Dangerous By Design: MLK Way’s Current Conditions

87 Replies to “Dangerous By Design: MLK Way’s Current Conditions”

1. At-grade trains on MLK and in Bellevue were such a mistake. It’s sad that Sound Transit took the cheap route in these areas. And now we’re talking about gold-plated extensions to Ballard and West Seattle.

   1. Sounds like we learned our lesson. I’d count that as a win, yeah?

   2. Surface running has its place just like streetcars have their place. But the advantages of being completely grade separated are huge because it goes along with being automated. Having an automated line means smaller trains running more often. This is cheaper to build and better for riders. To a certain extent you want one extreme or the other. Maybe a streetcar/light rail system with lots of surface running to keep down costs. Some grade-separation, but not a lot. Not necessarily faster than buses but with more capacity. Or you want the other extreme — a completely grade-separated line that is automated. Such a system should be geared towards quality, not quantity. In other words, lots of stations in the urban core (even if it doesn’t stretch out to the suburbs very far). Add bus-rail connections for the express buses from the suburbs and you’ve got yourself a really good system. For a city like Seattle bus/rail integration is crucial, even in the city (even if we built everything really well and spent a fortune it is unlikely that trains would carry more riders than the buses).

      We kind of stumbled into our system. We were never quite sure what we wanted to build. Is it a metro? Sort of. It has plenty of tunnels (like a subway) and lots of grade-separation. But other than downtown itself it has terrible stop spacing. There aren’t enough stations. Is it light rail? Again, sort of. Down Rainier Valley it feels like it. It runs on the surface in SoDo but that seems like like an anomaly than intent. But if you look at the overall system it doesn’t feel like light rail. It is, technically the most expensive light rail system ever built because it really doesn’t operate like light rail. Is it commuter/regional rail? In some ways, yes. This gets back to the stop spacing. There are miles between stations in some areas. Yet unlike commuter/regional rail, they didn’t take advantage of existing tracks. Thus it is this weird hybrid. It is as expensive as a real metro but far less effective. It is light rail but really expensive. It goes for miles and miles outside the urban core — just like commuter/regional rail — yet again, it is extremely expensive. It feels like it was created by a committee that really didn’t know much about transit. That’s because it was.

      1. When Sounder was rolled out, the tracks were upgraded, I believe. That is to say, they were ripped out and replaced by more modern tracks. I don’t know if the right-of-way was wide enough to just build a new set of tracks next to the freight tracks. Or did BNSF take advantage of ST’s weak bargaining position and only give ST the option of replacing the freight tracks?

         Did any cities block additional track lines?

      2. This must be satire. The Sounder tracks weren’t upgraded wholesale. The reason Sounder was included in ST1 was because it could be implemented quickly on the existing tracks, so it was seen as too big a low-hanging fruit to miss out on. There have been spot incremental track improvements over the years that both WSDOT and ST have been involved in, but the entire corridor wasn’t replaced for Sounder, and no additional tracks were added. That’s still a live issue: the state could fund adding a track in the South Sounder corridor to allow more passenger trains without conflicting with freight, but so far it hasn’t done so.

      3. As shocking as the current accident numbers are for MLK, it was worse before Link was built. Three lanes in each direction, nominal speed limit of 40 mph, very few crossing signals and very little pedestrian activity along the street resulted in frequent fatality accidents. As a teen, I worked in the strip mall at Graham Street and had friends in the apartment complex behind the mall. MLK was very dangerous and it was hoped that building the Link infrastructure would make life safer along the corridor. Today, it’s amazing to remember that we spent billions of dollars on the Rainier Valley segment and, except for a few bike lockers and racks, didn’t upgrade the bike infrastructure at all. There were some pedestrian improvements included, but they haven’t resulted in a safe environment.

      4. @GuyOnBeaconHill,

         “…….it was worse before Link was built.”

         You hit the nail on the head. MLK today is actually far safer than it was before Link was installed, with fewer accidents, injuries, and fatalities. Ya, it isn’t zero, but no street with a mix of pedestrians and cars will ever be 100% safe.

         And there is nothing “shocking” about the accident rate on MLK. MLK is actually far from the most dangerous street in Seattle. Even the stretch of Rainier Ave just south of I-90 is more dangerous than MLK.

         Don’t get me wrong. I’m all for safer streets. But if you want maximum bang for your buck, look elsewhere.

      5. Re: accident rate before:

         If we just elevated the tracks, we’d eliminate even more and speed up the trains through there.

         I would be surprised if we couldn’t squeeze in another station and still get trains from downtown to SeaTac in the same amount of time or faster.

   3. The segment in Bellevue is way better, I’m not sure they’re even comparable. If MLK was a single lane road with crossing guards it wouldn’t have nearly as many crashes.

      1. The segment in Bellevue is also much, much lower-traffic than MLK. It’s absolutely better designed, but that helps a lot.

   4. SoDo also has surface running trains. I’m not sure how many accidents occur there.

      1. I believe not nearly as many. Surprisingly (to nobody except SDOT and potentially SoundTransit), crossing gates work.

      2. Link is surface-running in SODO but doesn’t run parallel with general traffic at any point – only parallel to the busway. That, plus the heavy-rail-style crossing arms, seems to greatly reduce collisions there.

      3. Based on the SDOT data, SODO has had 32 crashes. Even on a per mile basis, there are significantly fewer crashes in SODO than on MLK.

      4. No left turns, no serious car crashes.

         If Link were going to be made reliable, Mlk needs 3 rights to left.

      5. @Psf
         For 3-rights-to-left to work, there needs to be a reliable grid of parallel streets a block back from the main drag, which unfortunately isn’t available in many stretches of MLK.

      6. For 3-rights-to-left to work, there needs to be a reliable grid of parallel streets a block back from the main drag, which unfortunately isn’t available in many stretches of MLK.

         Like where? If you banned left turns, what streets couldn’t you access? I see places where you would have to drive farther but I don’t see any places you can’t actually access.

      7. @Ross Bleakney This isn’t to address the question of what you couldn’t access, merely the three-rights-for-a-left thing.

         In some cases you can’t realistically do it, like the turn from southbound MLK to S Orcas eastbound (S Juneau is cut off from MLK by an apartment complex). Similarly if you’re trying to go from MLK northbound to S Myrtle westbound and access the seafood market or all those apartments up there, thre’s a whole bunch of misleading stub streets that look like they should go through, but end in cul-de-sacs or near-miss grid-breaks.

         In some cases you CAN do the three-rights-for-a-left thing, but there’s no reliable, and people who don’t live in that area won’t be able to navigate the non-gridded streets with confidence that they’ll end up back at their desired intersection. For instance, at the southbound MLK turn to eastbound S Graham, you go past it, hit S Morgan St and have to go like four blocks west to find a reliable way to turn back north and hit S Graham again.

         Maybe the answer to the latter category is to have a series of signs guiding people through the neighborhoods to get around to a proper crossing, but of course then that’s a question of increased commercial traffic through residential areas.

         It could be that so long as the tracks are at street level, there just isn’t a band-aid to the problem big enough to fix it.

      8. Here are some left turns off of MLK that can’t be replaced with (only) 3 right turns:

         – Southbound to Edmunds
         – Southbound to Kenyon
         – Both directions at Merton
         – Both directions at Norfolk

         Many other replacements would require sending a lot of traffic down narrow residential streets or large loops adding mileage (and increased risk of collision) which might negate the benefit of cancelling left-hand turns.

      9. @Nathan Dickey,

         * Left turns from southbound MLK to eastbound Edmunds St are currently banned.
         * Left turns from southbound MLK to eastbound Kenyon can be replaced by three rights starting from Thistle. It’d be ugly, as @ctishman says, but possible.
         * Merton and Norfolk, yes, would be a problem.

      10. On further inspection, it turns out southbound left turns onto Kenyon are prohibited as well.

      11. Here are some left turns off of MLK that can’t be replaced with (only) 3 right turns:

          Who cares how many turns it takes. The point is you can still get there.

          – Southbound to Edmunds — https://maps.app.goo.gl/tae5Lj5A6qBG1fAv6
          – Southbound to Kenyon — https://maps.app.goo.gl/hvit7gNZjdUFP3767
          – Both directions at Merton — https://maps.app.goo.gl/yf2XcoQB9LHFQkPr7 (you get the idea).
          Orcas — https://maps.app.goo.gl/AsHybghpvyFFtHT1A

          Yes, signs would help. If only there was a way to navigate via some sort of computerized system that the vast majority of drivers have access to. Sorry for the snark but I think people can figure it out.

          But yeah, this would mean more traffic on the side streets. But you could pick and choose how you want to handle that. For example look at the Orcas example again. Assume for a second that 32nd is cut off somehow (via a sign that says “local access only” or something more physical). There is an obvious alternative: https://maps.app.goo.gl/AsHybghpvyFFtHT1A. Obviously this is a lot less convenient. That is life in the big city. There are a lot of things we should consider — driving convenience should be at the absolute bottom of this. Not only because it is generally less important than things like safety but also because it ends up inconveniencing someone else.

      12. It is also worth noting that not everyone is starting there trip from someplace on MLK. They just happen to be using MLK. For example Orcas was mentioned. You can access Orcas from Rainier or Beacon Avenue. So if you are up north you just use Rainier (https://maps.app.goo.gl/DvjbSk9DNfTMohBQ9). If you are up by the VA you go via Beacon Avenue (https://maps.app.goo.gl/E9wSTo345hJaqmHPA).

      13. Rainier Valley is unusual in that the two through streets are diagonal. This makes it more challenging to do things that work in a straight street grid. Even if it’s physically possible to do some three-right combinations, it may be unreasonable to expect drivers to know the block-by-block conditions of which intersections they can do it at.

      14. Certainly SDOT could look to eliminate some traffic movements. However the big drain on time is the crosswalk time. It can take about 45 seconds to cross MLK in some places. No one else can do much during that interval. In contrast, the time allocated to left turns is pretty minor — so eliminating them won’t do much. It would help if SDOT would fix broken loop detectors though; way too often everyone from Link trains and 106 buses to pedestrians and bicyclists to local traffic ends up waiting 30 seconds for a green left turn light when there are no cars turning left!

          The Link station crosswalks are mostly at major streets too (Alaska, Othello, Henderson). That mixes Link riders with traffic doing all sorts of movements as well as Link trains slowly moving into or out of a station. It’s not a very smart situation to mix everything at the same intersection.

      15. The entire SODO surface section is separated from the busway by Jersey barriers except at the intersections. For that reason it can’t be compared to MLK.

   5. At grade can be done much more safely than ML King. The Chicago L has a few remaining segments of median running. They’re extremely safe, especially compared to ML King.

      Sure, elevated or underground everywhere would be great. It also means expensive. Some places, even in Chicago, it’s too expensive to build.

   6. It’s hard to correct stupid initial rail design. Technically, the raw design was legitimately “safe” on paper — but those in charge were trying to be cheapskates and did not seem to care about the effects of the layout on the environment that existed. Keep in mind that there were two large public housing projects near Columbia City and Othello stations that were vacated about the time Link construction started.

      After LA got the Blue line (now the A Line segment from Downtown LA to Downtown Long Beach) , those folk figured out things like elevating at stations when trains move slower or in places where major streets cross the tracks. They built the Exposition line (now the E Line between Santa Monica and the Blue line tracks) by applying lessons learned.

      I’m hoping that there is some review of how other light rail systems have designed for safety when there is a crossing. It seems ST and SDOT keep trying to design light rail or later to solve the safety problem doing whatever they can envision — rather than to first canvas and see how other places have done it.

      There seem to be strategic ways to modify the corridor to provide some separations or crossing treatments but not spend several billion and wait decades for a funding opportunity to fully separate the line. Other places have used different ways both successfully and unsuccessfully. (I’m reminded of how San Francisco rashly put the Central Freeway ramps to land at Market Street and its streetcar tracks, instantly creating SF’s most accident-prone intersection.)

   7. ST has had different attitudes to at-grade alignments over the years. In the first phase when MLK and SODO were designed, ST was emulating previous American light rails (Portland, San Diego, San Jose) that were 90+% surface to keep capital costs down. It said it couldn’t justify elevation or tunneling in flat areas without physical barriers.

      The next segment after MLK was Tukwila. The City of Tukwila objected to the surface alignment on Tukwila Intl Blvd saying it had just beautified the street, and it objected to taking a corner of Southcenter’s property. Since Tukwila had the power to withhold construction permits, ST redesigned it elevated. Because it had to go over all the highway infrastructure and couldn’t have steep inclines, that forced it all the way to just south of Rainier Beach.

      Between then and ST2, there developed a consensus among ST and the cities that all future segments should be grade-separated by default. (This includes segments in freeway ROW that are technically on the surface but don’t have level crossings, so they don’t have the collision issues.)

      ST2 was completely grade-separated at one point. But then the City of Bellevue wanted a tunnel in front of City Hall, and begged ST to economize elsewhere in East King to pay for part of it. That’s what led to Bel-Red’s lowering to the surface. Bellevue also begged North King to take on the cost from Intl Dist to Judkins Park station, which it agreed to. Bellevue itself paid for half the tunnel’s cost. But that’s why Bel-Red is surface. It wasn’t ST cheapening out; it was ST ceding to Bellevue’s request to add an unbudgeted tunnel. ST usually cedes to cities’ wishes as much as it can.

   8. Don’t you think it’s the driving behavior that needs to change?
      How many of those accidents occurred without moving violation?

      1. Perhaps speed cameras could be installed along the length of MLK. That would be a great starting point but a complete non-starter in Seattle. Ditto for added police presence.

2. 🌈🚈Imagine a world where there’s an elevated or cut-and-cover grade-separated guideway, and Link trains run the entire way without ever hitting a car🚙🌈

   1. Thanks for this series, Michael!

      Are there statistics on crashes involving trucks, crashes involving buses, and crashes involving scooters?

      Also, do the train/car crash statistics indicate the direction from which the car was coming?

      Are there any school zones along MLK, and if so, is the data broken out in those sections?

      For context, what is the train’s speed limit on MLK?

      1. 35mph limit on mlk

      2. Trucks, buses, and motorcycles are grouped with cars a “motor vehicles” in the data. There have been four reported crashes with scooters.

         The car/train crashes don’t indicate direction. That would be nice to have to see if the cars are turning left or going straight through the intersection.

         I don’t think there are any school zones on MLK. There are several schools nearby, but I don’t think their school zones reach MLK.

         As D M said, it is the train speed limit is 35.

      3. I am curious about the time of day too. It feels like many crashes are during daytime hours.

         I’m also curious about how many crashes are a result of illegal movements — whether it’s a car running a red light, a bicyclist ignoring street rules or a pedestrian wandering on the tracks but not paying attention or having suicidal motives. Designing to minimize conflicts seems a bit different than designing to make illegal or stupid behaviors much less common.

      4. Thanks for the answers, Michael!

         Should we be surprised traffic parallel to the train is going as fast as the train, or even faster?

      5. Crash reports are digitized very well on these days almost in every state. You might be able to submit a public record request through Crash Data WSDOT to get a detailed crash record with as many attributes as you can imagine. They can pull data along a road and date range of your choice, but I don’t know how easy to request through WSDOT Public Disclosure Request Center.

         Time of day and weather are among the easiest attributes to process, directionality could be tricky because crashes often involved multiple vehicles and it is not 100% consistent whether the responding officer is documenting the direction after collision or direction of travel in the event of an overturned crash.

      6. @Brent

         Not at all. In fact, the signal timings encourage drivers to drive alongside the train at 35mph. Whenever the train gets a ‘green’ light, drivers on MLK do as well. If a car stays in Link’s shadow, it can cruise down MLK and only hit a red light when Link is stopped at a station.

      7. @HZ
         The data used for this series is from SDOT (see the note at the bottom of the article) and it is quite detailed. I’m sure WSDOT has similar data available for their roads and highways.

         It sounds like another post that dives more into the crash data may be warranted in the future.

      8. Having some 4-way stops while the train gets the vertical white line to proceed might strike some as weird, but we have some intersections downtown that have all-way-pedestrian-crossing in their cycle.

         Of course, crossing the tracks has to be disallowed at that time, so this could only be done at Columbia City and Othello. But hey, we may as well take advantage of the side platforms.

         I’m still skeptical about whether the 2-Line-style gates improve pedestrian safety, or could turn into a death trap. I suppose the implementation details matter.

      9. @Brent White, the gates all open outwards, so I’m pretty confident they’re good things. The only problem I can imagine is if they get stuck, but I haven’t seen that – and anyway, there’re two of them on each side of the tracks.

3. Hi Michael,

   Love this! Looking forward to the rest of the series.

   I want to make sure you have a copy of the road redesign alternatives analysis Toole Design did for SDOT back in 2020 that has been sitting on a shelf collecting dust ever since waiting for a mayor to support it: https://drive.google.com/file/d/1YzeLTZFSYJVhKDE_WEuaZ9BcsvkBkW23/view

   1. Thanks Gordon! Yes, we discuss that report in tomorrow’s article and have a link to it.

   2. The problem I see with the Toole report is that it does not narrow MLK and reduce the time it takes to walk across the street. The objective of the report is to design a PBL but not to reduce pedestrian-car, pedestrian-train, train-car or car-car crashes. There may be reductions as a conceptual impact but it’s not the primary objective of the design.

4. I’ve mentioned this before, but I think that the excessive width of MLK affects the impatience of people using all modes since the signal cycles are so long. I hope that you look at that aspect of the street design.

   Solutions that make the street narrower to cross are going to reduce the long waits and thus probably improve safety for everyone.

5. Given how SDOT has erected bollards and fencing on so many streets in an effort to calm traffic, it’s amazing to me that most of the Link tracks have no visible barrier between them and the traffic lanes. The area around Oregon Street has a low fence in the middle and that intersection has a much lower accident rate than the other intersections do.

   Of course trains run on guided tracks so low fencing does not disturb the operation. A treatment would be inexpensive. I wish ST and SDOT would look into installing low fencing on the corridor to create a better visual awareness.

   1. I recall how someone managed to wrap a stolen car around a power pole between the tracks, causing thousands stuck in jam-packed trains to miss the Super Bowl Parade.

      One would be tempted to ask for the extreme solution of fencing off the track areas between intersections, but the simpler (albeit not necessarily cheaper) solution would be to move the poles to not be in the middle of where streets let out.

6. How feasible is it engineering-wise to still upgrade MLK to elevated rail, ignoring costs for now? To my untrained eye, it does seem like there is space to construct a new set of elevated tracks to the left or right of the existing tracks. Once they are complete, connecting the tracks would be a challenge. Then with Link running on the side, aftering removing existing at grade Link tracks, MLK Way can just be a normal street with some train pillars next to the sidewalk, like near Angle Lake. How much disruption would current 1 Line riders suffer for such a project?

   1. I’m not sure where you’re seeing that space? Looking at Google Street View, which lines up with my memory of the last time I rode there, there’s extra space in the median sometimes but not most of the time.

      Or are you suggesting to construct the elevated tracks on top of one direction of travel? Say, close the northbound lanes, make MLK a two-lane street in the existing southbound lanes during construction, and then eventually tear down the in-street track and build new travel lanes there? That’d be possible, but I’d argue not a good use of money.

   2. IF there was money, some strategy could be developed to gradually add aerial tracks. It would almost certainly require permanently closing a traffic lane in at least one direction to drop the ground supports in. But as you say, the aerial structure would allow for the traffic lanes to be put next to each other and make the street easier and safer to cross.

      The big challenge is not only funding but also in phasing any transition like building new stations or tying back into existing tracks even if all the neighbors supported it.

      1. @Al S,

         “…the aerial structure would allow for the traffic lanes to be put next to each other and make the street easier and safer to cross.”

         Elevating Link on MLK would reduce safety and increase the accident, injury, and fatality rates. And the reason for that has to do with the types of accidents that are occurring on MLK.

         The vast majority of accidents on MLK don’t involve Link at all. Most accidents are vehicle-vehicle and vehicle-pedestrian accidents. It’s not even close.

         So anything that increases vehicle-vehicle or vehicle-ped accident rates will have far more of an effect than eliminating the fairly rare Link involved accidents.

         And adding elevated structures, with their obstructed views and reduced sight-lines are guaranteed to increase accident rates. You simply can’t reduce visibility and expect anything but more accidents, ore injuries, and more fatalities.

      2. @Lazarus, two things:

         1) Elevated structures will not reduce visibility. Have you tried looking around Angle Lake station near the elevated tracks? They are so high up that they do not affect visibility whatsoever. You could argue that elevating the tracks actually increases visibility because the two directions of car lanes are no longer separated by 30 feet of tracks.

         2) To be honest, the injury rate of MLK Way is high, but comparable to other arterials in the area. According the the High Injury Network statistics, MLK Way has a score of 9.5/mile, Rainier Ave has a score of 10.0/mile, Aurora Ave has a score of 11.6/mile, and N 85th St has a score of 12.7/mile (lower is better). So the most pressing issue unique to MLK Way is definite the incidents involving Link, delaying transit for thousands of riders once a month, though we should certainly be trying to solve the vehicular collisions for all high injury roads.

         Street view near Angle Lake: https://www.google.com/maps/@47.4270697,-122.2977757,3a,75y,341.93h,88.29t/data=!3m7!1e1!3m5!1sOjySGOhmbCraQt7zX46qkA!2e0!6shttps:%2F%2Fstreetviewpixels-pa.googleapis.com%2Fv1%2Fthumbnail%3Fcb_client%3Dmaps_sv.tactile%26w%3D900%26h%3D600%26pitch%3D1.707861097367072%26panoid%3DOjySGOhmbCraQt7zX46qkA%26yaw%3D341.93001108494786!7i16384!8i8192?entry=ttu&g_ep=EgoyMDI1MTAyMi4wIKXMDSoASAFQAw%3D%3D

      3. @ Lazarus:

         “And adding elevated structures, with their obstructed views and reduced sight-lines are guaranteed to increase accident rates. You simply can’t reduce visibility and expect anything but more accidents, ore injuries, and more fatalities.”

         I consistently and frequently point out (and you are ignoring) that any strategy to take Link off the surface median needs to be accompanied by a significant narrowing of its street width. Narrowing the street from 90 or 100 feet to less than 60 would vastly improve the field of vision for any user — whether driver, pedestrian or bicyclist. Signal cycles can be shorter too, reducing the tendency for impatient people to endanger themselves and others by breaking laws.

         While going to an aerial track won’t by itself probably have much impact on safety, getting the street width down would greatly improve visibility within one’s line of sight.

         It’s all about the resulting narrower street width!

         Today the street is so wide that it’s almost impossible to visually identify everything going on at an intersection. I cross MLK frequently and I can say that it’s almost impossible to visually scan an intersection to see everything that’s happening. One just has to take their chances today.

      4. What if instead of narrowing the roadway, we kept the track space as an enormous median and turned each intersection into two synchronized intersections like Beacon Ave at Columbian way? Then you only have to worry about one half of MLK at a time, much more manageable for all.

   3. I’m guessing one of the other articles will focus on solutions. Based on what I remember, it would be much cheaper to bury the line instead of elevate it (roughly a billion dollars). The biggest issue is that it would be very disruptive.

   4. the most recent study had the cost of elevating the link on mlk way around 2 billion dollars

7. “I’m hoping that there is some review of how other light rail systems have designed for safety”

   How is MAX’s track record on level crossings? How does it mitigate them? The original vision of Link was like MAX but with the existing DSTT and extending the tunnel to the U-District because of the hills and Ship Canal. It would have gone surface from Intl Dist skirting around the northeast side of Beacon Hill, down MLK, and continue surface to SeaTac and Tacoma. That would have meant a lot more level crossings. The MLK surface alignment was a legacy of that thinking.

   1. Most of the places where MAX is running in street medians are on streets with lower volumes than MLK has. They tend to also be much narrower than MLK.

      MAX is often running adjacent to wide, high volume roads rather than in the median.

8. One thing that doesn’t get discussed much is how ST may have a train car overcrowding problem on MLK sometime in the future. The opening of Federal Way Link in just a few weeks will add more riders to the segment. What kinds of excess capacity left after opening remains to be seen. I don’t expect a Tacoma Dome Link Extension to add many more Link riders on MLK beyond that — and station area development at stations from Federal Way northward will likely add more riders in years to come than that extension likely would.

   It’s train car overcrowding that could tip the scales in designing a long-term solution. If 1 Line trains are too crowded leaving Downtown Seattle in the Beacon Hill tunnel, there would be heightened regional interest in running more frequent trains and thus replacing or augmenting the current line with something separated. Otherwise, North King would be paying the bill all by itself.

   Since ST is emphatic about needing DSTT2 and their forecasts in the past have shown more crowded trains through the Beacon Hill Tunnel per train car than even through Downtown Seattle, overcrowding may be a problem. I don’t think it will be — but it’s worth mentioning as a risk.

   By next summer, ST will have a robust, connected two-line Link system and will have a real-world picture of how much excess train rider capacity exists in the system. I suggest that ST not start any new major projects until an analysis of crowding happens with next summer’s actual ridership as a basis. It may or may not add another reason to do something.

   1. The Ballard-DSTT2-Tacoma Dome plan increases the line’s frequency to 6 minutes.

   2. I really doubt that Federal Way Link will lead to capacity problems in Rainier Valley. It would if they eliminated the express buses (and Sounder) but there just aren’t that many people going from Federal Way (and the other stations) to Seattle. Way fewer people are commuting long distances. Most of the increase will be to SeaTac (from the south) as people switch from buses like the 574 and A Line. We are running four-car trains now and while they aren’t running every six minutes they could in a pinch.

      1. @ Ross:

         Yeah I expect some new riders from FWLE next year but I doubt it will be huge, similar to your expectation. As I mentioned, additional riders beyond that will probably mostly come from station area redevelopments than the extension openings.

         And ST needs to revisit their overall ridership badly. They probably don’t want to do that since Ballard LE and Everett LE EIS’ are in process. But once those are on the street this next year ST should recalibrate their models to 2026 data and modify their forecasts of expected station area development.

         ST has really backed off providing public updates to their ridership forecasts even to the point of increasingly scrubbing them from extension web pages and fact sheets. It makes us armchair observers uninformed about what to expect. Unfortunately, it also makes it hard to comment on how best to spend the limited remaining ST3 funds strategically. It’s a de facto way to keep legitimate data from guiding how to move forward with ST3 system expansion — probably so that backroom discussions and deals with land owners and developers can continue unquestioned. (Sorry to be such a cynic but it’s very disturbing how the ST Board is now rejuggling many billions of dollars without telling the Board and public what the forecasted Link ridership is. Very disturbing.)

      2. As someone who lives in the south end and has thought through the travel-time math, I have to disagree that backtracking downtown to catch express buses is competitive with taking the train, especially during peak hours, and certainly for anyone going south from Beacon Hill Station or further south.

         All that said, I expect new transit ridership induced by the Federal Way Station connection will be larger than current express bus ridership going that way. Backtracking downtown is a huge deterrent to making transit trips to the distant south for any reason other than employment.

         Potential ridership may be lost if ST does not take the Federal Way transfer connection seriously, or waits too long to do so.

   3. Even from Tukwila International Blvd, the advantage Link had over the buses to downtown Seattle is only about 10 minutes. At Federal Way, the time advantage is reversed.

      Once you do get overcrowding on the trains, you then have to ask what is the better investment: ML King improvements or something else? Making extremely expensive changes there (as opposed to less expensive and needed safety changes) really doesn’t buy you that much. $1 billion could buy a lot of other stuff, that might be better, depending on the desired result.

      1. $1 billion could buy a lot of other stuff, that might be better,

         Burying the line would actually get you quite a bit. You would transform the street. It could be narrower *and* have bike lanes and maybe even BAT lanes. It would mean the trains could run more frequently and faster in Rainier Valley. If you added overpasses in SoDo you could automate the line. This would save quite a bit of money on operations. So yeah, there might be better projects we could spend our money on but this would be one of the better values in ST3 (if not the best value).

      2. Is the requirement that a line be completely grade separated to automate still the case? I find it hard to believe that we can have self driving taxis on city streets, but a self driving train on a fixed route is infeasible because of a few at grade intersections.

      3. “I find it hard to believe that we can have self driving taxis on city streets, but a self driving train on a fixed route is infeasible because of a few at grade intersections.”

         Trains take longer to stop. Plus, people stand on trains so slamming on brakes is more dangerous. And trains can’t steer around track obstructions as they run on fixed tracks.

         But the technology is improving rapidly so it seems likely in the next few years. It will probably start with lower speeds and get faster over time.

      4. @Al S,

         “Trains take longer to stop.”

         What does this have to do with self operating trains?

         Self driving trains have nearly the exact same stopping distance as human operated trains. And often they actually have better (shorter) stopping distances because of reduced brake application time and better brake modulation. Stopping distance is absolutely not the reason we don’t have self operating trains.

         But there is a reason, and it can be summed up in one word, “capitalism”. Or maybe more accurately, “profit”.

         Self driving tech is being developed by private companies for profit, and there is simply more profit to be made in automating the roughly 1.5 B worldwide cars than there is in automating the relatively minuscule number of urban light rail trains.

         “Plus, people stand on trains so slamming on brakes is more dangerous.”

         Total BS. Modern LRV’s are designed to keep north acceleration and jerk (second derivative of velocity) within manageable limits. Modern transit vehicles don’t behave the same as your dad’s old jalopy with its mechanical brakes.

         But hey, there are some demonstration programs starting with self driving LRV’s, so eventually we will get there.

      5. I think you could automate Link right now but the trains would go slower. It is common for most automated cars to go slower so it would be the same idea. At the same time you have to allow longer stopping distance. Think of an unidentified object. The computer isn’t sure if it is a bag or a person. Moments later it reveals itself — it is a person. With a car you would have more time to stop. What this means from a practical standpoint is that in both cases the vehicle would slow down when it isn’t sure what the object is but it would have to slow down sooner if it is a train. That being said I guess Siemens is working on it (and has been for a few years): https://www.mobility.siemens.com/uk/en/portfolio/rolling-stock/trams-and-light-rail/autonomous-tram.html. So it is a possibility.

         But this still doesn’t give you the full advantage of being grade-separated. It reduces operating costs which means that it could lead to better frequency. But headways in Rainier Valley are limited in part because of the crossing traffic. The city won’t allow trains every five minutes. There are other issues. As D M mentioned (https://seattletransitblog.com/2025/10/28/dangerous-by-design-potential-improvements-for-mlk-way/#comment-970397) with full grade-separation you can implement tried-and-true technology that improves reliability, speed and frequency.

      6. “What does this have to do with self operating trains?”

         My original comment response was wondering about driverless rubber-tired taxis compared to driverless trains — not human-driven compared trains to driverless trains. Your apparent desire to make me look stupid kinda backfires.

9. I think some signal timing adjustments at the intersections with Link stations could go a long way. Current signal timing is optimized for crossing all of MLK in one go, even though most ped crossings are only crossing halfway to access the Link station.

   Furthermore, there should be scramble phases so people can cross both east and north at once, and also, the signal should permit ped crossing during left turn phases when there are no conflicts.

   I’ll have to think a bit how to sequence everything, but it seems like it should be doable.

   1. Yes this is one of the curses of all Columbia City station riders. You quickly realize how idiotic the signal timing is. A big contributor is the lack of ped signals at the median at this station. As a result, the half-crossing during left turns is impossible to signal. So hundreds cross against the solid red hand anyways. If a driver were to hit a light rail rider performing this maneuver, chances are they would be on their own in terms of liability, which is really unfair to riders, imo. As far as I know SDOT still hasn’t committed to upgrading the ped signals though.

10. Remember this is what the people of the area wanted when they were all at the meetings telling ST what was best for them. No consideration or engineer mindset for creating anything that actually works. Most Americans haven’t been to cities which have working high capacity transit and road it consistently. Less then 1% would build a proper system ask them to draw a bike and see if it would be a working bike in real life.

    1. It doesn’t help that ST usually deliberately conducts meetings staffed almost entirely with PR hosts but no senior engineers to discuss technical things. Then any technical questions or suggestions seem to never get recorded or addressed in meeting summary documents.

11. Light rail in the median of avenues still works for many other cities. Many european cities and even some american ones. I fear forcing light rail to never be built at grade means ( built elevated or underground only) means building only next to freeways for the foreseeable future making transit much less useful.

    For elevated alignments, almost every chance to build it Sound Transit has shown it is unable to build it. For lynnwood link instead of on aurora ave elevated instead the alignment was on i5. Same for federal way link, instead of pacific highway, they chose to build next to i5.

    Even for this central link section itself on mlk way, sound transit itself did offer to build it elevated but (some of) the community desired it to be at grade instead.

    https://archive.seattletimes.com/archive/19990127/2940938/the-light-rail-transit-plan—-on-wrong-track—-many-people-who-live-and-work-along-martin-luther-king-jr-way-south-believe-the-current-transit-plan-unfairly-favors-residents-of-north-seattle

    > In 1992, the city persuaded planners to put the line through the Rainier Valley. Planners first discussed a subway through Southeast Seattle, but Johnathon Jackson, a Sound Transit manager for community relations, said local residents supported surface or elevated tracks to spur development. … Those were the two options included when the transit plan was approved by voters in 1996…. In 1997, Jackson said community groups and residents pushed to get the elevated-track option eliminated. Rainier Avenue business owners got the route shifted to the wider and less developed MLK Way.

    for underground alignments, honestly sound transit doesn’t know how to tunnel cheaply. It can’t tunnel any cheaper than a billion per mile.

    Basically if we can’t build at-grade light rail, the seattle region is only going to be building transit next to freeways for the next couple decades

    1. I fear forcing light rail to never be built at grade means ( built elevated or underground only) means building only next to freeways for the foreseeable future making transit much less useful.

       Unless, of course, you decide to build a real metro. Ballard to UW would not run next to the freeway. A Metro 8 subway would not run next to the freeway. West Seattle Link would not … oh wait, yes it will. Anyway, it isn’t that hard to avoid running close to the freeway. Just choose routes that don’t mimic it.

12. Elevated light rail in the Rainier Valley along one of the current travel lanes on MLK would be the best option. SDOT would have the opportunity to create a linear park underneath the guideway similar to how the Skytrain did in the 80s with the original Expo line segments. It would also give the rainier valley some much need bicycle infrastructure love, and a well used cycleway would discourage unintended uses of the trail like camping.

    It also would give SDOT the opportunity to reconfigure MLK into a 3 lane street with center turn lane, planted medians, etc. It would also allow for any future trolley improvements that are currently not able to be done. It would provide a good experience for passengers on link and a safer experience for residents of the rainier valley, both in and out of automobiles.

    I propose constructing it in the south or northbound travel lane because it would reduce disruptions to the light rail until it would have to be connected to the guideway at either end of the valley.

    1. Why would you elevate it if burying it is cheaper? I have no problem with elevated rail but it seems like being underground would be better. It allow you to more easily narrow the street and/or add bus/bike lanes.

       1. The issue is that is an open trench, not full burying. It would retain the barrier effect that link currently has on the neighborhoods rather than allow people to pass under it. Bridges across would be more expensive than just letting people walk underneath.

13. Grade separating the RV segment of Link may turn out to be the only ST3 “project” that Sound Transit can afford after the Lollapaloozas to Everett (pop 113,000) and the Tacoma Industrial District (pop. 49). It could be a “constellation” [sic] prize, because it would be such a star!.

    Who wants to go to Ballard of West Seattle anyway?

    Enquiring minds want to know!

    1. Grade separating MLK is not in ST3. ST can’t spend money on on-voter-approved projects until it has finished the voter-approved ST3 projects, unless it has another vote to authorize it. Tacoma Dome station is not just for the industrial district, it’s for the city of Tacoma (pop 228,000) and the Pierce subarea (I can’t get the subarea’s population but Pierce County is 941,000).

       1. Yes, you would probably have to have another vote. I think there is a limit to how much leeway the board has in switching projects. Maybe we could call it the “second downtown tunnel” — it would just be a little bit farther south. Even though I’m joking there are some similarities. The second tunnel was billed as a “regional resource” (or whatever term they used). It is seen as something that benefits everyone. The point being that even if you don’t ride through downtown you would benefit from the improved frequency and reliability. The same can be said for burying the line in Rainier Valley. Even if you don’t take the train through Rainier Valley you benefit from the improved frequency and reliability. Meanwhile, unlike the second tunnel, existing riders would be no worse off. No additional transfers, no bad downtown stations — at worst you don’t notice the difference. But those in Rainier Valley (and those who travel through it) would be much better off.

          But it would benefit some areas more than others. It would benefit the South King County subarea a lot more than Pierce County. It would make Link a lot more attractive from Federal Way. It still might be faster to take a bus in the middle of the day but Link would beat the bus more often during rush hour.

          From the Tacoma Dome, I don’t think Link will ever be able to compete with Sounder. In the middle of the day the buses would be faster. It would improve trips to Rainier Valley or Beacon Hill but my guess is there aren’t that many riders taking that trip. But then again, the same thing can largely be said about Tacoma Dome Link. It improves trips to those same areas (while Sounder or express buses are faster to Seattle). Maybe if Tacoma Dome Link served downtown Tacoma you could make the argument that it saves people a transfer (for trips to say, SeaTac) but more likely it will just mean people driving to a different park and ride lot (Tacoma Dome instead of Federal Way).

    2. P.S. The rhetorical about Ballard was snark.

14. I don’t have a great understanding of using MLK for transit, save for a few train rides through the area. However, I do have a fairly good understanding of biking north of Rainier and driving its length N-S. I also live directly on the street so I encounter issues quite often.

    There are many people with a very clear disregard for traffic rules because they simply need not be bothered with them. Traffic laws are simply not enforced or pursuit rules mean people can drive recklessly without fear of consequence.

    Last night, in Othello, I watched a speeding car blast through a red light while narrowly avoiding pedestrians in the dimly lit crosswalk.

    Last month, someone totaled my vehicle parked on MLK by hitting it so hard it they launched the 3.25ton SUV up and over the curb and caused it to knock the truck behind it into a completely different parking spot. I’m guessing this was one of our top speed record holders and not someone doing 35.

    Almost every day, I see people driving in the center turning lanes to pass vehicles or beat the line at a red light.

    There is a lot of nuance to digest here on how effectively the train runs through that segment, but there is a much larger problem in South Seattle that is causing MLK to be as dangerous as it is. Even with transit or road improvements, the care free nature of drivers in the area pose a serious risk to others.

    1. Sounds like MLK could really use speed cameras. A road diet would also help.

15. Im aware this risks starting a vicious set of cocomments but umm maybe don’t ride your bicycle on rhe sidewalk. .walking on the walk is difficult enough without having to dodge you guy zipping by at very close distances.

Comments are closed.