Transit & Streets:
- The Urbanist: Sound Transit Board Rubberstamps West Seattle Link Route, Amid Cost Issues
- Seattle Bike Blog: SDOT set to start Aurora-Licton Springs healthy streets upgrades on N 100th St, Fremont Ave N
- West Seattle Blog: Bus service modifications (feedback wanted!) and community updates at HPAC October meeting
- Kent Reporter: ‘Drivers going too fast’ led to 45-vehicle collision in Kent on I-5; also on KOMO.
- Seilbahnen (Cable Car) International Magazine: Study shows: cable car is more sustainable than the electric bus for point-to-point transit with no intermediate stops.
- Bloomberg CityLab: Robotaxis Are No Friend of Public Transportation
- The Seattle Times ($): Lynnwood light rail is super popular — but there’s a problem; the times presents anecdotes from Lynnwood commuters who say they drive to downtown if there isn’t enough parking at the station. Sound Transit and Community Transit suggest taking the bus.
- The Urbanist: Bellevue Reveals Initial Designs for Pedestrian Bridge over I-405
- Streetsblog USA: Why America Has So Much Road Safety Research, But So Little Actual Safety
Housing & Land Use:
- Real Change News: Seattle feels the crunch: Affordable housing and human services sees cuts while police funding grows in Mayor’s budget
- Urban Land Magazine: Downtown Office-to-Residential Conversions; three case studies in Calgary showing how a city whose office core began declining in 2014 was able to become the world leader in office tower conversions.
- Planetizen: Book Review: Combining Sustainability and Smart Growth, a review of The Sustainable Urban Design Handbook by architects Nico Larco and Kaarin Knudson.
Commentary & Miscellaneous:
- The Urbanist: The Seattle Times’ Credibility on Transportation Levies Has Run Out
- The Urbanist: Op-Ed: Vote No on Initiative 2117 to Ensure Mobility for Washingtonians; see our own electoral recommendations here (we agree).
- Ron Davis: Harrell’s Halfhearted Housing Plan
- Seattle Now & Then: Alfred’s Cafe (Tacoma), 1910
- Mass Transit Magazine: OP-ED: Happy 120th anniversary to NYC subway system
- Streetsblog USA: Opinion: Who Does Passenger Rail Serve? “In short, passenger rail serves everyone – even the people who don’t meet the profit margins of airlines and car manufacturers.”
- The Seattle Times ($): The dash through the airport has become a marathon
- MinnPost: The uneven geography of Halloween sidewalks
This is an Open Thread.
East Link starts testing on Mercer Island. ($)
@Mike Orr,
That is the start of testing to clear the dynamic envelope. It’s a great milestone. Glad to see it.
Opening 2-Link all the way to Lynnwood is going to be awesome. Can’t wait.
Thanks for posting. I hadn’t seen this.
Don’t have access to the article.
Assume this is just testing from South Bellevue to Mercer Island. Would be great if Link could be extended to Mercer Island station until the full opening especially if its Spring 2026 full 2 line, would be a very easy bus transfer and would gain the benefit of more traffic free Link for several more miles.
Yes, from South Bellevue to Mercer Island.
ST will make orders of magnitude more people happy about a surprise early opening than the number of people unhappy that it wasn’t mentioned in a monthly Progress Report.
Truly. I think we’re seeing the impact of an effective CEO of Megaproject delivery; less focus on reporting, more focus on product.
Who is the CEO of Megaproject delivery? I saw it was recommended but didn’t see who it was.
Terri Mestas, technically “Deputy CEO” – https://seattletransitblog.com/2024/10/29/an-extended-interview-with-terri-mestas-sound-transit-megaproject-delivery-chief/
Having multiple “CEO“ positions is weird. The projects they oversee are more aptly referred to as gigaprojects, unless they are from the UK.
“ it’s hard to interline some combination of 1 line going every 8 min and say 2 line every 12 min”
This is definitely not something TriMet found difficult, at least not pre-pandemic. Eg, blue line trains had a few cases where they were only 4 minutes apart, while green line only operated at 15 minute headways as there has never been enough demand to justify more. Occasionally the orange line would get a couple of green line trains as well as yellow line trains, so that it would sometimes get 2 minute headways during the several times per day it needed that.
TriMet is hardly the best rail operator in the country, and if they can manage variable headways (or at least were able to prior to 2020) I would think SoundTransit wouldn’t have a problem.
Sorry, that was supposed to be below.
“Perhaps this is more obvious to anyone else who has lived in a city with a branch rail line corridor but that is generally how it works.”
Just a note that I’ve lived 22 years of my life within walking distance and using a Muni Metro or Boston Green Line station. Muni Metro had 5 branches and the Green line had 4. Both lines have asymmetrical branch operations and always have.
The concern here is with two branches. Maybe 3 someday. Systems across the world operate fine with two and three branches — including NYC. Study a subway map of Manhattan.
The most major frequency constraint affected by design is the turn around. I lived with those nutty few months in 1996 when Muni Metro tried to terminate every T-Line train at Castro Station without a siding — and when trains could not loop around at the end east of Embarcadero station. True it was a nightmare. The source of the problem: naive operations planning when they originally designed the tunnel. ST has wisely been building sidings (yay)! Because reversing a train is the challenge, the rest of the system only bottlenecks if there is heavy station platform activity like a ball game or a train is forced out of service.
So then it next comes down to the number of operable vehicles and drivers. That’s the problem the Board was presented with. I’ll even note that automated trains don’t have to worry about staffing drivers for every train run.
But in practice, most riders don’t get frustrated if their branch runs every 10 minutes unless the segment is too crowded or the operations are unreliable.
And most riders don’t care about clock face scheduling if a train comes every 10 minutes or less. Even 12 isn’t too bad. Clock face scheduling is however much more important when frequencies drop to every 15 minutes and especially 20 or 30 minutes for a branch.
When a rider gets to a platform, they care about when their train arrives. They will look at the electronic next train arrival signs at the platform for only their next train. They aren’t OCD and expect every branch and arrival to be evenly spaced.
So someone should mix up the negatives of branching by mixing the biggest drawbacks of a high frequency operation and a low frequency operation together. They each have their time and place and they don’t occur simultaneously.
“So someone should NOT mix …”
I totally missed the interview post when I asked my question – sorry!
WL, my main focus is convincing ST to do what it can but isn’t, not on worrying about what ST wants to but can’t because of physical limitations. We’ve all experienced periods of substandard service because of maintenance or single-tracking or a collision or these mysterious “signal issues”. I just put them down to ST isn’t perfect but at least it works some of the time. That’s better than when there was no Link at all and I was transferring between the 71/773X, 550, and/or 132/132 downtown, or taking the 62 for most of its length, or transferring to hakf-hourly buses.
I’m sorry if my argument has seemed inconsistent but that’s my overall concern. Try to get ST to make better decisions, and just get out of the way when something unintended occurs and wait for better days.
If you do write an article about this, please address the difference between intended and unintended limitations, and temporary vs permanent ones, and where this issue lies.
Your argument that adding branches lowers frequency on the existing branch seems to imply branching is bad so we shouldn’t do it. We could avoid this potential frequency drip by not opening the full 2 Line but I don’t think that’s a good idea. You don’t seem to have another solution, so you’re complaining about an inevitability, and not one ST wants. But maybe it won’t be as bad as you’re thinking.
@Mike
> Your argument that adding branches lowers frequency on the existing branch seems to imply branching is bad so we shouldn’t do it.
I understand it is a bit complicated of a position to explain but let me reiterate that opening east link is still of course awesome. However, it is also true that with the branching of line and maintaining same frequency on both lines that line 1 rainier valley segment will tend towards lower frequency than it would otherwise have. Both can be true.
> Try to get ST to make better decisions, and just get out of the way when something unintended occurs and wait for better days.
Well first off is to recognize the issue. It seems many of y’all keep convenient forgetting about the section south of CID. Most didn’t even know sound transit was planning on decreasing frequency south of CID originally. And many kept forgetting how the ‘extra’ frequency on the lynnwood side from doubling the line given the same operational money is really from lowering what could run on the rainier valley section.
> You don’t seem to have another solution, so you’re complaining about an inevitability, and not one ST wants. But maybe it won’t be as bad as you’re thinking.
There’s northgate turnbacks and also having not matching frequency between the two branches. Or perhaps later on having turnbacks at federal or lynnwood if extended farther.
if you look at ross’s statement “Folks from the East Side would have a major hissy fit.” from separating out the frequency. Well why exactly is rainier valley supposed to appreciate lowered frequency?
You all (in general not saying just you) understood that there is a limit to resources when it was for the rapidride G restructure, and then somehow just magically don’t understand it when the same concept is applied to the rainier valley section?
Even worse when Dijibell said the statement, all of you guys were trying to gaslight them that it would of course never happen — when that was the opposite of what sound transit was planning. Even look at Lazarus trying to handwave the issue as “temporary”.
The last two comments belong to the long subthread below.
if you look at ross’s statement “Folks from the East Side would have a major hissy fit.” from separating out the frequency. Well why exactly is rainier valley supposed to appreciate lowered frequency?
They are not. You completely missed my point. You made a claim that was simply not true. You made two claims:
1) Once East Link opens up, there will be a shortage of trains.
2) The problem is due to branching.
The first is true. The second is false. It has nothing to do with branching. Nothing. Zero. Zilch. You would have the same demands on the system if East Link to Lynnwood trains ran on entirely different tracks.
Ah, but if they were different tracks, then Sound Transit could decide to reduce frequency on East Link while keeping frequency in Rainier Valley the same. Yes, they could. But why the fu** would they?!! Why should folks from Federal Way get better service than the East Side? The only possible reason ST would even consider that is if they were short on trains and Federal Way was way more crowded.
But our trains are not one size! If there is a demand mismatch — and a shortage of train cars then they can just run smaller trains to one branch or the other. They could run three-car trains on one branch and four on another. Or they could alternate between three and four car trains.
Another option is to turn back some of the trains at Northgate.
But wait, is there some other reason why ST would reduce frequency in Rainier Valley? Yes! To save money! ST used to run trains every six minutes during peak. They don’t any more. They run less frequently to save money. They could easily run trains every 8 minutes all day long but they don’t. They run less frequently to save money.
But why be cheap now? Because the lines are so damn long and ridership per service hour is not that high. If Rainier Valley gets screwed (again) then it won’t be because of branching, it will be because Sound Transit is building a very long system and is not that interested in running trains frequently.
@Ross
> Ah, but if they were different tracks, then Sound Transit could decide to reduce frequency on East Link while keeping frequency in Rainier Valley the same.
> But our trains are not one size! If there is a demand mismatch — and a shortage of train cars then they can just run smaller trains to one branch or the other. They could run three-car trains on one branch and four on another. Or they could alternate between three and four car trains.
Ross you’re literally agreeing with me here, yes that’s what Sound Transit is choosing to do so then in general it is will tend to run trains less frequently on the rv segment than it otherwise would because of branching.
> But why be cheap now? Because the lines are so damn long and ridership per service hour is not that high. If Rainier Valley gets screwed (again) then it won’t be because of branching, it will be because Sound Transit is building a very long system and is not that interested in running trains frequently.
The branching and the longer transit lines are tied together. The branching combined with the long lengths on both lines, means it cost a lots a more to run even slightly more frequency.
> You would have the same demands on the system if East Link to Lynnwood trains ran on entirely different tracks.
It’s about maintaining the even frequency that Lazarus talked about, if one doesn’t want to maintain the same frequency on both lines, then yes one could increase or decrease frequency independently.
> Why should folks from Federal Way get better service than the East Side? The only possible reason ST would even consider that is if they were short on trains and Federal Way was way more crowded.
Has literally no one opened the slide? thats literally what it talks about how the rv segment has more ridership than the east side line
> Why should folks from Federal Way get better service than the East Side? The only possible reason ST would even consider that is if they were short on trains and Federal Way was way more crowded.
Has literally no one opened the slide? thats literally what it talks about how the rv segment has more ridership than the east side line
Have you not literally read the rest of the comment? Come on man. You are way to smart to keep ignoring what I’ve been writing. I’ll put this as simply as I can. In terms of frequency and the total number of train cars are concerned:
It. Doesn’t. Matter.
If you have a shortage of trains cars you just run smaller trains more often. That’s it. The fact that these are two different lines gives us more flexibility. If the RV segment has more crowding than the East Side segment they will run bigger trains on the RV segment. If the East Side segment has more crowding than the RV segment they will run bigger trains on the East Side.
The only reason to run less often is to save money on operations — something they do now!
Our long regional nightmare is over! From the article: “The October train towing marks another milestone — that concrete track ties known as plinths passed quality-control checks in the Mercer Island area”
There’s a million other things that could further delay, but that one’s been weighing most heavily on my mind. Being a rainier valley rider, opening the full 2 line will lower train frequency for me, but my excitement for Seattle to have two connecting light rail lines overwhelms any disgruntlement I feel at that.
And I can only imagine there are a *lot* of north king and snohomish county riders that will breathe a sigh of relief – If they can figure out a way to get to the stations without driving.
Rainier Valley is getting lower frequency when the full East Link opens? How is that possible? The frequency isn’t very high to begin with. Looks like every 8 minutes during peak.
The RV is not getting lower frequency when the 2-Link opens. The frequency through the RV will be the same as it is today, 8-mins at peak.
IDs to LCC will get 4-min frequency, but that is because the 1-Link and 2-Link will be interlined on that section.
Correction: Rainier Valley is not getting lower *scheduled* frequency,
But the incidences of single points of failure over a given period of time will likely double.
ST will need to raise its game for dealing with breakdowns.
> Rainier Valley is getting lower frequency when the full East Link opens? How is that possible?
It is something to watch out for though. Rainier Valley may get less frequency when more branches are added to the south. It’s why I don’t actually support the idea to through-run West Seattle line into the existing tunnel — it’ll mean even less frequency for Rainier Valley.
> Looks like every 8 minutes during peak.
Did people also forget that the 1 line used to run 6 minutes at peak? That ran through rainier valley. Also frequency at late night was also decreased to run trains farther to lynnwood.
Of course the extension to lynnwood and new east link is worth it but it is something to check for decreased frequency on other segments the longer we extend. It’s something that happens with BART and DC Metro for some lines given their long length and branching.
Rainier Valley may get less frequency when more branches are added to the south. It’s why I don’t actually support the idea to through-run West Seattle line into the existing tunnel — it’ll mean even less frequency for Rainier Valley.
The train runs every ten minutes most of the day. With absolutely no work done in the tunnel the trains could run through there just fine. During peak, when the trains run every 8 minutes, it would mean trains every 2 minutes 40 seconds, just below when ST worries that it “wouldn’t give our ridership as reliable a service”*. My guess is the vast majority of riders would clearly accept an occasional ten second delay in exchange for retaining the far superior downtown station and the one-stop trips to Capitol Hill, the UW, Roosevelt, Northgate, etc.
Of course you would want to invest in making the system more reliable. But building a second tunnel makes it worse! The main reason we have so many problems with our system right now is because it is so big. There is a lot of track to maintain. We are adding riders (and thus money for operations) but not as fast as we are adding track. Speaking of which, the big thing that folks in Rainier Valley have to fear from a reliability standpoint is further extensions to the south. The Tacoma Dome is very far away. A lot can happen between there and Seattle, which means that running the trains from there is much worse than sharing the tunnel with trains from West Seattle.
Of course the best thing for people in West Seattle — and people in Rainier Valley — is to avoid building West Seattle Link and improve the bus system. Same goes for the second downtown tunnel. Put more money into maintenance to improve the reliability of Link. It is really kind of crazy how ST pleads poverty when it comes to adding things like gates in Rainier Valley (to reduce collisions) but then turns around and says it is essential that we run rail to places like Fife.
* https://seattletransitblog.com/2015/03/21/capacity-limitations-of-link/
Of course the extension to lynnwood and new east link is worth it but it is something to check for decreased frequency on other segments the longer we extend.
Yes, absolutely. It is the same idea. We are spreading ourselves too thin. Expect lower frequency and more breakdowns.
Link was originally 8 minutes peak, 10 minutes off-peak, 15 minutes after 10pm.
When the ride free area ended around 2012 or 2013, it was increased to 6 minutes peak
due to capacity/logistical issues preparing for U-Link and waiting for the U-Link and Northgate Link trains to arrive.
In 2016 or some time after it revered to the original schedule.
From 2020 to 2022 there was less service due to the pandemic.
In 2022 it restored the original schedule, although 15-minute serve may have started earlier for a while.
The full 2 Line plan has the 1 Line keeping its current schedule, and the 2 Line running in between it at the same frequency. That will give double-frequency from Intl Dist to Lynnwood, or 4 minutes peak, 5 minutes off-peak. (ST had earlier considered 3 minutes peak, but that was withdrawn as unneeded.)
In ST3 the 1 Line will run Tacoma-Ballard in DSTT2 at 5 minutes all day. That will double off-peak service. (Previously ST had said 6 minutes was the limit in Rainier Valley to avoid blocking cross traffic, but it now things SDOT will allow 5 minutes.)
The 2 Line will run Everett-Redmond, and the 3 Line Mariner-Alaska Junction, both in DSTT1. As far as I know they will keep the ST2 frequency, each line running every 8-10 minutes, for 4-5 minutes combined frequency.
WL, I *had* forgotten that frequencies had already dropped and figured ST was just doing a bad job of hitting 6 minutes.
Cool, don’t have to look forward to worse frequencies because I’ve been living it all along! Unless, as you note, running two lines causes more problems…
” Also frequency at late night was also decreased to run trains farther to lynnwood.”
That’s for a short-term maintenance project, not permanent.
ST was going to reduce peak frequency in the gap between Lynnwood Link and the full 2 Line, but in the end it managed to avoid it.
The bottleneck is not late-night, it’s peak hours, when all the trains available are deployed, and some of them spend the night in makeshift locations, because the SODO base is full, and westside trains can’t access the Spring District base yet, and the extra trains have to be rotated in at SODO for maintenance.
“The 2 Line will run Everett-Redmond, and the 3 Line Mariner-Alaska Junction”
I got those reversed. The 3 Line (West Seattle) will go to Everett, and the 2 Line to Mariner.
“It’s why I don’t actually support the idea to through-run West Seattle line into the existing tunnel — it’ll mean even less frequency for Rainier Valley.”
It could mean less frequency for West Seattle instead. To keep within the 20 trains per hour in DSTT1 ST is comfortable with (3 minute frequency), lines 1 and 2 could run as in ST2 (8-10 minutes each), and line 3 (now West Seattle) at half-frequency (16-20 minutes). Or ST could upgrade DSTT1, which would surely cost a tiny fraction of building DSTT2.
I think the Rainier Valley frequency confusion stems from ST running Link at 6 minutes before Covid. It’s 8 minutes now.
It’s worth noting that 6 minutes is about the top frequency that ST can run on MLK. That’s because people could not get across MLK as pedestrians as long as Link gets signal priority if it was any higher. The SODO crossings aren’t nearly as wide as MLK and ST proposes full grade crossings in SODO anyway so SODO could handle both lines if ST wanted.
That said, Northgate Link opening after the Covid shutdown changed the train loads. The loads between Downtown and Northgate are substantially more. Meanwhile, the Rainier Valley stations have lower daily boarding than they did in 2019 (although I believe that SeaTac is higher).
Of course, Federal Way Link may change the loads of people already on the train . And Tacoma Done too — but to a lesser extent. We will know how significant the MLK train loads are by 2026.
And some people who transfer from or to Route 7 may switch to Judkins Park (2 Line) away from Mt Baker (1 Line). That’s especially true if the rider already knows that they want Route 7 and a 2 Line train comes first when at the southbound platform Downtown.
Train loads are certainly interesting puzzle pieces!
At its core, the issue for me is flexibility in operations. ST is not building SODO for enough flexible operation and frequency. Trains stall. Service gets disrupted. Trains hit cars and people on MLK. Traffic accidents that don’t involve a train can block Link tracks. It’s why I remain concerned about tracks and platform layout in SODO. It could really be a rider disaster when things fail. And that’s “when” rather than “if”.
ST seems focused on going more places. However, by 2026 they will have to shift their primary focus more on managing the system every day.
It’s like building versus living in a new house. The building period is exciting and optimistic; home maintenance not so much.
@Mike
> It could mean less frequency for West Seattle instead. To keep within the 20 trains per hour in DSTT1 ST is comfortable with (3 minute frequency), lines 1 and 2 could run as in ST2 (8-10 minutes each), and line 3 (now West Seattle) at half-frequency (16-20 minutes). Or ST could upgrade DSTT1, which would surely cost a tiny fraction of building DSTT2.
I am not talking about the max number of train per hour in the tunnel but in about maintenance and operating costs. If there is a train from west seattle heading to everett that’d be three train lines from everett (or lynnwood/northgate) heading south. Whenever there is that much branching the other branches will have decreased service as the transit agency isn’t going to duplicate that much service.
> The bottleneck is not late-night, it’s peak hours, when all the trains available are deployed, and some of them spend the night in makeshift locations, because the SODO base is full, and westside trains can’t access the Spring District base yet, and the extra trains have to be rotated in at SODO for maintenance.
No, there was a separate bottleneck for maintenance itself. That is what I am talking about as well. The longer the train route and more branches the more it costs sound transit to run trains more frequently and the more likely the rainier or other branch segments will have lower frequency.
“ If there is a train from west seattle heading to everett that’d be three train lines from everett (or lynnwood/northgate) heading south. Whenever there is that much branching the other branches will have decreased service as the transit agency isn’t going to duplicate that much service.”
It needs to be noted that short-turning sometrains has an advantage to riders gettin on in North Seattle. If every train comes from Lynnwood or Everett, the likelihood that riders can’t get on a Roosevelt because of overcrowding is the same. However, if some trains reverse at Northgate, then riders have a much better chance of finding room on those trains.
@Al
Sure, that is true, but the main point stands that more branching can lead to lower frequency on the other branches. We through-run West Seattle and then Sound Transit will likely just run each branch with lower frequency. And both them and ya’ll here will say but the shared segment has good enough frequency.
Mike talked about the maintenance costs — well when there is an operator shortage or higher than expected maintenance costs what else would happen. Of course they’d decrease the frequency on the branches, they’re not going to run high frequency on three branches.
Again, just to clarify East Link and lynnwood extensions are exciting but
Dijibell’s original point of “Being a rainier valley rider, opening the full 2 line will lower train frequency for me” is an actual concern not something all of you are trying to handwave away as if it wouldn’t happen.
“… is an actual concern not something all of you are trying to handwave away as if it wouldn’t happen.”
What is the “concern”?
In my mind once a train or bus line runs once every 10 minutes, the concern shifts from service to crowding. 8 minutes is nice and 6 minutes is great, but with the average wait going from 3 to 4 or 5 minutes, the inconvenience concernsis relatively negligible.
Not being able to get on the next train is what I think would be a much bigger concern.
And I will say that a short-turning train operation requires the appropriate siding track and signaling. That’s especially true if drivers have to walk to the opposite end of a train. So it’s a tradeoff like all other operations options — and it should be designed at the outset.
I see the key word being “flexibility”.
I think there’s an important distinction between “will” and “might”, but maybe that’s just me.
I mean let’s give an easy example. Let’s say off peak times sound transit decides to run after 7:30pm rather than 10 minute frequency on both lines until 9 pm lowers it to 12 minute frequency on both lines citing that there’s the shared 6 minute frequency in the cid to lynnwood segment.
Are you guys going to really say it’s not because of branching that there is lowered frequency for the rainier valley/seatac segment?
We’ve literally already seen an example with extension to lynnwood they lowered frequency during night time.
This a really common practice amongst transit agencies worldwide or we can also just look at american transit agencies. If this needs to be explained further I can write an article about it but I really didn’t think it would be necessary.
When does Downtown Redmond Link begin simulation? Could Mercer Island be added into that?
I highly doubt that ST could open Mercer Island to South Bellevue on the same day that Downtown Redmond to Redmond Tech opens — and it’s not going to add that many riders.
I think it would be better to open the Downtown Redmond segment ASAP and open Mercer Island with the rest of East Link as has been planned. It would not make sense to me to delay Downtown Redmond’s opening just to add in Mercer Island.
Regardless, it seems potentially doable.
A another idea would be for ST to merely use the crossover tracks south of South Bellevue but not open Mercer Island. Right now the crossover is just south of East Main. If those could open with Downtown Redmond I could see that having some value to the 2 Line operation.
I think the contention is over the assumption that ST will reduce frequencies on the 1 Line when the 2 Line opens (which is not in any plans I’ve seen, and in fact opposite of plans to run both lines at 10 trains per hour per ST3), versus the concept that ST might reduce frequencies on the 1 Line once two or more lines are running through the DSTT.
I agree that the increased frequencies in the DSTT will likely result in delays cascading down through Rainier Valley, but that just tells me ST will have to get good at turning trains back at SODO (and Judkins Park) to maintain operations outside the trunk line.
@Nathan
It was specified that they were planning on decreasing frequency on line 1 when opening line 2. I think I’ll write an article on this to explain how branching works in general, and also how sound transit chooses frequency given so many do not understand.
https://www.soundtransit.org/st_sharepoint/download/sites/PRDA/FinalRecords/2023/Presentation%20-%20ST2%20Light%20Rail%20Service%20and%20Passenger%20Experience%2009-14-23.pdf#page=15
Page 15 in 2026 peak frequency on Line 1 will drop from 8 minutes to 10 minutes given the insufficient train vehicles to serve both line 1 and 2.
To clarify they might fix that issue if they have extra train vehicles in the order they placed, but if they don’t have enough Sound Transit definitely has drawn up a couple plans about decreasing frequency on Line 1. There were a couple more powerpoints in the past.
Again I really do not understand why so many here refuse to understand branching when it comes to rail. It is not that different from bus branching.
Mike, I think that it would be Everett-West Seattle and Mariner-Redmond, to keep the run lengths and times roughly equal between the two lines. And that is ideal, because neither Everett nor West Seattle will ever need service more often than every ten minutes even at the peakiest peak of peak.
The one argument for Mariner-West Seattle is that it could be automated because it will have no at-grade running except within freeway envelopes.
I think most of the people understand branching. What they are saying is that for Rainier Valley it doesn’t matter. The most they were ever comfortable running trains was every six minutes. If the three lines from the south combined in one tunnel, that would be trains running every two minutes. We should be able to do that.
Realistically, we will probably never see six minute frequency in Rainier Valley. Sound Transit moved to eight minutes frequency (at most) quite a while ago, and doesn’t appear to be interested in running more often. Eight minutes on each line works out to 2 minutes and 40 second (for three lines) through the tunnel. It is fairly simple math. https://humantransit.org/2011/02/basics-branching-or-how-transit-is-like-a-river.html
Page 15 in 2026 peak frequency on Line 1 will drop from 8 minutes to 10 minutes given the insufficient train vehicles to serve both line 1 and 2.
That is a different issue and has nothing to do with branching. The number of trains you need is dependent only on how long it takes to get there and how often they run. For example imagine we give up on East Link, but extend the main line to Marysville to the north and Olympia to the south. This means no branching. But it means we need a lot of trains to go that far (assuming we run them every 8 minutes during peak).
Of course one option is to simply run smaller train sets. That applies here as well. If ST lacks enough train cars then the obvious solution is to run smaller trains more often. But it is quite likely they won’t do that. That is because it is clear that ST does not care about frequency. They consider ten minute frequency to be standard. The only reason they run more often than that is because of crowding. Otherwise how do you explain the fact that we don’t run every eight minutes all day and night?
This also explains why we ran trains every six minutes way back when. We didn’t have enough train cars. ST was trying to avoid crowding. Thus it is quite possible that we will run every ten minutes in the future all day long. Not because we branch. Not because we don’t have enough trains cars. But because Sound Transit is being cheap and doesn’t want to spend the extra money running more trains than are necessary to avoid crowding.
Fundamentally, ST is not a transit agency but a transit-construction agency. Their goal is to build more and more miles of rail, not improve transit as much as possible. Once you wrap your head around this idea everything else makes sense (eight minute headways in the middle of the day; rail to low-density distant destinations; a second tunnel downtown that doesn’t add any coverage). The branching is fine. The lack of commitment to quality transit is not.
> That is a different issue and has nothing to do with branching. The number of trains you need is dependent only on how long it takes to get there and how often they run.
That literally is proposed because of the branching. if there wasn’t the two branches sound transit would never have suggested it.
“Are you guys going to really say it’s not because of branching that there is lowered frequency for the rainier valley/seatac segment?”
If DSTT1 can handle 20 trains per hour, the reduction in frequency isn’t caused by the branching, but by other limitations.
I think that’s what people are missing: there’s other constraints than just sheer tunnel capacity.
If the reason is a shortage of equipment that hasn’t been delivered yet, SoundTransit should be asked about possibly leasing some of the surplus MAX cars. MAX ridership has been slow to recover, and the spares ratio is some 50%. The static converters would have to be changed for the operating voltage, but I think that’s just a matter of changing the tap configuration (Eg, a multi-tap transformer can be configured for multiple voltages). TriMet could easily spare 20 or so cars for the next couple years and not notice. In fact, it might help ease the storage problem TriMet has.
“… neither Everett nor West Seattle will ever need service more often than every ten minutes even at the peakiest peak of peak.”
I agree. I’ll just mention that 1 four car train has the same capacity as about 6 or 7 60-foot buses. With a train every 10 minutes or 6 trains an hour, that’s a directional capacity of 4,200 riders in a single hour at a conservative 700 riders per train. West Seattle boardings are forecast at less than 13,000 a day so a single hour will get nowhere near that 4200.
While not known yet, it would not surprise me if the Lake Washington crossing of 2 Line also won’t need more than 10 minute service even at the peak. ST is saying 50,000 2 Line riders on East Link in 2030 or 25,000 in one direction. Given the commute peak being bi-directional across Lake Washington, I doubt that a peak hour will approach 17 percent of a full day. And that 50,000 seems improbable and it includes Bellevue-Redmond commuting that I’m ignoring in the discussion.
So 6 trains from West Seattle, 6 trains from Bellevue, that’s 12. That leaves 8 trains to reach 20 trains an hour for Rainier Valley (or 7.5 minutes on average).
Or maybe each of the three lines gets 6.6 trains an hour at peak times (19.8 total) and that may be enough service and it’s still less than 20 trains an hour.
It comes down to actual field data. Forecasting can only inform so much.
If I had my druthers, we first should not decide on DSTT2 until the full 2 Line opens. We can’t afford its construction anyway. If the loads across Lake Washington don’t merit higher frequencies it should be known by 2026.
@Glenn
Again I am not talking about the tunnels max frequency…
The limitation that we are talking about is number of trains and available operators. with sound transit running two lines it will be running a lot more trains and since there’s relatively high frequency in the middle it encourages it to decrease it on the branches.
Y’all need to stop focusing so much on most amount of frequency possible in the tunnel and think about what is the operationally possible amount of frequency given number of trains, operators and maintenance. These latter issues is why branching will (and was proposed) to decrease frequency on the rainier valley segment and again has nothing to do with the tunnel
Ah, I’d forgotten that the train shortage was still going to be significant after the bridge connection opens and East Link is running full-length.
I do remember a significant part of the issue being a need to increase the spare ratio, and longer-than-expected runtimes on the 1 Line. Perhaps they’ve made progress toward resolving those issues?
I don’t think we can say that 1 Line frequencies will decrease solely because of the 2 Line crossing I-90. It may be a contributing factor but it certainly isn’t the only factor.
The linked presentation talks about 2026 after the Federal Way extension opens as well – adding ~8 miles of new track will also require more train cars (or less frequent headways). I believe that presentation was also given at a time where ST also highlighted a need for a higher spare ratio due to maintenance issues on the new fleet and a slow zone near Royal Brougham Way. If those can be resolved by 2026, it would help with the train car shortage.
I also believe they also moved up an order of a couple trainsets to help with the shortage.
Okay let’s put it this way if we had it flipped around and instead of i90 we used 520. When instead the lynnwood link extension frequency drops are you guys going to say it isn’t because of the branching?
Or perhaps to illustrate it another way currently from lynnwood to SeaTac would be around 60 minutes. Lynnwood to Redmond town center would also be 60 minutes. To provide the same frequency across the lines sound transit is running around double the number of trains.
“We’ve literally already seen an example with extension to lynnwood they lowered frequency during night time.”
As I said, it’s because of maintenance. Maintenance happens periodically regardless of whether we build a branch or not.
@Mike
Look frequency = # of trains / length of the route
When the increased the length of the route they had insufficient trains to maintain the same frequency. The branching and having two lines will double the effective length of the route as I explained above.
The number of trains is capped by the amount of physical trains, train operators and the amount of maintenance for available actually running trains. The maintenance leading to insufficient trains is just one example of insufficient trains for a longer route. It could be not enough drivers or not enough trains etc…. But it’s all stemming from a longer route
“Realistically, we will probably never see six minute frequency in Rainier Valley. Sound Transit moved to eight minutes frequency (at most) quite a while ago, and doesn’t appear to be interested in running more often.”
It appeared in the Ballard Link plans this year. No, I don’t have a citation but I’ve seen it multiple times. ST has been saying 6 minutes is the limit in Rainier Valley, but now it’s saying 5 minutes is doable (or at least might be doable), and that’s what it’s planning to run Ballard Link with.
WL: but what is actually their limiting factor?
What I propose could solve a lack of cars problem until they receive more, if that is the limit.
If it’s operators or maintenance, then that’s another kettle of fish.
@Glenn
It is no different from running a longer bus line or more bus lines. It just takes a lot more buses and costs a lot more money.
Sound transit is no longer running a short 30/40 minute light rail line but will be running two 60 minute long light rail lines. It will in general cost a lot more to run the same frequency. With the branching/shared segment sound transit will just optimize for whatever capacity and it will definitely lead to decreased frequency on the branch segments.
Sound transit will need more train operators, larger train omfs, maintenance staff etc…
Again this is not that different from how it works with bus branches, I am not sure why everyone here who understands how it works with buses is magically not understanding the same concept when it comes to trains.
There is some extra complication with branches where typically the easiest branching way is to split every other train so it goes from 5 to 10 or 6 to 12 min as it’s hard to interline some combination of 1 line going every 8 min and say 2 line every 12 min.
ST has a specific problem with the full 2 Line opening four years late, and therefore being unable to access all its trains or the Eastside maintenance facility. Normal operations are also requiring more trains than expected. That doesn’t mean we’re doomed to lowering frequency; it means ST should order more trains. As for the ST3 future or the three-lines-in-one-tunnel concept, the alignment isn’t even committed to yet so there’s no specific number of trains ST should/will/won’t have for it. Yet you’re assuming ST won’t have enough when we don’t even know what that “enough” is yet. That’s an issue for a decade in the future, not now.
As to whether longer lines and more branches will cause even more than the baseline of reliability problems and force a frequency cut, well, that’s speculation. The logical conclusion of that argument is that we shouldn’t have ST2 or ST3 because it’s impossible to have a network more than one 15-mile line that’s reliable/frequent. It’s rather late to make that argument about ST2/ST3, the time was 8 and 16 years ago. Other cities have 15-line networks that are pretty reliable. ST has a surprising number of reliability problems even with one short line, and it will probably remain that way until ST improves its quality, regardless of whether we add extensions or branches or not.
@Mike
It’s was proposed in two years in 2026
https://www.soundtransit.org/st_sharepoint/download/sites/PRDA/FinalRecords/2023/Presentation%20-%20ST2%20Light%20Rail%20Service%20and%20Passenger%20Experience%2009-14-23.pdf#page=15
Page 15 in 2026 peak frequency on Line 1 will drop from 8 minutes to 10 minutes given the insufficient train vehicles to serve both line 1 and 2.
> The logical conclusion of that argument is that we shouldn’t have ST2 or ST3 because it’s impossible to have a network more than one 15-mile line that’s reliable/frequent
That’s not what I said, what I said was agreeing with Dijibellls statement that the branch will lead to lower frequency on the rainier segment.
Perhaps this is more obvious to anyone else who has lived in a city with a branch rail line corridor but that is generally how it works.
Regarding the maintenance and not enough vehicles etc… that’s how it always is for all transit agencies. There will also be a competing pressure between providing service/frequency versus the cost. When there’s a branch and combined segment that will always lead to the branch receiving less unless there’s some great demand on it.
> It’s rather late to make that argument
No there’s still debates about the exact service patterns remember? Like whether to turn back trains earlier at northgate rather than running them all of them to lynnwood. And also for st3 how far to run the trains.
Again as I said many times, I’m not as worried about the max frequency in the tunnel but rather about how much frequency operationally sound transit can actually fund given the lengthening transit lines. It’s an actual concern I mean look at bart or wmata they had some awful like 20 minute frequency sometimes on the weekend midday. I don’t think sound transit will be that bad — but it can tend towards that if ignored.
I mean we keep talking about the peak frequency right now what about the off peak. Is sound transit going to run degrading down to 12 or 15 minute frequency for the branches and 6/7.5 minute past 7/8pm. There’s really nothing suggesting they wouldn’t and they probably would. I mean I don’t know why some are expecting them to act differently from other American transit agencies
Gee. Figuring out a miscounting of the needed vehicles in 2023 for a system approved in 2008 really speaks volumes about the naked truth of ST:
It’s been run by too many people who don’t have enough experience in running a rail transit system.
Please ST! Quit hiring people from other fields and actually hire people who actually have years of experience in planning, building and operating a multi-line, high-demand rail system!
Not FTA. Not an airport. Not a city DOT. Not a smaller bus-only transit system.
@WL,
“ It’s was proposed in two years in 2026”
Nope. You should read your own documents.
That doc clearly states that ST intends to run @ 8-min headways. It says that on almost every page.
And the 4@10 stuff? That is not a “plan” to do something. It is just contingency planning. Everyone knows ST has had trouble with LRV availability. It’s not a mystery, nor a surprise at this point.
And that doc is obsolete anyhow.
How do we know? Because it talks about the same problem existing for LLE, and proposes the same 4@10 solution for LLE. Yet everyone of us knows that the 1-Link, with LLE fully open, is running day in and day out at 8-min headways. Just go stick your head in any 1-Link station. You’ll see.
And, as stated in your doc, IF this problem did occur after full 2-Link opens, the 4@10 solution would be only temporary, and only until 2030 anyhow. At that point ST would go back to 4@8.
But that is a really big *IF*. And it is not something I am going to lose any sleep over.
And it is certainly not a technical feature of “branching”, which is not the correct terminology anyhow.
@WL,
“ it’s hard to interline some combination of 1 line going every 8 min and say 2 line every 12 min.”
No it’s not. It happens all the time, all over the world.
It’s usually not the optimal solution from a user point-of-view simply because the trains will arrive unevenly on the interlined section. It ends up being sort of like bus bunching, and that is considered a bad thing with rail.
And Link is not a “bus”, and it shouldn’t be thought of as a bus. That might be your first mistake.
@Lazarus
> “ it’s hard to interline some combination of 1 line going every 8 min and say 2 line every 12 min.”
> No it’s not. It happens all the time, all over the world.
You’re right, let me correct my statement. It’s currently “hard” for Sound Transit to do so as every option they list only running equal frequency on both line. If they are able to break that assumption then they could run higher frequency on line 1 without having to increase on the line 2.
> And, as stated in your doc, IF this problem did occur after full 2-Link opens, the 4@10 solution would be only temporary, and only until 2030 anyhow. At that point ST would go back to 4@8.
As I stated earlier there will always be some ‘temporary shortages’ of xyz whether not enough train operators, trains, or maintenance. That is exactly why I am bringing it up now.
> And that doc is obsolete anyhow.
What was Dijibel”s original statement “Being a rainier valley rider, opening the full 2 line will lower train frequency for me”. I’ve literally provided a document example where sound transit was proposing such a service pattern. And we don’t even know the what the off peak service pattern would have been like.
Honestly it’s leaning towards a bit insulting how all of you are treating rainier valley riders as if they don’t matter and are trying to gaslight the situation. If the same thing was happening with Lynnwood extension segment with say northgate turnbacks I doubt I’d have to repeatedly explain the situation over and over again.
> And Link is not a “bus”, and it shouldn’t be thought of as a bus. That might be your first mistake.
If it was a bus there wouldn’t be such a problem with increasing the line 1 frequency. The entire problem is that Sound Transit wants to match the line 2 frequency with the line 1 frequency. Check out the slide 15 i showed earlier. line 1 seatac south of cid has overcrowding in the original proposal, but they can’t increase the line 1 frequency without also increasing the line 2 frequency.
If the 2 and 1 line ran at different frequencies, you’d have problems with uneven passenger load between North Seattle and downtown.
@asdf2,
“you’d have problems with uneven passenger load between North Seattle and downtown.”
ST isn’t pursuing asymmetric interlining for 2 very good reasons:
1) it’s a monumentally “dumb” idea, particularly for high ridership, high demand lines
2). Doing so wouldn’t meet ST’s service goals of providing 8-min service on both lines.
It’s a total non-starter.
@WL,
“If it was a bus there wouldn’t be such a problem with increasing the line 1 frequency.”
Oh man, way to make me laugh. Points awarded.
You need to sit down and do a little math.
A Link LRV at full load carries 200 people. In 4-car trains that represents 800 people per train. And those trains run every 8 mins currently. Going to every 4 mins interlined.
A Metro artic carries around 60. So that is 13 buses to replace one Link train. But let’s say 12 for easy math.
So that is 12 buses to replace every Link train. Or in terms of frequency it is one bus every 40 secs, going to one bus every 20 secs to match Links eventual 4-min headways. That is a lot of buses!
But wait! It gets worse! Capacity is not thru-put.
The last time I took a CT commuter bus from Lynnwood Station to DT Seattle it took over an hour. And that is an express bus with no intermediate stops.
Link takes less than half that long. So in order for to replace Link a bus operator would need to run an artic about every 10 secs. That is around 360 buses per hour!
Could a bus agency do such a thing? Sure! Is it smart? No! Is it economical? No! If Metro tried to do such a thing would it push them over the financial cliff? Absolutely!
This is why we built LR. Because buses hit their limit, and they just aren’t the right approach on certain, high ridership corridors. Link is.
@Lazarus
that is not what I am talking about but about the branching and turning around a vehicle to increase the frequency. They cannot do that easily with the trains compared to buses.
Sigh has no one here lived/used a train system with branches.
Look again at slide 15. Why wasn’t sound transit able to increase frequency on line 1 even when it predicted to be overcrowded. It was because they didn’t want to turn around trains at northgate and had to run trains all the way to lynnwood. Why can’t they just increase the lynnwood to federal way trains then, because they’d also need to increase the lynnwood to redmond trains.
To increase frequency on just one leg they end up having to spend 4 times the frequency on the doubling from lynnwood to downtown seattle and also from seattle to redmond.
what I said was agreeing with Dijibellls statement that the branch will lead to lower frequency on the rainier segment.
Perhaps this is more obvious to anyone else who has lived in a city with a branch rail line corridor but that is generally how it works.
That suggests we don’t understand the fundamentals, which is ridiculous. We get it. I even came up with a reference for it. What we are saying is that it won’t happen in Rainier Valley because the trains are already infrequent. Furthermore, you haven’t explained why it would happen. There are couple reasons why it could:
1) The trains can’t handle running twice as often north of CID. This is simply not true because again, the trains are infrequent. The trains can easily run every four minutes north of the CID. If trains were running every five minutes on Rainier Valley then this would be an issue, but they aren’t.
2) As a money saving tactic they will reduce frequency during peak (the only reason they don’t do that now is because the north end is crowded). OK, I suppose this could happen, but it seems highly unlikely. The last thing ST wants to do is needlessly increase crowding from the East Side the day it opens. Federal Way Link will increase ridership for the main line as well. I just don’t them cutting frequency at this point.
I mean we keep talking about the peak frequency right now what about the off peak. Is sound transit going to run degrading down to 12 or 15 minute frequency for the branches and 6/7.5 minute past 7/8pm.
Yes, but they could do that now. It is the same dynamic. If you are focused only on crowding (and saving money) then you don’t run trains as often as we do most of the day. Branching has nothing to do with it, really. At this point Sound Transit is willing to provide a certain base level of service that has nothing to do with crowding. This may or may not continue as the system gets longer, but branching has little to do with it.
If the 2 and 1 line ran at different frequencies, you’d have problems with uneven passenger load between North Seattle and downtown.
Not necessarily. What folks are talking about is an uneven pattern. So rather than a conventional 50/50 branch you do something unusual. For example, imagine this is the schedule for a southbound train at Westlake:
Lynnwood to Redmond: 8:00, 8:04, 8:12, 8:16, 8:24, 8:28 …
Lynnwood to Federal Way: 8:08, 8:20, 8:32 …
If you are trying to get from Northgate to downtown you have a train every four minutes (it doesn’t matter which one it is).
In any event I think this is highly unlikely. I don’t expect a huge mismatch between the two lines. You also don’t get that much out of it. You are making things much worse for the those in the south, while making things only a bit better in the north. The only reason I could see doing something like this is if things are too limited because of Rainier Valley and the trains are too crowded from the East Side. But we’ve run trains down Rainier Valley every six minutes — I see no reason why we wouldn’t do that again. That would mean trains every six minutes from the East Side, which should be more than enough.
> ST isn’t pursuing asymmetric interlining for 2 very good reasons:
> Yes, but they could do that now. It is the same dynamic. If you are focused only on crowding (and saving money) then you don’t run trains as often as we do most of the day. Branching has nothing to do with it, really
Branching has literally everything to do with it, especially if Sound Transit is sticking with symmetric interlining.
> That suggests we don’t understand the fundamentals, which is ridiculous.
Yes, I am not suggesting but outright claiming one does not understand the fundamentals if so many do not understand how train branching works and how it will lead or pressure to lower frequency on the branches.
Here I’ll explain it step by step:
Oversimplified schedule:
Lynnwood to Seattle (cid) ~30 minutes
Seattle to Redmond ~30 minutes
Seattle to Federal Way ~45 minutes
Northgate to Seatac ~45 minutes
headway = travel time / vehicles
Let’s say we have 9 vehicles
45 minutes * 2(for north/south) / 9 vehicles ~= 10 minutes headway or 6 tph
We now extend from lynnwood to federal way
75 minutes * 2 / 9 ~= 16 minute headway or ~4 tph.
You’ll need to run a lot more trains or around 15 trains for the same frequency
75 minutes * 2 / 15 trains ~= 10 minute headway or ~6 tph.
Then we have the two lines of course so now it is
(75 + 60) * 2 / 15 trains ~=18 minute headway or ~3 tph.
So now we need to run even more trains almost doubling at 27 trains
(75 + 60) * 2 / 27 trains ~=10 minute headway or ~6 tph.
For the RV segment while it used to only take 9 trains for 10 minute frequency now sound transit is running 27 trains through the entire system as a minimum given the symmetric interlining.
For northgate to Seatac to increase from 10 minute headway to 8 minute headway by itself only used to take 2 extra train cars from 9 vehicles to 11 vehicles. Or even 6 minute head way was from 9 to 15 vehicles.
45 minutes * 2(for north/south) / 9 vehicles ~= 10 minutes headway or 6 tph
45 minutes * 2(for north/south) / 11 vehicles ~= 8.18 minutes headway or 7.33 tph
45 minutes * 2(for north/south) / 15 vehicles ~= 6 minute headway or 10 tph.
> Yes, but they could do that now. It is the same dynamic.
Look at slide 15, there is overcrowding on line 1, but they still can’t run more trains. Why is that.
For them to increase from 10 minute headway to 8 minute headway it is no longer just adding 2 trains. They now need to add a train running longer to lynnwood and to federal way and then with symmetric branching must also add a train running from lynnwood to redmond.
(75 + 60) * 2 / 27 trains ~=10 minute headway or ~6 tph.
(75 + 60) * 2 / 34 trains ~= 7.94 minute headway or ~7.5 tph
They need to add 7 more trains instead.
The same thing applies for off peak.
Train schedules: (older estimates)
http://stb-wp.s3.amazonaws.com/wp-content/uploads/2015/08/14151500/Screen-Shot-2015-08-14-at-8.12.45-AM.png
http://stb-wp.s3.amazonaws.com/wp-content/uploads/2015/08/10143005/Screen-Shot-2015-08-09-at-9.07.39-PM.png
Yes I know the run time/headway calculations are more complicated with turn around time, I’ve simplified it here.
“The only reason I could see doing something like this is if things are too limited because of Rainier Valley and the trains are too crowded from the East Side”
There could just as easily be more demand to Federal Way and SeaTac than the Eastside.
Either way, the Rainier Valley trains seem crowded enough I don’t think you’d want them any less frequent.
> There could just as easily be more demand to Federal Way and SeaTac than the Eastside.
That is literally what the slide 15 states… that there is more demand to rv side than to the eastside segment. That is why sound transit is recommending to run smaller trains there…
Anyways I’ll probably write up an article, this would be easier with some pictures for others to follow
@WL,
“ that is not what I am talking about but about the branching and turning around a vehicle to increase the frequency. ”
I’m not really sure what you are attempting to say there, but you seem to be missing the main problem here.
The problem isn’t that ST can’t find a solution to deliver the required capacity and avoid overcrowding. There are literally a plethora of operational scenarios that can deliver the required capacity.
No, the problem is actually LRV availability. ST’s preferred solution of providing 8-min headways on both lines actually works perfectly well per capacity. The problem is that if there aren’t enough available LRV’s at any given time to provide 8-min headways, then ST needs to back away from that approach. And that is what causes crowding issues.
And adding frequency on Northgate-FW is exactly the wrong approach. That is nearly the longest leg in the system, and it is certainly the slowest. Adding frequency there, no matter how you do it, just raises the LRV requirement and makes the problem worse. And the data clearly shows that adding capacity oat FW is just wasted capacity, and wasted LRV service hours.
No, the first option is exactly what the ST presentation calls out. Get the LRV availability up. If LRV availability can be increased to 85% (or more), then a lot of this problem goes away.
And if it doesn’t go away completely, then the next step is to look at spare ratios and gap train requirements. Cutting back on those areas will also move ST closer to a full 4@8 capability on both lines.
And you seem to forget that ST has another knob they can turn. They can adjust train length too. If 2-Link doesn’t require 4-car trains when operating at 8-min headways, then ST can cut back to 3-car trains and redeploy those LRV resources to the 1-Link.
But the solution is not adding overlays to the two lines. The solution begins and ends with LRV availability for regular service.
And those charts are obsolete anyhow. I wouldn’t get too wound up about obsolete information. ST has clearly moved on from where they were a few years ago.
@Lazarus
The main point is the same as what was stated from the beginning the branching leads sound transit to lower frequency on the rainier valley segment lower than what sound transit would otherwise run.
@WL,
“ branching leads sound transit to lower frequency on the rainier valley segment”
That is a fundamentally false statement. Interlining has exactly zero impact on base frequency up to the point at which the interlined segment becomes maxed out.
The important thing here is that ST moving from 8-min headways to 4-min headways will NOT max out the interlined segment. So such an operational change will not impact base frequencies on either line.
It’s just the way the world works.
@Lazarus
> That is a fundamentally false statement. Interlining has exactly zero impact on base frequency up to the point at which the interlined segment becomes maxed out.
The powerpoint referenced literally proves you wrong. Given the interlining and even with higher demand sound transit is unable to run trains more frequently in rainier valley even though it can now precisely.
If you are going to claim it is false then explain why Sound Transit couldn’t just run more trains on the segment in that scenario.
Again I am not talking about the max frequency in the tunnel but about the constraint from number of available trains whether due to physical trains, maintenance or number of train operators.
How about this for the simplest analogy I can think of that will get the point across:
Imagine that every time one wanted to increase the Route 7 bus frequency with an additional bus per hour, instead of just increasing route 7 one also had to increase RapidRide E with an additional bus. But not only one extra bus but also another RapidRide E and a 550 bus as well because they are interlined.
Line 1: rapidride E + route 7
line 2 : rapidride E + route 550
Now every time one wanted to increase the frequency of route 7 it is tied together to the cost of increasing 3 other bus routes. This is not generally bad for the shared segment but will lead to situations where the agency is forced to run less route 7 buses than it typically could because it has to run so many rapidride E buses and run more than needed 550 buses just to maintain consistent frequency.
Yes, I am not suggesting but outright claiming one does not understand the fundamentals if so many do not understand how train branching works and how it will lead or pressure to lower frequency on the branches.
Isn’t it possible that we understand the idea just fine, but that you are simply raising issues that at best are only tangentially related to branching? You dismiss this idea without offering an explanation as to why. Then you engage in a lengthy discussion focused on the number of trains which fundamentally has nothing to do with branching itself!
Consider these scenarios:
Scenario 1: We build a system that is similar to what we’ll soon have, but without branching. From Redmond a train goes across the lake to downtown but instead of going in the existing tunnel it goes in a brand new one. Then it goes over Aurora all the way to Lynnwood, ending a mile west of the Lynnwood Transit Center. This means absolutely no branching. It takes just as long to get from CID to Lynnwood as the existing line. This solves the problem, right? No! The trains have to travel just as far, which means if you want them to run just as often you have to have just as many trains.
Scenario 2: We build things the way we have now, but much smaller. Link does not go north of the UW, south of Rainier Beach or East of Downtown Bellevue. Does this change anything? Yes! Absolutely! Same amount of branching, but suddenly we have more than enough trains.
Scenario 3: The tracks are built as we planned, but the trains from Federal Way turn back at Northgate; trains from Redmond go to Lynnwood. Would this fix the problem? Maybe.
Scenario 4: We run as planned, with Federal Way and Redmond trains going to Lynnwood. However the trains from Redmond have three cars, while the trains from Federal Way have four cars. Does this solve the problem? Again, maybe.
The branching itself isn’t the issue. Fundamentally it has nothing to do with it. Holy cow, you even cite the current situation — Lynnwood Link — as an example of this sort of problem. But guess what: Lynnwood Link no branching! It has nothing to do with it. There are only three key parameters:
1) Frequency
2) Time it takes to run the trains end to end
3) How many trains we have
There are several solutions:
1) Buy more train cars.
2) Run smaller train sets (see scenario 4)
3) Run less frequently.
4) Run shorter lines (scenario 3 is an example of this).
Oh, and then you say this applies off-peak, which makes no sense. Off-peak the trains run less frequently.
You do raise the issue of dependence. If we want to increase frequency on Line 1 we can’t do it unless we raise the frequency on Line 2. Fair enough. But why would we do that? That suggests a major mismatch in service demands that just won’t exist. It is ridiculously presumption to think that ST would run trains every eight minutes from Federal Way and every ten minutes from the East Side during peak (if not for that damn branching). I don’t see that at all. Folks from the East Side would have a major hissy fit. You’re years late with this thing *and* you aren’t running as often as you said you would (or as often as the folks from Federal Way)? WTF?
But consider scenario four again. How does that compare to running four-car trains from Redmond every ten minutes (which apparently is Sound Transit’s dream scenario): Four-car trains every ten minutes means 24 train cars an hour (6 per hour times 4) . Three-car trains every eight minutes means 22.5 (7.5 per hour times 3). That is practically identical. Better for riders, but the same amount of train cars in use.
The point is, there are real issues involving branching — they just don’t apply here. For example bunching can be a problem — but we are talking about running trains every four minutes (at most) on a system that can easily handle trains every three minutes. Another is that we could be forced into running trains less frequently. For example it could be that the tunnel can only handle trains every three minutes and we are currently running trains every four minutes down Rainier Valley. In that case running East Link on the same tracks would result in a loss of frequency for Rainier Valley. But again, that isn’t an issue. We are running (at most) every eight minutes. We can’t run trains that often even if we want to.
There are more issues related to branching — I just don’t think they apply here. The issue you mentioned (lack of train cars) is a different issue entirely.
Imagine that every time one wanted to increase the Route 7 bus frequency with an additional bus per hour, instead of just increasing route 7 one also had to increase RapidRide E with an additional bus. But not only one extra bus but also another RapidRide E and a 550 bus as well because they are interlined.
You keep going over the same point while ignoring the fact that we get it. We get it man! There is a dependence between East Link and Federal Way Link. So freaking what? It doesn’t matter. ST is not going to arbitrarily run one line frequently and the other line infrequently — even if it could. There is no reason to!
Hell, if nothing else it violates the spirit, if not the letter of subarea equity. Why would we arbitrarily run one line way more often, given they two lines are likely quite similar? Oh, and if we did have complete independence, what makes you think it would be the line to the south that would benefit? It seems just as likely that East Link runs more often (at certain times of the day).
@Ross
What was the original thing that Dijibell stated that opening east link will lower frequency on rainier valley. I have explained how it can over and over again.
1) there was literally a proposal by sound transit to do so
2) outlined how high maintenance costs from lengthening the route can lead to lowering frequency and the lynnwood extension is
3) showed how the branching and symmetrical frequency means it costs a lot to run the higher frequency
4) people here talk about “temporary until 2030” didn’t we just talk about sound transit running out of money? Are you all expecting sound transit to prioritize higher operating costs
You say east link would throw a hissy fit about their frequency being lowered, well are you expecting rainier valley riders to just be happy about their frequency being lowered?
Anyways if yall really don’t believe me why don’t we just email and ask sound transit. If there is such a supposed guarantee they will not lower frequency on the rainier valley segment. If there isn’t then stop pretending there is such one
@Ross
I had to keep going over it because people kept talking about the max capacity of the tunnel when that was completely not what I was talking about.
What was the original thing that Dijibell stated that opening east link will lower frequency on rainier valley.
But you claimed it was because of branching! That is the part that is completely wrong. Again, if East Link was just as long, but didn’t branch they would have the same problem (not enough train cars). The problem is that the lines are too long and we may not have enough train cars to run four-car trains as often as we want, or as far as want. But again, branching isn’t the issue. Otherwise solutions (smaller train sets, turning back trains, buying more trains) simply wouldn’t work. There are real problems with branching — this isn’t one of them.
@ WL:
There does not have to be synetrucal frequency. Period. San Francisco and Boston don’t have it. Go stand at Civic Center Ststion in SF or Copley station in Boston and you won’t see synryrucal trains arrive during the commute hours. SF Muni even only describes frequencies on its main page; the actual schedules are buried in a PDF and they aren’t symmetrical.
https://www.sfmta.com/routes/k-ingleside
Boston won’t even publish a full timetable. All the train frequencies are in ranges:
https://cdn.mbta.com/sites/default/files/media/route_pdfs/batch_7116/SUB-S4-P4.pdf
Chicago’s Pink and Green lines are branches and they don’t have symmetrical schedules either:
https://www.transitchicago.com/schedules/
I admit that it’s nice to have symmetrical schedules. And it’s important if trains branches get 15 minute service or less. But it’s not required as part of the trunk service.
Finally, ST even has unpublished reliever trains on reserve at peak times and for stadium events. A transit operation needs that kind of flexibility to manage crowding. It’s just how things are done in practice.
What I was trying to say and apparently didn’t articulate clearly, is that ST has specific problems with insufficient trains due to the 2 Line infrastructure not being fully active yet, and the actual service level requiring more trains than predicted. The solution to this is to acquire more trains. This would obviously take a few years and require some money. ST may do it when it’s not so busy finishing the 2 Line. That may lead to a temporary substandard frequency, but it doesn’t mean that would necessarily be permanent.
@Mike
I understood what you’re talking about, but attributing it to insufficient resources is the incorrect one. It’s like the swiss cheese model and blaming the last factor.
If one switched from a fuel efficient sedan to a suv blame the car not traveling as far on not buying enough gas, or the initial switch to a fuel inefficient car?
Or let’s say for Lynnwood, if we had a turnback at Northgate are you going to tell Lazarus that the reason why the lynnwod section’s frequency dropped from ~8 minutes down to 10 minutes is not because of the turnback but because of insufficient frequency?
“temporary substandard frequency” also Mike you and others keep implying some sort of pledge by sound transit to keep higher frequency for rainier valley. Can this be cited somewhere?
@Mike Orr,
“ ST has specific problems with insufficient trains due to the 2 Line infrastructure not being fully active yet”
Yep. I think you get it.
Supporting LLE without access to OMF-E has been a major task for ST, and the problem has been made worse by higher needs to support their target service level of 4@8, and by having too many LRV’s out-of-service at any given time.
ST has solved the LLE problem with distributed storage, and by improving their LRV in-service availability. That is as it should be.
When Full ELE and FWLE open the goal is exactly the same – support both lines with 4@8. They will do this with a combination of strategies, but increased LRV in-service availability and reduced spare and gap ratios will be a key part of the solution.
Another part of the solution might be to reduce train lengths on either the 1-Link or 2-Link, but those details are TBD. And might not even be needed. But frequency will be maintained.
But the key takeaway is that ST has no plan to reduce frequency in the RV. And reducing frequency is in no way a direct result of opening Full ELE, or of interlining. That is simply not true.
“temporary substandard frequency” also Mike you and others keep implying some sort of pledge by sound transit to keep higher frequency for rainier valley. Can this be cited somewhere?”
Current service. That has been the minimum service level since Link started in 2009, except during the pandemic. ST obviously things Rainier Valley should have at least 8-10 minute service because that’s what it has always had.
Link is different from other American light rails and BART because its minimum frequency is 10 minutes instead of 15 minutes (or 30 minutes on weekends). That makes it higher quality in that aspect, more like a European rapid transit or Chicago/DC/New York. I don’t see ST throwing that away, or the North King representatives letting it. It would be throwing away one of the good things Link has.
It may not be written in an ST2 document because it’s an underlying assumption, “Of course it will keep its frequency.” How else can it deal with rising population growth, rising ridership, and adding KDM and Federal Way riders. Peak hours it’s close to full, so I wonder if it will have capacity problems with KDM and Federal Way. ST wouldn’t lower frequency in that circumstance.
It may have to lower frequency if it doesn’t have enough trains, but it can solve that by ordering more trains. Then there’s the question of whether it can afford to since it wasn’t budgeted in ST2, but ST has plenty of money for more trains or frequency if it delays or scales down ST3 project(s).
“I understood what you’re talking about, but attributing it to insufficient resources is the incorrect one. It’s like the swiss cheese model and blaming the last factor.
“Insufficient resources” usually means a lack of money. That’s not the case here. It’s just that ST underbought trains and undercalculated the train requirements. If would be like if Metro didn’t order enough red buses for a new RapidRide line.
“Or let’s say for Lynnwood, if we had a turnback at Northgate are you going to tell Lazarus that the reason why the lynnwod section’s frequency dropped from ~8 minutes down to 10 minutes is not because of the turnback but because of insufficient frequency?”
No, because you’re the one saying turnbacks or branches cut down frequency, not me. As far as I’m concerned, it ST adds a turnback or branch, it will add trains to compensate to maintain frequency. Except when it miscalculates travel time and turnaround time, or the trains need more maintenance than expected. All three of those happened here. That’s more of the “specific problems” in this situation. It doesn’t mean all situations will necessarily be like that.
You can always write to ST and ask for confirmation of future frequency.
> But the key takeaway is that ST has no plan to reduce frequency in the RV. And reducing frequency is in no way a direct result of opening Full ELE, or of interlining. That is simply not true.
I literally cited sound transit documents that disprove that statement
“I literally cited sound transit documents that disprove that statement”
And Lazarus thinks you’re misinterpreting the document.
Page 1: It’s dated 9/14/23. That’s over a year ago. Things may have changed since then. It’s not a board-voted policy decision; it’s a staff status report from a year ago.
Page 3: “Light rail feet is capped at 214 LRVs until the early 2030s when more fleet and base capacity becomes available.”: So ST has already ordered the additional vehicles needed for full service, or at least is planning to.
Page 3: “There are not enough LRVs to support originally planned service levels”: So it’s a limitation ST is working around, not an intended reduction.
Page 4: “We committed to come back to the board with impacts and options for the board to consider”: So staff are trying to fix the problem. And they’ve been trying for over a year, so maybe they’ve done something since then. A year ago ST also thought it would have to reduce frequency on Lynnwood Link because it didn’t have enough layover places for extra trains, but in the end it found the space and Lynnwood Link opened with no reduction. Maybe something like that will happen with the Rainier Valley situation: it can close the gap, or at least close it partway.
Page 12: “Strategy: Improve fleet availability. Work to fix warranty and preventative maintenance issues.”: That would help fill the gap. Maybe it can be finished by 2026: that’s over a year away.
Here is a link to the Urbanist news item about ST ordering 10 more vehicles:
https://www.theurbanist.org/2024/04/06/sound-transit-greenlights-10-more-link-cars/#:~:text=Sound%20Transit%20will%20purchase%2010,the%20way%20for%20the%20procurement.
The item was written in April of this year — well after the 2023 Board presentation.
@Mike Orr,
“ It’s not a board-voted policy decision; it’s a staff status report from a year ago”
Exactly. ST has been churning in this for a while now, and these studies are not final plans. And nothing has been baked in yet.
And the docs are clearly out of date as they don’t even align with what ST is currently doing with LLE in service. The implication of that is that a lot has changed since the doc was written, and that ST went with a solution path that isn’t even listed. Namely distributed storage.
But the bottom line remains that ST is committed to providing 8-min frequency, and that is the goal they are working towards.
And, in absolutely no way shape or form, does interlining force a frequency reduction in the RV. Any changes would be driven by fleet availability and not by interlining itself.
And that is the other key takeaway from this. Any temporary operational changes driven by fleet availability would be exactly that – TEMPORARY.
The sky isn’t falling. There isn’t a secret plan to screw the RV. ST isn’t doing something nefarious here. None of that is true. ST is only trying to work around a temporary limitation in fleet availability. That is the kind of problem transit agencies work with all the time.
@Al / Lazarus
I don’t think you guys have been following the sound transit news closely or listening to what I’ve been saying enough to understand that the 10 trains didn’t actually fix the issue. The past Urbanist articles have literally talked about the same issues I’ve been referencing about the lowered frequency.
If you read the article it says
> Sound Transit intends to hold Siemens to contract terms for fully delivering and getting the vehicles accepted for passenger service by March 2028
Secondly if you guys thought about it for a couple seconds, how would going from 214 trains to 224 trains fix the issue. Going from 10 minute to 8 minute frequency on both lines requires like ~20% more trains.
If you guys read the other linked articles by the Urbanist
> [once the 2 Line is able to start running through Seattle] …Overall though, peak frequencies won’t be able to attain the planned eight-minute headways — every 10 minutes may be the best case per line — and crowding could be a common problem in the system core.
Also remember how I talked about OMF’s earlier
> Sound Transit settled on 10 extra cars because that’s the number of vehicles that likely could be stored in increasingly tight operations and maintenance facilities (OMFs) in Seattle and Bellevue. The agency told The Urbanist that they could have exercised an option for 16 vehicles, a further six beyond what’s currently contemplated. Those extra trains, however, would need to find storage outside of OMFs, which evidently is why they aren’t being pursued.
If you guys were to then actually read the most recent document https://www.soundtransit.org/sites/default/files/documents/st3-light-rail-service-presentation-april-2024.pdf
> OMF Central will exceed design capacity by 35%
> Off-peak frequency and “service span” reductions are required to meet peak demand
> Schedule adjustments include moving from 10-minute to 12-minute
headways from 8pm-10pm.
Lastly, now you guys are trying to shift the goal posts. First I showed that sound transit had plans to decrease service and then you guys shifted the goal post to “but now it isn’t”. Then when I prove that it still is decreasing the service now you guys shift the goal post to “its only temporary”.
@Lazarus how about you actually prove that is temporary? If you thought about it you’d realize sound transit needs to build or expand an omf and then send in an order to siemens to procure more vehicles and then on top of that operationally spend more money. It could be a decade at that point.
Anyways, this is why I keep saying I do not worry about the max frequency in the tunnel, but about what Link can operationally provide.
> There isn’t a secret plan to screw the RV. ST isn’t doing something nefarious here. None of that is true. ST is only trying to work around a temporary limitation in fleet availability.
I never said it was a secret plan to screw RV. At the same time it is true that the branching will lead to lowered frequency on RV.
You guys just have not been actually following the sound transit presentations, probably just reading about the sections north of CID. And failing to realize that the sound transit board has repeatedly talked about decreasing the peak/off peak frequency south of CID relative to what currently exists.
Any changes would be driven by fleet availability and not by interlining itself.
Exactly. Furthermore, it is highly likely that any changes will result in the same frequency, just different sized trains. There is no shortage of trains. There is a shortage of train cars. Right now we have enough hardware to run two-car trains every five minutes (we just don’t want to).
If anything, all the interlining does is encourage ST to run the trains more frequently. Imagine for a second that these trains are not interlined, but we still have a shortage of train cars. Also imagine that East Link is not nearly as crowded during peak. Assuming ST wants to run both lines to Lynnwood, they have a couple options:
1) Run three-car trains to Redmond and four-car trains to Federal Way. Have both trains run every eight minutes.
2) Run (four-car) trains to Redmond every ten minutes, and (four-car trains) to Federal Way every eight.
The second option saves money, but the first option is better for riders. But remember, the money saving options isn’t actually possible. Thus, if anything, the interlining is forcing ST to run trains more often. As long as one branch can justify eight minute frequency, the other branch gets it too.
This is especially true in the north end. The only reason there will be a big increase in frequency on the section that gets the most riders per service-hour is because of interlining. Right now, trains from downtown to the UW are running every ten minutes in the middle of the day. The heart of our system — the section that every transit expert in the world said should be a subway — is finally going to get good frequency (five minutes all-day long). This is happening because of interlining.
This brings up another issue. You mentioned how a branched line creates a dependence. This is true, and it exists for our bus system as well (3/4, 10/12, 31/32, etc.). But there is another dependence — the rest of the line. Consider that section of Link again: from the UW to downtown. This is a very cost effective section from a ridership/service-hour standpoint. It is also very urban, with strong all-day demand. It would be quite reasonable to run trains there more often (e. g. every six minutes all day long). But that is not the case of trains to Lynnwood or Tukwila. The destinations are not as cost effective (because they are farther away). Demand is more heavily oriented towards peak. Thus you have a major service mismatch. You want to run trains from downtown to the UW more often, but not trains to Lynnwood more often. But you are stuck, because it is all part of the same line. The only solution: Branching! Yes, that’s right. The very thing that creates a dependence can also be used to reduce dependence. Some of the trains could turn back at the U-District (theoretically).
Again, we see this with buses. It is not as common anymore (since it complicates things) but it still exists. In the Central Area there are two versions of the Metro 4. One goes to Judkins Park, while the other stops short (at Garfield). This can also be considered a branch, with the version that ends at Garfield being much shorter. Of course this sort of thing also happens with rail systems (especially regional rail).
Branching typically creates a dependence that is based on fractions. Half the trains go this way, half that way (or one-third, two-thirds, etc.). But lack of branching creates a one-to-one dependence. The reason trains run so infrequently from downtown to the UW is because the trains also go to Angle Lake and Lynnwood. These are all just trade-offs. You could have each section be its own subway line, but that would be worse in many ways.
If anything, I would say the future setup (after East Link) is actually closer to ideal. Yes, it is overkill for Lynnwood. But at least the core our system will get really good frequency. The rest of it will remain unchanged. (Unless, of course, ST is feeling a pinch from all of these long-distance lines and decides to cutback on service — but that has more to do with length and destinations than branching itself.)
@WL,
“ If you guys were to then actually read the most recent document”
Actually, I would suggest that you read the documents you link to before you comment on them.
First, the 35% shortfall in OMF-C storage capacity doesn’t refer to 2026, it refers to the here and now. This is the problem that ST faced when opening LLE without access to OMF-E. And this is the problem that has been discussed multiple times on this blog before.
ST solved the OMF-C problem using distributed storage. It isn’t the most elegant solution, but it works. And all of us who have ridden 1-Link lately can attest to that. It works.
But the other part of your doc is even more informative. It says that after FWLE opens, 8 to 10 min headways will be possible on 1-Link. That is very good news, because it means that ST can meet its promise of 8-min headways even with interlining.
So, ya, your doc says no reduction in frequency is necessarily required. That is very good news.
And it should also be noted that there are other, more creative solutions, that might actually do better. With potentially even better frequency.
But there are still potentially up to 2 years before FWLE opens. Hopefully ST continues to make solid progress, but until FWLE opens everything is still TBD. But your doc is very encouraging. Progress has been made!
> Branching typically creates a dependence that is based on fractions. Half the trains go this way, half that way (or one-third, two-thirds, etc.). But lack of branching creates a one-to-one dependence. The reason trains run so infrequently from downtown to the UW is because the trains also go to Angle Lake and Lynnwood. These are all just trade-offs. You could have each section be its own subway line, but that would be worse in many ways.
> If anything, I would say the future setup (after East Link) is actually closer to ideal. Yes, it is overkill for Lynnwood. But at least the core our system will get really good frequency. The rest of it will remain unchanged. (Unless, of course, ST is feeling a pinch from all of these long-distance lines and decides to cutback on service — but that has more to do with length and destinations than branching itself.)
@Ross
That is exactly what I’m talking about, the dependence means the lynnwood segment and the redmond segment will get more frequency than it “normally” given the same number of resources, while it is harder for the rainier valley segment to even maintain it’s original frequency. In order to increase the rv segment means it has to increase the other two segments at the same time.
Even when there is demand on the rv segment they have to run more trains to lynnwood and redmond before then can run more trains on the rv segment
The Mike Orr Link headway history does not seem quite correct. I recall that it was six-minutes in the peaks until Covid.
It isn’t much, but the report to the System Expansion Committee on 10/10/2024 is reporting that the project is “On track for Q4 2025 Revenue Service”.
From viewing the progress courtesy of the 550, it appears that the eastern high rise of the bridge is getting most of the attention right now. Even when I’m coming home past 6:00, a crew is out working under lights.
https://www.soundtransit.org/st_sharepoint/download/sites/PRDA/FinalRecords/2024/Presentation%20-%20East%20Link%20Budget%20Amendment%20and%20Contract%20actions%2010-10-24.pdf
Picnic spaces on a freeway ped bridge
I wonder how loud the noise would be
Or the view
Or building all the new TOD up next to the freeway. Ah, the fresh air on the top deck!
picnic on new bridge
freeway roar ends conversation
better than nothing?
It’s too loud here
Let’s go to the Downtown Park
Or Meydenbauer Park
Where are they?
On the Grand Connection trail this bridge is part of
See ya, freeway
Wouldn’t want to be ya
That’s a looooong bridge too. About the same size as the ped bridge at Northgate. I mainly see the ‘nodes’ being used as places people will rest their weary bones as they hike to/from downtown.
That’s a looooong bridge too. About the same size as the ped bridge at Northgate.
Yeah, it looks quite similar. It crosses over about as many lanes of traffic — a wide part of the freeway. In both cases it makes for a very good bike path. The improvement for biking is huge.
In the case of Northgate the station is close to the freeway but the bridge parallels the freeway for a while, which means it takes longer to cross. With Bellevue the station is a bit farther from the freeway. In both cases the other end of the bridge is a ways away. In the case of Northgate there is nothing much in between. In the case of Bellevue there is, which is why it looks like they plan on adding elevators at various places (which makes sense). Plenty of people walk across the bridge at Northgate for a couple reasons. First it is the main way to get from the college to the station. Second because it is one of the few pedestrian crossing in the area (alternatives require a long walk). In contrast I see few fewer people walking across the bridge in Bellevue — at least not the whole way. It is often closer to Wilburton and there are nearby alternatives for walking across.
But yeah, unless they plan on completely covering the freeway I think a picnic area is silly.
I think Bellevue was talking about a lid on 405. That may be longer-term.
Has anyone noticed that there is a potential RapidRide corridor identified as a high-tier project in the STP? It’s under “38 | Eastlake to Rainier Beach | Transit + Multimodal Improvements” and looks like it connects the 49 to the 36.
https://seattle.gov/transportation/about-us/seattle-transportation-plan
See: STP Part 1, page 120; Appendix A, page 44
I’ve been meaning to write up a summary of transit projects included in the Transportation Levy; there are a handful of corridor projects that are not-quite-rapidride and probably more in alignment with the arguments from our “No More Rapidride” article earlier this year: https://seattletransitblog.com/2024/03/03/no-more-rapidride/
Hi John, I talked about it previously in https://seattletransitblog.com/2024/07/30/rapidride-corridor-1064/ as 1064A if you want to read about it more.
King County Metro decided not to move forward with it, though STP is funded by Seattle the city itself. It is a bit unclear what/if such a rapidride will still be built.
Ash Way P&R is an excellent option for catching a quick and frequent bus to Lynnwood Station. There are four express bus routes from the P&R to the station during AM peak: CT 201, CT 202, STX 512, and STX 513. They use the dedicated freeway ramps on both ends, so no stop lights.
CT 166 is also pretty quick and straight, but off the freeway. The Swift Orange Line takes 16 minutes, with 3 quick stops along the way. Just be sure to catch the one going toward Edmonds College. CT 112 is a little slower, but will also get you there. All these options, except STX 513, are also available off-peak, 7 days a week, late into the evening.
.
All that said, I wish CT would re-organize the bus bays at Ash Way. All the buses headed to Lynnwood ought to serve the same bay. Riders should not have to choose between waiting at Bay 3 for the 201 or 202, or at Bay 1 for the 512 or 513, more than half a football field away.
All the buses exiting I-5 northbound ought to get to serve Bay 4 along the exit-ramp road. The one missing bus is ST Express 532, which, like ST 512, is headed to Everett. Instead, the 532 serves Bay 1, about a football field away, and does a pointless bonus loop within the bus transfer area.
That would leave just five routes for the miscellaneous CT-only bay.
I would move the Lynnwood-Station-bound buses to Bay 3, across the bus street from Bay 4, to minimize the effort new riders have to make to find their correct bay. The miscellaneous CT buses would move to Bay 2. Bay 1, so very far from the others, and a scary place to wait, could be shut down.
Ash Way P&R is an excellent option for catching a quick and frequent bus to Lynnwood Station.
I agree. I think people will figure this out. It is not very intuitive. People are used to driving to a park and ride and catching their bus to their destination. Or at the very least catching their bus a long ways (downtown Seattle) and then making a transfer (to First Hill). But for people approaching from the north it seems like parking at Ash Way is a better option.
All that said, I wish CT would re-organize the bus bays at Ash Way. All the buses headed to Lynnwood ought to serve the same bay.
Agreed. It appears they are organized by agency instead of destination, which doesn’t make much sense. The agencies do charge different amounts, but fairly soon ST will charge the same amount as Link. Since CT charges less than Link riders who transfer to Link will be charged $3 for the trip no matter which bus they take. Again this might take a bit of education.
It isn’t clear to me which buses serve each bay. Community Transit has a nice map, but they don’t list the buses (https://www.communitytransit.org/images/default-source/services-images/park-and-ride-images/ashway.webp?sfvrsn=8db72796_2). Google maps lists them like so:
Bay 1: 512, 513, 532
Bay 3: 109, 112, 119, 121, 166, 201, 202
Bay 4: 201, 202, 512, 513
Bay 5: Orange Line
It doesn’t list Bay 2. Bay 4 is basically for people getting off the bus (from what I can tell). It is the first stop for the bus when it leaves the freeway. Consolidation doesn’t matter as much when you get off the bus. The only advantage to serving that stop is that people north of there (e. g. by the apartments) have a shorter walk. I assume that all buses make a loop before continuing.
The 109, 112, 119, 121 end at Ash Way. The 166, 201, 202, Orange Line and ST buses keep going. For buses that go in a loop like this it is better if they serve one bay one direction and another bay a different direction (to avoid confusion). Riders just use the proper bus stop — they don’t have to read the writing on the bus (to determine which direction it is going). It is also best if they are grouped by common destination. It think something like this would work nicely.
Bay 1: 201, 202, 512, 513 to Lynnwood (via I-5)
Bay 2: 112, 119 (which start there) along with the southbound 166, 532 and Orange
Bay 3: 109, 121 (which start there) along with the northbound 166, 201, 202, 512, 513, 532 and Orange
Of course the Orange may want to stand alone (because it is “special”). Whatever. The point it that you have various buses that work for the same destination. The big one is the one you mentioned: all those buses that run express to Lynnwood use the same stop. There are other groupings. The 109, 201 and 202 go to Mariner. The 201, 202, 512 and 532 go to Everett. The 121 and 166 share some stops along 164th.
The Orange Line has its own on-street stop (Bay 5) that serves Swift buses going both directions. The buses then do a u-turn at a traffic circle.
There are actually riders boarding at Ash Way to go to Everett. That’s the main reason to group the northbound 532 with the northbound 512. The other reason is that any buses coming off northbound I-5 that skip Bay 4 then have to do a loop within the P&R to serve a different bay. The ones that serve Bay 4 and then get back onto I-5 to continue north simply head straight out to the street, turn left, left, twirl 270 right, and are back on I-5 north with about 300 yards of deja vu.
Perhaps the 166 could have an on-street Bay 6 adjoining the Swift Bay 5. Of course, that would partially split up the Lynnwood-bound routes. A variation would be to just have Bay 6 for the eastbound 166. Perhaps the 121 could start at Bay 6 as well, but it would have to do the traffic-circle u-turn before heading south.
I get your point about 109, 201, and 202 all heading to Mariner P&R. Rather than have northbound 201/202 do a loop before turning north, I would suggest an on-street Bay 7 just north of Ash Way P&R.
Another variation:
If the transit agencies want *all* the Lynnwood-bound routes together, then have all of them at the hypothetical Bay 6, next to the Swift Bay 5.
Both directions of CT 166 would be there, along with CT 112, which would start there. Those two routes would use the traffic-circle u-turn.
The other four routes (201, 202, 512, and 513) would turn right onto the dedicated I-5 entry lane, completely avoiding the inner bus loop.
Between that new Bay 6 and moving 109, 201, and 202 northbound buses to hypothetical Bay 7 north of the I-5 access road on Ash Way, and northbound 512, 513, and 532 at Bay 4, that would just leave 119, 121, and southbound 532 serving Bay 3 in the bus loop. Bays 1 and 2 could be mothballed.
The ones that serve Bay 4 and then get back onto I-5 to continue north simply head straight out to the street, turn left, left, twirl 270 right, and are back on I-5 north with about 300 yards of deja vu.
Oh yeah. That makes sense. Since these buses are coming from the I-5 bus-ramp but can’t reenter the ramp the stop saves the bus from having to do an extra loop.
The Orange Line has its own on-street stop (Bay 5) that serves Swift buses going both directions. The buses then do a u-turn at a traffic circle.
OK, so it doesn’t go into the main loop (with all the other buses). That makes sense. With those two ideas, here is the plan then:
Bay 1: 201, 202, 512, 513 to Lynnwood (via I-5)
Bay 2: 112, 119 (which start there) along with the southbound 166, 532.
Bay 3: 109, 121 (which start there) along with the northbound 166, 532.
Bay 4: 201, 202, 512, 513 northbound
Bay 5: Orange Line (both directions)
That has a really good balance, which is nice. The Orange Line is the only bus where you have to read the front of the bus (to figure out which direction it is going). You lose some combinations, but not that many. The 109 has a different bus stop than the 201/202 even when the buses are all heading to Mariner. The 532 to Everett is separate from the 201, 202, 512 (all of which go to Everett as well). I think this is the best we can do, unless…
We move Bay 4. Right now it is slightly east of the loop. We could move the stop a few feet to the west. That way a bus that goes through the loop (to turn around) could stop there (instead of Bays 1, 2 or 3). That would mean:
Bay 1: 201, 202, 512, 513 to Lynnwood (via I-5)
Bay 2: 112, 119 (which start there) along with the southbound 166, 532.
Bay 3: 121 (which start there) along with the northbound 166.
Bay 4: 109, 201, 202, 512, 513, 532 northbound
Bay 5: Orange Line (both directions)
That would be ideal, but is probably not worth it. You might introduce congestion for the buses leaving the loop. I probably wouldn’t bother, seeing as the other proposal seems good enough to me. There are probably only a handful of people going from the parking lot to Mariner. More people would benefit from combining the northbound 532 with the other buses, but notice I put the 532 at Bay 3, which is not that far from Bay 4 — a basketball court not a football field :). A savvy rider would wait at Bay 4 and if they see a 532 approaching from Ash Way (the street) turning into the parking lot, they should be able to easily catch that bus (since the bus has to do a loop). The 532 isn’t nearly as frequent as the 512 or 201/202, either. Overall I think the proposal above is good enough.
In the long run it makes sense to build the ramps from the north, which would mean that Bay 4 wouldn’t make sense any more. Coming from Lynnwood, the 201, 202, 512 and 513 would make the loop and then just head back onto the freeway. Thus they could use one of the regular bays when heading north which means it would be trivial to add the 109 and 532 to them.
I suggest you go check out Ash Way P&R’s current bay configuration.
I suggest you go check out Ash Way P&R’s current bay configuration.
I’m not sure if I ever will. I used to live in Lynnwood, but didn’t have reason to go to there. I have parked there once (for a weekend hike) but people have moved around, so we stopped doing that. I can imagine myself taking a bus through there, but I probably wouldn’t get off the bus there. In any event, is Google right (to the best of your knowledge)? They have it like so:
Bay 1: 512, 513, 532
Bay 3: 109, 112, 119, 121, 166, 201, 202
Bay 4: 201, 202, 512, 513
Bay 5: Orange Line
My guess is that in terms of direction, I assume it is:
Bay 1: 512, 513, 532 (all southbound)
Bay 3: 109, 112, 119, 121 (which start there) 166 (both directions) 201, 202 (southbound)
Bay 4: 201, 202, 512, 513 (all northbound)
Bay 5: Orange Line
Is that right? What confuses me is that there is nothing listed in Bay 2. In any event, what we have sketched out would be a huge improvement and goes back to your original suggestion. Buses heading to the same location should serve the same stop.
I’ll ask my friend about the Ash Way bay configuration or I’ll check when I go there for Thanksgiving. The last time I was there the Swift Orange station was on the street, not inside the P&R, and both directions stopped at the same station. So I don’t understand the part about the Orange sharing bays with other routes or having a different bay for each direction.
The Swift Orange Line stop is the only on-street stop, and serves both directions, which certainly can cause riders to occasionally get on going the wrong direction.
There are no routes serving Bay 2, and close to a football field’s length between Bays 1 and 3. I have no idea why they did that.
As much as Ross is confused by the loops, at least he has the excuse of being an armchair analyst. Why CT and ST would collectively come up with much-less-sensible bay assignments than Ross is a real head scratcher.
At any rate, the lowest-hanging fruit here is moving the northbound 532 to Bay 4, so it does not have to pointlessly do the loop, and riders going to Everett don’t have to sprint the football field when they see the 532 approaching from I-5 and basically outrun the bus.
Second-lowest fruit, but the greatest positive impact, is moving southbound 512 and 513 up to Bay 2 and southbound 201 and 202 down to Bay 2 , or all together at Bay 1 or Bay 3 — pick any of those three bays. Sprinting from Bay 1 to Bay 3 when a Lynnwood-Station bound rider sees “201” or “202” is even more of a losing race.
The only thing that ought to be controversial here is whether STX 512 skips Ash Way P&R altogether. The combining of bus routes with identical outbound paths at the same bay (the 512 and 532 going north to Everett via I-5 and the 201, 202, 512, and 513 going south to Lynnwood Station via I-5) is a no-brainer,
While I am on the topic of bus bays,,, Did ST look at the feasibility of having on-street bays for the 201/202/512/513 close to a station entrance/exit? If that could be done, most of the routes remaining in the old transit center could move closer to the south station entrance.
I’m pretty sure I understand it, based on Brent’s explanation. Google was largely correct. The reason they didn’t show any buses for Bay 2 is because it isn’t used. What Google doesn’t show is the direction the buses are headed. This is where Brent’s explanation helped immensely. So correct me if I’m wrong, but this is how it works:
Bay 1: Southbound: 512, 513, 532
Bay 3: Start there: 109, 112, 119, 121 — Both directions: 166 — Southbound: 201, 202
Bay 4: Northbound: 201, 202, 512, 513
Bay 5: Both directions: Orange Line
The Orange Line does not enter the Park and Ride area. The busses serving Bay 4 skip exit the bus-only lane of the freeway, serve the stop, then head straight over to Ash Way, skipping the loop. The other direction those buses go through the loop. The 109, 112, 119, 121 all start and end there. I assume they layover inside the loop. The 166 and 532 go through the loop both directions. Thus if I’m correct, riders of the 166, 532 and Orange Line have to make sure they get on the right bus (it isn’t enough to get to the right bus stop).
Again, correct me if I’m wrong here. Google Maps also has a good approximation of the distances. This being the USA, I’ll list them by feet:
Bay 1 to Bay 3: 190 feet
Bay 3 to Bay 4: 150 feet (requires crossing the crosswalk).
In terms of getting to Seattle or Everett it is a mess.
Heading to Seattle: You want the 201, 202, 512 or 513. This means Bay 1 or Bay 3 — a distance of roughly 190 feet.
Heading to Everett: You want the 201, 202, 512 or 532. This means Bay 1 or Bay 4 — a distance of roughly 340 feet (we are talking Canadian football field now). Think about that one for a second. You are at Bay 4, waiting for that bus to Everett. The 201/202 works, but it is not as fast as the 512 or 532. In the distance you see a Sound Transit bus (you recognize the colors). It is traveling north on Ash Way, about to turn into the transit center. It could be a southbound bus though — you have to wait until you can read the numbers. As it makes the turn you can see it — the 532. Great! Except now you have to cross the crosswalk, and walk end zone to end zone before that bus gets there. It is quite possible that the bus isn’t picking up anyone at that stop (because you are generally better off with Bay 4). So you have to cross the crosswalk and walk (or jog) while waving your hands so that the driver stops for you.
It just isn’t good.
At any rate, the lowest-hanging fruit here is moving the northbound 532 to Bay 4
I don’t think you can do that. Someone who rides the 532 can correct me, but I don’t think the bus uses the bus-only freeway ramp to Ash Way Transit Center. A northbound bus has to be in the right lane of I-5 once it leaves 405. At that point there isn’t a lot of space to move over all those lanes. It is my understanding that instead of attempting that maneuver, the bus simply stays in the right lane and exits at 164th (the way a normal car would). The schedule map has a detailed portrayal of the route and shows it doing that: https://www.soundtransit.org/ride-with-us/routes-schedules/532. If I’m wrong, and the bus does manage to get into the bus-only exit (on the left) then the 532 should definitely use Bay 4.
Otherwise it can’t. Bay 4 can only be used by buses that come off of the bus-only freeway ramp.
That is why I suggested putting the northbound 532 at Bay 3. It is less than ideal, but at least minimizes the distance. Another alternative would be to move Bay 4 (as I suggested) but that has drawbacks. It is worth noting that the 532 only runs peak and is not that frequent. It runs roughly every half hour while the 512 runs every fifteen minutes. The 201/202 also runs every fifteen minutes, making them a good combination, especially for going to Lynnwood which brings up your second point:
Second-lowest fruit, but the greatest positive impact, is [grouping southbound 201, 202, 512 and 513 onto the same bay]
Agreed. That is basically what prompted this thread in the first place. This would be trivial and improve things immensely. All those buses are identical in the southbound direction from Ash Way. They should serve the same stop.
The only thing that ought to be controversial here is whether STX 512 skips Ash Way P&R altogether.
Yeah, that is a completely different subject. In the past I came up with some ideas, but they are definitely controversial. You are right, moving the buses around in the bays should be a no-brainer.
Oh, and STX 532 Bellevue-bound could join the miscellaneous bay.
Skagit Transit says they may have significant impacts if 2117 passes (repeals climate action):
https://www.skagittransit.org/CCA/
A tour of East Link and the Microsoft campus, from the YouTuber underwhelmed by Lynnwood’s walkability. https://youtu.be/VUG-eefNDCg?si=L0x3cewwyIai1Oqt
Does anyone think that platform screen doors at Link stations will ever happen? Would it allow for more trains in the DSTT? How much deterrence do we need to avoid trains hitting people who accidentally or purposefully end up on the tracks in front of a train or maybe brushed by a train when it pulls into a station?
I used to think it was superfluous — but it seems to be increasingly commonplace around the world. It could help provide more frequent train service. On the other hand, it would be one more layer of coordination and mechanization to go wrong and door positions could not move when buying a new series of vehicles in the future.
Thoughts?
I don’t think it’s necessary and will make upgrading trains to different trainsets more complicated as well.
How often do people go on the tracks? When I was living in Tokyo and in New York, I heard about people getting run over on the tracks somewhat regularly, but those places get so much traffic that I always assumed it happened relatively rarely.
It happens an average of every forty day according to this: https://www.theurbanist.org/2019/08/06/every-40-days-train-collisions-happen-in-the-rainier-valley/
These every 40 day collisions are not happening inside stations for the most part, so platform screen doors would not change those numbers significantly. They would allow trains to approach at a faster speed, but ST doesn’t seem interested in speeding up the trains. Better isolating the MLK corridor with fencing and crossing arms would allow faster speeds and increased safety, but ST doesn’t seem interested in spending the $$.
It’s actually done partly for a suicide prevention and a homicide prevention (getting pushed onto the tracks) concern in addition to preventing accidents. However, the gates can also allow for things like trains that can skip stations without crawling through them.
Agencies don’t like to discuss the suicide aspects. They rightfully don’t want to put ideas in people’s minds.
@Evan — Correct, sorry. I read your comment in isolation. I was thinking about collisions in Rainier Valley, not people falling (or being pushed onto) the tracks. Different subject (and a much bigger problem in places like New York than it is here). Sorry about the thread drift.
I don’t like the idea of adding more mechanical systems to stations with the level of escalator downtime ours have.
I feel like people make so much hullabaloo about the escalators and elevators being constantly broken even though they’re working more consistently and ST is actually taking initiative on fixing the long backlog of DSTT maintenance that Metro deferred for decades. My only complaint is just that the elevators are inconsistent speed wise. Some definitely need an upgrade so they operate much quicker.
@Zach B,
“ I feel like people make so much hullabaloo about the escalators ”
I concur. But I’ve sort of taken it as a sign of how well things are going at ST. It’s sort of like people complain about escalators when there is nothing else to complain about, but the implication of that is that things must be going pretty darn well then.
And remember, when ST took over the old bus tunnel escalator availability was horrible. Escalator availability was only 22% in Pioneer Station at the time. It was really pretty bad, but people hardly mentioned it.
Now escalator availability is much, much better. Yet people seem to complain about it more. Personally I’d give credit where credit is due, but I guess this is Seattle.
Just call it “the Seattle Whine”.
I feel like people make so much hullabaloo about the escalators and elevators being constantly broken even though they’re working more consistently
The only real flaw was at the UW, where they didn’t have a backup. Otherwise some level of failure is likely. Folks tend to make a big deal about it, even though I agree — I wouldn’t consider it that important. But I’m likely in the minority. Greater Vancouver voted down a very good mass transit proposal in large part because of unhappiness with things like escalators and elevators. It is worth noting that Vancouver likely has one of the most cost-effective systems in North America — especially impressive given the size of the city. But agencies are often damned if they do, damned if they don’t. Spend more money on maintenance and people complain about the cost. Expand the system and everyone cheers, ignoring the fact that this means more to maintain. It isn’t just transit either. It is very difficult to get agencies to focus on maintenance — the leaders want to cut ribbons. It is only when things go wrong that they freak out and complain about the lack of maintenance.
Is L the only one here living in a parallel universe in which trains have not been breaking down at least once a week, since the service change, forcing single-tracking for part of the 1 Line, and roughly 25-minute headway for the rest of the line, once trains get moving again?
Somehow, it must be Metro’s fault. If the breakdowns started after the service change, it must be operator error, by new operators!
The solution to vertical conveyance breakdowns is not to have fewer vertical conveyances, but more, for redundancy. Even when there are two escalators, ST policy allows them to be broken down at the same time, without fining the contractor, so long as each is operating properly at least 95% of the time. So, the contractors install the cheapest escalators that meet that low bar. ST does the work of fixing them or figuring out someone hit the emergency stop button. There may also be a need for a state inspector to be involved, but ST has been lobbying to reduce the downtime due to a shortage of state inspectors.
I want to get people’s thoughts on the West Seattle and Ballard Link opening dates. The official ST opening dates are still listed as 2032 and 2039.
I just don’t see it.
West Seattle: Tacoma Done Link is already pushed to 2035 and that has no tunneling and deep stations. Northgate Link with its tunneling was 9 years between groundbreaking and opening. Assuming that the groundbreaking happens in 2026 (the earliest that I can see happening), the 9 years makes it 2035 at the earliest. Keep in mind that assumes no roadblocks. I’m thinking 2037 is a probable date.
Ballard/ DSTT2: it’s going to take a money miracle to have the project fundable as planned into at least 2035 or 2040. Plus the difficulty of digging subway stations in the middle of Downtown and SLU will add probably two more years to construction so say 11. Even if groundbreaking is 2035, opening would likely be 2046. If it takes longer to arrange the funding. It seems likely to get pushed past 2050. I’m thinking 2051 as a probable date.
What do others think will be the realistic opening dates (assuming the full funding comes through)?
It’s possible for west seattle asssuming it just takes all of the ballard link money for now (aka heavily delays ballard link), above ground construction doesn’t take that long and the tunnel portion is a short one. It wouldn’t be waiting a long time for a tbm to dig like how northgate had to
Yeah, I fear we’re slipping towards an outcome where not only does West Seattle take all of that Ballard Link money, but West Seattle doesn’t even get a line to downtown for that, only to SODO.
Even the cable-stayed bridge high over the Duwamish will take longer than a typical track section on columns. Keep in mind that trains cannot run through West Seattle unless they cross the bridge. So ST will need at least 9 more months once the first test train is in the bridge if not more.
The Tillicum Crossing Bridge in Portland took about 5 years between the construction award and the transit line opening. However that bridge was able to restrict boat usage during the construction and the soil seems less problematic. Elliott Bay is quite deep and the soil around there appears to be volcanic or dredged fill. Given other soil-related Link delays in Kent as well as the problems we’ve seen in the literally adjacent West Seattle bridge, the bridge will be a challenge.
To get to 2032 and allowing 5 years, the bridge plans would need to be fully designed and bid by 2027. That’s really pushing it.
As far as the Alaska Junction station goes, it’s not so easy to create. The TBM will likely be aimed, brought to the site from the Port, then repositioned and turned around as the second bore. But the real time and effort is digging the 100-foot football-sized hole in West Seattle for the station.
It would be a stretch to get it only to Avalon by 2032 but ST has not yet mentioned opening it in two phases.
When ST3 was presented in 2016, the “excuse” for West Seattle Link opening 5 years before Ballard Link was that no tunneling would be involved. The project now has a deep tunnel and a deep subway station. And ST is already pushing Tacoma Dome Link to 2035 for soil reasons in Fife, which seem analogous to those around where the bridge is supposed to go.
>To get to 2032 and allowing 5 years, the bridge plans would need to be fully designed and bid by 2027. That’s really pushing it.
Based on the interview with Mestas, it sounds like she’s going to push for a Progressive Design-Build (PDB) process for WSLE, and we’ll know what the project will actually cost once they hit 60% design later next year. PDB also allows for pre-construction during the design process, but the Board will have to authorize that expenditure. I imagine the headache right now for the WSLE team is the work on shaving cost off the project and trying to keep it competitive for FTA grants. Perhaps they’re waiting for the results of next Tuesday to figure out much trimming they’ll have to do.
@ Nathan:
As glamorous as progressive design-build is as a concept, I am hard pressed to see how that works well for a tall cable-stayed bridge. I would think that a bridge has to be fully designed and funded before any construction can begin. Bridges like this have too many things that can wrong and testing for seismic, wind and soil performance must be determined first.
Plus, the station excavation off of Alaska Street will add considerable time onto the project. It’s not like the way that the tunnel segment was built in Bellevue. It’s more akin to Northgate Link, which had two deep stations. This extension has one deep and one shallower station. Plus the two extensions are almost the same length.
As I noted before, I think there is a disconnect between the original schedule (no subway stations) and the current one (two stations underground).
Finally, let’s be clear about how ST has not kept on schedule on anything opened in the past 10 years. East Link in particular was supposed to have opened in 2021 then 2023 — and now ST is trying its best to get it opened across Lake Washington by Christmas 2025.
>I would think that a bridge has to be fully designed and funded before any construction can begin.
I suggest reading up on the Progressive Design-Build process; the “Progressive” part is the ability for an owner to off-ramp the design-builder if needed between the design and build stages, after the designer (who is also the presumptive builder) provides a maximum cost to build to the Owner (in this case, Sound Transit). Pre-construction activities like utility mapping, site investigation, etc., happen during the design process. Relocations can also happen during the design process if there are elements common to all likely designs for construction, but that requires coordination between the Design-Builder and the Owner.
Conversely, the Design-Bid-Build process is slower and often excludes the designer from bidding, and the Design-Build process forces the Owner to commit to the Design-Builder at the start of the process.
>Finally, let’s be clear about how ST has not kept on schedule on anything opened in the past 10 years.
It’s unfair to fail to recognize the various circumstances out of ST’s control which delayed each of these projects; primarily COVID-19, and then a 6-month concrete driver strike in the middle of each ST2 project’s concrete-heavy construction stage. While the East Link plinth issues should have been handled differently, and the FWLE soils issue is questionable, those issues weren’t the only reason these projects were delayed.
Construction timelines are mostly a question of cashflow, where more money = more parallel work = faster build. It’s fairly obvious we can’t afford WSLE at its current price tag and it would be surprising if BLE comes anywhere close to its current estimates. I think we’ll end up with both projects progressing to building the Minimum Operable Segments (MOSs): WSLE building the Duwamish Crossing and terminating at Delridge, and BLE building DSTT2 and terminating at Smith Cove.
For DSTT2, I think the closest comparison might be LA’s Regional Connector project, which took about 10.5 years from pre-construction to opening, including COVID delays and discovery of poorly-mapped historic utilities during construction. If BLE gets to 30% design in 2028, 11 years of construction gets us to 2039. I think we’ll get to Smith Cove in 2039, and Downtown Ballard (which better be upzoned to U-District heights) in 2045 or so.
For WSLE, it will be interesting to see what the final construction timeline ends up being. They’re really pushing for the cable-stayed bridge design, which might be a first for the Puget Sound but not first in the PNW. Construction firms should be able to design-build it fairly effectively, if we can afford it. As many have speculated already, finishing WSLE to Alaska Junction will come at the direct expense of BLE’s construction timeline, unless…
An interesting future might be one where the Board kills Subarea Equity to infinitely delay Issaquah-Kirkland (a project which has no visible cheerleaders today and is currently estimated at $3.8B based on the representative alignment) while diverting those funds (mostly) to North King to finish BLE and WSLE in the early 2040’s.
A WSLE stub to Delridge is a ridiculous project. There’s no “there” there for walk-up ridership and never will be, because of the steel mill and Pigeon Point. Do they really believe that the people of West Seattle will stand for riding a bus to Delridge, transferring to one train for a mile and a half to SoDo, and then transferring to another just to ride to downtown?
What a stupendous waste of the public’s money!
I think there’s a good chance that West Seattle, and Ballard especially, never get built at all. Within the next decade or two the world will become even more chaotic and some supply chains will probably break down completely. The costs of these megaprojects will spiral ever more out of control, and honestly, Seattle will be doing well in 2050 if they manage to keep lines 1 and 2 running reliably.
I’m not that concerned about that, because as I’ve said many times, ST has made so many passenger-hostile changes since the vote (long transfers downtown, potentially a Ballard 14th station a long walk from Ballard Ave) that it’s debatable how much better it is than nothing. Construction won’t start for several years, both because ST2 bills are still coming in and the status of planning. Who knows what the economy or federal government or overseas shipping will look like then. The worst outcome I think is if West Seattle starts construction but never finishes, or if we get an AJ-SODO line or Delridge-SODO line but nothing else.
While walking downtown today, I saw that a lane of 6th has been closed off between Westlake and Virginia, containing a small trailer-mounted drilling rig and a lot of metal tubes. On the sidewalk next to the trailer there is a stack of narrow wooden crates, every one stenciled “WSBLE”.
Twenty or thirty years from now, if I am still alive, perhaps I will be able to smile beatifically at the opening of some new bit of Sound Transit infrastructure and tell the young whipper-snappers how I was there when it all began, with a soil survey for a new tunnel. (Or whatever it is they are doing: nobody will remember by then, anyway.)
You don’t have to wait twenty or thirty years to witness the next little bit of Link opening. You only have to wait a few months!
DRLE will open in a few short months, and will be a game changer for residents of DT Redmond.
Then, in late 2025 or early 2026 full ELE will open, with trains cruising pasting traffic on the floating bridge. Another amazing improvement.
And then FWLE will open toward the end of 2026. About time.
By the time we get to 2027 the transit world around here will be completely transformed. No need to wait for WSBLE to witness the future.
“trains cruising pasting traffic on the floating bridge”
Only during congestion. Otherwise trains will be running at 55 mph while cars will be flying by at 65mph (or more if they speed).
And please note, that is 27% over 2019 levels.
Still pretty impressive given how new the extension is.
Something I’m curious about. Is there a reason that the bus bays at the new Lynnwood Link stations are set up so that buses running in both directions use the same bay. I’m not counting buses that terminate at a station.
For example, at 148th, the 365 towards both 185th and Northgate use the same bay. At 185th, the 348 towards both Richmond Beach and Northgate use the same bay. At Mountlake Terrace, the 331 to both Kenmore and Shoreline CC use the same bay.
It seems that this is more likely to cause confusion. Someone sees the route number and just gets on, not realizing they boarded the bus heading the wrong direction for them. Contrast that with Northgate where, for example, the 61 to Greenwood and Lake City stop at different bays.
I think it is a flaw, similar to the ones we’ve been discussing at Ash Way (https://seattletransitblog.com/2024/10/30/midweek-roundup/#comment-944222). That is a long thread but it boils down to two issues:
1) Buses going to the same place are not using the same bus stop.
2) Buses running in both directions are using the same bus stop.
Both are mistakes. You are correct, Northgate does it right (or at least mostly right). Here is the map: https://kingcounty.gov/en/dept/metro/routes-and-service/schedules-and-maps/boarding-maps/northgate-station. The 61 serves different bays depending on which way you are going. Not only that, but they are grouped quite well. If I’m heading northeast (towards Lake City) I use a bus at Bay 1. Lots of people take advantage of that and take the first bus that arrives if they are heading towards 5th & Northgate Way. Same thing applies to going southwest. Before Lynnwood Link, one of the bays had the buses heading to Snohomish County.
Even then there appear to be flaws. The 303 uses the same bay (both directions). It is unlikely folks will get on the wrong bus though, since it only runs peak-direction. A bigger issue is the northbound 322. It should use Bay 1 (since it goes to Lake City). I’m not sure why they did that. There are a couple unusual things about the bus. First of all it uses the express lanes. Then, after using Bay 4 it turns east, serving 100th & 4th (https://maps.app.goo.gl/DYFoYnXLT7uGpder5). It does this without any additional detour (serving that stop and Bay 1 and that bus stop would require an extra loop. They may figure that during peak (when that bus is running) there is enough service headed towards Lake City — might as well have one of the buses provide a little extra coverage. In any event they’ve been doing that for a while — since the bus had a different number (and slightly different routing).
Larry,
yes, it is a flaw.
At Mountlake Terrace, Route 331 trips going in two directions serve the same bay.
At North Shoreline, Route 348 trips going in two directions serve the same bay.
At South Shoreline, Route 365 trips going in two directions serve the same bay.
I have written ST and Metro.
This is a new mistake. It is not done at any other facility that I can think of.
There were also typos in the ST area maps at all three stations.
addendum
It will lead to confusion by intending riders and could delay the departure of coaches.
They do that at the South Kirkland P&R.
Yes, south Kirkland.
In other news, ST has released new ridership data, including Link ridership data for the month of September (post LLE opening).
Link ridership was up an astounding 27% in September. And that is only the first full month of ridership with Lynnwood Link in operation, and only half of that month was with the bus restructures in place. One would expect ridership for October to be higher.
This is all very good news.
Also, ELSL continues its puzzling ridership fluctuations. I’m to exactly sure what is causing that, but given the magnitude of the fluctuations I would guess it is a data acquisition problem.
*One would*, not *Wine would*.
Fixed.
And please note, that is 27% over 2019 levels.
Still pretty impressive given how new the extension is.
Great to hear. Looks like ~8400 daily boardings at the 4 new LLE stations; Northgate ridership is down 50% MoM to ~4100 daily boardings.
On the STX side, no data for the 515 yet.
512 ridership is down 55% to ~1250 daily boardings.
513 ridership is down 65% to 107 daily boardings. I wonder how many of those are at Seaway/Eastmont vs. Ash Way.
@D,
I have it in high authority that the remaining 500 series I-5 buses have been pretty dead, except for when Link has had service disruptions.
When ST fixes their current issues under UW it might be time to cut the 500 series back even more.
@D — I simultaneously wrote the same thing (except I used a lot more words). There is a net gain for Lynnwood Link of about 3,000 riders. Things may go up as the bus changes settle in.
When ST fixes their current issues under UW it might be time to cut the 500 series back even more.
I think the 515 is bound to be eliminated soon. Other than that, I’m not sure what I would cut. The 512 is still an express to Everett. Sure, the 201/202 does the same thing, but it is pretty hard to say we should build Everett Link and then turn around and cut buses from Everett to Lynnwood.
One option would be to do a restructure. For example the 512 could be even more of an express and skip Ash Way. The 201/202 would continue to make the connection from Everett to Ash Way to Lynnwood. That would reduce frequency from Ash Way to Lynnwood though. Given the apparent crowding at the Lynnwood Station Park and Ride that seems like the last thing we want to do. Ash Way is the obvious place for people to park and take a bus to the station — the more buses we run from there to Lynnwood Station the better. Then again the 513 and 532 could still serve Ash Way. This could lead to very good frequency (at least during rush hour) if we ran those opposite the 201/202.
Then there is the 513. It should be bidirectional, since it serves Boeing. If nothing else the service for the 515 should be shifted to the 513. My guess this is the first thing they do. Even that probably won’t happen for months. A bigger restructure is likely much further off.
I hope the 513 gets the bi-directional service it was supposed to. As is, I don’t see many people taking it from Seaway. Looking on Google Maps, it looks like a pretty industrial area, so if anything I’d expect people to be commuting to Seaway in the early morning. There’s no P&R there so your only options to get there are a handful of CT buses and ET buses.
Of the CT buses that connect, there’s Swift Green, the 103, 270, 271, 280, and 907.
I can maybe see some people taking Swift Green / 103 to Seaway if that works out well for them, rather than the longer local trip to Lynnwood. The 270, 271, and 280 all stop at Everett Station too so anyone Lynnwood-bound would’ve probably transferred there instead. Folks coming from Stanwood/Marysville that are headed further south would likely take the 905 instead of the 907 too.
So that just leaves the ET buses that run to Seaway – the 3, 8, 12, and 18. Unless you live closer to Everett station, if you’re headed south you’d likely head to Seaway on these buses. Looking at the latest ridership numbers from ET, December 2023 was their best month last year, and these buses combined for almost 800 boardings a day.
On the topic of bus restructuring:
ST didn’t address this in the original 2024 STX service plan, but IMO some service hours should be invested in adding 535 service on Sundays. It’s a little absurd that there’s Saturday service on that route but not Sunday, especially when Stride will replace it in a couple years.
It’s also interesting to see that the 535 ridership has almost reached 2019 levels – it’s just 13% under as of last month. The 532, on the other hand, is still far away, 64% under.
I found this really detailed ST Service Implementation Plan from 2020. It breaks down ridership by STX route/trip/stop which is really nice.
So in 2019, out of the 306 people riding it northbound daily, only 7 or 8 got off at Seaway. 150 got off at Eastmont, and 28 at Casino/Evergreen. 112 got off at Mountlake Terrace, which is also served by other routes, and the remainder at other points in Seattle.
Too bad ST doesn’t publish these anymore, this was a fun read.
Yeah, the Service Implementation Plans had a lot of good data. Not only did they have the ridership at each stop, but they had directional data as well (for both the buses and trains). So you could tell how many people took the train from Rainier Beach north and how many took it south. I think this was the last one of those reports: https://www.soundtransit.org/sites/default/files/documents/2020-service-implementation-plan.pdf
IMO some service hours should be invested in adding 535 service on Sundays. It’s a little absurd that there’s Saturday service on that route but not Sunday, especially when Stride will replace it in a couple years.
Agreed. It would also be nice if they ran it more often on the weekends. Ridership in general (across the board) is far less weekday-oriented than in the past. The main reason some buses have much better weekday ridership is because they run the buses a lot more often then.
I hope the 513 gets the bi-directional service it was supposed to. As is, I don’t see many people taking it from Seaway. Looking on Google Maps, it looks like a pretty industrial area, so if anything I’d expect people to be commuting to Seaway in the early morning.
The problem is the lack of stops (and the routing). Casino Road has plenty of apartments. It is one of the more densely populated parts of Snohomish County. The bus only makes one stop, next to SR-99. If nothing else I would stop at Casino Road and 5th. The bus is a bit unusual in that it makes a giant clockwise loop from there, thus avoiding having to turn around at SeaWay. If it instead kept going on Casino Road it could make a couple stops along the way (to serve the cluster of apartments there). Everett Transit also serves the street (with the 12) which means that riders in the middle of the day could still get home. According to OnTheMap, about 4,000 workers live in the area and about and about 600 commute to Seattle. The 513 only gets about 100 riders now, and undoubtedly some of those are folks from the Eastmont Park and Ride. It wouldn’t take many riders from those apartments to bump up the ridership on the route.
It is also worth noting that both the Everett Transit 12 and the 513 run every half hour. So if the bus shared a few stops along the way it would benefit riders who are heading to SR-99 or Cascade High School. The 513 needs every rider it can get. Stopping along Casino Road would be a good way to get them. Making a few extra stops is also dirt cheap. In contrast running more often is not.
That being said I would definitely make this a bidirectional peak-oriented route. SeaWay is a transit center, but I just don’t see people taking two buses to Lynnwood (followed by a train). Not given the frequency of the 513. The main value is that it is close to surrounding employment centers (mostly Boeing) and Boeing does run a shuttle. It is also far more likely that someone from Seattle doesn’t own a car (or doesn’t want to use it when commuting) and would be more tolerant of an additional transfer, especially if there wasn’t much waiting.
“some service hours should be invested in adding 535 service on Sundays. It’s a little absurd that there’s Saturday service on that route but not Sunday,”
ST was planning to start Sunday service on the 535 in 2022, along with 15-minute Sundays on the 550, 15-minute frequency on the 522 to go with Northgate Link, and maybe 15-minute service on the 594. All those got swallowed by the driver shortage.
“Then there is the 513. It should be bidirectional, since it serves Boeing.”
@Ross: I agree there should be two-way service. But the reverse-peak routing should probably be along Airport Road to serve more employment sites than Boeing. Perhaps using Ash Way to Mariner; then 128th St SW/Airport Way to Seaway. This would be better than reverse-peak service that operated to Eastmont and one stop on Casino Road. Call it a 514, if you will.
But the reverse-peak routing should probably be [different]
Yeah, good point. I was thinking that too. There is bidirectional demand, but people are going to different locations. The problem though is that employment is fairly spread out. Other than Boeing (which itself is spread out) there are no good clusters or even corridors. I could see the bus stopping at Airport Road & Casino Road though (along with a couple more stops along Casino road close to the apartments). There are few businesses around there plus a Goodwill and a technical school.
It is worth noting that Community Transit runs a peak-only express to Seaway (the 907) and it doesn’t make any additional stops in the area. The 280 also has a version that continues to Seaway (in the morning) and it too just ends there. This suggests that from a peak-standpoint the most efficient thing to do is just go to the Seaway. (Both CT and Everett Transit run all-day service in the area, but it is doesn’t seem focused on employment at all.)
But that doesn’t mean the bus has to be symmetrical (even though it is bidirectional). I would skip everything between Ash Way and Seaway in reverse-peak direction (i. e. heading to Boeing). I could easily see it making a live loop. In the morning it would go express from Lynnwood TC/Ash Way to Seaway then head over to Casino Road, making several stops along the way (including the existing Casino Road & Evergreen Way stop) then continue to Eastmont, Ash Way and Lynnwood TC. So this: https://maps.app.goo.gl/hTdCneU3btbzWU8d6. In the evening it would reverse itself. Making a live loop saves on service hours. This way most riders get a fast trip. At the same time, a rider on Casino Road could take the bus going the wrong direction in the evening, simply because once it is done serving Seaway, it will get to Lynnwood fairly quickly (much quicker than any transit alternative).
It would also make sense for ST to coordinate with Everett Transit, since it runs much of the transit in the area. They may have plenty of demand in certain areas, and it would be good to piggy-back on that. That is why I like the idea of additional service along Casino Road. The 12 is one of the more frequent buses in Everett Transit’s system, and it runs every half hour. Giving folks along there a fifteen minute connection to SR-99 (and a half hour express to Lynnwood) could very easily net a fair number of riders.
The fluctuation per station is odd. For example ridership at Beacon Hill is up quite a bit for September. It was their biggest month since the pandemic. But Columbia City ridership is down for the month. That may be data acquisition (as you mentioned) or just random fluctuations.
Interestingly enough this is the second month in a row where Saturday was the biggest day. This may be a combination of people out and about with the nice weather along with Husky football (three out of the four Saturdays had home games). With a fairly nice October (and the new waterfront) it would not surprise me if this is true in October as well. In contrast, this looks like a typical November (not so pleasant).
This is our first glance at Lynnwood Link ridership and it is quite similar to what we saw last month (when it was first reported and then removed). Basically Lynnwood Station dominates. Lynnwood has 5,000 riders weekdays, 5,500 on Saturday and 6,000 on Sunday. The two Shoreline stations have about 1,000 a piece while Mountlake Terrace has 1,500 on Saturdays and around 1,300 the other days.
As you mentioned, this is with only half the restructure though. That should definitely increase ridership. Like the station at 130th, many of these stations are highly dependent on connecting bus service. Unlike 130th, they also get a lot of riders from the park and rides. Hard to tell exactly how many, but the Lynnwood park and ride is huge (and apparently full most weekdays). Meanwhile, Northgate ridership has dropped in half (to about 4,000 riders a day). I have to assume this is due to both of these factors. More people are taking buses and driving to stations north of Northgate. It is worth noting that last year we saw a bump in ridership in September (presumably as more students took the train to North Seattle College).
Not surprisingly, ridership on the 512 is way down. It lost about 1,500 riders. This suggests that not that many people were riding the 512 from Everett, which is consistent with what previous reports showed. Similarly, the 513 is way down, losing about 2/3 of its riders, or 200 people a day. (The 511 was suspended a while ago.) No numbers for the 515 (they are probably tiny).
By my count within the ST system Lynnwood Link had a net gain of roughly 3,000 riders. Some of those were new riders, and some of them (undoubtedly) came from Community Transit. Unfortunately Community Transit — to the best of my knowledge — does not release data on particular routes or even groups of routes (e. g. how many took the 400-series buses). But there were a fair number of express buses to Seattle and a few were running frequently enough to suggest they were at least moderately full. To be clear, the express buses were a tiny compared to what they before the pandemic. My guess is this accounted for 1,000 riders (at most).
It appears that there are definitely a significant number of new riders, but the vast majority of Lynnwood Link riders are folks who are switching modes (or stations).
That isn’t necessarily a bad thing. Riders may be better off even if you have just as many people riding transit. From a system standpoint there may be big improvements. Community Transit has now shifted much of its service to trips within Snohomish County (having been freed from running buses into Seattle). This may be leading to significant increase in overall transit ridership in the north end. It is hard to measure, as they have simultaneously been improving the system (e. g. by adding more Swift lines).
There has been a partial closure of the transit parking garage at Northgate. I have not heard whether the remaining stalls still fill up.
There’s a technical glitch on the Sounder data page.
Among the longtime stations, I find the surge at SeaTac notable. It’s the second brightest number of boardings (above 9 K on average) and barely lower than Westlake (above 10K in average), right?
Clearly the longer the line, the better for SeaTac ridership. Why should anyone ever drive again between Snohomish County and SeaTac?
I predict that ST will get lobbied for later trains at night and maybe earlier ones in the mornings.
Why should anyone ever drive again between Snohomish County and SeaTac?
Because most of the time it is a lot faster. It isn’t clear how much (if any) of the ridership increase is due to the extension. Since April (long before Lynnwood Link) ridership has been setting monthly records. It really took off in May and June. Since then it has shown a slower irregular increase (down in August, up in September). It is quite possible this is all due to an increase in employment, although I don’t have the numbers. But SeaTac is setting records for flights and there have been significant additions (with more retail than ever).
It also might be due to people switching from Uber/Lyft. But there is no reason to assume that the extension made a big difference. Most of the people who work at SeaTac live closer to the airport. Federal Way Link will probably have a lot more riders heading to SeaTac (especially if the bus to SeaTac is truncated there). They will still need to keep late night service though (with the bus).
My wife and I took the train from MLT to SeaTac and back last week for a short trip. For the record, the total time from home to terminal was about 85 minutes. That compares to 35 minutes by car (off peak of course). We’re retired now so our time is not that important and there’s no employer paying for the parking, otherwise driving wins easily. There was also the issues of having to arrange drives to and from link, and also the lack of space for luggage on the train. Not an issue for our weekend trip, but next time we travel we will have luggage for multiple weeks and also be traveling at night. There is no way my wife would feel safe on link at night, and neither would I, since the luggage is a dead give-away to the criminals that we are not armed.
Where I see Link being better between Snohomish and SeaTac is when it involves dropping off or picking up fliers. I-5 can be quite congested several hours a day in North Seattle — and North Seattle I-5 rehab has started. There are often delays on the approach road to the terminals as well; I’ve seen it take 15-20 minutes of crawling to get from 518 to the terminals.
Link seems still reliable although I keep reading about service problems lately.
Would someone want their spouse to drive all the way to SeaTac and back (1.5-2 hours), or ask their spouse to merely stop at the nearest Link station (like 10-20 minutes)?
Or maybe someone hops a bus to the station but gets a ride when returning home. The Swift frequencies are pretty good and so there’s not much waiting for a connecting Swift bus.
Certainly safety may be a concern. I’ve ridden back from a flight landing at SeaTac many times and seen some things like homeless people sleeping on the train. However, what could be interpreted as a “scary rider” at first glance is often just driven by my own silly prejudices about people who don’t look like me. I chastise myself on this feeling once I observe that “scary” rider’s dress and demeanor. I think that fears ease if a few other fliers are in the same train car too.
If someone really feels unsafe, I recommend boarding on the first car so the driver can better hear any disturbance. I am a bit surprised how empty the first train cars often are in the evenings.
” I’ve ridden back from a flight landing at SeaTac many times and seen some things like homeless people sleeping on the train. However, what could be interpreted as a “scary rider” at first glance is often just driven by my own silly prejudices about people who don’t look like me. ”
Yeah, exactly. Took the train from Minneapolis airport to downtown at 1am last week. Everything was fine until all the homeless left, and it was just me and what I guess was a tweaker right in front of me, who just turned and stared at me for 10 minutes saying nothing. “Come back, friendly homeless folks!”
Sounder data is still broken. Ridership on the 510 is down, and my guess is this is people switching to the train. The drop isn’t that big (only about 50 riders compared to last year or last month) but North Sounder doesn’t get that many riders.
When the full 2 Line opens, how will ST handle the following situation? A Lynnwood-bound 1 Line train is a couple of minutes away from the downtown tunnel. A Lynnwood-bound 2 Line train is also a couple of minutes away from the downtown tunnel. Because one of the trains is a few minutes late, one train will arrive at the tunnel about one minute ahead of the other train. In that situation, will ST just let them both run all the way to Lynnwood a minute apart?
We’d have to speculate on their operating plan, but I suspect the last train to arrive at Stadium Junction will have to wait, then follow behind the first at whatever minimum spacing is allowed by the tunnel signal system.
It really depends on the train control and the dispatching. Each train’s position is known. I can’t claim to know how advanced the specific train control logic is, but these systems can flag a late train or an early train. They’ll probably send the late train through first unless it running really, really late.
Already, the train control holds a train in a station on MLK with doors open until the signals allow (dependent on the train getting green signals). So I imagine that the prior station doors (Stadium and Judkins Park) won’t close until the train order if established automatically or by a dispatcher. Keep in mind that a rail switch has to be activated and the system won’t send a train through until the prior train has left the segment of tracks controlled by the signal for the switch.
Keep in mind that even today, ST switches tracks for the end stations. Like each arriving South Bellevue Station train alternates between the west and east platform. So it is already routine to control the switches in synch with the train control.
Yeah, this happens now. They try to keep the trains synchronized, but there is some wiggle room. Four minute headways (through the shared section) is huge, really. It is unlikely that the train will be delayed often because of bunching.
To get back to your scenario Sam, it is likely that one of the trains would have to wait thirty seconds. I think the minimum headway on the trains in the tunnel is 90 seconds. For safety reasons they don’t run them any closer together. But this would be rare, since it would mean that a train is three minutes late (and the other train is right on time). The lines are long, but mostly grade separated. Furthermore, they build in a lot of float for this very reason. If a train is running faster than usual — or even at normal speed — it will run fairly slowly through Rainier Valley and SoDo. If a train is running behind it will run quicker than normal through there. The type of delay you are talking about will be quite rare, if not unheard of. Mechanical problems are far more likely.
We know the length of signal zones in the tunnel because they’re marked on the tunnel walls as fire zones. There’s one or two between Westlake and Capitol Hill. It takes a train 3 minutes to travel that distance. So if a train from SODO is just behind a train from Judkins Park, it might wait 1-2 more minutes for the first train to clear the zone it wants to enter. If the zone includes a station stop, it seems logical to me that that would be included in the waiting time. So if the train takes 20 seconds to load/unload and 20 to accelerate/decelerate, you can take 1 minute and subtract 40 seconds for the additional time the second train has to wait (since it would have to wait for loading/unloading anyway).
Ross, I’m not sure if I’m understanding you. You aren’t saying it is quite rare for a Link train to be 2 or 3 minutes behind schedule, are you? Because, I believe it happens multiple times on a daily basis.
As others noted, it’ll probably be normal headway management. Tell the train too close to the next train to slowdown or wait longer and tell the next train to speed up if possible.
The more complicated problem happens if say one train from eastside is so delayed that it ends up being behind the train from seatac it was supposed to be ahead of.
It’s difficult because of the distance from Judkins Park to IDS. Typical block signals lack any sort of predictive ability, so the best it would be able to do is make the train wait just south of the junction for the first train to clear.
Ideally, IDS would have four platforms and 3 tracks, and if one train were delayed, the closely following train would go into the center track, open its doors to the platform on each side, and wait it’s needed time there before taking up behind the leading train.
Unfortunately, rebuilding IDS to work that way would be an undertaking,
The best solution is likely going to be to have an actual signal box and observer keeping track of the trains specifically for this junction. CTA does this at its busiest junctions.
We should rebuild Intl Dist long-term anyway to make it center-platform, so could that and triple-tracking be done at the same time? That would actually bring the station closer to its original role, when there were several bus layover lanes.
I don’t think it’s quite wide enough to allow for a full center platform and third track, which is why they have that ridiculous stub platform there now. They’d need space to the north to join the center track to the main line on either side.
Of course, while expensive, all of that would be cheaper than DSTT2…
We should rebuild Intl Dist long-term anyway to make it center-platform
We should have done that with East Link. That being said it isn’t clear how many people will reverse direction there. If and when West Seattle Link runs through there it would have more (West Seattle to the East Side).
I don’t think we need a third track. A train that just sits there is irritating, really. You aren’t sure what is going on. You run to catch it, only to sit there. Is it broken? When will it go again? Then you notice a different train on a different track going to the same place. WTF? In contrast a train that goes slow is just a train. Folks may wonder why it is going slower than usual or has such a long dwell time at each stop, but most people don’t even notice.
I can only assume that this system is modern enough that they know where the trains are. So assume that the train from Federal Way is much slower than usual. It is supposed to reach Mount Baker at 9:00 am. Instead it is 9:04 am. At that point the train should try and reduce the dwell times to compensate, but there are only three more stations until CID (Beacon Hill, SoDo and Stadium). We can assume that train will be considerably late. It takes about ten minutes to get from Mount Baker to CID, so instead of getting there at 9:10 it looks like it will get there at 9:14. Meanwhile the East Link train is supposed to get to CID at 9:14 as well. This looks bad. Except remember, all of this is happening at 9:00. The East Link train is 14 minutes away from CID. It is quite likely the East Link train hasn’t reached South Bellevue yet, let alone Mercer Island. It begins slowing down. It arrives at Mercer Island later than usual. It also sits there longer than usual. Then it goes over the lake a little slower than usual. If the train from Federal Way hasn’t made up any ground — or worse yet, is even slower — it spends a huge amount of time at Judkins Park (folks in Rainier Valley are used to really long dwell times).
There could be a delay closer to CID. SoDo is not grade separated. But a train going from being on time to being even three minutes late between Beacon Hill and CID would be quite unusual. Even if that does happen, the other train can compensate by waiting a much longer time at Judkins Park. Judkins Park to CID is not that far. Again, this seems like it would be quite rare.
Trains being late are common. Trains suddenly being really late (over a short stretch) are not.
Ross: dispatchers are only given an indication to which block the train is in. You’d need something that not only gives a much more precise train location, but also a predictive arrival time. This is something that isn’t available.
This is why Chicago uses staffed signal boxes at such junctions.
TriMet’s solution at Gateway is a 3 track station. The idea isn’t to have a train wait for long periods, but to allow simultaneous arrival when needed. TriMet also uses the center track for reverse moves, just like SoundTransit plans to use the existing IDS center track. Whatever train arrives first, that train leaves first (unless it’s a reverse move train).
Other options would be stopping the train south of the junction at no platform (TriMet does this at the junction of the green and blue line south of Gateway) or have a separate station platform (TriMet does this at Rose Quarter with the yellow line having a separate station before joining the red/green/blue lines).
I supppse another alternative would be to extend the platforms south, so that two trains could enter the station at once. It means the seconds train has to stop twice, but at least when it stops, it does so at a platform, rather than an unexplained stop in the middle of nowhere.
I don’t think SoundTransit is wiling to build stations on slopes, so I don’t imagine a Line 2 Stadium station on the elevated section south of International District is possible. That eliminates the Rose Quarter yellow line solution.
Either way, until ST has a solution to the reverse move need, they’re not going to put a center platform at ID.
“We should rebuild Intl Dist long-term anyway to make it center-platform”
“We should have done that with East Link.”
ST said in 2016 it would consider it, but then it didn’t.
“I can only assume that this system is modern enough that they know where the trains are.”
That’s just the problem; it doesn’t. It does some of the time but now always. That’s why the next-arrival displays are sometimes accurate, sometimes not, and sometimes so wrong ST turns them off.
Real time arrival systems are almost always based on GPS data, but GPS alone isn’t reliable enough to use for train signalization and control.
That will happen, and it does mean that when they both get to Lynnwood, the delay of the later train will have to be addressed.
If ST is wise, they’ll have a couple of replacement operators waiting at Northgate on an ongoing basis, either to take a short-turn in-fill train south if crowding starts to clog operations or, in this instance, to board the rear car northbound in order to ready the control cab for an rapid reversal. That would allow the train to depart southbound under the control of the double-cab operator, who would then hot seat with the assigned operator at MLT or one of the Shoreline stops and exit the train back at Northgate.
By forfeiting some layover time, the train will have gotten back into its proper slot.
This was supposed to be a reply to Wesley’s comment directly above.
You make a great point about what would happen at Lynnwood, Tom! I suspect that the dispatchers will prefer to keep the train sequence intact if the delay is just a few minutes. That’s just to keep the Lynnwood sequencing even.
If a train gets really, really late, then some bigger solutions have to be considered.
The first would be to simply change a 1 train to a 2 train on the way back from Lynnwood. At the end of the day, the train could switch back to the original line.
They might be able to reverse the out of sequence train quickly to jump back ahead, like you say.
If delays are major in one line, then the dispatchers have to look to something more major. There are options like:
1. The late trains become expresses and riders are told to use the other line for upcoming stations.
2. Late line trains stop at Northgate (or even ID-C) until the sequence can resume. Riders would be told that the late line isn’t going all the way to Lynnwood for awhile and that they should use the other line.
Interestingly, you have described a weakness of branching only at one end but not the other. If both ends of the line are branched, the disruption is less problematic. It’s an argument to have Paine Field and Everett as separate branches — although the Snohomish interests wouldn’t like it because of a forced transfer to get between the two branches.
The Sounder ridership tab was still not working today. However, I found a memo on Sounder North ridership in September here:
https://www.soundtransit.org/st_sharepoint/download/sites/PRDA/ActiveDocuments/Memo%20-%20Proposed%202025%20Budget%20Questions%2011-01-2024.pdf
The average seems to be 548 two-way weekday ridership in September.
A comparison reference showing 1300 daily Sounder North riders in 2019 with a drop to 300 during the much more limited schedule last spring is here:
https://www.theurbanist.org/2024/07/20/doubling-sounder-north-service/#:~:text=The%20N%20Line%20has%20long,the%20coming%20months%20and%20beyond.
What does this mean? Certainly the added trains made the service more attractive. Given the overall Sounder South drop in ridership too, it doesn’t appear the Lynnwood Link opening diverted many of Sounder North’s faithful riders. However it’s hard to be more definitive until ST fixes the Sounder ridership data tab online.
Yeah, Sounder North never had a ton of riders. It is not surprising that ridership crept upward given the added runs. An increase of somewhere between one and two hundred sounds about right. Here is the transit competition and the other changes for each station:
Everett: 510 is still running. Ridership on the 510 is down about 50 riders (last year it went up in September). Riders can also take the 512 and transfer but it isn’t clear if people are choosing that over the 510 or Sounder. Ridership for for the 512 is way down (having lost the key Lynnwood to Northgate section to Link).
Mukilteo: Riders lost their one-seat express (the 417). Some of those riders may have switched to Sounder instead of taking the 117/Link. The 117/Link combination is a huge improvement in midday service to Mukilteo/Whidbey Island though. It is much faster, and timed with the ferry. Since the area around the ferry dock has only a handful of people, this makes a big difference. It is quite likely that some riders take Sounder one direction and ride Link/bus the other way. In the past they would have just driven.
Edmonds: They also lost their one-seat express (the 416). There are a couple buses from Mountlake Terrace, but neither is especially fast or convenient. One bus runs every half hour, one every 50 minutes (timed to the ferry). Unlike Mukilteo, the area around the train station is quite developed. This means the bus timed to the ferry has less of an impact. For most of the riders, getting back to Edmonds in the middle of the day remains awkward. My guess is most of the increase in Sounder ridership here is due to the train running more often and the 416 being eliminated.
I think it’s worth mentioning that the impact of Lynnwood Link on Sounder North ridership is still too early to assess. Still, while demand could shift it seems like much of it would have been immediate. Granted the CT restructuring didn’t kick in until mid-September, but that likely isn’t pulling away many more Sounder North riders than the new train would.
I could see that if the Lynnwood garage keeps filling up early that Sounder North will have demand for the service. Even adding the 2 Line service won’t attract many more riders if the garages are already full.
I was thinking that more Sounder North riders would shift. But now it appears they won’t. So unless the situation unexpectedly changes, I don’t see Sounder North going away any time soon.
I don’t see Sounder North going away any time soon.
It is worth noting two things:
1) Sound Transit has a shortage of Link train cars and is very much concerned about crowding in the north end. This will not be the case once East Link starts.
2) North Sounder is ridiculously expensive per rider. The cost per rider probably just went up substantially.
3) Then again, I think the contract lasts a long time. So unless ST wants to renegotiate with BNSF, North Sounder will stick around.
I doubt Link has much of an effect either way, with the exception of riders coming from Whidbey Island. But in that case I think it will increase ridership (since riders know they have a pretty good way to get home in the middle of the day). In general the dynamic hasn’t changed that much. Consider each station again:
1) Everett. The folks who took the 512 to Northgate just take it to Lynnwood instead. The people who are headed downtown continue to take the 510 or Sounder. Hard to see why you would switch to the 512 from either Sounder or the 510.
2) Mukilteo. The express bus was a good option and now it is gone. Some of these riders switch to Sounder and some to bus/Link. While the local bus is faster than the old local bus the bus/Link combination still isn’t fast compared to Sounder. I can see why someone would switch to using Sounder (since the express bus is gone and Sounder runs more often) but I don’t see why someone would do the opposite.
3) Edmonds. Same dynamic. The bus/Link combination is quite slow compared to Sounder.
That doesn’t mean that a ton of people are switching to Sounder. It has its weaknesses. But the trade-offs really haven’t changed that much. Sounder runs more often and some express buses are gone. Link doesn’t have much to do with it.