SDOT recently announced RapidRide J Line construction will start in October, with service planned to launch in 2027. RapidRide J will travel from downtown Seattle to University District via Eastlake Avenue replacing Route 70.
According to SDOT, the two years of construction will build the following:
- 15,130 feet of repaired sidewalks
- 2 miles of repaved roadway
- 3.7 lane-miles of protected bike facilities
- 2 miles of bus priority lanes
- 177 improved crosswalks
- 33 intersections with traffic signal improvements, including 253 new traffic signal devices
- 190 newly planted trees
The project heading to construction has evolved somewhat since planning for a high-capacity transit corridor on Eastlake started in 2011, which eventually became RapidRide J in 2019.
Final Design
Given it’s been some time since the initial draft, the following will be a brief overview of the final design. Visit the SDOT RapidRide J Line page for the full details; the final design plans are here.
Generally the design consisted of converting the 5 existing lanes of 2 general lanes on each side and one center turning lane into either 1) protected two-way bike lane on one side with bus lanes on both sides or 2) one way protected bike lanes on both sides of the road with intermittent bus lanes where space was available.
Also, RapidRide J will be King County Metro’s first trolleybus RapidRide route, as RapidRide G couldn’t get five-door trolleybuses and RapidRide R is delayed)
University District
Starting from University District, the first southbound bus stop is on NE 45th Street with another southbound stop on NE 43rd Street. The last northbound stop is at NE 43rd Street and 12th Ave NE one short block from the U District station.
The bus will then travel on Roosevelt Way NE southbound and 11th Ave NE northbound. There are no bus lanes implemented on the couplet instead SDOT added a right-side protected biker lane and maintained 2 general lanes on each couplet road.
University Bridge
A subtle but important change for bike lane users on the south end of the University Bridge: the southbound center turn lane will be removed in favor of continuing the bike lane buffers through the intersection with Fuhrman Ave E, reducing the southbound general traffic to one through-lane.
South of University Bridge under I-5, SDOT will add short bus lanes in both directions serving upgraded stations. A northbound bus lane on Harvard Avenue will also be added serving Route 49.
Eastlake Avenue
On Eastlake Avenue from Harvard Ave to East Galer Street there are no bus lanes but instead one-way protected bike lanes on both sides of the street. The bike lane will go behind the in-line bus shelters, and
Fairview Avenue – E Galer St to Valley Street
Fairview Avenue from E Galer Street to Yale Ave will consist of a northbound BAT lane, while the one-way protected bike lane on both sides also transitions to a two-way bike lane on the west side of Fairview Avenue. (The previously rebuilt Fairview Avenue bridge has two-way bike lane on the west side.)
Note at the intersection of Fairview Ave and Eastlake Ave they moved away from a through-movement bike lane design shown above due to concerns with right-turning cars and through-moving bike conflicts.
Approaching Aloha Street, Fairview Avenue will have bus lanes in both directions in a somewhat complicated arrangement to accommodate the streetcar tracks. The street will be widened west by one lane in order to fit two-way bike lane, two southbound general lanes, southbound bus/streetcar lane, center turn lane, northbound general lane and northbound bus lane.
Fairview Avenue – Valley Street to Boren Ave
Fairview Avenue from Republican Street to Valley Street, crossing Mercer Street, consists of a northbound bus lane straddled between the 1 general northbound lane and the 2 right turn freeway on ramp lanes.
From Valley Street to Denny Way, Fairview Avenue will have business access lanes for both northbound and southbound directions.
The additional bus lanes are possible with the absence of bike lanes as
Virginia and Stewart Streets
On Virginia Street RapidRide J Line will use new northbound curbside BAT lanes, while on Stewart Street the bus will use the existing southbound BAT lanes implemented in 2016.
History and Future
 |  |  |
The 2015 open house showed RapidRide from Seattle to Northgate Station, the 2017 presentation had the line truncated to Roosevelt Station, and then finally the 2020 presentation the ended the line at U District Station. A large hurdle was the high cost of extending trolley wires north of U District to reach Roosevelt and Northgate.
The future is unclear given many competing priorities for transit capital dollars, however future extensions to Roosevelt and Northgate stations are still possible. For now, Route 67 continues to serve the Roosevelt Way NE corridor from U District to Northgate. And maybe the removed transit/bike improvements could be implemented on Roosevelt Way even without extending RapidRide J.
For downtown, the Seattle Transportation plan proposed the “Virginia St & Stewart St | Multimodal Improvements” referenced:
– Redesigning the street to better support transit service, including converting Virginia St into a two-way transit street and potential bus-only lanes
– Adding a protected bicycle lane on Stewart St
Other improvements, like a bi-directional cycletrack on Stewart Street which were determined to be “out of scope” for RapidRide J may return in future work.
This is just a bike project using transit dollars.
It will simultaneously speed up the 70 and finally provide a safe, practical bike route for the corridor. I think it’s OK for things to be win-win. I also think it’s neat they were able to get Federal funding for it; it’s unlikely the project would have happened without it.
Yeah, this turned out to be a pretty good project. The 70 will be significantly faster, even if it isn’t super fast. It is an essential bike corridor and it will be good to get the bike improvements. The alternative for bike travel (to hug the shore) would have cost a bundle (I’m guessing). Overall I think it works, it is just a shame it took so long.
Oh, and it is a shame it doesn’t go farther north. This makes things awkward. It runs on Roosevelt longer than it should, but not long enough to really provide value. It is ready to be extended, but that might take a while.
I have to assume they knew the fight to take parking off of Eastlake would be a major battle – I’m just glad SDOT won it.
I wouldn’t have the issue if Eastlake was defined as a multimodal corridor improvement, it was defined as a Rapid Ride project yet most of the infrastructure and benefit is non-transit. Similarly I’m not opposed to bike projects except when they sabotage transit projects and transit service. People forget bikes are still private single occupants vehicles (for able bodied people only) most concerned with getting from A to B as fast as possible and viewing everything in between as an impediment to their travel, just as motorists do.
I wouldn’t have the issue if Eastlake was defined as a multimodal corridor improvement
OK, but it should have been defined as a multimodal corridor. It is obviously important as a bike corridor and it is easy to argue it is the most important bike corridor in the state. It connects the UW with downtown via a flat pathway (next to the lake, in an urban environment). If it was part of the Burke Gilman it would likely be the most traveled bike pathway in the state. It is not only a “missing link”, it is a bigger missing link than the missing link.
It is nothing like Aurora, Phinney Ridge or 130th (where they are consider or have added bike lanes). Those are not only important transit corridors, but have clear bike alternatives that are better. You can’t say that about Eastlake.
Now some have mentioned some sort of alternative involving going along the water. That might work — I don’t know. But it is clear that the city should have a safe route around Lake Union, especially that part of it.
# Poncho is incorrrct to characterize this as a bike project using “transit $.* The two are complementary, not competing except to get people out of cars. I used to ride from Coleman Dock to UW because it was faster than the walk and bus ride on the 70. Bike improvements are long overdue as is reducing parking to speed up the buses and improve cyclists’ safety.
It was marketed as a major transit improvement. We’re tired of arguing unsuccessfully for transit improvements for decades, and then when one finally happens like RapidRide or Link or transit-priority lanes on streets, they get watered down into half as much or sometimes almost nothing.
I mentioned in the previous open thread that Metro is tripling the battery capacity of its trolley buses, which has interesting implications for future J-line extension. If extending the line to Northgate could be achieved with just a small section of trolley wire at Northgate to charge the batteries during layover, that would be much cheaper than needing to build new wire for the entire 3 miles or so between the U-district and Northgate.
Service-wise, I don’t think an extension would cost that much, since the area has to be served anyway by the 67. Of course, with an extension, the jog to U-district station becomes a 2-block detour, which begs the question of whether or not it’s actually necessary. I personally almost always favor straight routes, but I can see pitchforks coming out and people insisting on the detour so that their walk to the UW campus is two blocks shorter. I guess if such an extension ever happens, we will have to see. But, at least the increased off-wire capability of the buses somewhat raises the odds that it will eventually happen.
Poncho also shared info about a route in SF which uses the off-wire running as part of its route.
Metro may be hesitant to consider off-wire operations as part of a regular route, but it could be an interesting avenue for strategic electrification of some route which overlap with existing trolleybus routes, or extensions as you mention.
That’s pretty good. If SF can do it, that sets precedent, so maybe we can do. Of course, testing would be required to make sure, and, hopefully, repeated connection and disconnection with the wire doesn’t cause maintenance issues. But, if Metro is really serious about electrifying its fleet, leveraging the trolley wire that’s already there with battery/trolley hybrid buses seems like the low-hanging fruit, far easier logistically than running entire routes with 20 hour/day service using only batteries.
Metro has some operational experience here, as a side effect of the 520/Montlake construction: the 43 goes off-wire between McGraw and Pacific Place, about 0.8 mile. Not technically permanent, but it’s been a long-term experiment by now.
It is not just San Fransisco. Dayton does this and I know of two cities in Europe: Zurich and Arnem.
That’s a good point. It might reduce the cost of extending the route. For sake of argument I will say “70th” but it could extend anywhere past 65th then turnaround and layover between 12th and Roosevelt.
Good point about the 67. If you simply truncated the 67 at Roosevelt it is essentially revenue neutral. The problem is the 67 through-routes with the 65. But that isn’t the only alternative. You could move the 67 to The Ave (joining the 45) and then have the future 77 just end at Roosevelt. That is revenue neutral as well (more or less).
In any event, I think we gain a lot from having the J replace the 67 on Roosevelt (south of 65th). There are trade-offs, but this is why I think it is better:
1) There are a lot more people on Roosevelt heading to South Lake Union than are heading to Maple Leaf (where the 67 goes).
2) You have a much better transfer to the 44.
3) In the U-District, it complements Link better. It is about a five minute walk from Roosevelt Avenue to the station (not counting the walking inside the station to get to the platform). This means everyone on Roosevelt or to the west would have an alternative to get downtown or get to South Lake Union. There is a lot of density northwest of the station.
4) It is faster.
On the other hand:
5) It is farther from campus and The Ave. This is less than ideal, but the neighborhoods to the east have grown considerably over time.
6) It doesn’t connect as well to the U-District Station. I don’t think this matters much. If you are coming from the north, you just transfer at Roosevelt (which would be just as fast). It doesn’t make much sense to take the train from downtown and then take the bus back south. That basically leaves folks from Capitol Hill. But even then it is often better to take the 8 or 49 first, especially if they were faster and more frequent (respectively).
7) It doesn’t connect as well to other buses going east. Eastlake to the middle of campus or U-Village requires extra walking. But it still connects to the 44, which means you can still get to the south end of campus. Eastlake to the East Side would require extra walking, but at some point it makes sense to go downtown and just take East Link.
It is a trade-off, but I think the best option is to extend it up to 70th. South Lake Union is a major destination. It is basically part of downtown that isn’t served by Link. It is better to make this line longer so that you can give more people a relatively fast one-seat ride.
You could always detour back and forth but as with most detours it really isn’t worth it. It is worth noting that I’ve changed my mind on the subject. I remember arguing the opposite. But as someone pointed out, the D does not serve the heart of Ballard. Of course it would get more riders if it turned on Leary, went over to at least 20th and then worked its way back. But the riders who were north of there (along 15th) would be hurt by that. You make the route much slower, which cuts into frequency. In both cases it is best if the bus just keeps going straight.
The 67 could through-route at Northgate after heading west on 100th, perhaps with the 348 or 75. That leaves a gap between 100th and Northgate Way, but would generally be better for riders on Roosevelt since they’d reach the station faster.
The 65/67 through-route is useful for going from the northern U-District or Roosevelt station to stops on Pacific Street, U Village, or Children’s. It allows one route to maximize ridership as people from Roosevelt gradually get off and people from U-Village or 35th get on.
Yes, Ross’ points 5 and 6 look to me as poor design choices.
On top of that 45th is a messy street for buses to use for such a short distance as a turn-around.
Even if it doesn’t go all the way to Northgate, going further north than 45th looks to be as a better operational approach. It should at least serve the dense residential district up to Ravenna if not go all the way to at least Roosevelt High/ Link. In particular, having a better spot so several buses can idle at the northern end of the route that isn’t smack dab in the middle of the U-District pedestrian zone looks to me to be a better operational choice.
The 65/67 through-route is useful for going from the northern U-District or Roosevelt station to stops on Pacific Street, U Village, or Children’s.
I’m not arguing against through-routing. But there are other ways of achieving that. For example, assume the 348 goes to the UW via Roosevelt Avenue (replacing the 67) but it is too long to through-route with the 65. Jut find another pair. For example the 77 could just start at Lake City Way and run to the UW and then become the 65. This is actually better. The 65/77 pair is shorter than the big loop of the 67. The existing route is pushing it (in terms of length) as it is (and it will only get worse as traffic builds around Northgate). I’m not saying that is the only option (or necessarily the best one). I’m saying the 67 (from Northgate to Roosevelt) is a fundamentally flawed route.
I also think it is better if we consolidate service when it comes to the through-routing corridor (which is essentially this: https://maps.app.goo.gl/X7P6sZzhKrPKYwG99). Right now it is a mess. The 45 runs on The Ave, while the 67 runs on Roosevelt. Eastbound the 67/65 goes on Pacific/Montlake, while the 45/75 runs through campus (Stevens Way). You do have the 372, which at least helps the 45/75, but the 372 doesn’t go north of 45th. So if you are at 50th & 12th and want to go to Campus Parkway let alone the U-Village, you don’t know which way to walk. There are two buses fairly close to you and both go exactly where you want them to go. But they run on different streets, which means that effective frequency is much worse.
This is yet another example of the same problem — buses going pretty much the same corridor, but just different enough to make it hard to catch a bus. If the 67 was shifted to the Ave this problem goes away. Consolidate.
You could make the case that none of the buses should run on Roosevelt Avenue between 65th and Campus Parkway. Not until we are running buses frequently there. I’m not going to argue with that. But if you are going to run a bus on that section of Roosevelt then it should be the J, since it is doing something the other buses aren’t. The various through-routing buses should form a spine from 65th & Roosevelt to Campus Parkway (via The Ave/Stevens Way) while the J takes an entirely different pathway.
” I can see pitchforks coming out and people insisting on the detour so that their walk to the UW campus is two blocks shorter. ”
That’s exactly what happened when some people mistakenly believed there would be no stops east of Roosevelt/11th, and the real fact that the last eastbound stop is 12th.
There has always been a dilemma between a faster trip on Roosevelt/11th or getting closer to campus — and now the Link transfer. This goes back to the 80s. The 66 was always “express” with only three stops on Eastlake, but it didn’t go east of Roosevelt/11th, so you had to choose between faster travel time vs going directly to UW/the Ave when the 71/72/73X weren’t running (which used to be all day Sundays).
On the whole the detour seems to have the stronger case because UW/U-District/U-District station are such large draws. But it does screw up the simplicity of remaining straight north-south on Roosevelt/11th, as well as complicating trips for people on Roosevelt between 43rd and wherever it returns to it (45th? 50th? Ravenna Blvd?).
If it terminates at U District station it should be better of a transfer. If it goes north to Roosevelt and Northgate, it seems to me worth keeping on the Roosevelt couplet and having this 3 block transfer to U Distruct station for the benefit of more a direct faster route to those neighborhoods and where that transfer connection becomes less important.
Hope everyone likes the recap.
In general it does have a bit less bus priority than hoped on eastlake. More importantly on 11th and roosevelt way it didn’t get bus lanes there. The bus lanes nearing mercer street are exciting though hopefully northbound it doesn’t get stuck in that traffic.
Additional items:
Any thoughts on extensions to Roosevelt/Northgate? It seems not as pressing anymore given the link extensions, but that was the original idea.
What do ya’ll think about using both-side door buses on this route in the future. It could maybe use the streetcar stop on fairview
Any other recaps other people would like to see/know about? I was thinking of doing an overview of route 7 transit plus improvements given the bus lanes added.
It would have been nice to get bus lanes on Roosevelt/11th, though I’m happy to see that the project was fairly aggressive with removing parking despite heavy pushback. The 11th/12th repaving project is also removing street parking along the potential extension. Hopefully that sets up the street up for bus lanes in the future, perhaps if/when the J line is extended.
Center running bus lanes would have been nice, but it does look like Metro did fairly well with the right-side boarding constraint; center-running along all of Fairview might be a bit better, but I think that would benefit the streetcar more than the J line.
Hopefully that sets up the street up for bus lanes in the future, perhaps if/when the J line is extended.
That is what I’m thinking. I think it should come with contraflow (as I mention below).
As far as center running is concerned, it takes up more space (since you need a center bus stop). It is hard to see how they manage center running given they don’t have enough space for BAT lanes for a lot of this route.
To clarify for the couplet roosevelt/11th I’d imagine it would be left-side bus lanes so it wouldn’t use any extra space. This would also allow keeping the bikes on the right side.
Removing left turn lanes at an intersection (probably at Thomas, since that is envisioned as a bike/ped thoroughfare anyway) would give enough space for a center platform. I’m not sure there’d be a benefit to center-running anywhere else. On the Roosevelt/11th couplet there are turns in both directions. It’d probably be a marginal improvement to move the parking to the right side and bus lanes to the left, since then riders wouldn’t have to cross the bike lane, but I don’t see that speeding up the buses much over right-side bus lanes
To clarify for the couplet roosevelt/11th I’d imagine it would be left-side bus lanes so it wouldn’t use any extra space. This would also allow keeping the bikes on the right side.
Sure, that could work. But then they you have the same basic issues with cars turning left as you do with cars turning right. These are either BAT lanes or you restrict the left turns. You only allow them when there is a left-turn arrow and the car makes the turn in front of the bus. This is the case with Madison (at Boren and Broadway).
In contrast, with contraflow the bus just runs in its own lane that direction and cars that turn left have to yield (the way they would with normal traffic). Since buses aren’t that frequent it really isn’t an issue the way it would be with normal traffic. For example let’s say someone is driving down Roosevelt and wants to take a left on 55th (https://maps.app.goo.gl/US3V6upxPKcKEwdv5). There is no oncoming traffic, so they can focus on pedestrians. It is no different than a right tun. Now assume that it is is a normal two-way street and there is traffic heading in the opposite direction. Now this left turn is a big issue. The driver is focused on finding the gap in traffic, and might not notice someone crossing 55th (especially if it is a long wait). That could happen with the bus, but since there is a huge gap within buses, it isn’t much of an issue.
Center running makes sense with two-way streets. Contraflow makes sense with one-way streets (especially if you don’t have to worry about buses passing buses).
I agree about bus lanes on the Roosevelt couplet. This is the biggest weakness in my opinion. More about that in a second.
I don’t think it makes sense to extend to Northgate. It would be nice to have service along Banner and 5th (https://maps.app.goo.gl/bD5yZTLaQi9zkH2U9) or Weedlin and 5th (https://maps.app.goo.gl/g5QotVjA4A1eQFWU7). There are apartments along there. But it just isn’t worth it. If we had a lot more service hours it would be good, but I see other potential routes (e. g. Boren) as having a lot more potential.
I do think it makes sense to go past 65th (e. g. to 70th) as I wrote up above. This could be revenue neutral, or at worst it means running more buses between 65th and 45th (which are bound to get a lot of riders). You would need to add more right of way on Roosevelt/11th/12th. I think the best option is contraflow*. In the northbound direction the bus would just angle left to go on Roosevelt. Southbound (coming from 11th/Eastlake) the bus would turn right on 11th and then south on Roosevelt (https://maps.app.goo.gl/TNmHb3Fw1Tm9dYGh8).
The turnaround north of 65th (e. g. at 70th) would be the same, as it is a right turn. Not only would the buses run faster, but you also avoid the conflict with the bikes. The bikes would go the same direction as general traffic (as they do now) but the buses would go the other direction. There would be no need for the island bus stops (https://maps.app.goo.gl/8BTBNWF8mFYfMR22A).
I don’t think you gain much with both-side door buses on this corridor. It is too narrow to add a lot of center bus stops. It could use the streetcar stop but that seems like a lot of work for only one stop.
I would also get rid of that streetcar and move the bike lanes. This would be better for bikes and transit. I’ll probably write about it soon (as a post).
* I know I sound am a big fan of contraflow but I am also aware of its limitations. It doesn’t work well with a spine (unless you have two or more contraflow lanes) since buses can’t pass buses. In this case though, you only have one route. The speed improvements (100% exclusive transit lanes north of Campus Parkway) more than make up for the disadvantages.
I like that idea of routing buses on Weedin & 5th enabling a direct thru route from Roosevelt to Northgate to hit the Green Lake commercial district.
Yeah, that is probably the best option for connecting to Northgate, but I just don’t think it makes sense as this route. It is not that quick, and I don’t think you would get that many people. You start competing with other routes (and Link) for various trips. The big attraction here is South Lake Union. The farther north you go, the more it makes sense to take Link (especially since this would run right by the station).
I could maybe see it as an extension of another bus coming from the north. For example I think it would make sense to send the the 75 to Bitter Lake. Then you need to backfill 5th NE, north of Northgate Way. I would send a bus from Northgate TC up 5th and then cross at 130th (by the station). At Meridian the buses would branch. That bus (or buses) would go south of Northgate Transit Center to the Roosevelt Station via Weedin. That could go along with sending the 348 to the UW (via Roosevelt). You essentially give people along 5th a different option for a one seat ride to the UW (not involving Link) while making the other route faster. You probably haven’t saved much money in the process but you’ve added coverage. I would only do that if we have a lot more money for service. There are way too many places that lack decent service.
A recap of the work for the 7 would be great. It is a very important bus, and anything to make it better is well appreciated. This is a case where both-side door buses would be good, although you would also need other Jackson buses to the do the same. Do that and you could turn the streetcar pathway on Jackson into a transit lane, saving lots of buses (and the streetcar) a lot of time.
I wonder what the minimum order for 5-door trolleybuses would be, and if implementing 5-door operations on the G, J, R, and other routes would be enough to justify the purchase.
I wonder what the minimum order for 5-door trolleybuses would be
Great question.
You’ve got the G to start with. Along Jackson you have the R, 14 and 36. I think that would be enough. The 36 is a bit tricky. I think it makes sense to pair the southern end of the 36 (serving Beacon Avenue) with the 49 (extended to Beacon Hill). Then send the 60 to downtown. The 60 doesn’t run under wire. One alternative would be to split up the 60. Have it end at Beacon Hill (like the 107). Then run the 49 to Beacon Hill and the 36 just keeps going downtown (as a trolley with doors on both sides). Works for me.
The 1/14 seems like a stretch at first, but historically it has performed quite well. There is another (really interesting) possibility: Run the 1/14 along First Avenue. That way it can take advantage of the center running bus stops they want to add. Now the buses are certainly justified.
That leaves just the 106, but it could eventually be sent to SLU (via Boren) while in the meantime just slog through the traffic. I could also see it doing something different (like back-filling service on 9th after the 60 is sent to Broadway). You would have to find layover up there, but that doesn’t sound too hard.
I’d live to see the EmX in Eugene as a trolley coach line. They want zero emissions by 2035, and I don’t see a way to get there unless EmX gets trolleybuses. It’s 28 miles round trip every 10 minutes.
That’d be another 22 dual sided.
I’m not as concerned about extending the J to 65th or Northgate as I used to be. It’s not just Roosevelt station that helped; it’s the rechannelization of Roosevelt Way between 65th and 45th, and routing and street improvements on the 45.
The 67’s remaining problem is its poky speed between 65th and Northgate. That can be addressed with street treatments; it doesn’t have to become RapidRide. Oh, and restore 10-minute frequency on the 67 and 45.
Extending the J north would reintroduce the dilemma of staying on Roosevelt/11th vs going to UW station closer to campus. With the current route, it naturally goes to UW station because that’s the terminus.
The 67 is fundamentally flawed. It makes a huge button-hook. You can get off the bus, walk a few blocks, and then catch the exact same bus as it rounds the corner. The 348 should run on Roosevelt Way instead (replacing the 67). If this bus is extended beyond 65th it makes sense to go to Lake City instead.
I still think the best bet is to extend just north of 65th. The more I think about, the best option is 66th. This offers several things:
1) A lot more one-seat rides to South Lake Union and Eastlake. At worst you can walk a little farther to campus, but the potential walk for someone at say, 60th and Roosevelt to this bus is a lot longer. There are also some trips that go from being three-seat to two-seat rides (e. g. from the 62).
2) You can add contraflow bus-lanes (not just BAT lanes) between Campus Parkway and there quite easily (as I explained above — https://seattletransitblog.com/2024/09/23/rapidride-j-line-final-design-and-recap/#comment-941688). North of there (along Lake City Way) you can’t do that. The transition between the contraflow lanes and Lake City Way would not be as elegant as just turning around (as I explain below):
3) If you turn around at 66th it is ideal. The bus would go north on Roosevelt (in its own lane) and then turn right on 66th and stop right next to the station. Riders would have a very short walk to the station (they wouldn’t even need to cross the street). Then the bus turns right (in its own lane) and doesn’t mix with traffic until it gets close to the bridge.
I think the biggest problem is ending the route in the middle of the U District. It’s a terrible area to lay over buses, and there are a ton of student apartments north of 45th.
I can’t stress the layover location issue enough. With traffic and the drawbridge, reliability becomes a big concern and buses will need layover space with enough room for at least 3 buses if not 4. As RRG demonstrates, buses get off schedule and bunch when a route has a high frequency and lots of signals.
Unlike others, I think it should run closer to Red Square, and use 15th or a nearby street to run further north to at least Ravenna or Roosevelt Link. It’s a heck of a lot easier to find room for bus layovers further north and the run would pick up a whole new set of UW student riders. Getting to Northgate via bus from the U District is no longer that important in my eyes now that Link is here.
Overall, I see the primary market for the service are SLU employees, Eastlake residents and UW students and staff. People who want to bypass the campus will mostly ride Link. I don’t see it needing to go further than Ravenna or Roosevelt Link Station or maybe Green Lake. Anyone in the U District going to Northgate will find Link to be a much faster option.
[Ed: this comment got caught in the moderation queue; apologies for the late approval -ND]
The center stop on Fairview really exists for the convenience of the streetcar, not the passengers – it was the easiest way to turn the streetcar around at the end of the line. For buses, it seems simpler and easier to just stop in the curb lane, as the 70 already does.
As far as the University draw bridge openings, are they planning anything different than the 70? Perhaps one approach is having a bus or two on standby in Pioneer Square so southbound busses that get delayed don’t affect northbound service as much. In the event the bridge gets stuck, I guess the only real option is to detour to the Montlake bridge.
I found some 2015 data on drawbridge openings here:
https://statsonthestreet.org/wp-content/uploads/2015/10/aveplot.png
The Coast Guard does allow for scheduling openings at many bridges across the country. But 4-6 times on a typical September day is relatively minor. Fremont Bridge was almost double in 2015!
Fremont is much lower than the other (ship canal) bridges. The other ones are similar, but Ballard opens more because it is the farthest west. It probably has high weekday openings because of industry (not just the recreational sailboats). Bridge heights: https://www.seattle.gov/transportation/projects-and-programs/programs/bridges-stairs-and-other-structures/bridges#vehicularmoveablebridgelist.
Routes 31 and 32 have more lateness as a result. The bridge openings are seasonal; there are more May through September with sailboats. The fall 2021 change caused them to lose connectivity at NE Campus Parkway and the UW. The new pathway on NE 45th Street and the Roosevelt couplet has more congestion and delay. A fix: new Route 30 could connect Children’s and the U District station via NE 45th Street; routes 31 and 32 could return to NE Campus Parkway and use the Pacific Triangle terminal.
The 31/32 shift always seemed odd to me. There are trade-offs, to be sure. If you are coming from the north on Link it is better to get off the train at the U-District and ride from there. But if you are coming from Capitol Hill it is worse. If you are coming from the north via a bus it is about the same. If you are coming from the south on a bus it is worse. Likewise it is a bit worse if you are coming from the middle of campus. It seems like the original pathway (with more red paint to make it faster) would be better.
It does get you a fast connection from the U-District to Children’s Hospital. There are other ways to achieve that, but none of them sound good. A stand-alone route works for fewer trips. You could send the 44 straight across, but you lose a lot of one-seat rides. You could send the 65, 75 or 372 that way, but then you inevitably lose some combinations (or potential combinations) even before you consider the through-routing ramifications. Oh, and a bus coming from the north can’t just turn and head east on 45th. I suppose it could loop around the station, but ick.
I think the best solution would require a major reshuffle of the buses in the U-District. I’m not sure if you end up with anything that is great though. The best bet might be exactly what you suggested.
In general the biggest problem is not the bridge opening itself, but the congestion that builds up behind it. For example it might take a couple minutes for the bridge to go up and down. But it could take ten minutes before traffic is free flowing again. Since the bridges don’t go up during rush-hour it didn’t used to be a such a big deal. But as midday traffic has increased (especially on the weekends) it can be a big deal.
The key is to allow the buses to basically skip to the head of the line. A bus might get unlucky and get stuck just as the bridge is opening. But it won’t get stuck in the terrible backup that occurs a few minutes later. I’m not sure enough has been done here to address it (unlike with the 40 or with the Montlake Bridge).
SDOT did not well consider Link in their G, J, and R line designs. They had too little consideration for the network as a whole. The Kubly SDOT seemed more interested in monuments.
RossB introduced the concept of consolidating routes 348 and 67 with neither serving Northgate station. Route 348W could connect Richmond Beach and north Shoreline; Route 348S-67 could connect north Shoreline and Roosevelt via North City, Jackson Park, Pinehurst, and Mapleleaf. The consolidation should yield shorter headway and waits. Route 348S-67 could be labeled Route 367. Routes 65 and 67 need not be paired. Route 65 could be extended to SCC via Link (130 or 148).
Per RossB comments above, the fall 2024 Metro change is very weak. Additional consolidations should be attempted around the G line and Judkins station. Yes, the new and better batteries on the ETB could be put to use. The R line as Route 7 should be reconsidered. Link can carry the radial load to/from downtown Seattle.
SDOT did not well consider Link in their G, J, and R line designs.
I wouldn’t say that. The R is extended to Rainier Beach Station — one of the most controversial aspects of the project. The J originally connected to three stations (Westlake, Roosevelt and Northgate). Each truncation ended at another station (first Roosevelt and now the U-District). If you are going to Roosevelt (or Northgate) it doesn’t make sense to deviate to the U-District any more than it would make sense for the E to get closer to Westlake State before making the connection to Link at Symphony Station. The connection to the U-District is poor, but that is likely due to the truncated nature of the project (it would be a lot better if it went to Roosevelt Station). The G is on Madison/Spring instead of Madison/Marion in large part because it gets it closer to Link. Of course the big problem is that there is no station on Madison.
But you can blame ST (when they built the line from UW to downtown) or the county (when they built the original transit tunnel) for that omission. The buses can only do so much. It doesn’t make sense for the RapidRide E to cut over to the freeway and then back just because ST decided to follow I-5 instead of Aurora. I don’t have the exact quote but I remember the commenter d.p. saying “the worst thing you can do is have the train people build their system and then tell the bus people good luck”. Yet that is pretty much what happened. Very little of our subway system was built to maximize the network and take advantage of the buses. Even something as simple as a station at 130th required fighting tooth and nail. So we are left with some stations that are great for the network (Roosevelt) and others that require buses to make a seemingly endless set of turns just to get to the station (Northgate).
“The last northbound stop is at NE 43rd Street and 12th Ave NE one short block from the U District station”
This SDOT choice is terrible. The current Route 70 pathway would be better for almost all northbound riders. The last outbound stop is to be nearside 12th Avenue NE. The block is uphill; consider what that means for those of limited mobility or sight. Riders wanting to reach bus routes on the Ave or 15th Avenue NE will have two or three block walks. SDOT and the FTA will spend capital to make connections worse. Transfers are where the transit network slows down with walking, wayfinding, and waiting. The SDOT pathway degrades Link integration and connectivity. The dorms on NE Campus Parkway will lose easy access to the service. Today, relatively few Route 70 riders are oriented to Link; many more are oriented to the UW, U District, and bus routes; they will have longer walks.
According to NACTO, transit riders are better off on two-way arterials for better clarity and wayfinding. Consider that electric trolleybus overhead is more costly on a couplet than on a two-way arterial.
Why did SDOT make this alignment choice? To save a couplet of minutes running time, even though it will slow the combined trips of northbound riders? To show the FTA that they are headed to Roosevelt? But then, the Jay would have a long transfer walk to/from the U District Link station and the heart of the U District forever.
The SDOT restriping of Roosevelt Way NE in 2016 slowed the arterial for transit. A lane was taken in the peak. The bus islands are only long enough for one bus at a time. Now, there are routes 31, 32, and 67; SDOT would add the Jay. The UW health science express also serves NE 42nd Street. If the Jay gets its own stops, transferring riders will have walks. The SDOT ROW manual implies that bike facilities should be placed on the left side of one-way arterials. Yet, they have been placed and are being placed on the right side of the Roosevelt couplet.
During its Seattle Transportation Plan discussion last year, SDOT floated the concept of shifting the Stewart-Virginia couplet to two-way with a transit focus on Virginia Street. That has merit. But the concept could be applied to the Roosevelt couplet and Pike-Pine as well. On Roosevelt, transit could have priority on 11th/12th avenues and the transfer walks minimized at the Roosevelt Link station. Two one-wat PBL could be provided on Roosevelt. A better Pike-Pine Renaissance would have placed two one-way PBL on Pike Street and two-way transit on Pine Street. Then westbound cyclists would not have to shift to Pine Street via Melrose and transfers between Link and eastbound transit would not require the 400-foot walk.
The last outbound stop is to be nearside 12th Avenue NE.
I always assumed that was the last stop you could board, but it does look like it ends there (they kick you off). Maybe there is no room to layover on Brooklyn.
Why did SDOT make this alignment choice? To show the FTA that they are headed to Roosevelt?
That is my guess, yes. The long term goal of service here is to reach 65th if not Northgate. They haven’t given up on that. So the wire and the fancy bus stops (with ORCA readers and kiosks) are added along the future corridor. There are two on Roosevelt and one on 11th. They also didn’t spend any money on stops outside of this. At least I don’t think they will. 43rd already has a reader board (I’m not sure if they have an ORCA reader for the buses). I’m not sure about 45th (I haven’t been there in a while). Even if they invested in those, it would likely pay off in the long run, as it is highly likely that the 44 becomes RapidRide at some point. In contrast, it is unlikely any RapidRide bus will serve Campus Parkway.
If your ultimate goal is to get to U-District Station, this is clearly not the way to do it. Just follow the existing pathway of the 70. But they are trying hard to lay the foundation for future extension while ignoring the fact that it won’t serve the present very well.
Probably also because the layover on 12th Avenue NE is already wired and have a siding for doing a layover there (that is the current layover for 49/70), so it make sense to reuse this layover. But I have to agree it is a bad spot for a terminal.
On Roosevelt, transit could have priority on 11th/12th avenues and the transfer walks minimized at the Roosevelt Link station.
If Roosevelt and 11th/12th become two-way, then they should just become contraflow. Buses (and only buses) would be allowed to go north on Roosevelt and south on 11th/12th. The J buses would turn around at 66th.
It is a little tricky north of there. Two way traffic between 73rd and 75th is problematic because of the Lake City Way ramps. If you don’t fix that, then the two-way traffic ends at 73rd (or more likely, 70th). That would mean a northbound 67 would have to turn right on 70th, left on 12th, left on 75th and then right on Roosevelt just to keep going.
Simplest thing to do is just end the two-way streets at 66th. The other buses would have two choices. One is to dogleg on 66th and take advantage of the bus-lane. This gets the rider closer to Link and makes for nice same-stop transfers with the J. The buses would only share that little section (between 66th and Ravenna) before they cut off to go The Ave. Another is just ignore the contraflow lanes and keep doing what they are doing. Same direction transfers would require crossing the street and walking half a block to the (laying over) J on 66th.
I would run bike lanes with the general traffic. So southbound (on the right side) along Roosevelt and northbound (on the right side) on 11th and 12th. The only possible bus islands you would need are for those buses I mentioned (if they don’t dogleg). I generally don’t like turns but in this case I think the dogleg makes sense. The left turn doesn’t involve incoming traffic, so it is just a matter of pedestrian traffic. You might need to add traffic lights, but that is about it. The turns to and from Ravenna look fine.
“ If Roosevelt and 11th/12th become two-way, then they should just become contraflow. ”
If both of these streets are two-way streets, there really isn’t any such thing as “contraflow” possible. Traffic would flow in both directions.
Contraflow is by definition two-way traffic. It is just that in one direction you only have buses.
To explain I think they are talking about contraflow from general traffic.
Aka Roosevelt Way would stay southbound general lanes but a northbound bus lane on the right (east). While 11th Ave NE would stay northbound general lanes but a southbound bus lane on the left (west).
Some advantages with this configuration is that the bus doesn’t need special boarding doors and little conflict with left/right turns.
A major disadvantage would be signal timing as the general traffic signals would be coordinated opposite of the normal couplet traffic direction.
I see, Ross. What confused me was the term about making the road “two-way”. To me that means all traffic as opposed to still just one-way for cars.
Regardless, as WL notes, it goes against the signal progression. A big advantage of one-way streets is when it’s possible to hit mostly green lights as a car or truck or bicycle travels down that street. A contra-flow bus lane would be hitting lots of red lights if it’s running contraflow unless the progression is abandoned.
With these streets, there is also on-street parking and bike lanes. There are lots of driveways (including a fire station with trucks needing room to maneuver) on the left side so cars would be crossing the bus lane all the time when turning into and out of these driveways. Pedestrians often don’t look in the other direction if traffic is flowing one way, adding another safety concern. Finally, buses could not pass each other if one breaks down. It would make bicycling these streets scarier too.
I don’t think in this instance the extra effort and cost of redesigning the street, changing all the signals and resolving all the local access (driveway) situations while making the street more dangerous just to have a contraflow bus lane gets better benefit than looking at other ways that create less hazards.
My suggestion as I mentioned would be to look to another parallel street to the east and make that the transit-priority street. Let bikes and cars stay in Roosevelt and 11th and out of the way of bus riders.
A major disadvantage would be signal timing as the general traffic signals would be coordinated opposite of the normal couplet traffic direction.
It is fairly common to set the traffic signals to work both directions. Furthermore, it is rarely an issue with urban transit as the buses have to stop anyway. It is practically impossible to time it so that the bus can make every light. At best you have signal priority (where the light stays green longer than it would otherwise or the light turns green early).
Another issue is that you only have one lane of traffic for the buses. That is true of any street with one lane of traffic going that direction. For example the 45 goes along Green Lake Way, which is one lane each direction (https://maps.app.goo.gl/8rAaccjuMxzoSkhC8). Buses can’t pass buses.
While Roosevelt is wide, it isn’t that wide. There are competing interests — cars, buses and bikes. It is hard to see how we could have BAT lanes both directions, another general purpose lane both directions and bike lanes both directions. That is four lanes of traffic as well as bike lanes. Oh, and the stops are island stops (like on Dexter) which takes up more room.
This would be a major traffic change. I’m not saying it wouldn’t be better, but you have issues. For example it means one lane of traffic on Roosevelt heading towards Eastlake. Drivers could also use 11th and then turn to get on Eastlake by turning on Campus Parkway, but it is hard to see how that is good. My guess is they would ban that. It is a road diet — sort of. Again, you have four lanes of traffic and bike lanes on Roosevelt. Cars trying to avoid traffic use 11th and 12th, which now has … one lane of traffic each way? Two each way?
Meanwhile, these are still BAT lanes. You still have plenty of cars turning right, and slowing down the bus as it waits for pedestrians. In contrast, with contraflow (on both streets) you avoid all of that. Run the buses (and only buses) north on Roosevelt and south on 11/12th. Put bike lanes southbound on Roosevelt and northbound on 11th/12th. In both cases the bike lanes would be running with the general traffic and opposite the buses.
This doesn’t preclude a future road diet. Eventually Roosevelt and 11th/12th could have one lane of general purpose traffic each direction. Great! Widen the sidewalk. Add more of a buffer for the bikes.
Of course there is another option. You could run the buses contraflow one direction and with traffic the other direction. This is what they do on 5th. This makes sense if traffic is much worse one direction. For example if Roosevelt is a lot more congested than 11/12th. In that case you just add a contraflow lane on 11th/12th. This has the advantage eddie mentioned a while back (the bus stops are on the same street). If they did that then it would make sense to add bike lanes (both directions) on Roosevelt. That would be an improvement, but I still think the best option is contraflow (and bike lanes) on both streets. BAT lanes are good. Bus lanes are better.
@Ross
it might be a good idea for a short article. I understand what you’re saying but I think it’s a bit complicated to explain without a diagram for others.
What confused me was the term about making the road “two-way”. To me that means all traffic as opposed to still just one-way for cars.
I understand, but my point is from a driving perspective and a pedestrian perspective it is the same as two-way traffic. If I’m standing on the corner I can see vehicles going both directions. If I’m driving down Roosevelt and see a bus going the other way I’m thinking “when did they make this a two-way street?”. It is only after a while that I go “Huh. Only buses are going that way.”
Regardless, as WL notes, it goes against the signal progression.
See my other comment. I’m not sure why this myth persists. Maybe I should write something about it. In any event, any signal timing should first favor the bus. Doing so is complicated (https://nacto.org/publication/transit-street-design-guide/intersections/signals-operations/transit-signal-progression/). Once you do that, you can consider the timing of cars (going the other way).
With these streets, there is also on-street parking and bike
Sure, just like a normal two-way street. The bike lane would be on the right. The parking would be on the right. How you handle that is up to the city. (I think it is best if the bike lanes are next to the sidewalk, followed by a curb, followed by parking.). The only thing that is a bit odd is that the bike lanes only go one direction (with the general traffic and opposite the buses). That is the case now, so you wouldn’t have to do much work.
There are lots of driveways (including a fire station with trucks needing room to maneuver) on the left side so cars would be crossing the bus lane all the time when turning into and out of these driveways.
Yeah, sure, but the same is true now. Cars turn right (and delay a bus). But it is worse. A car is trying to turn right. But it can’t, because of all the pedestrians. So it waits. This delays another car in the right lane and a car behind that. So now you have a line of cars backed up in front of the turning car. The car finally makes the turn and traffic opens up. In contrast, it is unlikely that a car would initiate a left turn (and block the incoming lane) until they find an opening. There is no reason to. The only thing coming that direction is a bus. It makes sense to do that if you have a line of traffic (it is aggressive and effective). You basically stop traffic going the other way until you have a gap in pedestrian traffic and can get into the driveway. But that makes no sense with a bus. Just wait until the bus passes. It is a bus after all and there is nothing behind it for as far as the eye can see. As a result you don’t ease into the oncoming traffic lane until you are reasonably confident that you can get into the driveway.
Pedestrians often don’t look in the other direction if traffic is flowing one way, adding another safety concern.
Except again, traffic is flowing both directions. There would be a learning curve to be sure. But the same would be true if we made it general-purpose both directions. For decades people crossed and didn’t have to worry about traffic coming from one direction. But it is also quite common for people to look both ways on one-way streets. I do. It is just common nature. It is also a good idea. People driving the wrong way on a one-way street are not unheard of. If anything, the likelihood of being hit (for someone who doesn’t look both ways) is much worse with general traffic going both ways. There are often gaps in traffic (caused by a signal change). The lead car is often driving way too fast and is usually an amateur driver. In contrast, a bus is being driven by a professional who doesn’t mind going 25 MPH. They are more likely to notice a jaywalker and stop in time.
My suggestion as I mentioned would be to look to another parallel street to the east and make that the transit-priority street.
The only street that would work is Brooklyn. At that point you might as well run on The Ave. You are way too close to buses on The Ave to provide any benefit. You still end up with the same basic issues. These streets have driveways too. Do you prevent people from using them?
Finally, buses could not pass each other if one breaks down.
Yes, I mentioned that. But the same is true with most rail, trolleys (until recently) or buses that run on streets with one lane of traffic. But this isn’t a streetcar. They bus can pass a bus, it just isn’t easy. You have to go into the oncoming lane. Late at night it would be trivial, but during the day you might need a traffic cop. Hell, give me a yellow vest and I’ll stop traffic going the other way and let the bus pass. But again, there are usually gaps in traffic and this is a bus. It gets noticed. By the time the Metro officials are there dealing with the broken down bus, there is a Metro (or city) traffic cop there as well.
The main disadvantage of single lane contraflow is that you can’t do what they do on Third Avenue. Buses skip stops. They routinely pass other buses. Not because the bus is broken down, but because one bus has a bus stop there and the other one doesn’t. That wouldn’t make sense for this corridor. Furthermore, the only bus on it (for most of the way anyway) would be the J — buses that are scheduled to run equally spaced. This would make bus bunching worse, but bus bunching with leap-frogging isn’t much better. It is best to avoid bus-bunching in general, and one of the ways you do that is to avoid traffic backups (which this would do).
It would make bicycling these streets scarier too.
Not really. It would basically be the same. The bike lanes themselves would be one of the few things that wouldn’t change. The bikes are in the right lane now, and they would remain in the right lane. The only cars you have to worry about are those turning right.
@WL — Agreed. Contraflow is not that obvious. The name is part of the problem. In this case you also have issues with bikes and all rest of it.
Saving a couple minutes of bus running time is not as beneficial if you are adding several minutes of additional walk time to the major destinations at the end of the line (UW, transferring to Link at U District station, to avoid backtracking to Link in downtown).
Easiest fix I can see: Have the bus go through the loop in front of U District station twice. Coming NB, the bus does the loop, and kicks off all riders in front of U District station. The bus can layover somewhere, or go right into a SB trip. For the SB trips, the bus picks up pax at U District station, as currently planned. The bus “looses” those extra minutes (not ideal), but the bus riders, less so.
In this case, the bus might even skip the extra turns to get to that last NB stop, and just go up 11th Ave. to 45th and turn right.
Yeah, that would work. While extra distance, it would be good for riders. It might be a bit confusing to see a bus (especially RapidRide) and get excited, only to realize it is the bus that is arriving, not the one departing. I agree, the bus should just keep going on 11th. The 31/32 goes up 11th and turns there, so that should be no problem. That would mean that you would get rid of the “last northbound stop” shown on the map. Replace it with a stop on 11th.
Regardless of what they do, they shouldn’t add wire on 43rd or 12th. The bus can always run off wire until it hooks back in at the first (southbound) stop on Roosevelt. If and when the bus goes farther north (up the Roosevelt couplet) they should add wire. They might not need much.
Can someone explain how, with this new design, a person coming south over the U bridge is supposed to continue on Harvard Ave without doing what we have to do now which is merge left across two lanes of traffic at the south end of the bridge? I do this maneuver multiple times a week and it’s not clear how this plan improves it.
I don’t think this plan improves it. I don’t think there’s any way to improve that without building a wider bridge or some sort of overpass, do you?
If changing lanes twice is too hard for you, maybe just stay on eastlake and take a left on Roanoke to get back over the highway.
Or Allison St. prior
> a person coming south over the U bridge is supposed to continue on Harvard Ave without doing what we have to do now which is merge left across two lanes of traffic at the south end of the bridge? I do this maneuver multiple times a week and it’s not clear how this plan improves it.
I don’t quite understand? If you in a car are traveling south on U bridge there is plenty of time and distance to merge over 2 lanes even with the current setup. Just merge one lane over to stay on the left general lane and then merge into the left lane once you reach it.
If you’re talking about biking southbound there’s two choices.
1) merge from the bike lane into the general southbound lane then merge into the left lane. There’s only one southbound general lane instead of two so it’s generally much easier.
2) continue south past the eastlake/harvard intersection and then cross the intersection. There’s a new bike/pedestrian crossing and I assume the traffic light will be retimed to allow for pedestrians to cross eastlake.
The only one southbound lane will really help with the merge. Good point. I didn’t notice that part.
Whoa hold up, they’re getting rid of all the stops along Fairview north of Aloha Street? That’ll be more than a full kilometer between the Aloha stop and East Howe stops, up in Eastlake proper.
There are a ton of riders who currently board the 70 every day at the Silver Cloud Inn. This will probably push many more of them to use Ubers. :(
That does appear to be the case. I have no idea why. It doesn’t look particularly difficult to serve that stop either direction. The bus runs curbside through there (and doesn’t have to change lanes to get to a stop).
Hi, yeah looks like I forgot to mention stop spacing. From a cursory glance that does look like the largest station spacing between aloha to garfield
There is a tradeoff between number of stops and speed but looking at google maps the “yale” stop near silver cloud inn does look important enough. Checking google it’s around 3200 feet apart or around 0.6 mile stop spacing. Adding another stop in the middle would bring it down to around 0.3 mile.
It is a bit late given they’ve started construction but if you got a couple others perhaps you could write in an letter arguing for keeping/adding that stop.
It is worth noting that RapidRide has the concept of “stops” (which are just regular bus stops) and “stations” which have ORCA card readers at a minimum. Thus it would be trivial to add a “stop” along here. It would be one thing if the bus was running in the middle of the street (to make a left turn) but that isn’t the case either direction. This is not a limited-stop express; there should be a stop there.