard is similar to Capitol Hill. It is a dense area with a mix of uses. The station placement (at 14th is terrible) but a station in Golden Gardens would be much worse. There is an obvious solution for Ballard Link — move the station west a few blocks (20th would be ideal).
Kent, Auburn, Sumner and Puyallup are all sprawling suburbs. Even if they someday resembled Ballard someday, not that many people would ride the train. It is just too far. Proximity matters. A lot.
Oh, and you can’t build on the west side of Sunset Hill. It is a steep green belt.
When I was calculating all the curve radii, cant, and speed limits for SPIRE back in the day, I briefly explored how the passenger railway could be expanded north from King Street Station to Seattle’s modern business district, which is an obvious expansion of a regional asset. In a better world, the S Line would have a stop near Westlake. Anyway, does the line get a new tunnel under the city like the RER system in Paris, or does it leverage the existing railway tunnel and shoreline route toward Ballard?
These are tough fantasy questions. It is also a good planning exercise. As I generally feel we are incapable of making strategic transit investments, I decided to go with the leverage option.
In that framework, I’d propose we build the first northerly station as close to Pike Place Market as possible, if not virtually underneath it, perhaps with a careful expansion of the existing tunnel near the portal. This would correspond with a review of fixed train lengths. Then consider: a station near Harrison/Thomas streets, a station under Dravus Street, having the railway curve around Nickerson St to a rebuilt railway drawbridge over to NW 41st Street, then curving north along existing rail ROW to a Ballard station and layover trackage near Shilshole Ave NW. That’s probably as good as it gets with the leveraging option.
They are going to have to replace the Ballard Bridge eventually. When the do, upgrade the RR D to complete dedicated lanes and eliminate or mitigate with queue jumps the few lights and crossings it needs to deal with.
They don’t even need to wait for a new bridge. Just as they are reimagining Aurora they can reimagine 15th/Elliot. I’m thinking center running buses from Holman road to Mercer. The bus would stop under Dravus (instead of exiting there). The only bus stop on Mercer is at 3rd West. The road is wide enough to support a center bus stop there. Next the bus uses Queen Anne Avenue (inbound) and 1st Avenue North (outbound). These are one-way streets. Don’t make the same mistake they made with the RapidRide G downtown. Have the buses (all the buses) run contraflow through there (between Mercer and Denny). It is a short enough section that the buses don’t need to leapfrog. They just need to avoid traffic and they only need one contraflow lane for that. That gets you to Denny. At that point the bus switches over to running curbside. It uses the BAT lanes that were added to the Metro 8 (along all of Denny). Then the bus uses the transit mall on Third Avenue (now extended all the way to Denny).
If we are willing to live with fewer lanes over the bridge, none of that is very expensive. You’ve got to add a pedestrian pathway down to the surface at Dravus (along with a new traffic light) but otherwise just a lot of red paint. You are probably looking at a travel time of around 25 minutes from 15th & Market to 3rd & Seneca (similar to how long it takes at 1:00 am).
I still support an automated line (to Westlake) but I can see some trade-offs:
1) Buses can go more places in Ballard. The rail line is only going to have one stop in Ballard. It could be extended but that would be very costly. In contrast the RapidRide D already serves a lot more places. It would also be trivial to run a bus down 24th and onto that same pathway. If a rail can be justified to Ballard, then a couple all-day bus routes along that same corridor can be justified.
2) Buses can extend deeper into downtown. The only way this can happen with Ballard Link is if they build a second tunnel or interline them. The former is very expensive and the latter means the frequency of Ballard Link would be limited (even if it is automated).
3) The buses can run on the surface. This saves riders time.
4) From a political standpoint it becomes easier to also eliminate West Seattle Link. A lot of the proponents of West Seattle Link make provincial arguments. They talk about “deserving” rail, as if it is automatically better transit. If Ballard retains a rail line and they don’t have one, a lot of people from West Seattle will be pissed.
5) In the long run it may be easier to build a better project — an automated line from Ballard to UW.
On the other hand:
6) The train would be considerably faster. For a trip to downtown, a train would save about ten minutes compared to a bus (even if you completely avoid congestion you still have to deal with traffic lights).
7) The train can easily be automated. Eventually we hope buses can be automated but that isn’t a given. (Yes, I know there are pilot programs but automated trains are now the standard — automated buses are not.)
8) Unless you build a taller bridge (which seems highly unlikely) you still have a lot of openings. You are not going to be able to ban sailboats and large industrial vessels from Lake Union. By running in its own lane you avoid the worst part of the opening (the backup). But you still have the possibility of a significant wait.
9) Small, frequent, automated trains still have more capacity (per train) than our (big) buses. This lends itself to transfers and the associated savings. For example buses from Magnolia would no longer go downtown if they ran a train. Riders would be expected to transfer. This means more buses from Magnolia to the UW and better service within Magnolia. Again, this is based on the assumption that we still have people driving the buses. Automated buses would probably change the equation (although it still may be advantageous to run fewer vehicles downtown).
So yeah, I could see it, but unlike West Seattle I could go either way. With West Seattle I feel like a bus solution is just better (regardless of cost). With Ballard I don’t see that. There are clear trade-offs. A lot depends on how each project is implemented. So far, the station placement for Ballard Link looks terrible (especially in Ballard). But they could make mistakes with a “BRT” solution as well (and chances are, that is what they would call this even though it should be designed so that any bus can take advantage of these improvements).
a rebuilt railway drawbridge over to NW 41st Street, then curving north along existing rail ROW to a Ballard station and layover trackage near Shilshole Ave NW.
Taking over the Burke Gilman bike path would be a tough sell. I think they looked at that for a Ballard-UW option and almost immediately rejected it. To be fair that was for the whole path, this would be for only a section.
Building a new bridge would be expensive. Might as well do it right. Build a bridge (or a tunnel) at about 24th. Then curve onto Market (or one block north). Add a couple stops in Ballard (one close to 24th, another close to 15th) and you are heading towards the UW. It is basically Ballard Link but with Belltown instead of South Lake Union and Uptown. That would likely lead to very big savings (although you still have that pesky bridge/tunnel over the canal as well as whatever stations you add in Ballard).
“Ban mega-yachts from the locks, and your cost overruns are pretty much taken care of”
Unfortunately, federal law says otherwise. Legacy laws, passed in an era when boats were the dominant form of transportation (cars and trucks did not exist yet), prohibit obstructions to navigable waterways. That’s why any bridge must either have a drawspan or be as high as the Aurora or I-5 bridge.
There is some wiggle room for negotiating with the coast guard, for example, in prohibiting drawbridge openings during rush hour for all but the largest ships that can pass through the locks. But, the coast guard is primarily concerned about boat operators – including owners of super yachts. They don’t care about the efficiency of public transportation because that’s not part of their mission.
Isn’t the preferred station at 15th and Market? That seems reasonable to me; it has slightly worse density than 20th and Market but better opportunity for bus transfers and a potential northward extension.
“3) Ballard. The problem is it doesn’t really serve Ballard. You thought the plans for Ballard Link were bad — this is worse.”
That really depends on where the decision is made to put the north station end.
Given the budget of Ballard Link, I would rebuild the old Northern Pacific bridge that is partially still across the ship canal east of the Ballard bridge, and re-connect the main line across the Fred Meyer parking lot to the Ballard Terminal Railroad. Build a parking structure above the new line at the Fred Meyer parking lot. Then, add a station at the Ballard end as well. You’d need a third station at Golden Gardens for layover, because nothing in Ballard or Fremont proper can hold a full length Sounder train.
Troy, BNSF is not going to give you frequent train slots through the Seattle tunnel which is one of the first segments of the Northern Transcon to Chicago. How many times have you caught Sounder South or Amtrak at King Street when there was no freight train passing some time during your wait? Not very often. That’s a busy railroad.
Ballard may have to be content with a juiced version of the D as Ross detailed, but in no case will South Sounder be extended to Ballard, though Glenn’s idea for a fishhook via the old NP bridge location and two stations is brilliant.
Should have said “Cam,”
Tom, I once managed King St. Station. I have seen the tunnel be infrequently used at all times of the day—and the opposite condition, too. I understand there is surplus capacity at present. Negotiations would be dependent on BNSF planning and operations to Cherry Point, over Stevens and Stampede passes, and through the Gorge, as well as the manner in which passenger trackwork and facilities were laid out (we should disrupt or block the freight mains as little as possible—if at all). I don’t presume to know if BNSF would be opposed to S Line slots through the double-track tunnel and I doubt the question has been seriously asked and entertained. Remember, this was just a planning exercise—a what-if.
I may be mixing up who you’re directing your comment toward. I proposed using the old Northern Pacific pathway to Ballard, via a rebuilt drawbridge, although I think a 500 to 700-foot island platform and separate layover track could fit along Shilshole with some sacrifices. I don’t necessarily agree that tracks need to go farther than that, but if they did so be it.
It’s all make-believe anyway.
You’ve had some interesting jobs, Troy!
That must have been a fun exercise.
And how do I get a yard sign…. ;)
Ross – Proximity can’t just be measured as the crow flies. You have to take into account resistance to flow caused by barriers and bottle necks. And Queen Anne hill is a massive barrier, creating 2 major bottlenecks.
If you are trying to get to King Street Station, would you rather travel the 7 mile trip from Ballard takes just as long as the 20 mile trip from Kent. And that’s by car. It’s substantially quicker from Kent by transit, at the moment. When the Sounder is running. Now I’ll grant you Ballard has more breweries than Kent. Which is not inconsequential. But Kent also has it’s own charms.
And then the question is: Proximity to what? There are a lot more destinations people want to get to besides a desk in downtown. Covid and WFH flattened and spread usage. Right now Kent is essentially inaccessible from Tacoma, Puyallup, Auburn and Sumner by transit. That should change.
@Cam
Westlake center, not CID, is roughly the midpoint of downtown Seattle
It would be great to get more Sounder S service but Kent is not Ballard. Central Ballard is a sea of 4-6 story multifamily housing; Kent’s station area is still primarily low-slung strip malls and SFHs
To some extent it is irrelevant that the Valley cities are not (yet) full of densely populated core neighborhoods. For one, they have the wonderful old bones needed to accommodate the type of urban growth that is encouraged by many people on this blog. For another, the tracks that run through them should be passenger-dedicated outright for a variety of profound safety and environmental concerns. We don’t want a Lac-Mégantic disaster in Puyallup.
We have a duplicative, underutilized freight rail mainline that literally parallels the BNSF corridor in this sensitive area. It is frustratingly obvious low-hanging fruit.
Agreed. But Kent is also the 6th largest city in the state. And Ballard is not. It might be bigger than Moses Lake, by a smidge. I think our housing crisis is the biggest threat to cities at the moment. We are teetering on the brink. So I look at transit with an eye towards potential infill development. And Ballard basically has very little potential. It’s already primarily a sea of townhomes with a few 5 over 1s on the arterials and a tall apt building or 3. But the focus should be on the sea of newish shoulder-to-shoulder townhomes that will simply not pencil for a developer to knock down and build higher. Ballard is pretty close to done with potential for infill housing near the station, unless the economics change substantially. And it’s basically a dead-end, where a train line would have to twist itself in knots to extend anywhere useful. The area around Kent Station is practically greenfield in comparison, and it’s connected to a string of downtowns just like it.
@Cam
There are massive housing projects starting and completing in Ballard right now. There are plenty of underdeveloped lots; look to the east of 15th and Market. It’s far from fully developed.
Besides that, how does it make sense to avoid destinations that are already dense and developing?
https://www.seattleinprogress.com/
“Ban mega-yachts from the locks”
Some of them are going to yacht-repair companies in Lake Union. The city doesn’t want to lose those blue-collar jobs.
Not avoid. Right size the transit. Don’t spend 11 billion on 29,000 people in a town with little opportunity for further growth in the middle of an existential housing crisis, and situated at the end of the line.
Instead, expand, improve and upgrade RR to True BRT level of service, and spend the 11 billion where it will do far more good than saving a few minutes for those fortunate few for which a train won’t actually be worse by forcing a long walk or transfer.
But Kent is also the 6th largest city in the state.
That is completely irrelevant when it comes to transit. Density matters. Proximity matters. How they arbitrarily draw the city boundaries does not. Think about. Would the case for Ballard be stronger if was still a separate city? Of course not.
Proximity can’t just be measured as the crow flies.
Agreed. It is more about travel time, not the actual distance But proximity is shorthand for that. Basically fewer people will take a trip that takes an hour versus a half hour. That is true all over the world. The Shinkansen is extremely fast. But it is less popular than slower mass transit. That’s because the trips within the city are a lot more popular (and even though the bullet train is faster, the trips take longer) .
There are other factors. Average speed is one. I can drive quite quickly to places like Kent and Auburn. They are close to the freeway. The average drive time is very fast. In contrast it takes a long time to get to Ballard. Ballard isn’t close to any freeway. This means that transit is more likely to be popular.
And then the question is: Proximity to what?
Exactly. There are a lot more destinations inside the city than outside it. I’ve never taken transit to Lynnwood. Or Mountlake Terrace. Or anywhere in Shoreline. To be honest I’ve been meaning to (I write for a transit blog after all) but it seems more like an assignment than anything else. My wife isn’t interested either (I tried to talk her into a road trip — no luck). In contrast, without prompting, my wife told me she was excited about Northgate Link opening since it would be easier to get to Capitol Hill. We’ve taken plenty of trips to Capitol Hill and the UW. This is anecdotal evidence but studies repeatedly show this to be true. Suburban destinations are simply not popular for those who don’t live there. That is why ridership between those suburban destinations (e. g. Shoreline to Mountlake Terrace) are so small even though they are relatively proximate (by every definition). Again, this is common around the world (but especially in the US).
Maybe if Kent, Auburn or Puyallup were major cities with urban centers that could compete (or better yet exceed) that of Capitol Hill then I could see it. But even then, it seems doubtful. I remember talking to someone who lives in New York. I mentioned that one of the cool things about the East Coast is that you can easily get to other cities. I asked him if he had ever just visited Philly for fun. He looked at me like I was nuts. It was clear what he was thinking. “Why the hell would I do that — I live in New York”. While not quite the same thing, a lot of people have the same attitude towards the suburban cities.
Last time I was in Kent it was for Jury duty. Last time I was in Puyallup I was visiting a friend in the hospital. I will try and take the grandkids to the fair this year though but in general I just don’t see me taking a lot of trips. I know that makes me sound like I’m a snob but this is just the way it works. More to the point it works within those areas. Not that many people travel between Auburn and Puyallup. Or Kent and Sumner. American suburbs are just not that popular for people who don’t live there. To be clear — I’m not criticizing people who live there. They need decent transit too. But it means that the travel pattern is very different and better suited for express buses, not a metro. History has shown this to be true. Even when they disregard this standard idea and make a system that is clearly geared towards serving the suburbs (e. g. BART) the vast majority of trips are made within the urban core.
For one, they have the wonderful old bones needed to accommodate the type of urban growth that is encouraged by many people on this blog.
Do they though? Take Auburn for example. Yes, it has a normal street grid. As a result, it looks like a small town (e. g. Darrington). But it looks like a small town that never really grew up. There are no old brick buildings. It certainly isn’t like a small city in Europe. It isn’t even like Edmonds, let alone Tacoma. It is more like Lynnwood. It is less than a mile from the intersection of two freeways (SR-18 and SR-167). It is within spitting distance of a big set of ramps serving SR-18. It has some new big buildings and over time will have more. It could transform itself into someplace more appealing but it will take a lot of work. Kent has started to do that. But that doesn’t mean it is leveraging something from way back when. You can do that just about anywhere.
I want to be clear. By all means we should try and improve Sounder. But we have to be realistic. The vast majority of riders will commute from these stops into Seattle. They will take the 8:15 into the city. If the train is on time they will get to work by nine. (Sorry — a little musical reference there.) You can look at ridership of the Sounder. Even *reverse peak* has low ridership. Think about that for a second. If I worked in Kent or Tacoma, you bet your ass I would take South Sounder. So would a lot of people. Seattle has a lot of people — many of whom don’t own cars. Then you have all those folks who want to visit Tacoma (a real city that does have good bones). Yet will all of that, you have hardly any riders making the reverse direction trip. It is almost all people taking the traditional trip to Downtown Seattle in the morning, and back in the evening. That is why the best option is to run complementary bus service in the middle of the day. Yes, if WSDOT buys out the tracks (or double tracks) to run trains to Tacoma then by all means we should leverage it. But until then it makes the most sense to just a few buses in the middle of the day since you they won’t be full.
Isn’t the preferred station at 15th and Market?
Right now the preferred alternative is officially 14th & Market. But yes, 15th and Market is a decent compromise. It isn’t as good as 20th, but it is OK.
I was unclear. By “old bones” I meant a valuable pre-auto street grid. All the Valley cities have them, although Auburn, specifically, has some unique limitations to its grid. Still, north of SR-18 there is a strong zone of walkability with significant capacity for residential growth.
There is lots of unrealized opportunity in South King and Pierce. And while that is true, what you wrote is also true, Ross. Without a muscular intervention to bring about this opportunity—and there is zero evidence of that happening—just run a few more trains around the peak period and express buses the rest of the time.
“They are going to have to replace the Ballard Bridge eventually. ”
I really think a second bridge with low-clearance should be considered even if it means it opens many times a day.
Currently, Ballard bridge is not bike-friendly. The new one will certainly have higher clearance to open even less frequent than today, which means more hill to climb.
The Ballard lock is too busy and you cannot ride through. There isn’t really an ideal bike crossing west of Fremont Bridge.
On the cross-overs downtown, it seems that since the trackway north of IDS for a block is inside a box structure rather than twin tubes a scissors cross-over could be installed there. Now that only shortens the distance to the south end of a single-track zone, but it is two stations closer than the reversing track south of Stadium and one station closer than the cross-overs at Judkins Park. Importantly, it maintains the opportunity for a center platform to provide convenient out-of-direction transfers at Pioneer Square.
I recognize that the fact that nobody has mooted this means that somebody sometime may have decided it was not possible and I just don’t know about that decision. It seems pretty self-evident that it’s the least invasive place for a pair of cross-overs. Yeah, sure, you have to cut both tracks in two places to install turnouts, but even ST has shown that it is getting better at doing that. Yay!
Building a pair of them between Pioneer Square and Symphony would mean building an entirely new “station box” around the existing tubes, then — very carefully — removing the top 3/4 of the compression rings through the box to expose the trackways for the connections. It would take several years start to finish, with the street opened for at least six months before excavation would be deep enough to deck and continue work.
This certainly can be done but it is super– invasive, both to the operating railroad and the surface above. However, it is worth the cost and disruption, because it preserves the Symphony center lane for an out-of-traffic “siding” for reverse moves to a future non-revenue connection between Ballard-Westlake and The Spine at Third and Pine. By “storing” short northbound trains headed into the connection in the center track until there is a two-minute window between southbounds, there does not need to be a turnout and level crossing added at the Third and Pine curve. A single-track connection to the southbound track becomes the sole needed connection between an isolated Ballard-Westlake system and “the outside world” [i.e. heavy maintenance at Forrest Street].
Even if the trains in Westlake-Ballard (and eventually UW/U Village) are shorter, automated ones, they can either be towed by a crewed car from inside the connection track to and from the MF or have some sort of fold-out operator station that is used only for such non-revenue moves on the non-automated spine. This would significantly reduce the needed Maintenance Facility on the automated line. This is especially for an automated system since a couple of trains would run all night long, making the parking requirement a bit less severe.
Let us not forget this is a short-hand for the Blog’s “consensus” solution several months ago to the high costs of Ballard-Downtown (from smaller stations possible with frequent automated trains) and the stupendous cost and small utility of the second tunnel between IDS and Weslake. We have to keep saying it over and over: “Automate, shorten, and isolate Ballard-Downtown” to keep it within the cost constraints that the system faces, especially in a time of rapdily increasing steel prices driven by tariffs and the opportunities they give American suppliers.
Ah, I see Ross more completely outlined the consensus solution in the post a few above. Thanks, Ross.
We have the center lane in the stations inherited from the bus-breakdown use, so we might as well use them for a crossover since we’re not using them for anything else.
Wyes at both ends of the holding track could be included to create as a double cross-over, albeit a bit more slowly than simple scissors. It would be like the pocket track at Northgate.
But please don’t advocate for sacrificing Symphony to a simple scissors. Yes, in some — maybe most — cases a scissors would be just fine for crossing over a reverse move to the connector, but without a holding section you would lose flexibility and reliability. A northbound train would have to leave Forrest Street at just the right time to tuck through the cross-over when both platforms are empty and there’s no southbound ready to leave Westlake.
Choreographing that without delays at three preceding stations would be challenging at best.
I don’t understand what you’re talking about. I don’t know the difference between different kinds of crossovers, or what a holding section is, or how it affects flexibility, or what Forrest Street three miles south has do do with it. I’m just going by what others have said that it’s feasible; it would just mean closing the station ST needs to single-track Link downtown. There’s a tradeoff with that in terms of RapidRide G transfers and walking to destinations, but since the crossover wouldn’t be used 99% of the time, it wouldn’t affect station functionality most of the time.
Mike, you are correct that simple cross-overs, whether scissors or a pair of opposite handed simple cross-overs, will not affect station operation 99% of the time. What I’m talking about is a holding track in the middle of the station for use by non-revenue trains moving between Ballard-Westlake and the Maintenance Facility (which is on Forrest Street).
Ordinary cross-overs — which do not allow for a train to be “parked” within them — would be absolutely fine for out-of-service time periods. But when DSTT1 is in full four-minute headway operation, getting a train from the south wall at Symphony to the north wall at the Third and Pine curve where the junction would be will be a hard thing to co-ordinate without interrupting the ordinary flow of traffic. Both tracks at Symphony and the southbound track between Westlake and Symphony all have to be open to make the movement. Any train already in the station at Westlake would have to wait for the non-revenue move to enter Symphony, make the cross-over, travel north to the connection (not yet existing) at Third and Pine and clear the turnout there before it could proceed south.
But if a holding track is provided — that is, the northbound diverging turnout is right at the south wall of Symphony, but the northbound merging turnout is at the north wall, that leaves at least three car-lengths of straight holding track between the two platform tracks. A train parked in the tunnel just south of the northbound platform can move immediately after a train occupying the northbound platform starts to move, regardless of what is on the southbound track in the station or between Westlake and Symphony. It then waits until any train that has left Westlake or one already in the station clears the southbound platform, and then has about a minute and a half to clear the switch at the curve.
By putting “wye” switches in the middle track at the north and south ends, the holding track can be used as a cross-over, although it would not be a quick as a simple scissors or and end-to-end pair.
Tom, if the station is long enough, it is optimal to do what you say with a median siding long enough for a train. There may be a structural reason why underneath, but there was no problem at CID Station for their median siding although it doesn’t connect on the north end. And it would possibly remove the need to close a full station.
Another option would to cross west to east at one station and east to west in the other. ST may be able to offset the waiting area by a few feet to get just enough distance for the siding too.
For reference, a simple crossover can be an “X” (double crossover) or look more like an elongated “>” or “<“ (simple pair of single crossovers). In either case, it wound be hard to keep the station opened as the needed length would cut into the rtrack where a platform end may need to be.
Al, yes, I don’t think any configuration of cross-over(s) can allow trains to serve the platform at a station being used as a single-tracking start or end, even the center holding track version. That’s an excellent argument for building the “new trackage” station box between PSS and Symphony. But man, that is a lot of money and disruption while it’s built!
When King County built the tunnel, it was really making a bet that buses were going to be the permanent solution. Otherwise they would have built short extensions at one end or the other of a couple of the stations that could host future scissors cross-overs. Penny wise, pound foolish.
So far as the center holding track, it would never be long enough to hold a full four-car Spine train. I’m assuming much shorter Ballard-Downtown automated trains would be using it, not the long ones. I think you get that I mention it to clarify for the larger readership who might not know of the complete “consensus” design for Ballard.
I took a walk around Marymoor Village Station the other day, and I couldn’t help but feel as though the who station was hideously overbuilt. The shear size of the parking garage is a monstrosity, as is the the fact that the exit ramp has *three* car lanes, plus a traffic light. Then, there’s the ground level of the garage, which has four bus bays for just one bus route, with service every 30 minutes, and no plans, to my knowledge, for additional routes. I also noticed that the multiuse path underneath 520, of course, has a fence to prevent people from walking the more direct path into the station – instead, they have to detour around and press a beg button to cross the 4-lane garage access ramp. Seriously, what is the point in this fence? Safety-wise, crossing a bus access road with almost no buses feels safer than crossing the car access ramp and having to deal with turning traffic.
I can understand some parking being there, but does it really have to be that big? Could they have simply built a surface lot on the same land footprint and built a small layover facility for route 269 on the street?
Another annoyance about Marymoor Village station I noticed. When a train is approaching the station from the west, the crossing gates go down, even though the train won’t actually reach the crossing area until after it finishes stopping at the station. This needlessly forces riders who should be able to run and catch to the train while it’s stopped to wait an additional 10 minutes for the next one, something I don’t think the crossing gates do at the Rainier Valley stations.
I always saw the garage as being massive. I’m not sure how much the cross-lake connection will add, but right now there isn’t much reason to park and ride there since you’d might as well just keep driving to your destination if it’s on the Eastside.
I think this could change if the 545 served it instead of Bear Creek P&R, though this would lose the stop at the Target/Home Depot shopping area. I honestly think Bear Creek P&R shouldn’t really exist anymore.
“I honestly think Bear Creek P&R shouldn’t really exist anymore.”
I agree. One thing I wish Metro would understand is that a park and ride lot is only really useful if it offers either Link service or an express bus that enters the freeway directly adjacent to the P&R and drives a considerable distance. Before COVID, route 268 made Bear Creek P&R somewhat useful. Route 545 is insufficient to make the P&R useful, as it’s much slower than the alternative of just driving to Redmond Tech Center to catch the same route 545 bus. Local buses are useful, but they do not make a park and ride lot useful – nobody with a car is going to park their car to ride the 250 to downtown Redmond or Kirkland, for example. Since running a route like the 268 is no longer a good use of bus service hours, having a park and ride lot just for local buses (and maybe a few vanpools) isn’t a very good use of land either. Especially with the monstrosity of the Marymoor Village parking garage so close, with at least 20X the parking capacity.
This also brings into question why Metro is so insistent in detouring route 250 to serve Bear Creek P&R. For people that live up on Avondale Road, this detour is very annoyance, as it adds over 5 minutes to every single bus trip in or out of your home. If you don’t want to pony up the money for your own car, this means every single round trip out of your home takes 10-15 minutes than it needs to, just to serve a P&R lot that has nothing there except local buses that nobody with a car has any reason to park for.
What I suspect the 250’s Bear Creek P&R stop does get some use for is people without cars, walking from the P&R to nearby big box stores, such as Kohl’s, Fred Meyer, Target, or Home Depot. But, if the purpose of the detour is to serve the retail stores, then the bus should be taking 178th to 76th, with regular stops, which serves the retail stores much better than the park and ride does, and *not* be wasting more time leaving the street to drive through the P&R’s bus loop. Even then, I’m not convinced it’s worth it. If I lived up on Avondale with no car, I would argue that it’s easy enough to shop online (or at a downtown Redmond alternative, in some cases), and cheap enough to Uber one direction (if I really need to buy a lot of stuff from Target at once) that it’s better to have quicker service to downtown Redmond, Link, and the entire rest of the Puget Sound region, than to have direct access to a cluster of large retail stores. And, it’s not like these stores wouldn’t have bus service from downtown Redmond without the 250 either. There’s still the 269, the 545, and I think the planned route 251 (Woodinville to Redmond) is supposed to extend there as well. And, if someone were to build a pedestrian bridge over Redmond Way, Fred Meyer would suddenly become a very short walk from Marymoor Village Station, which feels like a much better way for transit to offer access to an area than a detouring bus.
So, yes, other than as a layover facility for buses, Bear Creek P&R really has no reason to exist anymore, and the 250 should definitely not detour to serve it.
The garage is for the entire city of Sammamish, which won’t be getting Link. Let’s wait and see how it performs after cross-lake service starts. I don’t know what ST’s intention is behind the bus bays. Three of them may just be to fill up the space under the garage. I agree the station shouldn’t have design flaws that make it less usable for pedestrian access.
Yep. The Marymoor Village layout is not pedestrian friendly. The out-of-direction walking, the inability to easily walk through the garage, the crossing gates required for a platform , the parking garage access signal crosswalk are all contributing to station area pedestrian hostility.
And the high height black chain link fence is openly messaging “we would rather be untrustworthy with open space than improve the walking experience.”
Dan Strauss’ predecessor Mike O’Brien suggested looking at Ballard to UW and West Seattle to SoDo, before Ed Murray’s SDOT head Scott Kubly* proposed the current ST3 two tunnel monstrosity without public comment.
Was O’Brien Copernicus or Gaileo?
*Former STB heads were ,at worst , friendly with Kubly, but refused to reach out to him on his plans, despite folks like me constantly bringing it up in open threads.
I don’t think permit streamlining or city staff reallocations are the right order of magnitude to address the budget gap for ST3. That kind of stuff will be laughably insufficient. Unfortunately, our collective approach thus far seems to be to avoid questioning any assumption ever approved by voters, no matter how misguided, rushed, or retrograde the planning process may have been that preceded it. The default approach seems to be to further dumb down the projects and push the schedule even more years out.
Part of the reason for the budget woes is the fact that we pre-announce the alignments decades out but preclude purchasing any real estate until after the entirely predictable run-up. Are we somehow obligated to waste public funds like this? Why do we keep doing it? You can see this problem miles away.
Next, Issaquah to South Kirkland is a joke. Come on, this emperor has no clothes. South Kirkland is not a place; it’s a park and ride. It won’t become a place, with or without a Link station, any more than, say, Magnolia Manor at the north end of Magnolia is a place. Meanwhile, most of Issaquah would be a long way from any station. There is no destination on this ride unless you count downtown Bellevue, and Issaquah to Redmond via Bellevue is a weird, roundabout route.
https://www.soundtransit.org/system-expansion/south-kirkland-issaquah-link
I would just fund shorter headways on line 2 and buses to Factoria, Eastgate, and Issaquah, and not build this silly line. Until we can have a reality-based conversation as a region about the discrepancies of value in these projects, we are going to be stuck paying for weak sauce for a long, long time.
I’m in Tokyo at the moment. Their train system seems a bit better than ours. #Goals
I think there is some value in looking at how ST did a lousy job with cost estimates in 2016. That’s particularly true with the DSTT2 segment. And the 10 percent contingency is especially laughable.
Right now the project appears 40 percent over the original estimate. Had ST drawn up a plan based on a 40 percent contingency like FTA recommends for early planning, the program would have turned out much closer.
And they can’t blame inflation. Downtown tunneling in San Francisco was already underway at a much higher cost than the ST estimate projected.
Basically, ST3 was sold with faulty costs.
Amazingly there remains some people that still portray this as a minor hiccup as they push for keeping ST3 intact. Will those still fantasizing about making that map a reality. To those people I saw GROW UP! There are three big choices: Pull back the track miles / stations getting built; Consider different rail technologies (especially automated); beg for a doubling what ST3 raises by a new referendum.
Just remove the added on portions
Aka the
1) fancy west Seattle bridge, use rhe cheapest design
2) the underground west Seattle station. Just build it elevated
3) for Ballard build a medium height draw bridge and elevated station
Changing the design would help sone on the cost, but the deficit is so massive that it’s going to need to be a much bigger axe in order to reduce the costs.
My bigger point is more on how ST and its many consultants lived in this magical world before the ST3 vote where costs and money were not taken seriously — and there are still many “advocates” who still believe in this world and can’t bring themselves to face how badly the ST3 package costs were assembled and cannot happen as approved in 2016..
And it’s important for ST to admit the mistake so that it doesn’t happen again and that ST can try to restore some credibility — which they need as the too-frequent disruption occurrences and extension delays continue to hurt their credibility.
Changing the design would help some on the cost, but the deficit is so massive that it’s going to need to be a much bigger axe in order to reduce the costs.
I agree. WL’s suggestions are good but they won’t get us close to where we need to be to build this thing. There are things they can do which are similar but very different than what they planned. For example:
1) Built a low bridge to Delridge. This is the only way to save a lot of money. It would skip Alaska Junction. This changes the nature of the project and probably wouldn’t be acceptable.
2) No stations at Avalon or The Junction, so the only stations are elevated (at Delridge).
3) A medium height bridge for Ballard is the best choice. But that still doesn’t save much money. Even if you go to 14th (which is unacceptable in my opinion) you are way overbudget. You can (and should) run on the surface. But you still have a lot of (expensive) underground stations in Uptown and South Lake Union. As with my suggestion for West Seattle, you have to change the project in a fundamental way to get big savings. At that point it may not be worth it.
That is the nature of subway construction. You really need to “go big or go home”. With bus infrastructure you can make small improvements in a piecemeal fashion. A great example is the Metro bus tunnel. It was quite short. Yet it saved an enormous amount of time for the riders and taxpayers. Metro was able to run a lot more buses downtown (and elsewhere). Now imagine the bus tunnel was basically a tiny subway line (from Westlake to CID). A few people would use it, but not that many. None of the buses would run more often. There really isn’t much point.
Of course there are instances where you can leverage the existing subway system. Link could have started with Downtown to the UW and then added stations one by one (on each end). East Link will leverage the existing line (when they finally get across the lake). But other than infill stations, none of the other ST3 Link projects are like that. With buses you can leverage the West Seattle Bridge, Spokane Street Viaduct and SoDo Busway. With trains you are building miles of new track from scratch before you add the first station. So even if you cut the cost by more than 2/3 (to “only” a couple billion) it is still a lot of money to get to Delridge. A single station on Delridge just isn’t worth it. So you decide to go farther (along Delridge). But that means the cost escalates. Maybe somewhere along the way to Westwood Village the project actually provides enough value to be worth it, but I doubt it. To avoid traffic you are going elevated the whole way — that isn’t cheap.
That is the fundamental problem with West Seattle rail. They have already built a lot of excellent infrastructure that can be used by buses. A train has to replicate all of that just to get started. At that point, the train has to go deep into West Seattle before the train is actually significantly faster. By then you’ve spent a fortune — and there is no stomach for that.
Ballard isn’t that different. Unlike SoDo, branching from the main line wouldn’t be cheap. Of course it is cheaper than building a new tunnel but it is still very expensive. Then what? Denny is very close to Westlake. The South Lake Union station is a misnomer — it is basically the “Aurora intercept” station. But the buses and the train are running the same direction. For those headed to downtown it isn’t worth the transfer. For those headed further south it saves a little time, but not a lot. Uptown would add a lot of value if not for the fact that we already have a monorail that connects Westlake to Uptown. To be clear, these are all worthy improvements over the current system. But you’ve built a branch and three very expensive stations for only a marginal improvement. Ballard Link really needs to go to Ballard to add a lot of value and that will be expensive.
I still think that Ballard Link should be built (as a stand-alone, automated line ending at Westlake) but I’m under no illusions that it is the best transit value in the city. Ballard to UW is a lot more important but it is also quite likely that we should just focus on the buses for a while. I would love to see a “Metro 8” subway. But for now I would just like to see the Metro 8 bus run a lot faster and more often.
okay. I like the idea of the shorter trains, automated, and smaller stations with the transfer at Westlake, and provisions for extension to first hill and beyond.
An expanded bus network to West Seattle also seems like a logical compromise with the budget deficits that ST has.
But.. Dow promised trains to West Seattle.
Might they end up looking at a West Seattle to Ballard line running under First Ave in a shallow cut and cover tunnel? ST could maintain their idea of splitting the rainier line off to go up first ave then to Ballard, while the west seattle line continues in the original tunnel to points north.
Lots of problems though I’m sure with this. It might not even be technically possible with whatever is underground there. it might even be just as expensive. Too, transfers would be a 2-3 block walk downtown (underground passageways maybe?), and South Lake Union would be bypassed, although you would then get Belltown.
Why can’t we automate longer trains? They provide better value as well. Less transfers.
Train length has nothing to do with number of transfers. Short automated trains can carry the same number of passengers per hour as long manually-driven trains, because they can run every 2 minutes without the cost of drivers. Automated trains also have lower capital costs because the stations and tunnels can be smaller. So passengers get less waiting time for lower capital and operating cost, a win-win.
Long trains can be automated as well. But if you automate the trains you can run them more often. If you run them more often, you can have shorter trains. If you have shorter trains you can have shorter platforms. If you have shorter platforms, it is cheaper to build the stations.
Toronto just went through that with their Ontario Line. To quote Wikipedia: By using driverless trains with automatic train control (ATC), Metrolinx expects the line to match the capacity of the existing heavy rail lines despite using smaller, lighter trains by increasing the frequency. In conjunction with ATC, stations will have platform-edge doors for safety, also allowing riders to exit and enter trains more quickly. The government also claims the alternate technology will reduce construction time and cost, as single tunnels rather than dual tunnels could be used.
Thus it is cheaper but better. Riders get better frequency while the agency saves money on construction. For Ballard Link it makes sense to follow Toronto’s lead and have smaller stations with shorter, automate trains.
There are other advantages to smaller platforms. Ballard Link is full of compromises. Those compromises are based on the platforms being a certain length. It is quite possible that if the platforms are shorter, the stations can be better (not just cheaper to build). You can squeeze them into places that would otherwise not be possible.
Then there is capacity. With an automated train there is no driver cabin which means there is more capacity (for the same size train). Thus a train that is half the size of our Link trains could carry more than half the riders. If we ran it twice as often it could carry more riders than our current trains.
On a related note, if the new Ballard line was independent then we could use high floor trains. This also increases capacity and decreases dwell times. Trains running every four minutes (which would be fairly easy to do) would carry quite a bit more riders even if they were half the size of our current trains.
In the long run it would be nice to automate the existing line. But that is trickier. You would need a system that could handle the dangerous Rainier Valley segment. Even if you installed such a system, you might not be able to run the trains very often. The city has been restricted the number of trains (to avoid traffic jams). Ultimately the best thing to do would be to build a trench and run underground down the valley. But that requires a lot more work.
“But.. Dow promised trains to West Seattle.”
How about rubber-tired guided trains — like those in Paris or Montreal or even inside SeaTac? They perhaps could run on the existing West Seattle Bridge and on an Alaska Street transit mall. They can operate at steeper slopes and at higher speeds. Maybe the cost savings could enable for them to run further south.
Rubber-tire train is almost just like heavy rail and it still needs metro-style dedicated guideway unless you are talking about this
The latter is no difference to bus and you cannot manufacture that in the US yet. Rubber-tire train running on high-speed feels like riding a bus and it shifts left and right constantly. I don’t think this solution will help mitigate current challenge much.
Also, I believe the main reason that a lot of airports choose rubber-tire people movers because rubber-tire train does very good job slowing down in short distance compared to conventional rail. Those airport systems often have crazy stop spacing.
But.. Dow promised trains to West Seattle.
Yes, and that is the problem. Current estimates for West Seattle Link are somewhere between 6 and 7 billion dollars. That is from West Seattle to SoDo. You can make various compromises and alterations but you aren’t going to cut the cost in half. At best you are looking at close to $6 billion (not 7). That’s if you get rid of the Avalon Station, run on the surface and do whatever else is necessary to save a few pennies. Meanwhile, you have made various improvements that allow West Seattle Link trains to run through the existing tunnel. Of course this is cheaper than a new tunnel but it still isn’t cheap. It will help us in the long run (more than a new tunnel) but now we are approaching $7 billion again and we haven’t even started Ballard Link. And yes it would be much cheaper to run smaller automated trains to Westlake but you still have to cross the ship canal. You are still doing a lot of digging and running a lot of elevated lines on 15th/Elliott. You still have revenue problems (not just spending problems).
If you commit to West Seattle Link (even with various compromises) I think you are looking at major delays to Ballard Link (even if Ballard Link ends at Westlake and has smaller stations). Or you are looking at building it in pieces which is a really bad idea. Ballard Link doesn’t really become worthwhile until it gets to Ballard. At some point the folks in charge should realize this and abandon West Seattle Link. Imagine this instead:
1) Spend a billion improving the connection from the Spokane Street Viaduct to the SoDo Busway and the busway itself. This is probably more than you need.
2) Spend a billion (over ten years) on service in West Seattle. Create a “West Seattle Transportation District” similar to the Seattle Transportation District. This means you spend money on West Seattle buses after the county and after the city.
Now West Seattle spending is actually close to the original estimate. Now you just have to make sure Ballard Link doesn’t cost a lot more than originally planned and that is a lot easier if it ends in Westlake (and has smaller stations). I really don’t think we can get out of this mess without abandoning West Seattle Link. It will take someone to explain to fans of West Seattle Link that not only can’t we afford the project but we can afford to build something that is better for West Seattle.
“But.. Dow promised trains to West Seattle.”
Dow is no longer an elected official. He now is running ST instead. And under his watch is the ST looming budget crisis, the East Link delay, the Board-needed West Seattle authorization that will jeopardize projects elsewhere, further studies on Everett, Ballard and at Tacoma Dome — all likely with new budget problems and serious environmental / engineering challenges, and a post-COVID demand scenario that increasingly looks permanent .
If he ever runs again, an opponent will blame him from day 1. By going to be the ST CEO he will not be able to extract himself politically from ST’s problems. So I think he will never run again.
Dow’s promises don’t matter much any more.
Yeah, it really isn’t about Dow. That was more tongue in cheek. The reality is that a lot of people voted for ST3 (which included West Seattle Link). This is often used as an argument for it. I think that argument is weak. The voters didn’t choose it amongst a range of options. It is quite likely they would have preferred to spend the money on other projects. The main reason it passed is because it had plenty of support in Seattle and lots of Seattle voters turned out to vote. But a higher percentage of Seattle voters supported Metro spending. Thus you can easily make the case that more people wanted to spend money on buses than light rail, but they weren’t given a choice.
You can also easily make the case that they wouldn’t support ST3 now. Obviously the cost overruns are a major reason why. But look at the cities that supported ST3: https://www.seattletimes.com/seattle-news/transportation/planned-light-rail-areas-big-backers-of-sound-transit-3/. Not that many cities supported it. There was plenty of support in Redmond. Now that there are trains running to Redmond, would they support ST3 with the same enthusiasm? Link will get across the water either way — hard to see why. Speaking of which there seemed to be a lot of support from areas that assumed that without ST3, the trains wouldn’t get there. The campaign pushed that idea. They didn’t talk about travel time from Everett to Lynnwood — it was Everett to Downtown (as if Lynnwood Link didn’t exist). With that in mind, why would Lynnwood riders support ST3? Again, hard to see it.
I’m not saying that voters are only motivated by self-interest but a lot of voters are. Even if they aren’t, the assume that a project is good if they can imagine themselves using it. Someone in Lynnwood can definitely imagine taking the train to Seattle. But to Everett? Might as well take the bus or drive. I think if there was another vote on ST3 it would fail (although it would still pass in Seattle).
Al S.
I wish the political problems Sound Transit has just Dow. It’s much, much worse than that.
Which mayoral candidate to going to break the hard news about Sound transit scaling anything back before the election? Or even afterward? Maybe Harrell? Katie Wilson has a huge “progressive” platform…. price tags be damned! Although the Transit Riders Union and Seattle Subway claim to support transit, both outfits have a history of not being realistic about the cost of projects.
I’m afraid the Sound Transit Board is going to fallow the path of Seattle Public School Board… the finances point to readjustments and hard choices that turn out to be politically impossible.
What is the best degree program at UW for an aspiring transportation service planner with Metro or Sound Transit?
I am looking at the transportation engineering program. I am interested in your thoughts.
Jeffrey Baldwin
Jeffrey,
There are several good ones.
Engineering has a civil and environmental track. Professor Scott Rutherford was national expert on BRT; he has passed away; he served on the first ERP for what became ST. Owen Kehoe, a great speed and reliability engineer at Metro, studied under him. Former Professor Mark Hallenbeck was a leader on tolling and the use of big data.
Geography might work. ST planner Ron Endlich had that degree. The founder of One Bus Away may have been in Geography.
The UW College of Built environments has several fields that are related: architecture, urban design and planning, and real estate (TOD).
The Evans School of Public Policy has produced some planners.
Others schools have produced great workers. Two friends got masters from MIT. The Portland State program has a good transportation focus. Eastern Washington has one. Some transit planners have unrelated degrees. Transit needs good logic, writing, and computer skills; they can come from many places.
[typo fixed]
What is the link for signing up for the relay race? I don’t see it.