At-Grade Crossing Plan

102 Replies to “At-Grade Crossing Plan”

1. Weird… Accidents along an at-grade, median-running light-metro alignment in a busy urban neighborhood? If only someone had said something…

   Saying “I told you so.” doesn’t feel as rewarding as I thought it’d be.

   1. Many cities have these sections though. The Chicago L even has a few sections of median running track, though on streets that aren’t as busy as Rainier.

      • Do these other operations have about the same number of these collisions, or is Rainier Valley worse? To me, it sure seems like there are more of these in Rainier Valley than on the entire MAX system.

      • If Rainier Valley is worse, then what is wrong with the street design that makes it worse?

      For one thing, except for the yellow line, I notice MAX was built with somewhat more space between the median and auto traffic than Rainier Valley. This is especially obvious at SE 17th and Holgate.

      Not that any of this helps the Rainier Valley much, but if Link must be built on the surface in other places, it would be good to know what to avoid.

      Such things may also be useful if median running BRT ever comes to such places as Aurora.

      1. There used to be a lot of accidents from trams, streetcars, and the EL in Chicago. They simply have had more time for generations to be raised with child-educated awareness of the train. That doesn’t justify transit agencies proposing more of these at-grade median alignments – This type of alignment is done because it’s cheap and generations found a way to survive and operate around such type, but in this day and age transit agencies should be constructing better quality, more efficient grade-separation.

   2. Well, the silver lining is that thanks to the unqualified disaster that RV has been, and will continue to be for the system, ST is fully elevating and/or grade separating everything south of angle lake and everything north of Lynwood based on the latest EIS drafts.

      So at least they are not doing the same thing and expecting different results. It sucks for line 1 riders that are simply passing RV on their way home or downtown because they go from basically a light metro to a crawling tram that has no business being that long.

2. Why would ST fund grade separation on BNSF mainline through the Puyallup and Green River Valley? With an elevated water table and rail lines cutting through the central business districts of numerous communities (Puyallup, Kent, Sumner) such investments in underpasses aren’t a good idea.

   Lastly, if we want to increase density near street level LLR, I don’t think residents will be in favor of louder train horns/bells.

   1. Where are you seeing that they want to do this?

   2. The street-level sections of the L are dedicated right-of-way, in alleyways, protected by fencing and crossing arms. The median sections are on freeways.

      The whole reason trains are grade-separated in Chicago is because at-grade crossings were deemed an unacceptable hazard and the city passed ordinances to eliminate them–over 100 years ago. Seattle is still catching up.

      1. @bidab,

         The Chicago L uses third rail power. As such, it represents a serious safety hazard when operating at grade.

      2. Lazarus is correct. Third rail is very unsafe for surface operation. But if you’re going to grade-separate anyway overhead contact is much more expensive and a permanent maintenance headache.

3. I feel like there is a simpler solution. Why not close about half the cross streets? Are all of them really necessary?

   1. For pedestrians, closing a single crosswalk across MLK can mean a 20-minute detour to go around. That’s very harsh, and many would just cross anyway where they’re not supposed to, increasing risk of being hit by a train.

      1. More pedestrian overpasses would fix this. A pedestrian overpass is at least cheaper than a roadway overpass.

      2. Pedestrian overpasses in Rainier Valley. Where I have heard that before? Oh yeah (https://www.seattletimes.com/seattle-news/seattle-police-to-review-tactics-officers-conduct-after-videotaped-punch/). For those who don’t remember it is quite common for people to jaywalk instead of using overpasses. So common that the Seattle Police Department put cops there to stop people from jaywalking. The cops caught some nearby students and one thing led to another and the student was punched.

         Anyway, the point is, overpasses aren’t a real solution if you have a lot of pedestrians. There are only a handful of places where they really make sense. Usually it is when there is a surface option as well (and the pedestrians doesn’t want to wait). This is the case on Aurora at 130th for example. One unusual example is over 15th NE just north of Campus Parkway. The pedestrian bridge is essentially level at the UW side which means people don’t have to go up and down. But that wouldn’t be the case in Rainier Valley which is why it wouldn’t be that effective.

         It also wouldn’t solve the problem with cars. A lot of the cross streets are closed to cars. They could probably close a few more but it would have to be part of a bigger effort (using the approaches discusses in other comments on this post).

4. I don’t like the idea of louder bells, as that impacts nearby resident’s ability to sleep at night, especially if they don’t have A/C and have to open windows in summer.

   Are the pedestrian Link crossing gated? If not, installing gates seems like a better option.

   1. I completely agree. In my opinion another cheap and effective solution is to fence off the entire right of way except for the pedestrian crossings. I doubt the fence will be able to survive a car running through it but it’s a visual reminder for drivers to not drive there.

      1. Do many vehicles incur on the ROW between intersections? It doesn’t seem like fencing the track would be worthwhile.

         It seems to me that the most common problem is drivers turning left across the tracks without looking for the train ahead or behind them.

         What isn’t obvious to me is why ST or SDOT can’t install short crossing arms that deploy across the left turn lanes only. Do they think drivers would go around them? Is the problem that the MUTCD (manual of uniform traffic control devices) doesn’t allow for intrusions into the roadway that only stop one lane of traffic?

      2. @Nathan, I was actually surprised how often pickup trucks try to just run across the tracks, not even at an intersection. Fencing would certainly help, or at least posts or higher curb. Yes, most accidents happen at the left turns, but there are also accidents with cars driving across the intersection, probably trying to outrun the red light.

      3. If they go with train gates blocking off the left turn lane while the train is approaching or stopping, it may make sense to have a stoplight there a bit upstream of the left turn lane, which turns red with No Turn (e.g., similar to some intersections along the Burke Gilman trail) when the train gates come down. That way, as far as MUTCD is concerned, it is just another traffic light.

5. I don’t think underpasses for cars or people are the best use of resources. MLK already feels enough like a highway. Instead, we should be focused on reducing car trips by improving ped, bike and transit accessibly. Redesigning MLK to slow vehicle traffic and encourage other modes will greatly improve safety for all road users. Quad gates are a good last resort to minimize train/car crashes when a driver runs a red light.

   1. I agree. A few underpasses doesn’t really add value. You want to be completely grade separated so that you can run the trains more often. It would also make automation easier. To reach that level a few of the intersections could be converted to “right turn only” but you still have to do major work on the main arterials. When you are done you still have pedestrians crossing on the surface. You can add pedestrian overpasses (or underpasses) but that doesn’t mean people will use them. Folks in Rainier Valley of all places are well aware of this (https://www.wsj.com/video/seattle-policeman-punches-teen-girl/945A013B-0AD7-4C00-B752-F75FE74E75A4). This means you still can’t run the trains at full speed and you still have to have the various things they are talking about (pedestrian gates, etc.). Digging a trench or elevating the train line is more disruptive but would be better. Most likely we don’t do any of that.

      Oh, and you would also need to deal with SoDo. That is one place where I can see adding more bridges (like Lander) but over the light rail trains and buses. There are far fewer issues (and fewer crossings) at SoDo then Rainier Valley.

6. As brilliant as the Caltrain grade separations are, they are very complex and expensive. San Mateo County has been adding a new one only about once every five or so years. They first got funding about 1990 and here in 2025 they still have major streets with Caltrain grade crossings up and down the county.

   1. A good rule of thumb is never to look at California as a proxy for affordable or timely public works.

7. I think more work needs to be done.

   Two possibilities are crossing gates for cars, or overpasses/underpasses to eliminate the level crossings.

   It seems like gates would be much, much cheaper.

   Another consideration: red light cameras. It is quite possible that one reason for people making illegal turns is they often get away with it. Consider this picture: https://maps.app.goo.gl/prvuSLBPnDJqfBMU6. Now imagine someone taking a left onto Orcas. It is obviously illegal but it also looks like you can get away with it quite easily. There are clearly no cars coming the other direction. There is no train coming the other direction either. But a train could be coming from behind and you don’t see it.

   It is the same idea for cars going straight. Running red lights is quite common (unfortunately) in part because it usually doesn’t result in an accident. But the accidents (when they do occur) can be really bad. Adding red-light cameras (and letting people know you’ve added them) can reduce the number of people breaking the law and ultimately causing collisions.

   1. The city’s hesitance to deploy traffic cameras has confused me for a long time. I haven’t been able to find a good explanation for it.

      1. Too many drivers believe it’s a constitutional right to be able to disobey traffic laws, so long as a cop isn’t there to see you do it.

      2. I think part of it is that red light cameras are pitched as being for preventing people from running red lights but in practice end up being mostly used to enforce the much more common, and less dangerous rolling right on reds which make people feel that they’re mostly for revenue.

      3. Can you tell me you run red lights regularly without telling me your run red lights regularly?

      4. For what it’s worth, just last week, I encountered a red light runner. I waited in the left turn lane for the green arrow. When it came on, I was all set to go, except the car going straight the other direction also decided that it was time to go, even though I had the arrow and he had the red light. Fortunately, I was paying attention to him, so I stopped, he went, and nothing happened. But, he really did deserve a ticket for what he did and, had there been a red light camera, he would have gotten one.

      5. Red-light cameras are sold as a safety measure, but they are usually operated as a profit generator. There has been a nationwide pattern where cities which introduce red light cameras juice the revenue they produce by shortening the yellow light durations at those intersections, which makes safety worse.

         When a court in North Carolina forced cities to donate the profits from their red-light cameras to the school system, the majority of those cities shut the cameras down.

         Most red light camera systems are operated by contractors who get a share of the profits. Many people find this objectionable.

      6. @Mars Saxman,

         If a private contractor can improve safety, reduce injuries, and save lives, and do it all while making a profit and not spending tax dollars, then what exactly is wrong with that?

      7. Well, that’s the sales pitch, all right. The reality works out somewhat differently, because the profit motive takes precedence over safety, and the shortened yellow lights cause more accidents than the threat of a ticket prevents.

         Many people are not OK with the idea of letting private companies make a profit by enforcing public laws, as a rule.

      8. @Mars Saxman,

         Photo enforcement doesn’t change the light cycles. The length of the yellow remains the same.

         And most states require that photo evidence of a violation be reviewed by a public employee before any citation is actually issued.

         It’s all very honest, open, and transparent.

      9. The length of the yellow remains the same until the city changes it, which they do because that produces more ticket revenue. Do you suppose that city governments are somehow immune to the profit incentives which affect other organizations?

         Municipalities have treated traffic tickets as revenue generators since forever. Everyone who has gone on a road trip through rural America knows that you have to watch out for small towns, which often hire a cop specifically to sit on the main road and write tickets because the fees make enough money to pay the cop’s salary with a healthy profit left over for the city budget. When you can do this with robots instead, the value proposition is even more appealing.

         None of this has anything to do with safety, except in the sales pitch. If you want fewer red-light crashes, what’s far better than a ticket robot is a better-engineered intersection. Here is a helpful US DOT FHA page about proven safety measures, including a section about intersections. Note that longer yellow-light intervals are shown to yield a 36-60% reduction in red light running and 8-14% reduction in crashes, while ticket robots are not mentioned at all:
         https://highways.dot.gov/safety/proven-safety-countermeasures

      10. To be clear we are talking about people taking left turns. A lot of left turns in this city are not controlled. People just assume that if there is no traffic going the other way they can take a left turn. This contributes to the problem. In this case it is controlled. They can’t just take a left. Because of the train and because of the prevailing north-south traffic it may take a long time before the left-turn arrow turns green. Look at that picture again. It is a bright sunny day. Visibility is excellent. You can see there are no cars going the other direction for a long ways. You can see there are no pedestrians to worry about either. I’m sure a driver thinks “Why the hell should I have to wait for this stupid arrow — I’ve got placed to go”. So they run the light. They get away with it. They do this over and over and eventually their luck runs out. They forget to check for the train coming from behind. This is quite understandable since you *never* have to check for a car coming from the left when taking a left.

          Another alternative is the one Brandon suggested below (https://seattletransitblog.com/2025/02/17/at-grade-crossing-plan/#comment-951763). The only way to turn left is from the right lane. This changes the dynamic. Now riders are confronted with a very unusual situation — now a train *or a car* could hit you from behind. Look at that picture again. Are their cars coming from behind? Yes! Most definitely. More than anything a driver realizes it would be a very risky maneuver to take a left from the right lane unless the arrow is green.

          Red-light cameras are sold as a safety measure, but they are usually operated as a profit generator.

          They can be both. Just to back up here, longer yellow cycles make an intersection safer but so do red light cameras. https://www.iihs.org/topics/red-light-running.

          So should we be bothered that agencies also make money? No. Congestion pricing pays for better transit in NYC. Does it actually reduce congestion? Yes! Same goes for cameras. Speed cameras in Europe are quite common. They reduce speeding *and* help pay for needed services.

          But the main benefit of cameras (as opposed to cops) is that they are objective and cheap. They don’t care who you are. In contrast a corrupt police department (say, in Ferguson Missouri) could be pulling over people and giving them tickets even if they aren’t speeding.

          In general this country has it backwards. It is part of our history. We think of ourselves as a bastion of freedom. Yeah, maybe in the 1700s but countries soon passed us. They got rid of slavery AND adopted the same rights we had for the white guys. We see cameras as a violation of our privacy. They see cameras as just another tool — and an effective one that can remove biases. A lot of people in this country embrace bias. They brag about talking their way out of tickets. If we become like Europe (and have a lot more traffic cameras) we will be a lot better off.

8. I think they need to restrict several streets’ access to MLK. Particularly those left turns. Keep the pedestrian crossings in those areas.

9. It sounds rather basic to ask this, but is there any breakdown of what are the most common accidents involving Link versus the street in general? How much of the hazard is because of Link trains?

   I ask this because there is this sense that it’s totally up to ST to address the problem. However, SDOT may be contributing to it too. Link trains are involved in some percentage of serious MLK accidents but not all of them.

   Also, any solution should be funded and implemented that addresses what the biggest causes or movements are. If it’s left turning cars from MLK crossing tracks the remedy seems to be small gates that drop when a train approaches. However, if it’s general traffic or pedestrians than the solution is big gates across the entire width of any street that crosses the tracks, for example.

   Is it at certain intersections or is the problem with midblock situations too? If it’s problem intersections are the issue than the solution is locationally specific. If it’s with random people crossing tracks than low level fencing seems to be in order (even a fence between the two tracks would deter mid block crossings).

   There have been many ideas on how to address the problem. What’s missing is a good analysis of whether any would have prevented a prior collision, or what the outcome would be by implementing them. How about some cost-prevention ratio assessment?

   1. The master plan’s appendix says: Figure 7-2 shows that the S Graham Street and S Dawson Street intersections experienced the greatest collisions, mostly due to illegal motor vehicle left-turn movements. Illegal left turns account for 63% of all collisions, while 15% involve pedestrians.

      1. Illegal left turns because they don’t have a green light, or illegal left turns because there is no left turn pocket with a signal?

         63 % is very high. That seems like the biggest problem. I’m not sure if louder sounds or flashier headlights on trains will deter anything if a maneuver is already not legal . It seemingly will take gates of some sort to deter lawbreakers.

      2. I believe all left turns along MLK have a separate turn lane with a signal.

      3. All left turn lanes do have a signal, but I’ve seen several drivers make a left where there is no left turn lane or signal, and there is a no left turn sign posted. That said, red light cameras at the signalized left turn lanes would catch most offenders.

      4. Sadly, the Figure 7-2 doesn’t differentiate between northbound and southbound at Dawson. It’s too bad because with the difference in design in the two directions it’s a good case study on whether the pockets are dangerous to have, a drivers should have to turn right first and the cross at the signal. .

   2. Dawson is an interesting case. There is a left turn pocket southbound but not northbound. Northbound drivers make a right, the make a left turn to get across Dawson so no left turn pocket is needed. In nit sure why they didn’t build southbound the same way. In curious which treatment causes more accidents. That’s a good case study right there!

      Evan is right that I’ve seen drivers illegally turn left where there is no pocket or signal. I may mumble “Are you crazy?” But people do it. And some of them die.

      Graham should have little gates for the left turn pockets where vehicles turn left off of MLK there. Those gates should drop every time a train approaches — even earlier than the train approaching warning signs and alarms. I suspect that there are stupid people who think that the warning sign merely means “beat the train”. So the gates should drop at least 5 seconds before the warning bells and flashing lights start.

10. All this proves is that bad designs do not get better with time. Anything over the current state is an improvement but the best solutions presented by many transit advocates is to simply bypass Rainier valley by routing the 1 line over marginal way and turning Rainier valley into a simple low speed tram line like the T line.

    Not only does rainier valley have massive safety problem, but it has low ridership stations featuring the lowest operating speeds that delay everyone coming from south of rainier valley all the way to Tacoma dome link. This is most ridiculous given the expense and effort ST is spending to make everything south of SeaTac match the rapid transit standards used on the north part of the system where trains travel safely at or near top operating speeds (55mph).

    RV is basically a poorly designed nightmare that is crippling the much more modern system south of SeaTac and any accidents cause rippled effects up to Lynwood and soon down to Federal way and beyond. It is just ridiculous and unsustainable and will get worse with questionable pet project infill stations.

    A bypass line would greatly improve travel times, on time performance, and reliability along the main spine of the system.

    1. Building a second line would be a huge overreaction and waste of money. Other than Georgetown there is nothing there to serve. You would be spending a fortune and do nothing for Beacon Hill and Rainier Valley. If you are willing to spend that kind of money for that area then elevate the line in Rainier Valley or put it underground. If we decided to run a new line then it should run on Rainier Avenue. The line already spends way too much time serving nothingness (it is already a massive express). The last thing we want to do is spend billions more on that.

       1. Somewhere in the comments section some years back, someone (I think dp) did some math on what a “Duwamish Bypass” might save, and it would up being something like 3 minutes. It’s also not a direct route, so it doesn’t save as much time as you might initially think, will have stations to be worthwhile, and in the end really doesn’t do that good a job at doing what Sounder and the express buses do.

       2. Everyone seems to think that the only possible “Duwamish Bypass” is a re-creation of Forward Thrust’s Renton line, with stations by Boeing, in South Park and in the residential area northwest of Georgetown.

          That would indeed be expensive and not save much time. But it’s not the only possible “Bypass”.

          If Airport Way were narrowed to three lanes south of the Air Terminal (two-way single lane with a center refuge/turn lane at the two or three intersections) there is plenty of room to run two tracks between the then-northbound street lane and the existing heavy rail freight tracks. Yes, just south of the terminal the trackway would have to go elevated for a way to just south of Albro Place. It would need to descend there to pass under Albro Place and a station could be placed between Albro and the Corson on-ramp, here: https://www.google.com/maps/@47.5483161,-122.3154016,3a,75y,349.99h,92t/data=!3m7!1e1!3m5!1s8Wy4YG5JjhsVIvp7lMagGA!2e0!6shttps:%2F%2Fstreetviewpixels-pa.googleapis.com%2Fv1%2Fthumbnail%3Fcb_client%3Dmaps_sv.tactile%26w%3D900%26h%3D600%26pitch%3D-2.004467354230414%26panoid%3D8Wy4YG5JjhsVIvp7lMagGA%26yaw%3D349.98685575596915!7i16384!8i8192?entry=ttu&g_ep=EgoyMDI1MDIxMi4wIKXMDSoASAFQAw%3D%3D.

          Leaving the station the train would rise again onto structure, cross diagonally over the BNSF trackway by about Orcas and descend to pass between the Olympic Foundry and I-5. North of the Foundry it would run in the spur right of way , rise up enough to cross Airport Way, run up the middle of Industrial Way and tie into the WSLE at its curve just south of Spokane

          Obviously, that junction would have to be designed into the structure when it is built.

          This would reduce the number of stops by five and the mileage by two. Even at slow-floor speeds of fifty-five this would save twelve minutes versus the existing routing at thirty-five miles an hour.

          Now, if they really can do the open trench then absolutely, do that instead. However, I doubt that they can and keep the system open. Where will trains run while the trench is dug under the existing trackway?

          I’m a big advocate for “decking” with temporary tracks like Muni had when the BART/Metro tunnel was built, but this is a whole order of magnitude more complicated because it has formal stations, not just trolley stops.

       3. Oh, I forgot to mention the south end connection. Just railroad north (compass east) of the curve from Marginal Way South to parallel to BAR, northbound trains would have a right turnout that leads to a down-ramp that continues to descend into a trench right next to the BAR-to-Airport Way ramp that would underpass BAR into the wide space to the north of BAR.

          Southbounds would simply rise up, overpass BAR and merge via a trailing-point turnout just west of Airport Way.

          This obviously means that “Boeing Access Road Station would actually have to be at 112th and East Marginal which is a better location for bus intercept anyway. However, it obviously doesn’t offer integration with Sounder.

       4. If the line is split then frequencies will be split. Does that mean RV only gets trains every 12-16 minutes (or worse)? That seems unacceptable to me.

       5. @John D,

          None of the lines are getting “split”. That is incorrect terminology.

          What is happening is that a new line is getting added. The 2-Line across the lake via Full ELE.

          So from IDS to Lynnwood the frequency would be doubled, with every other train being either a 1-Line or a 2-Line train. So the frequency would be a train every 4 minutes.

          South of IDS in the RV, and East of IDS on the 2-Line, the frequency would remain at the current base frequency of 8-mins.

       6. Now, if they really can do the open trench then absolutely, do that instead. However, I doubt that they can and keep the system open.

          I agree, but that is still a better approach. To quote the document:

          The construction would be phased into four sections and be from station to station as shown in Table 1 [south to north]. Temporary bus bridge service would be provided to allow passengers to bypass the construction and continue service on the light rail line. Construction phasing alternatives for maintaining a single track in service during construction could be warranted. This would include studying constructing the trench in halves, or as a whole with a shifted alignment, however substantial challenges would be anticipated.

          So you would have single tracking (maybe) and full closures with bus bridges (at times). Depending on what stage you are on you would still run the trains to various places. So theoretically the train could run to Othello while they work from the south past Rainier Beach. To do that though you would need turnback track and I don’t think there are any. But this could also be added as a mitigation measure. It wouldn’t be that expensive (since it is surface work) but it also wouldn’t last. It is the usual trade-off. How much money do you want to spend maintaining the current service. Personally I would do this:

          Stage 1.

          Dig a trench from south of Mount Baker Station (MBS) all the way to the end of the project. While this is happening:

          1) Add turn back track south of MBS (but north of where they are digging the trench).
          2) Continue to run the trains from Beacon Hill to MBS.
          3) Continue to run trains from Tukwila International Boulevard Station (TIBS) to the south.
          4) Add express buses from SeaTac/TIBS to SoDo/downtown. This is a bit of an overkill but since SeaTac is a bigger destination than TIBS (and downtown is a bigger destinations than SoDo) it avoids some transfers.
          5) Add a different express bus from Rainier Valley to SeaTac/TIBS. Hard to say where this would start.
          6) Reroute the 106 to Rainier Avenue. We might need to run some more 7 buses along with increasing frequency on the crossing buses (107/50). I could also see special express/truncated versions as well.

          Stage 2:

          Work on the section north of there (basically Mount Baker Station).

          1) The buses remain the same (you are still running an express from SeaTac/TIBS to SoDo/downtown).
          2) The train from the north turns back at Beacon Hill if there are turnback tracks between there and MBS. If there isn’t then it turns around in SoDo.
          3) If the train turns around at SoDo we add more 36 runs (this could include a truncated/express version of the 36 which just runs between downtown and Beacon Hill Station making only a few stops along the way).

          The express buses from SeaTac/TIBS to SoDo/downtown would be quite fast from TIBS to SoDo. So folks, there is your bypass :). It is on rubber wheels and won’t last, but there you go.

          Of course this is less than ideal but the people for which it is most disruptive are the people that would benefit from it in the long run. I’m not saying it is worth it but it is clearly a better value than a bypass which wouldn’t even help Rainier Valley (or Beacon Hill). It is just bizarre to me that folks point out the flaws in the trains in Rainier Valley (which are real) and their solution is “let’s basically just ignore them and build a very expensive but lower-ridership substitute instead”. Oh, and let’s just hope that ST (which has historically been stingy when it comes to running the trains) will just run both lines as frequently as they do now despite the fact that you are basically splitting ridership-per-mile in half. Meanwhile, trains will branch at SoDo and then meet again up in TIBS. Both will go to Federal Way magically spaced well apart from each other even though the trains have varying headways throughout the day. Sure, nothing could go wrong with that.

          Sorry but that is just BS. It would be a giant F. U. to Rainier Valley on top of the giant F. U. that is the second downtown tunnel.

       7. @John D,

          None of the lines are getting “split”. That is incorrect terminology.

          John is referring to the Duwamish Bypass idea (not what we are actually doing). The Duwamish Bypass idea is not reality (thankfully).

          You are correct. It is better to describe the line from the east and the line from the south as being merged. There is a split (or branch) but they aren’t decreasing the frequency on either branch. The trains on each branch will run just as often which means on the combined section (north of CID) the trains will be running twice as often. It will be a good thing for everyone.

       8. the existing routing at thirty-five miles an hour.

          That only applies to when the train is on the surface (on MLK) which is only about four miles. So not counting the time it takes to slow down and speed up (for each station) you save a couple minutes. Other savings come from skipping the stations but as I’ve pointed out before this is not a good thing.

          Just to back up here while it would be good from a speed standpoint to have the trains run elevated or in a trench in Rainier Valley there are bigger benefits. The trains become more reliable. You avoid the relatively rare accident that delays service (even if people walk away unhurt) as well as the occasional traffic-signal delay. You can also run the trains more often. This bypass would not help that in any way. In contrast if we buried the trains they could run the trains as often as we want (based on the rest of the system) and the nature of the street could be transformed. It would be extremely disruptive but when all is said and done MLK would be a much nicer street and everyone who takes the train south of CID would be better off.

          So the only significant advantage of the bypass is that some riders get to skip a few stations. We could probably save almost as much time if we reduce our excessive dwell times.

          Consider a future system like so: The line is buried in MLK. We end up not building West Seattle Link. Ballard Link is a stand-alone, automated line. The other lines are automated as well despite potential lack of grade separation in SoDo and Bellevue. The trains from the East Side and Rainier Valley run every six minutes all day long. They combine to run every three minutes to Lynnwood every three minutes as well.

          Would it be cheaper to build a bypass? Probably not. Would a bypass be better? No, of course not. Other than the new stations — and there wouldn’t be many — it is basically just a nuisance. You have to deal with the fact that trains have a split then a reverse-split on the same line. This does happen but typically in areas where each branch takes the same amount of time to cover. In this case the whole point is for one to be faster. I suppose you could just run the bypass spur to TIBS but there really is no point in that. TIBS is not a significant destination. They would end up transferring to the same train (coming from Rainier Valley) that they could have caught in the first place.

          If you do the reverse (end the Rainier Valley trains at TIBS) you are forcing all of those additional riders to transfer. Quite a few riders take the train from Rainier Valley (and Beacon Hill) to the south end. Somewhere around 20% based on old data* and it is probably higher now. The only reason to build the bypass is to serve Georgetown and with all due respect it isn’t worth it. There just aren’t that many people there.

          *ST has never released trip data. So we have no idea how many people ride the train from say, UW Station to Rainier Beach. But they used to release stop data along with direction data. So we knew how many people took the train heading south from Rainier Beach and how many took the train heading north. For Rainier Beach about 20% of the riders took it heading south. For the other Rainier Valley stations it was a bit more. This suggests that some were taking it for trips within Rainier Valley while somewhere around 20% were doing the same thing (taking the express train towards SeaTac) at other stations. Reference: https://www.soundtransit.org/sites/default/files/documents/2020-service-implementation-plan.pdf#page=84

       9. > If the line is split then frequencies will be split. Does that mean RV only gets trains every 12-16 minutes (or worse)? That seems unacceptable to me.
          > None of the lines are getting “split”. That is incorrect terminology.
          > What is happening is that a new line is getting added. The 2-Line across the lake via Full ELE.

          A new line is getting added but yes it will add pressure to lower frequency off peak times for the original 1 line section south of cid.

       10. it will add pressure to lower frequency off peak times for the original 1 line section south of cid.

           I’m not so sure. Now folks in the south end are in the same boat as those on the East Side. That will likely perform well (from a ridership per mile standpoint) and the riders have more political clout. I would be more worried about the extension to the south. It is quite likely that will result in worse ridership per mile. Unless they are prepared to turn back after SeaTac that might result in a reduction.

       11. John, the RV is limited to ten TPH TOPS, or once every six minutes, unless it’s grade separated. South of BAR the line is 100.0000000000000% grade-separated or freeway right-of-way. A person intruding would be very rare, and a vehicle close to impossible. So trains can run at “track speed”.

           ST wants to four-track the busway between Stadium and SoDo, so capacity there is NOOOOOOOOOOO problem. West Seattle will never be more the six TPH because there’s no “there” there to generate ridership. Especially if they do the sixty-five foot deep monster at the Junction.

           So there will be plenty of capacity on the West Seattle line; if they’re going to spend a billion and a half for this completely unnecessary four-tracking, AT LEAST get some use out of the second set of tracks. Maybe the Sea-Tac / Federal Way / Tacoma trains go into the new tunnel while the Rainier Valley / Sea-Tac trains go to Mariner or even just Lynnwood as an “overlay”. With even minimal automation north of CID Junction and platform doors, trains can run with 90 second headways in the old tunnel.

           Now, yes, this is all grandiose and depends on the country holding together (these days a long odds bet) and people fleeing from climate catastrophes (a MUCH better bet) and that the Puget Sound region has twelve to fifteen million people by 2045. In the absence of all that, we probably don’t need more than ten TPH on the RV line. And, as I noted in the original comment, if Ross is right that they can do the open trench for a billion and change, then yes, get out the bulldozers.

           Even two billion and change would be a good investment. I just don’t see how they keep it open and dig a trench under it. Yes, of course, you deck, but you have to build the decking, and that means shutting down for a loooooong time while the decking is installed.

       12. Ross, I’m not dumb enough to propose terminating the RV trains at TIBS. Of course they’d go to Angle Lake to reverse so they’d serve Sea-Tac. There are a lot of folks who live in the RV who want to go to Sea-Tac / Angle Lake because they work in the airline industry. From the merge point along BAR every mile to Angle Lake is separated and could run at least twenty trains per hour (three minute headways).

           There is no need for “reducing service”. In fact, if South Link / Tacoma Link turns out to be more successfull than we currently project, there’s not gonna be enough room on the trains for folks wanting to board at Graham, Columbia City, Mt. Baker or Beacon Hill. Folks in the RV will be agitating for either the trench or the bypass so they can use the system.

           Running along Airport Way would be cheap! That may or may not be a good argument for it, but just because Forward Thrust had a Duwamish crossing between South Park — which will never be a dense place because it’s ringed by weird stuff and directly under the Sea-Tac approach — and Georgetown (ditto) does not mean we have to build a bypass that way. The East Side has a four mile “express” section just a mile and a half from downtown Seattle and it’s not nearly as long.

           As I said, if ST can actually pull off the trench — or even the elevated — yes, of course that’s better. I just don’t see how it’s to be done without a long closure — three months? six? — while the decking is put in place. Plus, of course, digging underneath the decking is a lot harder than digging under the blue sky.

           The “bus bridge” for three or four years is a non-starter. ST should be ashamed of themselves for even considering it.

       13. I do not see the Duwamish bypass as a replacement for RV line, but a replacement for the WSLE and a way to extend the RV line towards Renton. I could even imagine going from SODO through Georgetown and then up Olson Pl and then along 4th Ave SW towards Burien and to the airport. That way it would serve the southern portion of West Seattle which is more dense than Tukwila. It gets close to Westwood and White Center, but still wouldn’t serve it directly. You could run 2 car trains between Federal Way and TIBS. Riders could transfer at Seatac. Or you could decommission the line between Rainier Station and TIBS and replace it with a TIBS to Renton along Southcenter Blvd with a mid-station and pedestrian bridge (or funicular) across I-405 to connect with the Sounder station.

       14. “the existing routing at thirty-five miles an hour. ”

           It may go down to 25 if SDOT applies to MLK its Vision Zero “25 mph on arterials; 20 mph on other streets” policy and says Link can’t go faster than the adjacent car lanes. What’s the speed on MLK now? I doubt it’s 35 mph; that was unusual for Seattle streets even before the policy.

       15. Link speeds on MLK are 35 mph which is unchanged since the line opened. The only changes to Link since the MLK limit was lowered to 25mph for drivers was to increase the volume of the train bells and add more warning signage for approaching trains.

       16. From the merge point along BAR every mile to Angle Lake is separated and could run at least twenty trains per hour (three minute headways).

           The problem is that you can’t time it. Consider some numbers here (which are not realistic but make it easier to explain). Imagine the bypass takes 10 minutes from SeaTac to SoDo and 15 minutes via Rainier Valley. Trains are running north through SeaTac equally spread apart — say, every five minutes. The train using the bypass actually catches up to the other train at SoDo. Oops.

           The solution (of course) is to run the trains 2.5 minutes apart downtown so they will be 2.5 minutes apart at SeaTac. That is not very good in either place and it is likely cutting it close. This particular problem and timing isn’t likely to happen but remember that we like to run the trains at different headways depending on the time of day (e. g. 7.5 minutes during peak, 10 minutes midday, 15 minutes night). Branching at both ends causes issues.

           Running along Airport Way would be cheap!

           Not if you are also running in Rainier Valley. You add very few riders while you dramatically increase the cost of operations. It isn’t just the drivers (and security) but also maintaining all of this additional track. One of the key things you want to do in any mass transit system is have high ridership per mile. This would move in the opposite direction by a huge amount.

           Again, I’m not saying we can’t do it but I’m saying why we would bother? There is so little benefit, the capital cost would be huge and operations would be more complicated. Would we even consider it if the line in Rainier Valley was grade separated? If not then why not address that since it would be much better and quite likely much cheaper. The only drawback to making Rainier Valley grade separated is that it would be very disruptive. But the end result is so much better (and probably so much cheaper).

       17. Dammit, Ross, I said “if it’s needed”. That would mean that ten TPH is not providing reliable service. Without an elevated, bored or trenched grade separation, ten TPH is all that’s going to happen. There’s still a lot of land in Southwest King County.

           I like Martin’s idea of swapping the southbound roadway and the trackway and putting the new tracks in a trench. Deck one lane over the trackway for local access and you have a reliable and reasonably speedy facility with the capacity to handle the load. It should be able to handle twenty TPH if needed.

           But it would be a huge political lift and I really don’t think you can do it for a billion; more like three or four. The bypass would be half that if done correctly.

11. This is not the only rail line in the world running at-grade. There are literally thousands of them. Put gates at every intersection. If the gates don’t work, close the intersections entirely and put in pedestrian overpasses so people can still get to the stations.

    If underpasses are deemed necessary, that’s not Sound Transit’s problem. That would be car infrastructure, not transit infrastructure.

    1. I agree. There is nothing fundamentally wrong with running on the surface. It is simply a trade-off. Like all aspects of a transportation system it needs to be managed properly. That is the issue.

    2. You can demand SDOT consider it “car infrastructure” and do it. but that doesn’t mean it will. It’s not in the transportation levy, so where would SDOT get the money? And where would it be on SDOT’s list of priorities?

12. While I feel bad for people in RV, I’m just so thrilled that once line 2 opens, the northern trunk reliability will improve drastically. Trains might go from one per 4 min to one per 8 min – but that’s not so bad. I have a feeling once that happens no one will care about this issue anymore.

13. The study you linked to — the preliminary study in 2023 on what it would take to elevate, tunnel, or trench the line or to grade separate at least the pedestrian access — is very interesting. I don’t see how see how anyone can seriously consider the Georgetown bypass after reading it. I guess if you want to serve Georgetown or South Park the bypass makes sense but it would likely be way more expensive than the options considered here. For example the open trench option is estimated to cost 1.2 billion in 2023 dollars. That is with soft cost added at 35 percent and a 30 percent contingency. By no means is this cheap but it would be much, much cheaper than a bypass.

    Yet it would be much better. You avoid all the issues with having branching (at two places). You still improve speed (and just as importantly headways) but you do that for riders in Rainier Valley and Beacon Hill as well.

    You also change MLK. You could add fully protected bike lanes, bus lanes and widen the sidewalk on both sides. Better yet we take a general purpose lane and convert it to a BAT lane (as we’re doing on Rainier) and then widen the sidewalk even more. This would completely change the nature of the street.

    1. As mentioned in the Graham St post’s comments, there have been discussions to reduce lanes along MLK. Adding a bike lane would be awesome. Having a BAT lane next to a Link line seems overkill. Just the lane reduction would make it far safer to reach a train in the middle of the road. If you add a trench, then adding a pedestrian underpass may also be an option though you may still need to cross tracks then. Alternatively you could cross the tracks at grade and then use a ramp and/or stairs to the station in the trench.
       https://roadsonline.com.au/designs-revealed-for-new-underpass-in-melbournes-north-east/

    2. I have to mention that a big issue with MLK is simply that it’s too wide. The width creates really, really long pedestrian crossing times. That creates really long waits for green lights generally — and that I think that it makes some people do dangerous things. That’s true for drivers, pedestrians and bicyclists.

       It appears as though these accidents are due to risky behavior. Changing the allocation of the overly wide street probably won’t help. What will help is to significantly narrow the street that has to be crossed. Then impatient people will be less inclined to take risks.

       These accidents are clearly mostly from foolish risk taking. These accidents may appear plentiful but they really are a very small fraction of 1 percent of the people on this street every day. Yeah every time one happens it disrupts Link service significantly — but there are tens of 10K to 20K users on MLK with another similar number crossing it every day.

       If a radical solution without rail grade separations should be studied, I think it’s simply putting all through traffic on one side of the tracks. The other side can be unsignalized and treated as a local frontage road. Then crossing the street gets much easier and faster.

       A three lane street on one side can operate like 23rd Ave. The light cycles would be much shorter. People can jaywalk more safely because they know that they can get across is 15 seconds rather than 45 seconds. And when a train approaches on the adjacent tracks, the left turn light just goes green after the through traffic rather than before it.

       There’s a reason it’s called a “road diet”. That’s because the street is too wide to begin with. Reallocating pavement use is not a “road diet”. It’s merely repurposing the pavement. You got to get that 90 foot crosswalk “waist” much much smaller. A street with reallocated pavement will still take just as long to cross, and the long waits will still inspire some people to take bigger risks and some of those will get injured or killed. And create Link delays.

       1. @Al S,

          “That creates really long waits for green lights generally — and that I think that it makes some people do dangerous things. ”

          There is also a frequency effect that also leads to dangerous behavior. Link operates at fairly high frequency in the RV, and drivers that drive that road daily are accustomed to continuous Link crossings at the intersections.

          So a driver waiting in a left turn lane might be tempted to attempt to beat an oncoming train by turning against a red signal. If a train is also coming from behind the driver then you will have an accident.

          This sort of “beat the train” behavior is particularly likely to occurred if the driver has already had to deal with multiple train crossings in a short period of time, such as are likely to occur in high frequency operation.

       2. The width is an issue but as Glenn pointed out, this is a very wide street as well: https://maps.app.goo.gl/3FkXmY8NBr3bEQqb8. It looks the same if not wider than MLK. But the biggest difference is that there is only one lane of traffic (each direction) in Portland. This makes a huge difference. People drive a lot slower and more responsibly.

          The other difference is the one that Glenn mentioned. The serpentine crossing: https://maps.app.goo.gl/7WAZ7rAV5FxAPPJd9. That is so safe they don’t even add a walk-signal there. But someone in a big hurry can’t just try to run across all the lanes. They are forced to treat these as three separate crossings. If nothing else we could do the same (but as two separate crossings) even if it just applies to those people jaywalking. In some of the crossings in Seattle they actually do that, it just isn’t that pronounced: https://maps.app.goo.gl/ZFLVtB84Su3Eb21p7. In that particular crossing they also made it worse for pedestrians by making the crosswalk angular. It is a minor thing but it also likely contributes to the high-speed psychology of the roadway.

          Basically Portland and Seattle both ran their trains down very wide streets but Portland changed the nature of the street. Google Maps doesn’t go back that far (of course) but my guess is the two streets looked quite similar before they added rail. But in Portland the road has changed dramatically but in Seattle is still treated like a major high-speed thoroughfare.

          My understanding is that this goes way back to when they were building each line. In Portland they were comfortable taking a lane. In Seattle they weren’t (they wanted the roadway to be similar). I could be wrong there (feel free to correct me on this point) but it is clear that the roadway (which is quite similar otherwise) is much safer in Portland.

       3. Interstate Avenue is also Highway 99W though that area. The nature before MAX was sort of an Aurora Avenue / Highway 99 environment, except with only 4 through lanes instead of 6. Also, before MAX it was economically depressed, so maybe add in an element of 1990s Crown Hill?

14. Lets compare street designs, maybe?

    Link on ML King:
    https://maps.app.goo.gl/o1PJfUcy37SA6j4n6

    Typical Interstate Avenue (Highway 99W) MAX makes it obvious you shouldn’t turn left in front of trains where there isn’t an intersection for doing so:
    https://maps.app.goo.gl/DysHiMMZCnXWVvnb6

    Typical ML King pedestrian crossing only allows crossing when the pedestrian light turns on, but also doesn’t have any refuge area for pedestrians:
    https://maps.app.goo.gl/PXPx74txC4sePX9K9

    Typical Interstate Avenue (Highway 99W) MAX allows pedestrians to cross whenever it’s safe to do so, but has a serpentine path when the pedestrian gets to the tracks to allow them to wait to cross the tracks if a train is coming. The serpentine path is used in many locations on MAX, and also at mainline pedestrian railroad crossings, as it forces pedestrians to look before crossing. The railroad industry actually allows this design at some crossings after grudgingly admitting the data shows it significantly reduces pedestrian deaths at crossings:
    https://maps.app.goo.gl/3FkXmY8NBr3bEQqb8

    Intersection design here doesn’t seem to be that different:
    MAX, Interstate Avenue / Highway 99W and Killingsworth:
    https://maps.app.goo.gl/FWEwPszXRGkA9nMc9
    Link on ML King:
    https://maps.app.goo.gl/nvbF1jSUSvoozd2E7

    Interstate Avenue MAX, however, doesn’t have any of these, where there is no left turn lane at all at the intersection. It’s either important enough to have a left turn lane and signal, or it doesn’t get a signal at all:
    https://maps.app.goo.gl/r1F2MYjAxj8HcLuGA

    Even minor cross streets, such as SE 17th and Shiler, got a through lane chopped out and a left turn lane added:
    https://maps.app.goo.gl/tFBbu8u44wStHXqV8

    I was mis-remembering SE Holgate and 17th. They have the same close track space to the left turn lane without the extra gap:
    https://maps.app.goo.gl/q3bFAoR91Fe8fh1d8

    Meanwhile, where the CTA has median running, they do full blown fences and things:
    https://maps.app.goo.gl/QEbPNwH1dR8MHwQP8
    not to mention crossing gates:
    https://maps.app.goo.gl/WDWkSCPgQbYLTufh7
    Of course, it does help that when they do median running, these streets are not particularly huge throughfares like ML King or Highway 99W / Interstate Ave. I’m not sure if the street slowly got downgraded over the years or if they always were that way and somehow the early L companies managed to find wide enough alleyways to run two mainline tracks down and still have two lanes left over for the alleyway.
    https://maps.app.goo.gl/ZxS6yjEGVHue9h5P7

    1. @Glen in Portland,

       Our Emergency Response staffs wouldn’t allow mid-block barriers. They insisted on being able to cross the tracks in emergency situations to reduce response time. I’m not sure why such things are allowed in Portland but not here, but it could be related to the fact that the Portland design allows many more legal crossing points, whereas ST closed more streets to crossing. Thus the travel distances for emergency response would be longer with the Seattle design if mid-block emergency response crossings weren’t allowed.

       That said, the data in the linked report does not indicate that mid-block crossing is a significant source of safety events. Almost all events occur at locations where crossings are allowed. And the greatest source of safety events is related to illegal left turns. Mainly turning against the light/signage.

       The linked report also shows that Link has half the number of injuries and less than have the number of fatalities as TriMet. However, the data is pretty sloppy and it is not clear what is actually being compared. I’m guessing that the comparison is actually “injuries since start of service”, which is pretty nonsensical since MAX has been around so much longer than Link.

       Additionally, Link has a much higher amount of grade separation whereas MAX is designed with many more crossings, so you would expect MAX to have a higher number of total injuries and fatalities. So again the data is somewhat nonsensical.

       The document does attempt to normalize the data by VRM’s, and the trends of Portland having a much higher accident rate still persists. But given the other data issues I’m not sure what “rate” actually means here. And again, Portland has many more crossings than Link.

       That said, the other side of the equation is never discussed in relation to the RV – namely the overall accident rate of the street. Supposedly the volume adjusted car-car and car-ped accident rates actually went down in the RV (pre-COVID) when the street was redesigned to accommodate Link. So the street supposedly got safer, but the trade was adding back a fewer number of LRV-car and LRV-ped accidents that get a lot of press attention.

       But the data in that report is pretty sloppy – use with extreme caution. I would never sign such a document, unfortunately Dow did. But he is a policy guy and not a technical guy, and therein lies the issue.

    2. @Glen in Portland,

       Oh, and one other thing with that report. Some of the data is references TriMet and Sound Transit, not MAX and Link. So that would need to be sorted out too.

       Messy data is often worse than no data. Someone needs to clean that report up.

       1. Yes, you’d need to compare street running miles per year or something.

          MAX has had quite a number of problems with the highway parallel segments. Eg: the first MAX fatality happened along I-84, when someone decided to jump from the freeway shoulder to the MAX line to get wherever they were trying to go.

          It seems like ML King has these on an almost weekly basis though, while based on the number of service interruptions it seems like MAX only has this happen 1-2 times per year.

       2. @Glenn in Portland,

          Yep, you really need a clean data set, and that data isn’t it. Got to get it down to some sort of nearly apples-to-apples comparison of similar systems.

          And Link and MAX are not the same. You’d expect more total safety events on MAX because it has been around so much longer. And MAX also has more total crossings in the system, even more total crossings per mile in many areas.

          But you’d also expect a higher rate of events on Link because it has higher ridership, operates in a much denser urban environment, and operates at higher frequency.

          And also, MAX/Portland Streetcar has been around a lot longer than Link, and the acceptance rate of the tech is a lot higher in Portland. Here we still have people who want to relitigate the mode wars of the late 1980’s. They still tend to come out of the woodwork every time there is an accident, a cost overrun, or even an out of service escalator. It gets sort of tiring.

          And don’t forget “The Seattle Whine”. It’s a real thing up here. It also gets tiring.

          But hey, it’s interesting that you called out the MAX line on Interstate. I’ve often used that line on my SC/LR journey from the Nob Hill district to StormBreaker. Great place.

    3. The most obvious difference to me here is that Link runs on a 2-lane road and MAX runs on a 1-lane road.

       MLK feels like a freeway. According to SDOT, the speed limit of the road is 25 MPH but people regularly drive 35+ (page 21 in the report below).

       https://www.seattle.gov/documents/Departments/SDOT/About/DocumentLibrary/Reports/2023_Traffic_Report.pdf

       1. I was just about to write this.

          For pedestrians this makes things worse. There is a higher chance of getting hit by a car and thus a higher chance that someone would be focused on the cars and not notice the train. The serpentine approach used in Portland also looks safer (although it looks annoying). It looks especially appropriate for the three-stage crossing in Portland since a pedestrian has to look both ways after the first stage. In the case of Seattle they look one way, get to the middle, then look the other way (at least at that crossing). I still think it would be an improvement as it would force jaywalkers to look one way then the other after getting half way across (it doesn’t make any difference if you have the light).

          For cars the two-lane road also makes a difference for the reason John mentioned. Cars drive faster and people treat it more like a freeway. Thus they are more likely to make dangerous, illegal maneuvers (like an illegal left turn).

    4. Glenn, Interstate is at least twenty feet narrower than King Boulevard. It has two tracks, two through lanes and an intermittent skinny parking lane in each direction where possible. MLK has three wider through lanes each direction with occasional left turn bays, no parking and two tracks. As you noted, both have about the same buffer distance between the trains and cars going the same direction.

       The big difference is that Interstate has been “calmed”. The driving lane wiggles all over the place to avoid station platforms, turn lanes and stretches of parking. There are those cool crossings like in your link every three or four blocks. Many of them have flashers, so it just doesn’t “pay” to try to make time on Interstate. The signal timing is gonna getcha.

       So, cars rarely exceed twenty-five anywhere south of Lombard. On the other hand, Seattle’s MLK lanes are proper stroadways, and the signals are timed for folks to maintain thirty-five for several intersections at a time.

       Pedestrians are at risk all the time along “Renton Way”.

15. This is off topic but I would like an answer. do you know of a map overlay tool i could use for a project? I am trying to overlay the community transit ptba and sound transit rta maps for Snohomish county.

    1. Esri might have them. If you are lucky they might already be in the public dataset and you just need to click to include and show it on their site

       1. Who is Esri? What is the site? How much does this cost? I am trying to overlay Snohmish county city boundaries, PTBA Boundaries, and RTA boundaries.

       2. ESRI (https://www.esri.com/en-us/home) is the maker of ArcGIS (https://www.arcgis.com/index.html), the industry-standard mapping software.

          You might be able to setup a personal ArcGIS Online account and access online data from there.

          You could also see if someone’s turned the Community Transit PBTA and Sound Transit RTA outlines into KMZ files and open them both in Google Earth.

       3. You might also have luck with Google My Maps (https://www.google.com/maps/d/u/0/), but you’d still need to find the outlines as a geospatial file of some sort to upload.

    2. Also, our latest Open Thread (typically a roundup, like the one that was posted today, or a sunday movie) is the best place to post off-topic questions like this.

    3. ArcGIS (desktop software tool) by Esri (software provider) definitely can do anything sort of geoprocessing as described, but if your organization is not licensed or budgeted to purchase license, QGIS is the free/open-source alternative.

    4. You can use data from the Sound Transit GIS resources in ARCGIS or QGIS.
       https://www.soundtransit.org/help-contacts/business-information/open-transit-data-otd/otd-downloads

16. Looking through the report, I realize that it’s easy to focus about the bigger problems. But I think there is some value to looking at where there are less incidents too. What was done right?

    Figures in Appendix 7 report no incidents at Willow, Holden, Elmgrove and Thistle. These are all places where traffic doesn’t cross the tracks. They are in residential blocks though. Some of these do have narrower cross sections because there aren’t any turn pickets at them. That seems to suggest that pedestrian-only crossing streets are pretty safe.

    Of the intersections with turning traffic, Columbian Way/ Oregon seems to notably have the fewest incidents. Why is that? The geometrics and the cross street volumes don’t strike me any better than elsewhere in the corridor. The area has on-street parking, adjacent retail and pedestrians. The intersection does have a curb extension and some green strip treatments that discourage jaywalking. Also, there is a short chain link barrier in between the two tracks. Could it be that the simple chain link barrier makes the intersection safer? Maybe it’s the sight lines? Maybe it’s the having more low shrubs near the corner that helps? Thoughts?

    1. Good point, Al!
       I bet all these treatments help. The bigger the intersection, the more difficult to oversee (such as Othello). Just making everything look more narrow using shrubs, chain link or fence, and curbs and using fewer lanes makes each intersection safer.

17. This might be a bit out there, but one solution might be using the rightmost car lane for turns, including for left turns and U-turns, instead of the leftmost lanes. The idea is a scofflaw driver would be much less likely to cut across multiple lanes of traffic to make a left turn (from the leftmost lane) that is illegal at all signal phases. You would have a dedicated signal phase for the turns, and have the signal be red for the rest of the cycle. I’ve been in several cities that do this, and we even have an example around here with the bus left turn in front of Husky Stadium. An additional benefit to this is it effectively also implements no turn on red, since usually you would be behind a car that wants to turn left or U-turn. Of course, you also install fences to further discourage cars from encroaching onto the train space.

    1. I like it. I’ve seen buses do this and it makes sense for cars. It also makes it easier for trucks to make a U-Turn (although I think it isn’t a problem since the street is so wide).

    2. That’s not far from the way that the MLK and Dawson intersection works northbound today .

       Another option is to eliminate left turns at busy crossings entirely and send the drivers past the intersection and make a U-turn instead.

       This gets at an earlier comment that I made, which is ST is not the sole agency culprit here. Accidents occur and pedestrians get hit even when there’s no Link train nearby. SDOT needs to be proposing solutions.

18. Some less disruptive way to building a trench in phases:
    1. Build a trench on the west side of MLK while the east side is turned into one car lane in each direction.
    2. Once the trench is ready, redirect Link into the trench, take out the middle tracks and replace it with a bidirectional cycle track and some turn lanes.

    1. The problem is how to maintain access to driveways along MLK.

       1. Turn the western sidewalk into a multi-use track. (walking/rolling/fire/access)

       2. I think people would prefer decking a third or half the trench in blocks that have fronting homes. If the single lane were always southbound, it would be pretty easy to turn off the southbound permanent lane at the next previous overpass and then drive (more slowly) down to one’s driveway. That might be three blocks, but it’s an easy movement.

          Coming from the south one would pass home and make a U-turn into the deck lane. Heading north from home, go south and make a U-turn to the main lanes at the first overpass.

       3. There are some driveways, but lots of the driveway access was taken away when ST rebuilt the street 20 years ago.

          And the trench would be narrow enough to accommodate pre-fab bridges above the tracks. A crossover could be installed in a day. A crane would just pick the bridge up off of a flatbed truck and drop it into place.

    2. This is a good idea. It may also solve the problem of “What about that 60″ sewer trunk?!?!” Let sleeping bogs lie.

    3. Yes that’s essentially how to phase a grade separation for MLK. The stations may be messy to add. The tracks would need to be closed for the transition. But even that may be able to be done with single tracking — and in an interim period one direction may be on the surface and the other in a trench.

       Certainly the entire project would need to be carefully designed with concern about adjacent property access.

       Finally, the final street would be much less wide to cross. Even if it was decided to have two through lanes and a left turn lane, it would still be much narrower that MLK is today — even if the current MLK was one lane in each direction.

    4. Yes, there are all sorts of ways of doing it. As I wrote up above it is a fairly simple trade-off: cost versus disruption. If you are willing to live with a lot of disruption you just shut everything down and go to work. This saves money. If you want things moving as well as possible it is the opposite.

       The biggest cost appears to be the sewer pipe. If you could build a trench next to it you would likely dramatically drop the cost of the project. I’m not sure you can though. But it is quite likely the project would be tied to the pipe. That pipe won’t last forever. When it needs to be replaced it would make sense to bury the train lines at the same time. Again this is a balancing act. Maybe the north end of the pipe is fine but the south end isn’t. Normally you would just replace that part. But it might make sense to bite the bullet and replace the whole thing while building a trench.

19. The problem that everyone is overlooking is its not the trains or where they run. The problem is the pedestrians that cross against the lights or Jaywalk, and the drivers that run red lights and turn left illegally. Maybe if there was consequences besides the fact they could get hit by a train, like red light cameras, or cameras that can catch the illegal left turn drivers and start issuing Large fines. Also, There is more then just an accident that occurs, no one thinks of the train operators that have to deal with hitting someone or something. That can be traumatic for some, and also put their career in jeopardy, just because people are impatient or just don’t care.

20. I think they are studying the feasibility of crossing gate now.

Comments are closed.