By JASON LI
This is part 3 of a 3-part series on route 8. Part 1. Part 2.
Even if Denny Way is solved, frequency for the route is subpar, especially at night, where it drops to 20-30 minutes.
| Headways | 7-9am | 9am-4pm | 4-6pm | 6-8pm | 8-10pm | 10pm-1am |
| Weekday | 12 min | 15 min | 12 min | 15 min | 20 min | 30 min |
| Saturday | 15 min | 15 min | 15 min | 15 min | 30 min | 30 min |
| Sunday | 15 min | 15 min | 15 min | 20 min | 30 min | 30 min |
A table of Route 8’s headways
At peak, theoretical 12-minute headways are rarely met as Route 8’s schedule only allocates 60 minutes to go from Queen Anne to Mount Baker, but oftentimes this trip takes more than 90 minutes. This means buses don’t arrive at their base on time, resulting in cascading delays as buses start their next trip late. Riders are often left waiting 20 or even 30 minutes for their next bus, especially during periods of bad bus bunching.
However, once bus lanes are added, the opposite would happen as buses speed through the route but spend the time savings sitting idle at their base until their next scheduled departure. Metro must be ready to adjust schedules accordingly and increase frequency to more efficiently utilize their existing resources to meet the exploding demand that would accompany bus lanes. Luckily, it seems like they are working closely with SDOT and will be well-poised to respond immediately to maximize benefits from any infrastructure improvements.
At night, current frequencies make relying on Route 8 to get you home at night untenable, especially if making a transfer, and also leaves it wholly unable to serve evening events in Capitol Hill and Seattle Center. Additional service hours during these periods is not only relatively cheap given low traffic conditions, but also extremely effective in improving frequency. Only one additional round trip bus per hour is required to bring headways from 20 minutes to 15 minutes, and only two buses an hour are required to to bring headways from 30 minutes to 15 minutes.
8 + 11 = ?
One method to resolve Route 8’s frequency issues without needing to invest large amounts of service hours into it is to re-route Route 11 onto Denny Way, fully overlapping Route 8 west of Madison. For most Route 11 riders, this wouldn’t even result in a slower commute after the full opening of the 2 Line, because doubling Link’s frequency to 4-5 minutes would make it always faster to get off the bus and board a Link train than to stay on a bus to Westlake. Although not ideal, routes 49 and 3 can also serve as transfers for the subset of riders doing more local trips between the two stations.
Map of Route 11
This may sound far-fetched, but it’s already in Seattle and King County Metro’s long-term plans. The Seattle Transportation Plan’s Frequent Transit Network assigns the lowest tier of frequency for MLK and Madison before Routes 8 and 11 intersect but gives the rest of Route 8’s corridor the highest frequency tier, including along Denny Way. The Metro Connects 2035 Interim and 2050 networks also show the 11 being re-routed to serve South Lake Union via corridor 1061, which is slated as a RapidRide corridor in the 2050 network. While we would love to see this corridor prioritized for RapidRide in the 2035 plan, there does not seem to be any feasible path to doing so based on our discussions with the County Council. However, all of this is likely dependent on Denny Way getting bus lanes as this would both extend Route 11 and pull it away from dedicated transit lanes on the Pike / Pine, requiring exceptionally high service hour costs if it has to fight traffic on Denny.
The Seattle Transportation Plan calls for the highest tier of frequency on Denny Way
Even without a re-route, Routes 8 and 11 already overlap on Olive and John, which is a mutual source of headache for both routes. Adding bus lanes to this corridor would speed them up, with Route 3 getting some minor benefits as well. However, doing so would not be a simple task as these streets are dotted with bus and curb bulbs, whose removal would simply throwing down some red paint to convert parking lanes into bus lanes would result in buses needing to merge with traffic every block, reducing their effectiveness. A full redesign of the corridor may be warranted to ensure they can run at full speed without harming pedestrian safety. Fortunately, the Seattle Transportation Plan includes multimodal improvements that will be able to completely redesign the entire Route 8 corridor west of Madison as a Large Capital Project, which calls for at least $10M in funding. In fact, transit improvements for Denny Way and Olive Way are included in the 2024 Seattle Transportation Levy, although it is not included in the 2025 Levy Delivery Plan and we urge SDOT to prioritize this transformational project.
The Seattle Transportation plan includes large-scale improvements for the western half of Route 8
Unfortunately, John Street and Thomas Street’s exclusion in the Levy means we will need to wait at least 8 years before design for that corridor even starts. Although not nearly as bad as Denny Way, westbound John St also serves as a major headache for both routes and we would encourage SDOT to look into possible spot improvements to mitigate traffic impacts in the meantime. For example, the existing westbound bus-only left-turn lane at John and Broadway for Route 43 could be extended a few meters to reach the nearby bus bulb to double as a queue jump lane for routes 8 and 11. Furthermore, that signal is currently triggered by personal vehicles and is extremely sensitive, often activating with an empty lane. Adjusting it so that it can only be triggered by buses would improve traffic flow continuing straight with fewer turn phases.
Either way, we are excited that Seattle has so many plans in the works to make meaningful improvements along much of Route 8. We just wish that they would come sooner. If you want to help the Fix the L8 campaign’s push to prioritize these changes, send an email to our local leaders and consider joining the campaign! We are also hosting a Race the L8 event on July 10th at 5 PM to highlight the issues that Route 8 is facing, more details here!
Historically, Metro has always shortchanged the 8 because they were skeptical of routes that didn’t serve the downtown core. Somebody involved even told me that, when the 8 was first created, there was a lot of resistance to even running it at all.
Today, with the Denny corridor effectively part of downtown, that attitude seems laughable, but, too an extent, it still persists, and is part of why the 8’s night frequency sucks (the other part being the MLK section of the route dragging down productivity numbers). Truth be told, the Denny corridor has enough nightlife that I would even argue that area should have 24-hour service.
Metro was long doubtful about crosstown routes. The Central District had to push hard for years to get the 48. Metro assumed it wouldn’t get much ridership. But when it opened it became popular. So it’s a similar story as the 8.
When I first noticed the 60 it was already going to Georgetown and had the 9th Avenue detour. I was told the detour happened when the route was created: Metro wanted to run it straight down Broadway but the First Hill community wanted it to detour to 9th to serve more of the neighborhood. But if there was a short 60 before that staying on Broadway, then either that’s wrong or by “created” they meant creating a much bigger route.
asdf2 and Mike,
“Metro” is complex. It includes a range of actors from operators to the Executive. There have been and are a variety of opinions and thoughts about crosstown service. Route 8 has been changed many times.
Over time, crosstown service has improved significantly. All types of service are constrained: hours, buses, operators, rights-of-way.
This makes a lot of sense.
Another possible improvement may be truncating the 8 at Judkins Park Station instead of Mt. Baker. Shorter route = less cascading delays, and this seems justified even without capital improvements once the full 2 line is running.
The opening of Judkins Park is imminent and KCM has already long approved the Route 8 change to access the station.
Keep in mind that Mt Baker is also a layover spot for driver breaks. There’s nothing similar for Judkins Park. Had ST constructed a layover and drop-off loop the situation would be different. (I’ve pointed out many times that the land for such a loop can be found within the I-90 loop ramp if any agency wanted to build one.) The nearest ready-made Metro layover spot to Judkins Park will still be the Mt Baker TC after the 2 Line opens.
At this point, I would suggest not changing this part of the route until we see how riders vote with their Orca cards and bus on-off door counters (APCs). It will probably take through the summer of 2026 to see what riders do when the station opens.
I don’t think there is a layover there. I think the best you could do is layover with the 4 but that really doesn’t save much. You might as well keep going to Mount Baker (and connect to the main line) or not bother with that part of the 8 at all (which would save a lot more).
It is worth noting that the 8 will change fairly soon. Once East Link is done the bus will run on 23rd between Massachusetts to Yesler. This part of the route — the part that overlaps the 48 — will get longer. This leaves two relatively small sections where the bus has unique coverage (on MLK)*. In my opinion the southern section doesn’t add much from a coverage standpoint. Riders are always five minutes away from a bus stop and quite often there is less walking. The walks are fairly flat as well. In contrast the gap widens to the north. While I don’t think it is worth covering I can understand the value from a coverage standpoint. That is why I think the best option (if we need to cover that area) is to branch the 27. That would cover that northern section entirely. Overall that would result in a minimal loss of coverage (similar to the East Link change). The biggest difference would be where the bus is going and how often.
*I’m largely discounting the tail of the 4 when I write about coverage for the 8. I don’t think the tail should exist either (for the reasons mentioned).
There have been some visioning scenarios that has a Metro bus running up Beacon Hill on College Street. With Route 48 being mostly redundant south of Judkins Park, maybe that’s a possible strategy.
I agree with Ross that the routing of Route 8 south of Judkins Park (and Route 4 too) is less useful when Judkins Park Station opens. The indoor tennis center is really the only redeeming land use near the the segment’s bus stops.
Again, the more basic issue is the lack of a bus layover facility at Judkins Park. Had ST and WSDOT done something as simple as added a one-way, turn-around semi-circle off 23rd Ave across from the station entrance on top of the I-90 lid, both Route 8 and Route 4 buses (and all the other drop off and pick up maneuvering) could much more easily terminate there.
There have been some visioning scenarios that has a Metro bus running up Beacon Hill on College Street.
I’m skeptical we could afford the frequency necessary for that to very useful. If it ran every fifteen minutes it would work but that would mean you are spending more on the segment south of John than you are today. (Although you would still save money from combining the 8/11). You could run it every half hour but then it is largely just a coverage bus and there are plenty of coverage buses (like the 17) that are currently suspended. I think the main area where you want coverage is where people would be losing service (and getting to the 48 is cumbersome). That largely happens to the north, between Yesler and Madison. Branching the 27 would cover that area. It is an obvious branch. The trunk (Yesler west of MLK) should have a lot more frequency then either branch. There is a strong cultural connection between MLK and Yesler as well.
Reroute the 11 to Seattle Center same as 8 is good idea since the G-Line will go to Downtown
Yeah, excellent point. It isn’t just the connection to Link. The transfer from the 11 to the G (to get downtown) is excellent. It is basically a same-stop transfer and the G is very frequent (running every six minutes midday). It is also faster than the 11, making the transfer penalty minimal.
As the author noted, combining the two routes saves money which would allow the 8/11 to run more often. Riders on the existing 11 would see their bus arrive more often even if they didn’t sink the savings into running that. It is crazy that the 8 runs every fifteen minutes but the 11 is worse (it runs every twenty). The increased frequency for Madison Park riders would more than make up for having to transfer to downtown, especially since South Lake Union and the Seattle Center/Uptown are big destinations now.
The issue is Pine-Madison riders would only have a connection via the 12, which runs every 20 minutes currently.
The issue is Pine-Madison riders would only have a connection via the 12, which runs every 20 minutes currently.
The 11 also runs every 20 minutes. You really don’t gain much from the 11 from a network standpoint. It adds some service along Thomas/John but not enough to really be useful (it can’t be timed with the 8). It provides some one-seat rides but at 20 minutes they aren’t worth waiting for. Even then there are alternatives. At 19th you can take the 12. At 15th you can take the 10. At Broadway (10th) you can take the 49 or Link. At Bellevue Avenue you can take the 3. For just about all of its routing it is redundant yet the only time it actually contributes extra frequency is on the Pike/Pine “spine” west of Bellevue Avenue. Other than that it is merely an infrequent coverage bus serving Madison Park (and area that should have better frequency and would have better frequency if it wasn’t for this redundant routing).
The other obvious flaw in the routing is that the 10 and 12 branch too early. They should branch at Thomas. That would extend the section that has ten minute frequency. Right now someone leaving Kaiser has to guess which of the three bus stops they should walk to before potentially waiting 20 minutes at any one of them (just to go to the same place). If they combined the 10/12 they would go to the same stop every time and wait at most ten minutes.
Ross
On your point about the split of the 10 and the 12, the split being later would make it harder to transfer to the G, as the 12 shares a stop with it at 17th by the Trader Joe’s. That being said, the transfer at 12th is only a block so not the worst by any means. The combined frequency of the 10 with the 12 is worth it anyway, plus the 12 has held up my G line bus several times there by occupying the stall to load and unload. There would need to be some changes to the trolley infrastructure at the intersection of Thomas and 19th as there is no way for busses to turn left if they are trolleys. Ideally Metro should be making the 12 back into a trolley route.
Metro intends to make the 12 a trolley route again as soon as it installs the 1.5 block gap in the wire between 15th & Pine and 16th & Madison.
On your point about the split of the 10 and the 12, the split being later would make it harder to transfer to the G, as the 12 shares a stop with it at 17th by the Trader Joe’s.
I don’t see a lot of value in that transfer. If you are coming from Pike/Pine or the north end of 19th and are headed to Madison Park you could just take the 8. If you are going the other direction and headed downtown you can always just stay on the 12. The only really valuable connection I see is going from 19th to First Hill or 5th & Madison. But that is unlikely to account for many riders given the infrequent nature of that part of the 12. As you pointed out, those riders can always just walk to the other bus stop.
This is very different than some of the other transfers involving the G. For example the 48 from Montlake to downtown. Those buses are going to two very different places — I’m sure a lot of people transfer there.
Overall the transfer just isn’t worth worrying about. As you noted there is also a drawback to having the 12 delay the G. The same thing happens with the 8 and 11 but that seems inevitable and would be mitigated by combining the 11 and 8. (Right now they overlap which means 7 buses an hour. If the 8/11 ran every ten minutes it would mean less conflict while a huge improvement for both 8 and 11 riders.)
Metro intends to make the 12 a trolley route again as soon as it installs the 1.5 block gap in the wire between 15th & Pine and 16th & Madison.
That is another advantage of branching later. There is already wire on Thomas between 15th and 19th (for the 43). It would take very little work to branch later. I would just move the southbound bus stop closer to Kaiser. Take the bus stop just south of Denny and move it block to the north. Either close to the tiny park or a little bit farther south. That would mean a shorter walk to get to frequent service. But that could also happen later. Even if they did nothing it would be a much shorter walk to get to the point where the buses combine. You save about five minutes of walking (on top of whatever other walk you did).
I think it would be even better to branch at Aloha but that would require a lot more work. The city would have to approve running buses on Aloha and run wire.
As far as frequency goes:
If route 11 moves to Denny to combine with route 8, they should clearly have the same frequency so that the combined frequency is consistent.
If the change needs to be revenue-neutral, it seems the 8 should come down to the frequency of the 11, because the new routing would take slightly more time than 11’s current route on Pike/Pine, I think. It would still be a huge net-positive for the corridor though.
Is either route run with trolley buses?
Neither one are trolley routes. The 11 was dieselized in the 1960s, and the 8 never had trolley wire.
Route 8 has not been ETB. It runs under wire of routes 43 and the QA couplet. Today, what is the battery range of ETB with their new batteries?
If funding were available, the gap on Denny Way would be a good candidate as it has strong ridership, steep terrain, and service density. But today, we should not expect much help from the Trump 2.0 FTA.
In the AWV Enhanced Trolley Scenarios, of about 2008, it was proposed get overhead as Route 8.11. Dreamland scenario.
If route 11 moves to Denny to combine with route 8, they should clearly have the same frequency so that the combined frequency is consistent.
Agreed.
If the change needs to be revenue-neutral, it seems the 8 should come down to the frequency of the 11, because the new routing would take slightly more time than 11’s current route on Pike/Pine, I think.
You forgot about the elimination of overlap. If the two buses were combined the new route would be more frequent than the current 8 or the current 11 (at no extra cost).
There are a number of ways of thinking about but imagine the 11 was simply truncated in Madison Valley. So instead of going downtown it just went from Madison Park to Madison Valley. According to the schedule it takes 8 minutes to get between those two points. The entire route takes 27 minutes. So by truncating the bus you can run it more than three times as often. That is a lot more than the 8.
Now split the 8 into two pieces (at Madison Valley). So one segment goes from the Seattle Center to Madison Valley; the other from Madison Valley to Mount Baker Station. Now connect that truncated 11 to the western segment of the 8. Now you have a bus running from Madison Park to the Seattle Center (the 8/11). Once we make the connection the two segments have to have the same frequency. So we reduce the frequency of the section from Madison Valley to Madison Park to match the 8. That means the new 8/11 bus is running as often as the 8.
But wait, there’s more! We clearly saved a lot of extra service. That extra service can then go into running the 8/11 more often. Thus we end up with a route that is a lot more frequent to Madison Park and considerably faster for the core section of the 8.
All of this could happen while still serving the section from Madison Valley to Mount Baker section just as often as we do now. As a standalone route that doesn’t have much value though. There are a number of options but Al mentioned a previous proposal which would be to send it to Beacon Hill (via College Street) instead of Mount Baker. That wouldn’t cost much more — even if it ran at the same frequency as the current 8. Yet it would provide more value (in my opinion).
Personally I would just get rid of that section of the 8 (and put it into running the 48 more often) but I can see why the Madison Valley to Beacon Hill route would be appealing.
What? I think we miscommunicated. If the 11 moves to Denny as proposed in the article, both routes still exist and each has its own published frequency, even though the effective frequency on Denny is increased, right?
And it follows that the respective frequencies of each route should be equal, so that the combined frequency is consistent, right?
And because the 8 and 11 currently do not have the exact same frequency, at least one will have to change frequency to match the other, right?
Under that context, I was saying that a revenue-neutral move would be to decrease the 8’s frequency to match the 11, acknowledging that the combined frequency on Denny would still be much higher than the 8’s frequency today.
You seem to be talking about a different proposal whereby all busses from Denny go to Madison Park at high frequency. But Madison Park doesn’t need high frequency. Nor does MLK. The only part that needs high frequency is LQA-SLU-Capitol Hill. That’s why it makes sense to run two overlapping routes.
The concept in Metro Connects since 2016 consolidates the 8 and 11 into a Denny-Madison route. So Madison Park would have a one-seat ride to Uptown, and would have to transfer to Link or the G or another route to get to downtown.
Metro didn’t implement this in the RapidRide G restructure last September, but says it still might someday.
The different concept of rerouting the 11 to Uptown, so that the 8 and 11 would overlap from the western terminus to MLK and then diverge, is new to this article.
I could support lowering the 8’s frequency only if the 11 is rerouted to overlap with it on Denny Way. But we don’t have to lower it. It’s OK if the 8 comes every 15 minutes and the 11 comes every 20 minutes. That still gives a maximum wait of 15 minutes, and sometimes less. But Ross makes two good points that it would be more efficient to just have one Denny-Madison route, and that the 8’s MLK segment is excessive given the parallel 48.
However, I think the 11’s original Pine-Madison route was best: that best serves all the activity centers in Midtown, Capitol Hill, Madison Valley, and Madison Park. So the old 11 should be the most frequent route from downtown east, something else should serve Denny-John, and the 12 can be deleted or become a branch or extension of the 10.
If the 11 moves to Denny as proposed in the article, both routes still exist and each has its own published frequency, even though the effective frequency on Denny is increased, right?
You are right. I was confused with what the author is actually proposing. I assumed the proposal was for combining the 8 and 11. This is what is on the Metro Connects website (https://platform.remix.com/project/7063754e?latlng=47.61762,-122.29599,12.932&sidebarCollapsed=true) as route 1061. What he is proposing is instead two routes as you described.
That is basically a branch. The current branch would be named the 8 and the new branch named the 11. Yes, both routes would have to run opposite each other west of Madison Valley. This means that the new 11 and old 8 would have the same frequency (just as the 3 and 4 have the same frequency).
It isn’t clear what that would be. Routing the 11 to the Seattle Center would mean the 11 takes longer to go from end to end. But not much longer. It does seem reasonable that both branches could run every twenty minutes midday (the current frequency of the 11) while the main trunk runs every ten (at no additional cost).
There are clearly trade-offs. First consider the two branches. In my opinion the service of the 8 between Madison Valley and Mount Baker is not justified. It overlaps the 48 much of the way and when it doesn’t it is very close. Riders could walk to other buses. The only place where walking to the 48 is cumbersome (and steep) is near Madison and at that point riders could just walk north to catch the 8 or RapidRide G. In contrast Madison Park should have bus service. It doesn’t take that long to get there and ridership is pretty good. Unlike MLK, riders they really have no alternative. I think it is underserved with 20 minute frequency.
If we implemented the two routes (the 8/11 branch) then Madison Park riders would likely complain about losing their one-seat ride to downtown and not seeing an increase in frequency. MLK riders would complain as well because they would be losing frequency (although I’m OK with that, obviously).
There is also less flexibility with a branch. Running the core of the 8 every 12 minutes would be an improvement over current routing. But that means running each branch every 24 minutes, which is awkward. I would be OK with service every half hour on MLK but not with service every half hour to Madison Park.
Now consider some of the alternatives. My first choice would be to focus on frequency. I would send the 8 to Madison Park and run that bus as often as we can afford. I would not serve MLK north of Mount Baker. I think it is possible you could run the new 8 every 7.5 minutes midday although that might be stretching it. If nothing else you could run the bus every 10 minutes and improve frequency later into the evening. As the author pointed out, evening/weekend service on the bus is poor. Even at every 15 minutes it is much better for Madison Park riders. This helps make up for the lack of a direct bus to downtown. Running the bus every 10 minutes may be overkill for Madison Park but it doesn’t cost much to go between Madison Valley and Madison Park because it is pretty fast. If we can afford to run the 8 every 7.5 minutes we could always turn back half the buses at Madison Valley. There is more flexibility with this routing.
Another variation is to run that bus from Madison Park to the Seattle Center and backfill service on MLK with another route. I see two options for a backfill. One would be what is proposed in Metro Connects as the 3997 (although I would skip the last bit over to CHS). That serves Beacon Hill instead of Mount Baker.
Yet another variation would be to be to branch of the 27. One branch goes on the current path and the other would head north from Yesler to Madison Valley. This provides coverage on that part of MLK while not redundantly overlapping the 48. It has the same weakness as any branch but it is shorter than running from Madison Valley to Mount Baker (or Beacon Hill).
There are a lot of trade-offs but I err on the side of side of minimizing overlap. I feel like the biggest weakness in our system — by far — is just bad frequency. Bus stops that have 20 minute service should have buses arriving 15 or 12. Stops that have 15 should have 12, 10 or 7.5 minute frequency. Just getting rid of the 11 and sending the 8 to Madison Park would be the best way to increase overall frequency while still covering the essential areas.
The SODO portion has so much cross traffic, that I don’t think red paint will fix its reliability problem. That portion would be a great gondola line with 8 seconds headways instead of 12 minutes and a grade separated line would be far more reliable.
SODO? Do you mean “SLU”?
If course