Last month Mayor Durkan decided to repair the West Seattle Bridge instead of replacing it. This is faster and cheaper, but means the next big bridge project in this corridor (except Link construction) will be decades sooner.

I don’t know if the war-on-cars people ever converged to a position on this subject, but this is the right call, for three reasons:

1. It’s a bad time for new capital commitments. We all have opinions on the post-pandemic future of commuting, but any honest person doesn’t know for sure what traffic patterns will be in 2025. It makes sense to wrap up projects that are almost done (like Sound Transit 2), or putter along in the planning phase of far-out stuff (like ST3), but it’s a uniquely bad time to start motion on a giant transportation project. In other words, it’s premature to take a big step back by canceling something, but also unnecessarily risky to add big new commitments.

2. A “rebuild” is never just a rebuild. I’m no expert on modern highway standards, but the viaduct experience suggests that any “rebuild” is a much larger monstrosity than whatever it replaces. Furthermore, big replacement efforts can’t help but become a Christmas list for community amenities (see the Green Lake Community Center ($) or the much-maligned North Precinct ($)). We’d be lucky to end up with a project with as little climate impact as the current bridge.

3. The future is bright. Bill Gates says that “Most people overestimate what they can do in one year and underestimate what they can do in ten years.” The most involved stakeholders aren’t about to accept a solution without a giant highway over the Duwamish. A few decades from now, after generational change and a longstanding light rail line to the Junction, attitudes might be different.

Kicking the can down the road is often an underrated strategy, and in this case circumstances are especially fortuitous. Luckily, for once, expedience has coincided with the right thing to do.

66 Replies to “Repair is the right choice”

  1. I agree. Once they released the cost estimates and the time estimates, a rebuild was clearly the best option. Buying us time is a good strategy, especially since there may be other parts of the freeway that need work in the future. For example, they could build another bridge, only to find that the West Seattle viaduct is falling apart. If that happens in the future (when the rebuilt bridge is close to its end) it allows us to rethink the entire system.

      1. Likely it won’t fail early, but a brand new WS bridge would presumably have a much longer expected life than the rehabbed viaduct. I agree with Ross’ general point that a repair better aligns the bridge with the other related pieces of infrastructure for a more cohesive rehab, replace, or remove project in a few decades.

      2. AJ wrote what I was getting at (better than me).

        Minor quibble: I mentioned the West Seattle viaduct, not the Spokane Street Viaduct. People generally don’t call it that, even though it is clearly a viaduct ( It is usually considered just part of the “bridge” or “freeway”. The thing is though, even if they replaced the “bridge”, they wouldn’t replace this part, which has plenty of potential problem areas.

        Putting in a new bridge is like replacing the transmission on an old car. Something else might break down, and in a few years you might not want the car anyway. If you can get by with a cheap repair that will by you a few years, all the better.

  2. Martin, here’s the main thing as I see it. For times and situations like the ones we’re facing, the best plans are the ones you make with a tool in your hand and dirt under your fingernails.

    Remember that Dr. Victor Frankenstein considered his creature a work in progress. Also a caution, though, that the trouble started when his planning and took on a life of its own.

    But warning. Anybody links “It’s Alive, Master, It’saliveitsaliveitsalive!” and I’m going to go [ah]!

    Mark Dublin

  3. The new 520 bridge didn’t add any general purpose lanes. It’s certainly wider, but all the extra width is in HOV lanes, shoulders, and a bike/pedestrian path.

    Of course, it could be restriped to have as many as four GP lanes per direction in the future (by elimination shoulders and the HOV lane), but so far, that hasn’t happened.

    1. Extra width also included capacity for HCT (thought that would require additional pontoons). The 520 rebuild also included at least 4 new lids, 2 on each side of the lake, with various transit stops, HOV exits, and bike/ped crossings. It’s basically a textbook definition of ‘replacement’ = something twice as nice and twice as expensive because standards are higher today than when the original was rebuilt.

      To unpack Martin’s point #2 further:
      or if you prefer podcast form, check out Episode 2:

    2. I agree that adding an HOV lane is better than not making it wider, but what would have been best would be to simply convert one GP lane to HOV.

      I’ll also note that the 520 project is spending a fortune on lids. I’m not opposed to lids necessarily, but it’s an example of the mission creep I’m talking about.

      1. Since they had to build a new bridge, my guess is it wasn’t much more expensive to add an extra HOV lane (and the pedestrian walkways). I give them credit for not trying to expand the general purpose lanes.

      2. Ross is correct – the margin cost of each of these (except perhaps some of the lidding) is likely well below the marginal benefit. I’m generally supportive of all the nice things the 520 added; the HOV and Montlake left-hand exit are excellent infrastructure for cross lake transit, the wider lanes (both SOV and HOV) are clearly safer, and the multi-use path is nice.

        But all these things add up and are all incremental costs in a replacement project that do not needed to be funded in a rehab project. Did we need the HOV lane to run the entire length of the bridge? For example, WSDOT early designs for the US2 trestle replacement, a comparable project to 520 in function, show dedicated HOV interchange at I5, but the HOV lane is only a ‘queue jump’ rather than running the full length of the trestle. Having an HOV lane the full length of the floating bridge is a good thing, but perhaps not worth the incremental cost (just speculating).

      3. Did we need the HOV lane to run the entire length of the [520] bridge?

        Hard to say. The old system worked fairly well. It got bogged down at the exit to the UW — something that is far more important now (as buses get truncated). So it is quite possible the HOV lane entrance and exit is more valuable than the extra lane across the bridge. It is also possible that cost more than adding the extra lane. Regardless, it would have been weird to have an HOV lane, then no HOV lane, then an HOV ramp (in the middle?). That might have worked most of the time, but I’m not sure it would have saved much.

        Replacing the US-2 trestle and adding an HOV interchange with I-5, but no lanes across the trestle seems like a bizarre choice. There is no question it would cost more — but without it I have trouble seeing the value of the interchange. My guess is that the trestle slogs both directions (unlike 520).

        Furthermore, most of the buses don’t go on the freeway — they go to downtown Everett. This is before Link gets to Everett. It seems to me like the first thing I would do is add lanes on the trestle, not the interchange (although I have no idea the cost of either project). It is possible that adding the lanes on the trestle would increase the cost dramatically. It just seems crazy to me that Everett is spending all this money on light rail through town but isn’t interested in improving bus connections from its suburbs. It is long ways from Lake Stevens to Seattle — very few people will commute that far, even with an express bus. The 425 probably only gets a few dozen a day (before the pandemic) while the other routes likely get a lot more.

      4. Yeah it would be odd, was just throwing it would there as an example of a Nice Thing for transit that perhaps wasn’t necessary. If 95% of the time the bus can marge into the GP lane and then merge back out, perhaps it’s not necessary? Dunno.

        For US2, yes the HOV (or otherwise managed) lane would access both I5 and local streets (via Hewitt).
        “A new trestle with three general purpose lanes, one of which becomes a short HOV/bus bypass lane at the I-5 interchange.”
        I think similar to how 520 will work? It’s 3 lanes until the Montlake lid, at which point 2 GP lanes continue onward and the left-hand lane has a dedicated exit.
        I don’t know why the 3rd US2 lane wouldn’t just be HOV the whole way? Perhaps it facilitates the 3 road merge on the eastern end of the trestle by waiting until halfway across the trestle to start “HOV only”

      5. @AJ — Thanks for the link. That actually makes a lot more sense to me. Their conclusions also make more sense to me. This is one of those “I could have told you that” studies. Westbound the trestle essentially acts like a giant on-ramp to I-5. In that sense it is like the West Seattle Bridge. You could add 20 eastbound lanes on the West Seattle Bridge and it wouldn’t make traffic flow faster. It just moves the point at which you have to wait (and merge).

        The four lane proposal makes sense as well. The HOV lane doesn’t end at the freeway, but ends in Everett (addressing my earlier concerns). That part looks very nice, really.

        As they pointed out, though, adding an extra general purpose lane doesn’t do much. Right now it is two lanes westbound. For the westbound study, they looked at 3 general purpose lanes, along with 3 general purpose lanes and an HOV lane. In neither case did general purpose traffic move much faster, although they found there is value in having an HOV lane to allow buses and carpools to bypass the traffic. They suggested they need to study two general purpose lanes and an HOV lane (which would be the obvious choice).

        I think the best option is to go with the HOV lane across the bridge, and the ramps to Everett. Even if it is an express bus to Lynnwood or Seattle, it makes sense to stop in Everett. You get off at Hewett, go a few blocks, then you can get back on the freeway using the HOV on-ramps into the HOV lanes. In contrast, the HOV ramps to southbound I-5 add value, but they simply merge on the right side. A bus still needs to merge and move across four additional lanes. That takes a while, and you skip Everett. Hard to see that really making sense.

      6. Yeah, to WSDOT’s credit they evaluated the 4 lane option (because it’s going to be asked for) and correctly concluded that without expanding I5 it does nothing because (as discussed below) chokepoints!

      7. I had a brief period (of a couple of weeks) where I was commuting from East of Lake Stevens into Seattle. This required a transfer at Everett Station to the 512 (there was no 425 at the time). I can definitely attest to the Everett detour being frustrating to no end. It basically adds a good 10 minutes to the whole trip even if you time the connection just right, so I would expect about the same of a detour without a transfer (it would be a timed transfer, most likely, so it would need to allocate time for delays due to traffic, give the operator a break since it’s a reasonable spot for it, etc.)

        All of which is to say that yes, there is value in adding the connection, but there is also penalty. If I were a rider of the 425 today, I would really really hate this, and probably would strongly consider changing to driving at that point.

      8. Good point about going through Everett is good for through-riders because you can then use the HOV ramps via Broadway, in addition to serving Everett itself. I didn’t consider that.

      9. All of which is to say that yes, there is value in adding the connection, but there is also penalty.

        Compared to today, certainly. But consider that a through-routed trip would add:

        1) More frequency, or more routes to serve Lake Stevens.

        2) HOV from the moment you get onto highway 2 until you get into Everett. Also HOV from the ramps on Broadway until you get to I-5. It is also possible that HOV lanes could be added along both Hewitt and Broadway, making the entire route via HOV.

        Very few people ride the 425. It only runs four times a day (towards Seattle). It doesn’t make sense to spend a lot of money saving so little time for so few riders.

      10. Yeah, HOV on Broadway would help a lot, since then you could use the HOV on/off-ramps back to I-5. Hewitt is probably too busy to have space for that. Broadway is weird, I don’t remember it ever being busy except at rush hour during Christmas season… when I-5 is jammed and so traffic spills onto city streets.

        And yes, I totally get the value of “local” traffic. I do wonder whether you would see that benefit though. It’s a question of where to invest the efforts to build ridership. What does Lake Stevens want for its service? Who are Lake Stevens riders today, and who are the potential riders? Why will those potential riders go to downtown Everett by bus instead of driving? Those are not questions I have the answer to. If you do, I would be very interested in knowing more, I think it is a very interesting topic. I certainly hope that CT planners and Lake Stevens + Everett city councils are at least including those questions as part of the planning process, also.

        FWIW, I think there are at least one (maybe two?) local routes going to Lake Stevens, the 280 (on the way to Granite Falls) and maybe something like a 271, I forget. The 280, whenever I have ridden it, has been a mix of low income people and teenagers, particularly skewed towards the latter from my limited (maybe 4-5 rides total?) experience. For better or worse, it’s unlikely that Lake Stevens city council will try to increase teenage ridership to Everett at the expense of people working in Seattle, because the latter are more likely to have a say in city politics, but I do think it’s an important topic to explore.

        Thank you for the discussion, as always.

      11. What if the purpose of those lids is to locate everything from amenities to small industries that’ll not only improve the transit experience and attract riders, but also help fund transit?

        Good metric for “Creep” might be how far you have to be wiling to “Crawl” to get what you want. Industry and service-wide, notice how much a worker’s worth is measured by their willingness to be on their feet!

        Mark Dublin

      12. I don’t have any data on Lake Stevens. I’m basing my comments on the bus network as well as universal commuting habits. Along with the 280 and 425, Lake Stevens also has service with the 109 and 209. (

        I think it is worth noting that there are other buses on the trestle, specifically the 270/271. Like Lake Stevens, Monroe has an express to downtown Seattle, the 424. Like Lake Stevens, very few people ride it (it only runs twice a day). In this case, the big problem is that riding a bus doesn’t get you anything — there are no HOV lanes for much of the way.

        In general, very few people commute as far as Lake Stevens to downtown Seattle. Even without any traffic, it takes 45 minutes to make that drive. The early morning bus takes over an hour, and that is from a transit center (most people have to drive there). Realistically, that means about 3 hours a day commuting, and there just aren’t that many people willing to do that, day after day, year after year.

        In contrast, it is quite reasonable to commute to Everett. The only way you are going to get decent ridership to Everett is if going across the trestle by bus is significantly faster than driving (otherwise you’ll have the same problem as Monroe). At that point, someone commuting to Everett will find it more attractive. It would be a detour to go to Everett and make a stop, but not a huge one (I wouldn’t loop around the transit center — I would simply go on the new ramp to Hewitt, then a single turn to Broadway). This would connect well to buses serving central Everett, which means that lots of people would get on and off there. The bus would keep going to Lynnwood (assuming Link terminated there). Of course if the train ends in Everett, than the bus would as well.

        Without traffic, it takes an extra three minutes to make the detour. You would encounter traffic on Hewitt/Broadway, but you would encounter traffic merging onto I-5. You would add a couple bus stops, but my guess is you are looking at a trip that would be about five minutes longer.

        To be clear, I wouldn’t rule out CT continuing to run express buses that bypass Everett. But if improvements are made are made on the trestle, then way more riders will be headed to Everett, and it just isn’t worth the extra money to save a handful of long distance commuters such a small amount of time.

      13. Yeah Ross has the right framing. Once you get out as far as Monroe and Lake Stevens, it’s important to first focus on more local destinations – Everett downtown, the various community colleges, Boeing, etc. – even if there are many Seattle bound commuters (>10K people according to this article:

        Once Lynnwood Link opens, this will be a moot point. Lynnwood TC will be a major transit node for CT and the terminus for pretty much all CT’s I5 routes. Once you get a rider to Link, CT doesn’t have to worry if the person is heading to Shoreline or Seattle.

        It will be interesting if the various US2 routes Ross highlights will be a 1-seat or 2-seat ride to Lynnwood TC. Probably depends on the route length? I could see a route like 270/271 or 280 end in Everett while a 425 continue to Lynnwood. But I think Ross is right – even if the route heads onwards to Lynnwood, the 5 minute detour through Everett city streets is likely worth it for the many connections it creates.

        For example, CT probably won’t ever run a bus from Lake Stevens to Everett community college, but I could see many a student catching a Seattle/Lynnwood bound bus across the trestle and then hoping off in Everett and catching a bus on the other side of the street to get to the CC.

      14. Thank you, both RossB and AJ, for the comments.

        I generally agree with the intuition – it makes sense. And I do think that terminating at Lynnwood TC would make a lot of sense once that option is in place. Which makes it that much more important to not require a detour through downtown Everett for the express-routes-to-Link, in my opinion (though it could be mitigated with HOV and queue jumping strategies, as RossB pointed and I concurred).

        I guess I would still like to see the CT planners do the analysis I suggested (and trust that they do). I know a number of people who went to ECC over the years, and none of them were reached by the 280 today. They could drive to meet it half-way and ride the rest of the route into Everett (and then switch to an ET bus), but if you’re already in the car, as Daniel Thompson keeps pointing out, the sensible option is often to keep going, if the distance is short and the impediment to driving is not large (the US2 trestle, to be fair, _can_ be a significant impediment, and marginal cost of extra driving is a thing, too). I am not saying that everyone will do this, of course, and it is entirely possible that there will be enough such transit users to justify the investment. But from the little I have seen of the 280 and other related CT local routes, the ridership is pretty low. I admit that my evidence is years old and very anecdotal at best, though :) So I am mostly looking for information, not trying to persuade anyone not to bother to invest here.

        Thanks again for the fun discussion.

      15. I think in low density, suburban environments, bus-lanes make all the difference. You’re already on the wrong end of the ridership-frequency cycle. By that I mean that the agency can’t possibly justify running the bus every 5 minutes. More like every half hour, even at rush hour. So you are already paying a big frequency penalty. You pay every other penalty that comes with transit (the extra walking, etc.). Making matters worse, Lake Stevens is especially low density, so most riders still need a car to reach the bus stop. The only reason to take the bus to Everett is just to save wear and tear on your vehicle (or to read while on the bus).

        Add bus lanes on the trestle and things change. Suddenly the bus flies by the cars. This makes up for the other disadvantages to riding transit. This in turn increases ridership, and you get more frequent buses.

        I want to mention that I don’t think bypassing Everett is a terrible idea. There are plenty of express buses. But it only makes sense if you’ve saturated the corridor. If you are running buses every 5 minutes from Lake Stevens, *and* can fill an express bus to Seattle (or Lynnwood) then by all means, run the express. I just don’t think you will be able to do that. As a result, you will gain a few riders that appreciate the savings that come from skipping Everett, but lose more that don’t like the infrequency.

      16. (Sorry for the late commentary. Just getting to this blog piece this afternoon.)

        “It will be interesting if the various US2 routes Ross highlights will be a 1-seat or 2-seat ride to Lynnwood TC.”

        I agree. As a SnoCo resident, I’ve tuned in and out of this project for a number of years (it’s been on WSDOT and SnoCo’s to-do list for quite a long time due to funding challenges). I think the current WSDOT plan is pretty solid even with just the partial transit lanes. I’ve been in those queues on the trestle westbound multiple times when returning home from visiting friends in Lake Stevens. (They had a big rural property and I would take my dog out there to run around and play with their dogs.) The I-5 interchange here is a textbook case of a choke point.

        Speaking of said friends, one of them at one time* did indeed commute from Lake Stevens to First Hill on a daily basis. He started out taking transit and, at first, enjoyed the quiet time to get a few extra winks in on his way to work in the morning. I remember him telling me that there were few other riders doing likewise and, based on the comments here (I haven’t looked at the ridership data myself), apparently that still seems to be the case. Soon enough though, he grew tired of the daily slog to and fro and started driving instead. The transit commute just took too long, particularly the late afternoon return trip which could be an absolute nightmare. They moved up there from Ballard in the early 00s because they wanted a sizeable property and at the time there was a discussion of his boss opening another clinic in the Everett area (which never happened though). As the saying goes, the best laid plans…

        *These friends eventually moved to another state once the area of Lake Stevens their property was located started to rapidly develop and subdivisions started going in. Once the developers started calling with ever increasing offers, they took the money and got out of town.

      17. “you’re already in the car” – that logic only holds if the remaining trip is (nearly) free. Tolls will nudge some to switch to the bus, and paid parking certainly will too. Lake Stevens is a great example of a community that would be well served by a string of small P&Rs that can generate the ridership to sustain moderate frequency. If your destination along I5 has free parking, then sure it probably does make sense to keep driving. But if it’s $20/day to park where you are heading, many people might choose to park and switch to a bus. Parking in Seattle is certainly expense, and as cities like Lynnwood and Everett (downtown) grow and densify, parking will become more scarce and therefore expensive, nudging more to transit in what will hopefully be a virtuous mode shifting cycle.

        Good discussion y’all.

      18. A focus on local destinations is ok to a point, but getting to Seattle should not require an entire day of bus riding.

        In the case of Lake Stevens, they already have a regular bus that goes to Everett, so it’s fine for the peak overlay to skip it and go straight to Lynnwood for more direct service. Remember, it’s not just about getting to downtown Seattle, but also Bellevue, the U-district, or anywhere else you can get to from a bus or train that serves Lynnwood.

        Monroe is, perhaps the most eggregious CT example of focusing too much on local destinations. They have all-day buses running to the tiny communities of Sultan and Gold Bar, while the SR-522 corridor to the greater King County eastside gets no service at all, except for a couple of peak-hour runs on the 424. A minimum viable level of service would consist of an hourly bus from Monroe to Bothell, where you can connect to the Sound Transit BRT routes to access the rest of the region (and get you to downtown Seattle much faster than going through Everett). I would pay for such a bus by discontinuing all-day service the Gold Bar. The communities are tiny. The ridership is very tiny. And the mere presence of having the bus fight traffic out there adds unpredictable service delays to the rest of the route – especially on weekend afternoons. (Not that Monroe ridership potential is big – it’s clearly not – but it’s certainly bigger than Sultan/Gold Bar).

      19. For whatever that’s worth, the bus to Gold Bar and Sultan seems to be primarily as a pure coverage run – the last resort connectivity. It’s hard to argue against that. I look at it as being similar to the Grays Harbor route that goes towards Humptulips and Quinault – even the locals barely ride it, but if you really need to get out of town (for example to escape an abusive spouse), there is _something_ to allow you to do it. I get why it’s a balancing act, but I guess I would err to keep some version of it along Highway 2, too.

      20. “Monroe is, perhaps the most eggregious CT example of focusing too much on local destinations….”

        Nah. AM has it right. The small communities served by CT along the SR-2 corridor are part of the agency’s community service network, which is part of the agency’s mission to reach all parts of the district with at least some coverage. This accounts for about 15% of CT’s passenger boardings annually. Here’s how this part of the network is described in the agency’s latest six-year Transit Development Plan:

        Community-based service feeds core service and connects outlying communities. Routes
        in this category are less frequent but more flexible than core routes, sometimes following a less direct path to link smaller scale destinations. While not considered to be trunk lines,
        community-based routes play a vital supporting role in the transit network and provide 15 percent of all Community Transit passenger boardings.

        “Feeder Routes-
        In southwest Snohomish County, the Marysville-Tulalip area, and the Highway 2 corridor
        from Everett to Monroe, local routes provide neighborhood connections and carry riders to
        core service routes. We call these “feeder routes”. Feeder service includes Routes 106, 109, 111, 112, 113, 119, 120, 130, 209, 222 and 271.

        “Rural Routes-
        In less-densely populated areas of north and east Snohomish County, rural routes provide important connections between outlying communities and the core service network. Rural service includes Routes 220, 230, 240, 270 and 280. Rural service accounts for 4 percent of all Community Transit passenger boardings.”

        My own neighborhood here in the SW SnoCo UGA is served by one such feeder route that only runs hourly and has very light ridership (the last I checked). While the frequency leaves a lot to be desired, I’m happy to pay my additional 1.2% in sales tax to CT to operate such routes to fulfill this part of its mission and serve the folks who depend on these types of feeder routes for their mobility.

    3. I opposed the plan to widen I-90 to four lanes in each direction because I thought the narrower lane width would be dangerous and cause a lot of accidents. I was wrong. The four lanes work pretty well, even during peak times, and eliminated bottlenecks where the old configuration would narrow, especially on I-90 in the middle of Mercer Island. There have been few issues with cars stopped in the narrow shoulders because cars are more reliable today. It hasn’t worked as well for buses however because the HOV lane is the far lane from exits and entrances.

      The irony is now traffic moves quite well across the bridge span, even during peak times, until it reaches I-5 and 405, one of which is very poorly designed with merges and entrances/exits on the wrong side (Mercer St., 520, the convention center), and the other simply over capacity. Since WSDOT controls ramp metering, and ramps are set at the lowest levels of service, and Mercer Island is the last stop in both directions, our ramp metering is slooooooow.

      The thing with HOV lanes on a bridge span is they are hard to enforce because there is no camera and no place for a police car to sit. HOT lanes work better for this. The problem with HOT lanes is they depend on traffic congestion to earn revenue, so widening 520 would reduce congestion and reduce toll revenue, and at the end of the bridge you still run into I-5 or 405. Plus the cost of 520 required all lanes to be tolled.

      One benefit of East Link is it will likely convert the HOV lane on I-90 to a general purpose lane because devoting an entire lane of an interstate to HOV only without transit makes little sense (and in fact today with no enforcement it is basically general purpose, and buses rarely use the HOV lane during peak times because it is too hard to travel across three lanes from the entrance to the HOV land and back by the time you reach Mercer Island).

      A general purpose lane across the bridge span was an option offered by the FHWA instead of the HOV lane but Seattle objected because Seattle thinks cars are evil. With working from home I wonder if Seattle will feel the same about commuter traffic.

      The FHWA also approved a HOT lane I supported because Mercer Island has the shortest distance hence lowest cost, it would reduce congestion in the HOT lane which benefits the “privileged”, would balance out congestion with 520 which is tolled, and would allow SOV access from Island Crest Way, but for some reason people — even on Mercer Island –go crazy when you mention tolls even if the benefits outweigh the drawbacks.

      1. For those who understand choke-points, it’s not ironic that resulting a capacity issue at one point in I90 simply removes the choke-point to somewhere else in the system. Widening 405 south of I90 should alleviate some of regular backup, but it mostly just moves the 405 capacity problem back onto 405.

        As a regular user of the I90 HOV lane, I would disagree that HOV serves no purpose without transit. While transit might be the majority of riders, it’s a small fraction of the HOV vehicles. As someone who regularly points out that transit is useless for most trips, I would think you’d be a big advocate of HOV lanes. WSDOT has built the vast majority of their recent Puget Sound highway spending on the logic of creating a cohesive HOV/HOT network (with hand-waving towards freight whenever they build 2-lane freeways), so I don’t seem them going back to SOV; even if you think WSDOT is operating in bad faith, politically it wouldn’t make sense.

        Converting I90 to a HOT lane, however, seems like a good idea IMO, particularly the direct access ramps on MI. But I’d rather see both bridges tolled equally, and the funds directed elsewhere like the bridge tolls in the Bay Area and NYC.

        Why would HOV have an enforcement issue but HOT does not? If anything, there’s a bigger incentive to cheat in a HOT lane (save money AND time).

      2. For those who understand choke-points

        I think a lot of people don’t understand choke points. As someone with way too much time on my hands who happens to enjoy going to the mountains on weekdays, I often think about choke points. As has been said many times — cars don’t scale. This means it is a zero sum game. For example, I remember driving home, and hearing on the news that there was an stall northbound on I-5, causing a big backup. Bummer for them, but good news for me. The merge from I-90 to I-5 went smoothly, and I had an easy drive home.

        This is why you can’t build yourself out of congestion. You just send it somewhere else. Ironically, the biggest beneficiaries of the new 520 bridge will be general purpose drivers. To be clear, things will get a lot better for HOV-3 riders and transit users. But the old design created a major choke point westbound in the evening. While there weren’t that many cars and buses in the HOV lanes (enough for that lane to flow freely) there were still enough to make the other lanes crawl. Essentially, all those cars had to wait for all of the other cars and buses to pass them. Three lanes both directions will solve that problem. It is likely the congestion will spread somewhere else, but that is the nature of choke points.

    4. asdf2, remember that lane-usage is a rule-maker’s choice, not an engineer’s. Also their employers who are also known as voters. Whose word really is, well, law.

      My own favorite example right now is the Tunnel Bored at just the right Depth to exclusively carry either or both trains and buses should it’s tenure as a parking lot expire.

      In the average “converted” lane, rails are laid with a jack-hammer, not a meeting-gavel. In the Pine-Street cut-and-cover near the Paramount, watched ’em do it.

      But word to Education: Making grammar boring is even worse for citizenship than doing it to history. But exchanging the Active for the Passive “Voice” is a very loud choice for Advancement.

      Mark Dublin

    5. the Portage Bay viaduct will expand to six lanes from four; that added capacity and cost was unnecessary.

      1. Yeah, but it is all HOV lanes. It is a reasonable thing to do, if you ignore Link. Prior to Link getting to the UW, it would be a nice bonus to have the 520 express buses go on HOV lanes right to downtown (my understanding is that the HOV lanes will connect to the express lanes). However, those buses should terminate at the UW. Ideally we would have both, but we don’t have money for both. It doesn’t make sense to water down our network so that folks have an express — better to increase frequency on the other routes (or extend them into more neighborhoods in Kirkland/Redmond). That being said, it is quite possible that Metro and ST will run buses similar to the 544, even though I think it is a poor value.

        The main beneficiaries of the extra lanes will be carpool users. Carpool use has gone down in years past, but it is still around 10%.

  4. “2. A “rebuild” is never just a rebuild.”

    Thank you for saying this. We use a lot of previously built public assets that aren’t perfect, but work well enough. But as soon as something wears out and must be replaced, that new asset costs many times the inflation-adjusted cost of the original asset.

    I look at the NYC Subway – it would never be built like that if it was built today. Narrow platforms and stairways, awful ventilation, no ADA access at so many stations and, dare I say, at most stations no artwork.

    1. Different cultural assumptions, [Another] Alex. To the people of those days, of all professions and classes, a magnificent creation like that subway was the definition of “A Work of Art!”

      Mark Dublin

    2. Well I agree with AJ on most of his points, although I never said transit is useless for most trips, and I don’t know where you got that.

      The reason HOT lanes are easier to enforce than HOV lanes is HOT lanes have cameras, and the funding and infrastructure to monitor the lane. HOV lanes normally do not. They require a police car parked somewhere monitoring the lane.

      I agree both I-90 and 405 should have been tolled to equal out congestion, but it is hard to have a rational conversation about tolling when you live on an Island. Lots of “equity” arguments.

      I drive the I-90 bridge span to and from Seattle five days/week, and the HOV lane is regularly used by SOV’s. So it might be called a HOV lane, but buses generally don’t have the time and distance to use it, at least to Mercer Island, and many SOV’s use it because there is no way to enforce it (enforcement was possible west of the tunnel until ST fenced off the area where the police would park). Now there is no traffic and no need to use the HOV.

      I-90 will be a little unique because there will be no transit using the HOV lane when East Link opens (unless express buses continue from areas not served by East Link).

      The FHWA has indicated it has no objection to a general purpose lane, HOV lane, or HOT lane at that time. With working from home this all might be a moot discussion.

      One of the negotiating points with ST over the bus intercept configuration on Mercer Island is ST’s help in converting the HOV lane to a general purpose lane, although a HOT lane would be much better for Mercer Island because it would allow SOV access from Island Crest Way and remove that traffic from the town center and same arterial the bus intercept will have to use, and limit traffic in that lane. I think it will come down to what Bellevue wants.

      I am all for a HOT lane but tolling in the past gets a lot of pushback from Islanders and those east of Mercer Island, even though there are not many HOV’s using the lane. I will have to wait for the two lane highways you mention, unless you are talking about US-2. Right now we seem to be moving in the opposite direction with I-90 and 405, and as I noted converting I-90 to four lanes helped with congestion quite a bit, without “inducing” car demand.

      1. converting I-90 to four lanes helped with congestion quite a bit, without “inducing” car demand.

        Wait, what??? I remember when they added lanes to I-90. People were excited, as congestion simply went away. I also remember traffic started getting worse and worse — to the point that it was worse than ever. It was a textbook example of induced car demand.

      2. You keep pointing out that transit is not useful for most (your word) east siders, which should suggest that an HOV lane across Lake Washington would still be useful for many carpools.

        For the island equity issue, it seems like the easy solution is to just toll the sections between MI and Seattle. That way, Islanders can still access various essential services in Bellevue, etc. without needing to pay a toll. Trips from MI to Seattle where Link (+bus) isn’t sufficient are very small.

      3. I’ve suggested there should be a 50% toll discount for cars entering/exiting Mercer Island. That would recognize that islanders use only half the road at a time, and address the fact that they have no other way off the island.

    1. That’s not accurate AJ. I have said transit has inherent issues due the the size of East KC, and the lack of density and steep topography, which makes first/last mile access difficult.

      I have also noted the Eastside is primarily a car culture in part due to the family structure, and transit on the Eastside is primarily a peak commuter issue.

      I have also noted ST subarea equity has twisted some of the projects on the Eastside, like rail to Issaquah, and that adding a seat to commuters who live south of I-90 when East Link will be unpopular.

      I have not drunk the cool aid on transit. I don’t think it will “kill” cars, or solve global warming, and that housing zoning desires will trump transit depending on the jurisdiction. Working from home is a good thing to me because commuters should not waste their lives on packed buses to earn a living.

      But if there is one point I try to make it is those who use transit pay a very small percentage of its costs due to large general fund subsidies. Revenue and funding will — and is —determining the scale and frequency of transit.

      I pay a lot of taxes to subsidize transit but don’t use it or demand transit services for my community, without too much complaint. Is there something more you would like me to do?

  5. The first demand by West Seattle residents was no loss of car capacity in any new bridge so that pretty much made repair the only option. It certainly didn’t help when the W. Seattle residents learned light rail would be extended five years, which raises concerns about the real timeline (and second transit tunnel). I don’t think building a new bridge would have accelerated ST’s new timeline for running rail to West Seattle, so residents could see a decade of lower car capacity but no light rail from a new bridge.

    At the same time The Seattle Times today noted the funding issues that are appearing for already very expensive bonded projects.

    Seattle is in a budget crunch, and Move Seattle at $1 billion was not seen as a good use of funds. Federal, state and regional funding was a big if.

    All the issues Martin raises are valid. If as many on the blog believe frequency is the critical component, and we have no idea if the peak hour commuter will return, and state, regional and federal funding towards a new bridge was unknown, a new bridge made no sense, especially to a community that some see as primarily white and “privileged”. West Seattle could actually see current transit levels lowered as transit service is reallocated to disadvantaged communities.

    What Bill Gates didn’t say is the future one year out is knowable, but the future 10 years out is not. It will take at least until the middle of 2022 to learn whether transit riders will return to packed buses and trains, general fund and gas tax revenues return, to what extent the peak commuter returns and what transportation they use and where they want to go, new forms of transportation and their cost, and many other issues.

    We have built a very expensive transit system (and roads too) based on peak hour travel capacity, and that peak travel funds a lot of our roads and transit. One proposal in the Times’ article is to raise tolls (although that is difficult if there is no congestion and there are alternatives, sometimes on the same road like 405), and increased fares on transit must be an option, but of course the best transit riders to raise fares on are peak commuters.

    Some see a transit only bridge in the future, some see transit completely differently in 20 years, most think travel/commuter patterns will change to some degree at least. In the middle of a pandemic and recession is not the best time to predict the future.

    If transit and Urban advocates are going to convince citizens in communities like W. Seattle to drop their no loss of car capacity demand for bridges then someone is going to have to truly convince them that is a viable option, they will get rail, there will be a regional system with decent frequency, and the route of light rail is where they want to go in the future.

    They are not going to buy the war-on-car dogma until they see some proof of a viable alternative, and extending rail five years and cutting Metro service 25% through 2040 certainly does not build confidence in a pretty non-dense community with steep hills like W. Seattle.

    1. Nobody knows if light rail will be delayed 5 years or not. That is simply a proposal at this point.

      I thought West Seattle light rail was not dependent on the 2nd downtown transit tunnel? The plan is to operate it as a spur for 5 years.

      1. The pre-covid schedule has a West Seattle-SODO line in 2030, and Ballard and DSTT2 in 2035. At that time, West Seattle would be connected to Everett through DSTT1. Tacoma would be connected to Ballard through DSTT2. This was always a silly plan because a stub to SODO would not be used much. The C would still run and absorb most of the riders. The stub was for the political goal of doing West Seattle first because it’s so privileged. It’s possible the board might reconsider this order now. But unlikely.

      2. That’s right about the original dates, Mike. I’ll add that the ST3 2030 date for West Seattle assumes no tunnel section, so that if West Seattle wants part of the line underground, it’s already probably too late for a 2030 opening (considering that environmental approvals are still a few years away and it takes about three to five years longer to build a tunnel station — or eight to ten years in total). Never mind that there is not enough contingency in ST3 to fund a tunnel section in West Seattle so additional funds (likely an unnamed local or state source) would have to be secured to get an FTA grant.

    2. Move Seattle at $1 billion was not seen as a good use of funds.

      Says who? It passed in an off-year election by around 58%.

      The Murray-Kubly fiasco that followed is a different matter. But that had a lot more to do with the leaders lying to us — saying they could built those projects for that amount of money, even though they knew they couldn’t.

      It is possible that Seattle won’t vote for infrastructure projects (unless they are heavily geared towards transit) but that really isn’t the issue. My guess is most of West Seattle wanted the repair, simply because it would be done sooner. That left very few people wanting a new bridge. You had those who favor big road projects (a minority in Seattle) but not West Seattle residents. That is a small portion of the city. The mayor made the right choice, both from a practical and political standpoint.

      1. That was my point too, Ross. Move Seattle promised way more than it delivered. I agree the time to repair vs. rebuild the bridge was a big factor. So was the cost, and as Martin noted building a new bridge always comes with unknowns, which is why these public projects often come with 20% cost overruns in the bids. West Seattle residents demanded no loss of car capacity, not me. Maybe that was just a vocal minority.

        Here is a April 17 2020 post by Dan Ryan on this blog. Dan notes the bridge normally carries 100,000 cars/day and 25,000 transit riders, (and kudos to Dan for questioning the 2030 timeline for light rail to West Seattle back then). I don’t know why you think the 100,000 cars that travel the W. Seattle bridge each day is a small minority.

      2. I don’t know why you think the 100,000 cars that travel the W. Seattle bridge each day is a small minority.

        Because I never wrote such a thing. Here, I’ll break it down for you:

        You had those who favor big road projects (a minority in Seattle)

        I’m making the case that a minority of Seattle residents favor big road projects.

        but not West Seattle residents.

        So within that minority, we can throw out those in West Seattle. My guess is West Seattle represented a considerable number of those who support big road projects, especially ones involving West Seattle. Thus it is a subset of a minority who would favor a big bridge. That’s why I wrote

        That is a small portion of the city.

        Got it now?

        Oh, and the 100,000 is for vehicles, not just for single occupancy cars. There are trucks and buses within that number.

    3. “Move Seattle at $1 billion was not seen as a good use of funds.”

      No, Move Seattle was not large enough to fully complete its projects. That doesn’t mean those projects were a bad use of funds. They were part of Seattle’s Transit Master Plan drawn up in 2013, which was a better network than what we have now. The problems in Move Seattle was that they weren’t clear that Move Seattle would only partly fund the RapidRide lines and the rest of the money would have to come from somewhere else, and then the Kubly SDOT had overoptimistic budgeting (like ST in the 1990s) so even its contributions turned out to fund less of the projects than they said.

      1. Ross, Dan Ryan wrote 100,000 “cars” and 25,000 transit passengers in his article I linked to. I simply repeated Dan Ryan’s statement.

    4. “The first demand by West Seattle residents was no loss of car capacity in any new bridge”

      One issue here. When did West Seattle residents make that demand? I don’t recall a vote on it, or a poll that could claim to reach “everybody”. Instead it’s a long-term assumption that West Seattlites always strongly oppose losing GP lanes because they’re so suburban-minded, like they were about getting the high-level bridge built in the first place. So the politicians are working around that assumption to avoid opposition. While that goes in the same direction, it’s not exactly the same as a majority demanding no loss of GP lanes in this project as an absolute first principle. West Seattle has gotten more diverse transportation-wise, and has stronger support for good transit lanes and light rail and other non-car modes than it did twenty years ago. Those people wouldn’t say no loss of GP lanes is their absolute first priority.

      1. One issue here. When did West Seattle residents make that demand? I don’t recall a vote on it, or a poll that could claim to reach “everybody”.

        A vigilante Facebook group led by a West Seattle based realtor, who has likely seen a drop in business and has a financial interest in opening the bridge ASAP and a structural engineer, who has apparently never worked on a bridge. They claim to speak for everyone because they got a bunch of “likes”.

        In typical Seattle fashion, the City decides to listen to a very vocal minority, so we end up with a bad end product. See actions by Safe Seattle and also the Missing Link Obstructionists.

      2. @RapidRider: There is a saying: “the squeaky wheel gets the grease”.

        I am mentioning this because it is common to blame broken Seattle for things that are in fact human nature, and fighting human nature is hard. I think that this is one of them. Yes, this is a minority, yes, they are loud, yes, the majority is silent. Yes, it is common for the majority to be silent, too. These are all things that happen everywhere, not just here.

        It is also important to internalize this, I think, from the perspective of the activist trying to change things. Change is hard, and slow. The best way to do it is, I think, through repeated engagement and education of the general public, which requires quite a serious time commitment. It also requires honing messages that _can_ reach and convince the general public, which is why I think it is important to practice messaging in places like this forum, where the eyes are (mostly) friendly and the criticism (mostly) constructive.

        Apologies for making a rather more general comment than perhaps applicable to the specific issue, but I think it is a message that bears repeating. And so I am just using my own approach of repeated engagement on my own pet topic ;)

        Thank you to both for the discussion.

      3. There is a saying: “the squeaky wheel gets the grease”.

        The issue is that while the squeaky wheel gets the grease, the rest of the car rusts away. The vocal minority is temporarily placated, no politicians have to make decisions that could affect their reelection chances and we’re left with a less than ideal situation.

        Missing Link: BGT moved to Market for a short stretch, rather than 54th public ROW where it had been planned for decades. The two or three industrial businesses will be gone within 5 years and we’ll be stuck with a multi-use trail that is less than ideal for trail users and the businesses along Market. Winner: two or three private, lawsuit happy businesses.

        Safe Seattle: Instead of implementing a bunch of solutions that are proven to help the homeless not be homeless, we get nothing. According to Safe Seattle, we should not help the homeless, because…it enables the homeless or something? Winner: Harley Lever’s 15 minutes of fame?

        West Seattle Bridge: (See summary for current scenario). Winner: …? Maybe Kevin Broveleit’s financial security until the quick fix fails?

        Honestly, as a North Seattleite, I couldn’t care less about the WSB. West Seattle let the vocal minority push for a rapid, unproven fix. If it fails, they can wallow in their own hubris as they deal with a (now) completely failed bridge AND traffic has returned to normal. But as a taxpayer, I want this bridge correctly fixed or replaced once. Instead, we’ll be paying for it twice, likely within this decade.

        The Ballard Bridge replacement is due in the coming years, something a little closer to home for me. I can only hope that a similar situation doesn’t play out where we trade short term gain into long term pain.

      4. @RapidRider: thank you for the response. I am in complete agreement with you that such outcomes are very unfortunate, and should be fought against. The points I am trying to add are that it is not a Seattle-only problem (which was strongly implied in your phrasing – if I misunderstood, I apologize) and that it has an easy solution (which was less directly implied – again, if I misunderstood, I apologize again). My sense is that this is very common and that the solutions are hard and require a lot of effort – in many directions, but primarily related to outreach and driving more community involvement in the process, is my sense.

        You mention you are a N Seattle resident – another project that you may be familiar with is the 35th Ave NE mess, with the “Save 35th” vs. “Safe 35th” competing efforts, and the mess we all ended up with there. The good thing, to me, about that process, is that there was strong community involvement. The bad thing is… it was not enough, as the vocal minority still got a lot of what it asked for. This is what I mean about the solutions being hard to find and requiring a lot of effort. I freely admit that I am guilty of being in the (mostly) silent majority in that particular project, too, and it is something that I regret, now that I have the benefit of hindsight.

        Thank you for the interesting discussion.

      5. @AM: You’re correct, I unfairly picked on Seattle. In my defense, I’ve only ever lived in the Seattle metropolitan area, so I have a very myopic view!

  6. West Seattle is just a preview of a conversation that is going to happen in virtually every Seattle neighborhood in the mid-to-late 21st Century. It’s already happening with the Magnolia and Ballard bridges. It will continue to happen as other major bridges reach the end of their lifespans.

    So far, there has been near-zero momentum to downscale any type of car infrastructure. Instead, we’re seeing a significant expansion of car infrastructure in several important locations including Mercer St, the Montlake/520 corridor and SR-99/Alaskan Way. Seeing the status of the Ballard-Interbay study suggests that car infrastructure expansion will continue in the future.

    The question is–will light rail fundamentally change the above? I have my doubts. But at least it should take some pressure off the constant desire to expand car infrastructure. But if current trends continue, I don’t know if I will ever see Seattle take away even a single lane of car travel over any major corridor/bridge in the city.

    1. So far, there has been near-zero momentum to downscale any type of car infrastructure.

      Oh, I don’t know about that. SR 99 is significantly smaller than the viaduct. There are fewer lanes and fewer ramps (including no ramp for Western). (So you have actually seen them take away lanes). It takes up a much smaller part of the above ground space. It cost more, but all of that money went into making the neighborhood nicer, as opposed to adding (or even retaining) lanes.

      It is highly unlikely that they will rebuild the Magnolia Bridge. It is too big for the number of cars in Magnolia. They may build another, much smaller bridge.

      Fixing the West Seattle Bridge was the cheapest option. I don’t think there was any option that would have had fewer lanes (maybe underground, but that didn’t pan out). Unlike a some projects, they wouldn’t have saved any money by making it smaller. My guess is Ballard will work exactly like West Seattle. It just gets replaced as is. There is the possibility that it will be moved or otherwise altered as part of an integrated light rail project, but that remains to be seen.

      I think is likely to continue. You will see small reductions, when it saves a considerable amount of money, but otherwise we will see the same amount of lanes. (For example, the proposal to build a second Montlake Bridge is pretty much dead, due to community opposition).

      1. Some of the viaduct’s lanes went to the boulevard. The tunnel was sized for what they estimated through traffic would be. Downtown-area users were diverted to Alaskan Way, which is growing from four lanes south of Columbia Street to four GP lanes, two transit lanes, and a ferry-queuing lane.

      2. You’re right in general, Ross, but the truth is that the Ballard Bridge cannot be replaced unless it is moved laterally. One can’t “shoofly” a 2/3 mile long, sixty-foot high opening bridge across an active waterway. The replacement must be completed while the original structure remains in service.

        In the case of 15th NW, that can happen directly adjacent to the existing bridge, though that would certainly impact Fishermen’s Terminal in the same way as an LRT bridge alongside the car bridge.

        Moving the roadway to 14th and then building LRT in the envelo oe of the existing road bridge, though creating some traffic flow issues, is by far the best choice. It is also difficult because two bridges would have to be completed before 2035.

      3. The replacement must be completed while the original structure remains in service.

        I’m not sure about “must”. They could just close off 15th, while they build a new bridge. There would be plenty of traffic on the Fremont Bridge (maybe the Fremont Bridge becomes “freight and transit, like the lower West Seattle Bridge”) but that could be the approach.

        I agree though, building a new bridge at 14th while the old bridge still works is by far the best option. Avoiding a terrible period of no bridge also helps politically. Those who would much prefer a new bridge on 15th at least avoid a long period with no bridge at all. Now we just need a new mayor (or someone) with the foresight to make it happen.

  7. Daniel Thompson, here’s the problem with controlling lane usage by enforcement. You would not handle landing and take-off at Sea-Tac with that tool, would you? At best, good temporary measure while the solution’s stuck in traffic known as Time.

    And morally, racially or not, income’s even worse for speed-control. Talk about pure putrid Privilege! With prejudice to Productivity itself, let alone schooling, health and safety. Two employees, one a hospital president, and one, a nurse in Emergency.

    If salary dictates priority, which one’s late arrival will kill fewer people? Enforce lane priority by Essentiality measured by Mortality-Prevention and we’ll talk.

    About Mercer Island, my intent is not to mess with you at all. The place is Regionally beautiful. With major possibilities for greatness as a just desert. (Entitlement, not mousse.) Until 1938, Roanoke Landing served the Fleet’s biggest ferry.
    With Shilshole at the other end, a hydrofoil will work much better.

    What I cannot see is why terminating bus-routes can’t be nothing but good for businesses like the place I went for lunch and coffee. And the former cafe across from Link and the 550’s that the city now owns, but should once again serve passengers. Who’ll be mainly also residents.

    If the problem’s civil engineering, can’t believe the talent isn’t there to fix it. But if it’s really a matter of which passengers are “unwanted” for clientele…..I think Restriction’s going to find itself Outvoted.

    You and me and all our readers? I think I’ve got a winning [ah] remedy. Though talk about a pandemic! Roll-Playing. One party pretends to be a driver forced to pick the 550. The other one, a Mercer Island passenger who’s just found their last parking-space “Tow-Posted.”

    In real life, what’s either one of us going to do but figure something out? Though the station-side caffe I do envision, will exert itself to make sure we’re both welcome there, and highly Free To Speak no matter what. Will definitely add ATMOSPHERE!

    Mark Dublin

  8. Problem with coverage on Jenny Durkan’s announcement yesterday: The idea that politics is a bad thing to govern with. Because across the years by all accounts, it beats clan violence and civil war by a mile.

    Can we look at politics as our own machine machine tool for creating useful things like transit systems? Anything of mine that valuable in my garage, if I’ve got problems operating it, I’ll get fixed long before I throw it away.

    And also get some training in its operations. Could be my model’s out of date. Exact same trouble as my computer. And also maybe switch my school from UW to Lake Washington Tech, Shoreline, or Highline Community Colleges.

    Mark Dublin

Comments are closed.