Ryan Packer has the scoop on Twitter: Yesterday (May 17), Sound Transit staff delivered a letter to the Sound Transit Board of Directors with an innocuous subject line: “Staff recommendation regarding the South Lake Union area alternatives for the Ballard Link Extension“
A brief history: last year, after the Sound Transit Board had selected preferred locations for the Denny Station (Westlake and Denny) and SLU Station (5th and Harrison), the Seattle Chamber of Commerce lobbied Seattle Mayor Bruce Harrell to propose a pair of new alternative locations for these stations in an effort to reduce construction impacts on Westlake and Harrison. In December, the Sound Transit Board approved a “feasibility study” to review whether these locations were technically feasible, and to determine how severe the impact would be to the project schedule if these alternatives were formally added for review.
After news came out earlier this year that formal study of these alternatives would delay the final Environmental Impact Study for the Ballard Link Extension by almost a year, potentially result in $500M to $1B in extra costs due to delay, the Seattle Chamber of Commerce kicked off a campaign to “Save South Lake Union” describing the late-stage alternative as a “Community Couplet” that would “avoid the worst impacts on Westlake Avenue and place a station closer to Seattle Center”, but withheld the fact that the reduction of construction impacts would be meager at best, and would have significantly worse connections to the bus network.
From the letter:
The staff recommendation is to take No Action, maintain the current Board-adopted preferred alternative and not carry forward the potential new alternative into environmental review, given that the current preferred alternative and the potential new South Lake Union and Denny alternative are reasonably comparable from a technical perspective, but the potential new alternative would delay delivery of the Ballard Link Extension by 10 months to 2 years, would add approximately $500M to $1B to the project cost, and would have overall system implications.
As Ryan Packer notes in subsequent tweets, having a direct recommendation is unusual, and apparently a direct result of a report from Sound Transit’s Technical Advisory Group (TAG) on speeding project delivery. In February of last year, the TAG delivered six recommendations for improving efficiency, the first of which was to “Rebuild trust and clarify the roles and responsibilities of the ST Board and staff” including the specific direction that “board members must be explicit in authorizing and expecting staff to bring forward timely information along with their best recommendations for Board decisions.”
Although it’s been over a year since the TAG published their report, it seems Sound Transit staff are finally feeling empowered to make recommendations against unreasonable political whims of the Board. Other visible changes in response to the TAG report include the hiring of a Deputy CEO for Megaproject Delivery and work towards adoption of a revised betterments policy.
This is good because it’s not just any “political whim”, it’s pushing back against stakeholders ST usually defers to: business leaders, cities, and individual boardmembers who defer to them. In a balanced world you’d expect these stakeholders might get 25% of what they want, and passengers and transit fans might get 25% of what they want. Instead it’s usually 98% what the cities and nimbys want and 2% what the passengers want. So this is a refreshing change.
Now, staff, please push back against the CID/N and CID/S alternatives. That’s an even more important issue. The 4th Avenue Shallow or 5th Avenue Shallow alternatives would have much better transfers between lines and put more of the CID within walking distance. That would benefit passengers and people who walk to the neighborhood businesses and amenities.
P.S. Not spending $500M-1B on SLU changes would cover the $800M cost of 4th Avenue Shallow over CID/N. Cost problem solved.
I second this. A proper connection at the CID is the lynchpin of this whole system. The benefits will be felt for decades to come. It’s worth the money and temporary construction disruptions.
“ Now, staff, please push back against the CID/N and CID/S alternatives!”
Yes please!
Of course the real game changer would be for staff to be backed up by a review committee of riders and drivers! Without such a review committee, the staff are the only ones who can officially push back against real estate development interests guiding the project development with little consideration of the thousands of users that will be paying the fares for decades.
It’s like having a bicycle plan without bicyclist review or a hiking trail plan without hikers. Sure some members are transit riders — but they are put in a stakeholder or elected official committee to reflect their other perspective and not because they are looking out for riders.
I agree with the position that Fourth Avenue Shallow, even at a higher price, is worth the investment. It gets the construction away from the International District, though plenty of congestion on Fifth Avenue will still occur. It massively imprioves the transfer between the 1 Line and the other Link lines. It makes practical any transfer between the 1 Line and Sounder/Amtrak.
So it’s a great thing, assuming DSTT2!
But we don’t want DSTT2, at least, most of us don’t. We want Ballard-Downtown to be an automated Light Metro as much like Skytrain as is possible with accommodation made for an extension into First Hill. So arguing for Fourth Avenue Shallow is self-defeating. It weakens our preferred position.
While we can certainly be flexible on station locations in SLU if the best simply can’t be attained, we want the project to focus on Ballard-Downtown and defer any second tunnel or extension to West Seattle.
A second bus-only lane in the West Seattle Freeway eastbound to SR99 northbound cloverleaf as Ross has suggested for years would provide the C and H lines excellent delay-free access to downtown Seattle. Add a version of the old 55 from Alki Point that doesn’t go downtown but instead heads straight to SoDo Station via the upper deck on Spokane to Fourth South for riders who want to get to Link faster or head south on Line 1. It’s a right turn from Fourth to Spokane that could be massaged into a two lane turn with a bus only lane on the left leading to the left-hand bus-only lane on Spokane as far as the busway.
Then turn the 125 into the same sort of feeder to SoDo for the areas southeast the plateau using the upper deck/Fourth South pathway to SoDo Station.
Metro can run that 55 stub and 125 more frequently for decades for what will be flushed away on West Seattle Link.
“But we don’t want DSTT2, at least, most of us don’t. We want Ballard-Downtown to be an automated Light Metro as much like Skytrain as is possible with accommodation made for an extension into First Hill. So arguing for Fourth Avenue Shallow is self-defeating. It weakens our preferred position.”
Ranked-choice voting. It’s better to get either single-tunnel or 4th Avenue Shallow than to get neither one. If we put all our eggs into single-tunnel and ST won’t do it, then we’ll likely get neither one. 4th Avenue Shallow is easier to convince the board of, because it’s just a station alternative and has the support of CID activists and groups like Seattle Subway. Single-tunnel requires ST to give up on DSTT2, reverse its decision in 2016 (when it selected DSTT2 instead of upgrading DSTT1), and admit it made a large mistake. It has never done something like that before, so what’s the likelyhood it will now? We’d need Seattle and other cities and large businesses to support it in order to get some influence with the board, and we haven’t found that yet. We need to support all good and OK choices, not put all our eggs into one long shot.
Oh I’d agree that ST is pursuing a much more expensive and less user friendly system than the Ballard Automated Metro + Automated Frequent Trunk would be. It’s a nostalgic clinging onto a light rail technology application that new systems today never choose from scratch — from Honolulu to Montreal to train control on the Elizabeth Line. ST refuses to revisit technology choices defined in its 2008 vote. All of this in a city where its own airport has actually had an internal automated transit line for decades that’s much more reliable than Link.
It’s notable that ST produced a white paper trashing a long distance gondola to West Seattle but remains silent on exploring automated technology applications.
And of course a big improvement would be the reduction in station sizes in SLU. It’s clear that those affected near the stations are increasingly concerned about the constructibility of these stations. Automated trains could be twice as frequent and platforms half the size.
One issue is that the staff itself lacks the interest or mandate to retool for automation. That may not be true for a new engineer, but for a manager trying to keep their bureaucratic job as easy and “safe” as possible they don’t want to revisit the technology. They may have to learn something new!
In my mind, the #1 issue in ST3 is to shorten the transfer distance between lines 1 and 2/3. It must be comparable to other metros. Single-tunnel does it best; 4th Avenue Shallow does it second-best. Distantly behind that, automated Ballard would be nice, and would probably be less expensive and 2-3 times as frequent.
A corollary of single-tunnel is Ballard would terminate at Westlake, so everybody would transfer. Or it could continue south or southeast, but in any case not be in DSTT1.
“In my mind, the #1 issue in ST3 is to shorten the transfer distance between lines 1 and 2/3.”
I must repeat my often overlooked point that simply revisiting the SODO Station layout can achieve an easy cross-platform transfer between Lines 1 and 3 at a much more modest cost than the 4th Ave shallow option could — and 4th Ave shallow still will require vertical elevation changes to transfer just like the awful CID-N / Pioneer Square scheme. (Admittedly it doesn’t directly address 2 Line or Sounder transfers.)
I just wish entities with clout could advocate for redesigning SODO for cross platform transfers. Easy peasy relatively speaking!
Tom probably meant Route 56 between Alki and SODO via Admiral. Route 55 was between north Admiral and downtown via the Alaska Junction. The routes between one-way peak-only shadows of themselves in fall 2012.
@Al — Improving SoDo transfers would be great, but that does nothing for East Side riders. Another issue is that trains converge at I. D. which means that riders going from Rainier Valley to the UW would transfer to a train running every five minutes at I. D. while a transfer at SoDo means transferring to a train running every ten.
eddiew, I did mean the 55, but added a flag turn out to Alki rather than North Admiral because there’s a real center out there. The main thing is to serve California between Admiral and Alaska with service direct to Link, so if you think the old North Admiral terminal is better, fine.
The beauty of this is that if folks on the new “55” or “56” want to go downtown they change to the C at Alaska. If folks from the C want to get to Link faster, they change to the “55”. In either case, the “55” would go direct to SoDo without the slow multi-light detour past Starbucks’ HQ on Lander. Folks headed to Starbucks from either could change to the 35 at the Triangle.
The point is to elevate the ride from West Seattle to SoDo Station to the importance it would deserve in the absence of West Seattle Link. Don’t make people take a “local” for that important leg.
Mike, you’re right that RCV would be ideal; it’s a no-cost way to indicate “second choices” when it’s available. But to accept Fourth Avenue Shallow probably would also mean accepting the cruelly deep New Westlake and the Spelunkers’ Paradise at Midtown. Or maybe even no Midtown; who knows how they’d make the extra budget for Fourth Shallow?
It would be the absolute height of folly for DSTT2 to travel under the greatest collection of high rises in the Northwest without stopping. CID-North is at least within walking distance of Columbia Center.
A lot more riders will switch between lines 1/2/3 and go to/from destinations outside downtown than will go to the specific Midtown station area. Over half of Link’s destinations will require a train-to-train transfer, and that’s many more people than go to Midtown highrises. So if we can have only one of good transfers or Midtown Station, it has to be transfers.
@ Ross
Even the shallow 4th Station will take a rider an extra 5 minutes to walk through and between the station platforms and use all the vertical conveyances. It might take longer if one is out of service or if there is a queue of riders trying to use the vertical conveyance.
At SODO a same platform transfer would take literally 6 seconds (24 feet) from 20 sets of train doors to the other 20 sets of train doors.
And it’s certainly possible to make the trains arriving in the same direction timed so no one has to wait. That assumes that there are four platforms at SODO.
Have you ever made a timed train-to-train transfer, Ross?
Oops. That’s 16 sets of train doors to the other 16.
Tom, there is room for both two-way all-day routes 55 and 56; Route 55 could be reversed; instead of serving Admiral to CBD via Alaska Junction, it would serve Alaska Junction to SODO Link via Admiral Junction; Route 56 could serve Alki to SODO via Admiral Junction.
OK, that would be fine. But the point is to have expresses to Link via the upper deck and the Fourth Avenue ramp from some places in Wedt Seattle that exchange riders at a common point with the C for the plateau and the H for the lowlands and SSC highlands.
I personally think that it makes the most sense to do the exchange for plateau riders at Alaska Jct or The Triangle where the lines to Downtownl and SoDo Station jump onto the freeway. Doing it your way would mean that folks transferring from the C south of Alaska to go to Link would have tovride up to Alaska before heading east to Link and coming back they’d have to do the same in reverse.
And, while folks from Alki would get a GREAT ride to Link, they’d lose their connection to Alaska Jct.
Of course more total people would transfer between lines than would go to any single non-transfer station. But Seattle would be a worldwide laughingstock if we dig a tunnel underneath the Financial District without some station serving it.
The beauty of this is that if folks on the new “55” or “56” want to go downtown they change to the C at Alaska. If folks from the C want to get to Link faster, they change to the “55”. In either case, the “55” would go direct to SoDo without the slow multi-light detour past Starbucks’ HQ on Lander. Folks headed to Starbucks from either could change to the 35 at the Triangle.
This brings up an interesting issue, but I think it makes sense in its own thread (and not on this post). So I added this comment: https://seattletransitblog.com/2024/05/16/midweek-roundup-open-thread-49/#comment-932160
Even the shallow 4th Station will take a rider an extra 5 minutes to walk through and between the station platforms and use all the vertical conveyances.
Yes, which is why the ideal approach would be merge all the trains.
And it’s certainly possible to make the trains arriving in the same direction timed so no one has to wait. That assumes that there are four platforms at SODO..
Well someone has to wait. For example consider the two northbound trains. One is going from Tacoma to Ballard. The other is going from West Seattle to Everett. Now imagine the trains are timed so that the train from Tacoma arrives a minute before the train from West Seattle. If you are trying to go from Tacoma to the UW, this is great. If you are trying to go from West Seattle to Ballard this isn’t.
Then there are reverse direction transfers. West Seattle to Beacon Hill or Rainier Valley. Same idea — someone wins, someone loses (if you try and time it). Of course things get even more complicated because you are trying to do this on top of the transfers going the same direction.
Trying to time trains in the manner makes sense when one of them can have a big layover. But in these sorts of situations they typically just run them frequently and hope for the best. Or they aim for having the trains run opposite each other even though they are on different lines. That means the worse case scenario isn’t that bad and it also means that if a train gets delayed a little bit you don’t have a long wait.
Of course more total people would transfer between lines than would go to any single non-transfer station. But Seattle would be a worldwide laughingstock if we dig a tunnel underneath the Financial District without some station serving it.
It isn’t that crazy, as long as:
1) The trains are frequent.
2) The transfer is great.
I can’t think of something exactly like this, but it is quite common for trains to come close to downtown but then deviate away at the last minute. Riders are expected to transfer to another train. For example the Millenium Line ends a couple miles from downtown and riders are expected to transfer. Eventually the line will go west, which means it will get even close to downtown, but riders will need to transfer.
If anything what makes this unusual is that the plan is to add downtown stations, but not significant coverage. It is neither here nor there. Riders who use the existing system and don’t want to transfer are worse off, no matter what they build. The downtown stations are worse (and there are fewer of them). At the same time, the plan is to build them so close to the other stations that it isn’t worth the transfer.
In contrast, consider the Ontario Line in Toronto. This is being built in part because the existing line is crowded (the same rationale used to build our second tunnel). But the station places downtown maximizes coverage. Other than the two stations that serve as connection points, every new station is nowhere near an old station. Thus riders on both lines are expected to transfer to get to various parts of downtown.
I get your point, but in general tunneling is not as expensive as building downtown stations. If our only goal is to reduce crowding then this works as well as any other alternative (that Sound Transit is considering). Which is to say it does reduce crowding, but a lot of existing riders will be worse off.
@Ross:
Again, have you ever done a rail to rail transfer? I’ve done it many times. Both trains arrive roughly at the same time. They don’t arrive sequentially as long as there are two platforms in each direction. They sit in the station an extra 39 seconds to give riders extra time to get between trains. So it works in both transfer directions.
It only works this way if there are cross-platform transfers possible with train doors only several feet away at the same level.
There is another situation that can work — mix/ matching the 1 line and 3 Line branches. Half the trains use the other track. While that seemingly reduces the frequency in half for direct service, a rider can choose to take the first train and transfer or wait for the second train and not transfer. That is better than making every single rider have to transfer to get to another airport.
So let’s use the MI to SeaTac example in this mix/ match operation. The MI rider rides to IDS only (no need to ride to CID/N). They cross the tracks to the other side (ideally with a down escalator added). A southbound train arrives. If the destination of that train is West Seattle they would wait for the second train. If the first train’s destination is Federal Way/ Tacoma they could finish their trip right away. But even if they got on the first train to West Seattle, they could hop off the tracks at SODO and walk to a third train waiting 20-30 feet away in the same level. Compare that to what ST currently proposes, which requires this horrific underground 3-D maze a further station out of direction that this rider would have to get through with luggage taking several minutes every single time.
A mix/ match operation or a cross-platform transfer is only possible if ST revisits the SODO station layout NOW! This station needs to be redesigned NOW! NOW as part of the pending final design of West Seattle Link!
I feel like few see how fundamentally important it is to revisit the station layout. With the terrible transfer setups at Westlake and either USS or Pioneer Square, the importance of redesigning has grown in importance from even two years ago. When the CID-N change was proposed, this station redesign should have been included as a mitigation for 1/3 Line transfers.
Again, have you ever done a rail to rail transfer?
Of course I have. I’ve ridden the New York City Subway. In fact I distinctly remember just missing our connection. Life in the big city.
They sit in the station an extra 39 seconds to give riders extra time to get between trains. So it works in both transfer directions.
Wait, what? You want to increase our already very high dwell times so that those riders have a better transfer? If the train from Tacoma is running a bit late (which seems quite possible given the distance) then the train from West Seattle is just supposed to sit there? This sounds like it is causing the exact problem that the new tunnel is supposed to solve. Even if the trains are timed perfectly you have a longer dwell time, which means that riders from West Seattle to Downtown are worse off. It also means Rainier Valley to Downtown riders are worse off. Imagine that scenario for a second. You live in Rainier Valley and commute to downtown. You get off at University Street. Now you take the train and it spends extra time getting to downtown. When it finally gets downtown there is no stop close to your destination (I guess you get off at Midtown, wherever that is).
Look, I agree that the SoDo transfer point should be better. But it is no panacea. It doesn’t address East Side riders, which will probably outnumber West Seattle riders. It doesn’t address reverse direction transfers at Westlake, which will likely outnumber every other transfer. There are a lot of issues here and the best way to deal with them is to just put the trains in the same tunnel and run the line from Ballard more often. By no means is that perfect, but most of the transfers are much better and gives us extra money (and more flexibility) to deal with reverse direction transfers. Oh, and more frequent service through the core. And the long term prospect of serving First Hill (a downtown location that is far enough from the other stations to be worth the transfer).
@Ross:
Of course I agree that one tunnel is best. I actually think that West Seattle is going to demand that their line runs through Downtown years and maybe a decade before a DSTT2 would open anyway. The main reason it hasn’t been pushed to date is because those West Seattle stakeholders don’t see that it’s operationally possible yet.
I’ll take an extra 30 seconds so people can change trains over a possible over five minute, two block, 150 feet plus of elevation changes that I would have to do half the time. It’s done across the world from BART to Singapore. Honestly, people barely notice if train doors stay open another 30 seconds at just one station.
And I trust that automation can make schedule coordination can only get better as automation improves. Those that don’t are Luddites.
“They sit in the station an extra 39 seconds to give riders extra time to get between trains. … You want to increase our already very high dwell times so that those riders have a better transfer?”
The point of subway frequency is that if you miss a train, you don’t have to wait more than 8-10 minutes, or 4-5 minutes where the 2/3 Lines overlap. This isn’t like BART at MacArthur on Sunday where everybody transfers because the north-south line doesn’t go to San Francisco so timed transfers are very important.
“And I trust that automation can make schedule coordination can only get better as automation improves.”
Schedule coordination doesn’t matter when trains are running every 2-3 minutes.
And you’re assuming that Link will coordinate the Tacoma-Ballard line and the West Seattle-Everett line, and that everybody walks the same speed and jumps off the seat at the same moment. Even if you fix it for one transfer, you can’t fix it for all transfers, because what works for one transfer and direction doesn’t work for another transfer and direction. And you’re assuming trains will always arrive at their exact time, when it often varies by 1-2 minutes from when the next-arrival board says it should be there.