Swift Blue Improvements

39 Replies to “Swift Blue Improvements”

1. I think that a bit more could be said about how Swift Blue has recently changed from a limited stop route inside Snohomish County to also being a Link feeder route. It’s now doing double duty!

   The Urbanist reported a 7% jump in Blue Line ridership as soon as it extended to Link at Shoreline North:

   https://www.theurbanist.org/2024/10/07/lynnwood-link-brings-huge-ridership-boost-to-community-transits-swift-network/

   Of course, these are just data from the first two weeks!

   It will be interesting to monitor what happens over the next year. The corridor has many new options — from lots more park and ride garages on Link to restoration of two Sounder North weekday round trips to new Link connections for both Orange and Blue lines. That compensates for the loss of Downtown Seattle express routes.

   For later in 2025, the addition in 2 Line trains from Lynnwood could change things more as Link becomes so frequent that it will feel like on-demand service.

   There are two markets that are in play. The first is how existing transit riders change their route choices. The second is how new riders may shift to riding transit. Each of these will play a factor on how the Blue Line performs.

   Finally, I think lane closures on I-5 in Seattle upcoming in 2025 and 2026 will have some ripple effects on both SR 99 and Swift Blue. Those bear noting too.

   https://wsdot.wa.gov/construction-planning/search-projects/i-5-yesler-way-northgate-vic-pavement-deck-joints-and-drainage

   1. The Snohomish 99 corridor hasn’t had good access to most of Seattle until now. It misses downtown Lynnwood and Mountlake Terrace where connections to Link or the previous express buses stop. I attend events at Edmonds Community College and had this issue the other direction. Before Link and Swift Orange, you could take the 512 to Lynnwood + an infrequent bus to 200th & 99, or you could take RapidRide E + Swift Blue which were more frequent but took over an hour. From other points on 99 E+Blue was the only feasible option.

      Now suddenly the Blue goes limited-stop to Shoreline North station, and from there you can take Link quickly to downtown, the U-District, Northgate, Roosevelt, Capitol Hill, south Seattle, or the airport. That’s a major increase in where you can get to in 30-60 minutes. It adds a lot of kinds of retail and activities that aren’t available in Snohomish County, especially on a 1-2 seat local CT ride from the 99 corridor. Want to go from 99 to the U-District? Think Blue+E+45 or Blue+E+44, now Blue+Link.

   2. @Al S,

      The Swift Blue extension via 200th and Meridian is awesome. An immediate 7% jump on ridership is nothing to sneeze at.

      And, as you state, that 7% jump in ridership was after only 2 weeks. But it was also before some signal priority improvements were completed at Meridian and 185th.

      Those improvements should be complete now, making the route via 200th and Meridian even better.

      But the Swift Blue extension is a huge improvement. Good for CT and SnoCo.

   3. As someone who was once chastised from afar for using the term “feeder” to describe bus routes that connect to Link, I have to encourage the use of “connector” in describing the Swift lines. In the Blue’s case, it is clear that a minority of riders are coming from or going to North Shoreline Station.

      When I have ridden the Orange, there is noticeable churn at Lynnwood Station, but also plenty of riders continuing on, both going toward and away from Edmonds College.

      CT Expresses 901-904 may more aptly be called “feeder” routes, since they are peak-direction only. But are they really getting much ridership compared to their CT commuter route predecessors? I’m guessing the new garage majals at each station are doing most of the feeding.

      1. At a recent CT board meeting, they mentioned ridership on commuter routes is down 75% from pre-LLE, but there’s much fewer routes than there were before, so it’s hard to make a direct comparison.

         Overall, ridership is up so presumably some folks have shifted to local buses instead and new demand has been generated from improved / new service.

      2. I agree Brent. It is worth noting that the Swift Green added almost as many riders as Swift Blue, and it doesn’t connect to Link. I’m not sure why it saw such a big increase. It is running more often (every 10 minutes midday instead of every 12) but that seems wouldn’t seem to explain all of it. The other buses are running more often, but I don’t see anything that would easily explain the increase. Maybe all of the publicity helped — people took another look at Swift and realized it had good frequency overall.

         In any event, the Swift routes that do connect to Link have seen an increase, but it is still relatively small compared to the overall ridership. The Orange Line (as expected) saw a big increase in ridership. Some of those are riders taking advantage of something never offered before (e. g. a good connection between Edmonds College and Seattle) but others had to switch from the express buses (the 413/415 were one of the few busy CT express buses after the pandemic). But even with the big increase, it is only 36% more. That means the vast majority of riders are *not* transferring to Link. Same goes for Swift Blue, which saw an increase of only 7%.

         I don’t want to belittle in the increase. This is still very good. It is also quite likely that those riders have saved a huge amount of time. But at the same time, it also suggests that the vast majority of Swift riders are using the bus for travel within Snohomish County (as they always have).

   4. It shows the latent ridership that existed in the 99 corridor that was waiting for compelling service. The same thing happened with Lynnwood Link access to north Seattle and central Seattle, the creation of a N 40th Street route (now 31/32), RapidRide H, the creation of the 8 and 48, etc. The jump may not have been as sudden in the Seattle bus cases but it was similar.

      This is why I think we need an optimum bus-transit level higher than current, because then the hard-to-measure latent ridership would become actual ridership.

      1. I think the ridership increase in Seattle was just as sudden, they just didn’t track it (or publicize it) the way they have here. It took a while to get the numbers, but they did turn out quite good (https://seattletransitblog.com/2017/02/21/the-ulink-restructure-was-a-bold-gamble-it-has-mostly-paid-off/). To quote the post, average weekday ridership on restructured routes + Link growing by 13.3%, from 149k in 2015 to 168k in 2016. This occurred at a time when most agencies were seeing a reduction in ridership.

         The big takeaway is that frequency matters — a lot. Northeast Seattle saw an increase in ridership despite the awkward transfer at the UW (when it was the terminus). For a lot of people trying to get downtown it was much slower. But for a lot of people trying to get to the UW it was much better (the buses were much more frequent). Swift is dominating ridership in Snohomish County because it is only set of buses that are really frequent (running every ten minutes in the middle of the day). That is much better than the average Community Transit bus. Hell, that is as good as Link. I think only the RapidRide E is more frequent.

         Meanwhile, in King County we continue to suffer from transit austerity and inefficient routing. This is why ridership is a shadow of what it was just a few years ago. Frequency matters — a lot.

      2. There are now 34 CT-branded routes. 8 are peak-direction-only, 13 are roughly hourly mid-day on weekdays. 1 is roughly every 45 minutes, timed to the Edmonds-Kingston Ferry schedule. 9 are roughly half-hourly mid-day on weekdays. And then there are the 3 SWIFT routes coming every 10 minutes on weekdays. Several routes are planned to become more frequent by the end of 2026.

         In no way is King County Metro in austerity mode compared to this.

         Ridership is still down (but growing back quickly, year over year) primarily because of COVID. There are people who went and bought cars, and don’t plan to return to public transit. Getting back to Before-Times frequency won’t bring them all back.

      3. “Getting back to Before-Times frequency won’t bring them all back.”

         But the population is increasing, dense apartments near transit hubs are being built, and other new riders will come, even if those people don’t.

         Swift Orange creates an east-west H-grid connecting Swift Blue and Green and Lynnwood station. That creates more trip combinations.

      4. In no way is King County Metro in austerity mode compared to [Snohomish County]

         I never said it was. Metro is in austerity mode compared to Metro of a few years ago. The opposite is true of Community Transit. I would guess that CT has more service within Snohomish County than it has had in a really time, if not ever (because it isn’t spending a lot of time running expresses to Seattle).

         Bus comparing the two agencies is like comparing a watermelon to a grape. Before the pandemic (and the austerity measures that followed) Metro had about 400,000 riders a day. I don’t think Community Transit has ever had 40,000.

         But the same rich fertilizer that works on a watermelon works on a grape. In both the agencies we’ve seen increases in ridership due to an increase in service. We’ve seen decreases in ridership because of decreases in service. This is not in the least bit surprising.

   5. All three Swift buses saw an increase in ridership, including the Green, which doesn’t even connect to Link. This suggests that the restructure was as important as Link itself. Of course Lynnwood Link enabled the restructure. But it does suggest that if they had somehow improved the network anyway (e. g. with an increase in service spending) they would have seen a lot of the same increase in ridership.

      Frequency matters and right now a lot of the buses in Snohomish County are a lot more frequent. But the connections are a lot better as well. Community Transit only ran a handful of buses to Northgate and they only ran them during peak. Unless you were on the Aurora corridor this meant a three-seat ride from most of Seattle destinations to most of the Lynnwood destinations. Now you can get from places like the UW to places like Edmonds Community College or Swedish Edmonds with a single transfer. Again you could have done that sort of thing before, but it would have cost more money (in terms of service). The buses would have also suffered from an awkward transfer at Northgate (just as the 512 suffers from the awkward service to Ash Way).

      It is tough to gather detailed information from Community Transit as they tend to avoid releasing all but the most promising data. It is great to see the numbers for the Swift lines, but what about the rest of the routes? It remains a big mystery.

      1. Consider that Metro had to move drivers to covering a 15-minute extension of the 1 Line, while CT was relieved of sending drivers downtown on 12 routes and to Northgate on 7 routes. I’m still trying to wrap my head around how CT managed to roll out the Orange Line in the spring, and fell so short of its original plan for the fall.

         The Blue Line extension was the only major extension of CT service, while everything else was truncations and path swaps.

      2. “all but the most promising data”

         This statement is unfair and needlessly pejorative. CT makes quarterly system performance reports to their board that contain the good, the bad and the ugly. You have to dig into their online board meeting material to find the info. It’s not easy, but it’s there. At worst they’re guilty of still being a 20th century digital practitioner.

         I’m told by a friend who works there they have shifted their data warehouse platform to Power Bi, and are working on taking an external facing portal live next year.

      3. CT makes quarterly system performance reports to their board that contain the good, the bad and the ugly.

         Where is it then? For example, how many people rode the 201 or 202 last year? I would really appreciate an answer to that because in the past I’ve found answers to questions like that to be extremely challenging — by far the most challenging of any transit agency in the area.

         For example here is where Pierce Transit stores their documents: https://www.piercetransit.org/documents. The route data is in the Transit Development Plans. Here is the one from before the pandemic: https://www.piercetransit.org/file_viewer.php?id=4085. It is a big document, but skip ahead to Appendix B. Right there — clear as day — is the ridership data of every single route. Not just the ridership, but ridership per service hour. Thus you can see that while the 41 does not have a ton of riders, it performed very well in terms of ridership per service hour (and thus might be a good candidate for improved service if ridership is your goal).

         If there is an equivalent in Snohomish County let me first apologize. But in the past I’ve never seen it. Here is the latest Community Transit Transit Development Plan: https://www.communitytransit.org/docs/default-source/pdfs/programs/adopted-2024-2029-transit-development-plan.pdf?sfvrsn=e05472b8_3. I can only find old ones on the Wayback Machine (and only because I know the URL). You can’t find route data in there. There is ridership information, but only in the aggregate. They do break it down between “Fixed Route”, DART and vanpool but that is it. At least in the past they would group by things like “Commuter” and “Local” — now they don’t even seem to do that. Either way we have no idea how many people are riding each route, and never have, unless they told us in a press release.

         Again, my apologies if this is out there and I just can’t find it. I would really appreciate it if you know where it is.

      4. I’m hopeful that they’ll share a public dashboard later, but agreed that they haven’t been that open in the past aside from broad categories.

         In the meantime, the most recent numbers I could find for ridership and service hours broken down by route is 2019 data from the Journey 2050 appendices on page 24/25: https://www.communitytransit.org/docs/default-source/pdfs/journey-2050/journey-2050-appendices-compiled.pdf?sfvrsn=3b6dbe24_1#page=24

         If anyone has anything more recent, I’d love to see it.

2. The Blue line isn’t the only Swift route that has a dearth of BAT lanes.

   The Green Line has them on Airport Rd from SR526 to I-5. For the rest of the route it travels in normal traffic.

   The Orange Line has frequent service but no dedicated lanes. It was intended as a feeder to the Lynnwood Link station however from where I live it’s quicker to take the ST512 from the South Everett P&R.

   1. The Orange line really needs BAT lanes, especially on busier sections like 164th St SW and around Alderwood mall. Especially during morning & evening rush hour, it takes forever for it to get from North Lynnwood over to Mill Creek.

      Apparently there is a plan to improve it in the Ash way area when the light rail station is built, with Orange line buses skipping the interchange entirely. But that won’t be a reality until well into the 2030s.

   2. Don’t you wish STX 512 didn’t have to do the loop-de-loop stop at Ash Way giant surface parking lot?

      Does the Blue still have to loop around at Aurora Village, or is that stop on-street in both directions now?

      1. My feelings on bus loops. If I don’t use or transfer in the loop, I like to complain about it being a time-wasting detour. But, if I do use the loop, by transferring to another bus, for example, then I remain silent, and don’t complain about it.

      2. The problem with the Ash Way loop isn’t the parking lot, it’s that there’s no direct connection to I5 on the North side, so buses coming from the North or, worse, going to the North get stuck in traffic on Ash Way / 164th.

         When I talked to ST folks recently, they said there weren’t any plans to complete the I5 direct access ramps so it sounds like 512/513 riders will be stuck with that unfortunately.

      3. > When I talked to ST folks recently, they said there weren’t any plans to complete the I5 direct access ramps

         ST & SnoCo need motivation for safe passage of peds / cyclists across I-5 between Martha Lake and Ash Way, i.e. something other than 164th.

         Leveraging the HOV ramp/island to connect to the IUT / Utility RoW seems feasible. Would also be a good time to finish the north side ramps.

         While they’re at it, a non-164th ped/bike route connecting Mill Creek and Ash Way would be good too…

      4. Andy, SnoCo intends to use the Ash Way bus ramps to extend a pedestrian/bike/bus path across I5 to the east, which would likely be the mechanism in which Swift Orange buses bypass the 164/I5 intersection as Chloe mentioned above.

         I found this SnoCo project list that shows the project, “I-5 (164th St SW Texas T)”, it does note something about building the north side ramps so perhaps it might happen after all.

         https://www.economicalliancesc.org/media/userfiles/subsite_201/files/resource-library/SnoCoRegionalPriorityProjectsRev12-8-23.pdf

      5. Thanks for the Link “D” (and bringing up the subject). I’ve had my eyes on that for years. Here is a quick rundown on the situation:

         Here is what it looks like from the air: https://maps.app.goo.gl/2PPnHd2LhJTwkvY28. Notice that there are no ramps from the the north end of I-5 to the park and ride area. Thus a bus from Everett has to leave the HOV lanes, move over into the exit lane, exit at 164th, take a right on Ash Way, then take another right to get to the transit center. There is no good alternative given the current roadway.

         However, if you altered the freeway just a little bit, you could save quite a bit of time for the bus. Zoom in on an aerial view of the ramps that connect the south end of I-5 to the transit center: https://maps.app.goo.gl/2PPnHd2LhJTwkvY28. Notice that there is a little stub to the north, leading to the freeway median. Unlike much of I-5 this median is wide for quite a ways (well over a mile). Thus building ramps from the north would not be very expensive.

         This explains why the project — which includes more than just the ramps — is expected to cost 65 million. This is a bargain for what it would provide. The 512 would run completely in HOV lanes and ramps from Everett to Lynnwood while also serving South Bellevue and Ash Way Park & Ride.

      6. The north continuation of the express lanes on I-5 decreases in value every day that Everett Link gets closer. But the safe foot path across I-5 … priceless!

         I find it a little disappointing that the new CT 905 skips South Everett. But there may be reasons.

         How much more frequent would the 201/202 have to be to consider the 512 stopping at Ash Way to be redundant?

      7. I found this application that details the project further with some maps and diagrams; it looks like the plan is to complete design by 2031 with construction in 2035: https://systemaccessfund2.showcase.infocommunity.org/wp-content/uploads/2019/07/SnoCounty1.pdf

         I agree with Brent, by that point the north-side ramps won’t be providing much value. Maybe the timeline is to ensure that there’s no conflicts with Everett Link, but it’s disappointing that improvements will have to wait that long.

         Re: CT 905, it does look like it stops at South Everett: https://www.communitytransit.org/docs/default-source/mappdfs/systemmappdfs/mapsystem.pdf

      8. > The north continuation of the express lanes on I-5 decreases in value every day that Everett Link gets closer.

         > I agree with Brent, by that point the north-side ramps won’t be providing much value. Maybe the timeline is to ensure that there’s no conflicts with Everett Link, but it’s disappointing that improvements will have to wait that long.

         Actually, given that link will detour to paine field and probably stops at more stations, an express i5 bus could still have quite some time savings if traveling from everett (and actually had tolled lanes)

         The current expected time from everett to westlake is ~60 minutes, while the same on 510 takes around 47 minutes in traffic. but if it was actually tolled a bus could probably do it in 40 minutes.

      9. > Actually, given that link will detour to paine field and probably stops at more stations, an express i5 bus could still have quite some time savings if traveling from everett (and actually had tolled lanes)

         > The current expected time from everett to westlake is ~60 minutes, while the same on 510 takes around 47 minutes in traffic. but if it was actually tolled a bus could probably do it in 40 minutes.

         I’m not convinced that an express I5 bus will exist by the time this gets built, considering that ST was originally planning on eliminating the 510 with the LLE opening. I’d bet that they follow through with that once the 2 Line connects across I90.

      10. > I’m not convinced that an express I5 bus will exist by the time this gets built, considering that ST was originally planning on eliminating the 510 with the LLE opening. I’d bet that they follow through with that once the 2 Line connects across I90.

          Yes, currently ST plans to eliminate all the express buses on i5 once link is built. I agree with that approach if there is no toll lanes given the inconsistency of the existing transit travel times.

          However, if there are actually toll lanes., one could hypothetically drive from downtown Seattle to Everett in 30 minutes, I could see an express bus with much more limited stops actually working somewhat well.

3. This is of course good, but systems like Swift should drive in the middle of the street and not on the outer lanes. This will really speed up the movement of buses + it is much safer than doing bus lines along the edges.

   1. I agree! In places like San Francisco the busses have a physically separated lane down the middle of the road, very similar to how the on-grade light rail travels through south Seattle. It’s much safer for everyone and makes less of a disturbance to traffic and businesses. It also allows the lights to be timed better to allow free-flowing transit apart from regular car traffic.

      1. If there are economies of scale, fleet purchases can become cheaper. I’m afraid CT will follow Seattle’s path of a stand-alone bespoke specialty design, that isn’t even battery electric.

   2. Center-running buses are being considered for the future Gold Line between Everett Station and Arlington.

      But we know the drawbacks of specialty livery being required for a route. CT should take a close look at Seattle’s orphaned streetcars to see how inadequate fleet, lack of light timing, and lack of attention to detail can ruin a 9-figure project.

      1. Center running transit is able to use right side platforms with the right street design. See MAX east Burmside:

         https://maps.app.goo.gl/y5nBzaw2JHSJetJJ6?g_st=ic

   3. Fyi for the gold line brt here’s the open house if people want to see, give comment

      https://engage.communitytransit.org/swiftgold

4. I think most commuter’s travel a long distance and having more non stop travel benefits commuters more. While fewer stops may increase “Last Mile” distances, these blue lines accompanied by E-bike stations and other personal endpoint transit may be worth paying a little bit more and worth partnering with employers as an employee benefit.
   I have commuted 9 years by bike (6 years pedal, 3 years E-bike) and I found that on some routes, I could get just about anywhere in 45 minutes under good conditions, which is not bad in the quality of life metric.
   Under the worst conditions, my daily total commute was 5 hours a day due to schedule conflicts.
   I liked counting the time I left my home till the time I returned to my doorstep and divided my daily net pay as an honest metric of my daily hourly earnings. By measuring my time this way, it kept me honest about how much personal time I have and really drove me to look at my daily earning potential.
   If I’m commuting 5 hours a day, I’m less likely able to work overtime and increase my spending potential.
   So what I suggest transit engineers focus on is setting records in trying to attain the lowest commute times across a zip code matrix to include best and worst times within each zip code.
   In some cases, there are geographical barriers which show a destination only a mile away but to navigate to that location, you need to travel much further to get there and might be limited by speed zones, restricted access areas, or natural barriers such as green belts, and water ways.
   I hope to see several transport options address these roadblocks to faster commuted.

   1. I think most commuter’s travel a long distance and having more non stop travel benefits commuters more.

      In general it is the opposite. This is why even the best commuter rail has poor ridership compared to even the slowest mass transit. Express service has its place, but you aren’t going to get a huge number of riders compared to shorter trips. Proximity is one of the key elements to transit ridership (https://humantransit.org/basics/the-transit-ridership-recipe#proximity).

      I’m talking in general. The corridor served by Swift Blue may have an unusual number of long distance travelers, making the stop spacing appropriate. But it is worth noting that serving the corridor is extremely expensive compared to most service in Snohomish County. It is served by two routes, which greatly increases the cost. It is quite possible that you would get higher ridership if you had more standard (international) stop spacing, like the RapidRide E. You would have only one bus (not two) but the bus would run more often. You would still have the 512 (an express from Everett to Lynnwood).

      A lot of it comes down to the corridor itself. It would not surprise me if it is a land of extremes from a density standpoint. There is basically nothing (for large sections) and then a major destination (like a hospital or college) or a large apartment complex. This would be suited for a limited stop express. But it should not be taken as an appropriate model for the region — quite the opposite.

5. Lynnwood station’s new P&R is full ($) and cars are circling for a space at 10am. People are parking illegally in diagonal-striped zones (wheelchair-maneuver areas) and in adjacent store lots (partly displacing customers to small businesses).

   Some would-be riders are going back to driving because they can’t get a P&R space, and are saying the garage should be twice as large. ST and CT recommend feeder buses to Lynnwood or Mountlake Terrace stations, but these riders aren’t interested in that (or they live where it’s infeasible).

   The article also says all the P&Rs on the 1 Line are full, at least on some days. ST will introduce $2 reserved spaces soon at several P&Rs including Lynnwood. ST estimates that will raise $6 million in revenue.

   1. There is ample parking available at Ash Way P&R, and super frequency of express buses in the peak direction during peak, among routes 201, 202, 512, and 513, with dedicated freeway connections on both ends. Oddly, those southbound runs are split between bus bays that are far apart from each other, a topic I will go into in more detail at another time.

      How many reserved vanpool and carpool slots does the Lynnwood Garage Majal have?

Comments are closed.