The Washington State Department of Transportation will close all but one lane of westbound I-90 for a full week from July 18-25 to replace expansion joints on the East Channel Bridge between Mercer Island and Bellevue. This closure will occur on a summer weekend AND the following full workweek AND a week in which the Sounders play a Tottenham friendly AND the suddenly not-terrible Mariners also happen to be in town. As this closure is all but one lane, no HOV/transit priority will be maintained, and no buses will be rerouted. So anyone who depends on the East Channel bridge and/or the 90/405 interchange should expect truly massive delays. The transit routes most affected are 111, 114, 210, 211, 212, 214, 215, 216, 217, 218, 219, 550, and 554, and to a slightly lesser extent 241, 555, and 556.

Stage 1, roughly Friday 7/18 through Monday 7/21
Stage 2, roughly Tuesday 7/22 through Thursday 7/24

What can you do about this?  For most people, not much. But here’s my advice:

  • If you are so fortunate, take advantage of any flexibility afforded by your employer (later start times, vacation, telework, etc). It’s a great week to take a couple days off.
  • If you are a Renton Highlands area commuter on routes 111/114, I’d strongly suggest switching to commuting via Renton instead on routes 101, 102, or 143, or via Sounder in Tukwila.
  • If you must commute over I-90, and you own a bicycle, use it. The 8-mile ride from South Bellevue to Seattle takes between 25-55 minutes depending on your level of fitness, and will surely beat vehicular and transit traffic across the bridge.
  • If you live on Mercer Island and commute to Seattle, enjoy your week of having the World’s Largest Ramp Meter.
  • If you’re an Eastgate, Issaquah, or Sammamish commuter and cannot bike? Bring a lengthy book. Maybe something dense and wordy like Dickens or Dostoevsky.

Follow WSDOT on Twitter for the latest updates, and commiserate with your fellow sufferers using hashtag #I90to1.

52 Replies to “Huge I-90 Closure Looms July 18-25”

  1. Love the animations! That needs to be the standard procedure for informing the public on closures. Now you just need to add a few thousand more cars trying to merge at 2 mph and it would be closer to reality.

    May also help to show an old lady with a walker zooming by.

    1. I don’t love the animations. Let’s face it, the average American is a dullard, and I don’t believe most people will know what’s going on when looking at those closure maps. They’ll just be confused. What does orange mean? What does light gray mean? What does dark gray mean? Don’t get me wrong. I understand the maps. As a genius, one of my gifts is I can imagine how an idiot thinks. The maps need to be dumbed-down so common person can understand them.

    2. Tim, thank you, that’s simpler. But even the word “avoid” might confuse a lot of mouth-breathers. I would substitute it with one syllable words like BAD or NO.

  2. Why doesn’t the state put up money for extra bus service during these closures? LA knows this works. We’ve seen from West Seattle / viaduct traffic that this works. Can’t the governor sign an executive order to spend the money needed for this desperately-needed extra bus service?

    1. Because the state “doesn’t do transit.”

      Remember how much Sturm und Drang it took to get WSDOT to actually extend the mitigation money for DBT construction until the DBT was completed? There still isn’t a final agreement on that, although they have agreed to it in principle.

      1. If they were really serious about mitigation, they could hire a fleet of charter buses for the week for park and ride specials.

        Transit agencies have a hard time putting out much extra; they are required to keep a certain spare ratio by FTA rules so they don’t have many extra buses. When there was a major viaduct closure for a week a year or two ago Metro was in the process of retiring a bunch of Gilligs. They held onto some in a contingency fleet and used them for extra service for that week only, then promptly retired them.

      2. They wouldn’t need more than 10 or 12 (split between extra 212 trippers and extra S Bellevue only 550 trippers) to make a significant difference. They can easily find that many. I’ve seen it done many times for all sorts of reasons.

    2. With what buses? I agree it’s a great idea, but all of the Puget Sound operating agencies — except ST of course — have been starved for capital funds for so long (except ST, of course) that there are no buses lying around to add. And, of course there’s also no surplus of available drivers.

      The state is going to RAKE IN THE TOLLS that week.

      1. There are enough buses and operators to run a small amount of extra mitigation service on I-90. It wouldn’t take that many.

      2. The shooting happened when afternoon trippers were starting to finish up and head back to the base. Since there are three bases right in the area, and the all call went out to all of us, finding 8 bodies in buses would be relatively easy around that time. With all the buses stuck in traffic that’s likely due to be horrendous well past the rush hour, it won’t be as easy to find extra buses/bodies to drive them.

        This is compounded by the fact that it’s summer (when people are taking time off) and Metro has been running everything really tight in an attempt to avoid layoffs when the cuts start in September. 10-20 trippers are open, looking for drivers, virtually every day.

  3. They should reverse the leftmost lane on EB I-90 and use it for HOV. You could do it pretty easily — you’d have to reverse the EB I-90 -> Bellevue Way HOV ramp (which you could do with some traffic cones on bellevue way). WB traffic would use the existing HOV offramp, and stay to the right up until the express lanes, which you could have permanently locked in WB flow for the duration (as a small mitigation of the massive backup.

    The only thing you’d really lose is the HOV offramp to Bellevue Way — but given there’s a GP offramp already it’s not that bad.

    The advantage is HOV’s and the 550 get an unimpeded route to I 90.

    1. And depending on how bad the backup is all of the service from Eastgate/Issaquah could just go to S Bellevue park & ride, turn around, and follow the same path as the 550. It would add a few mins, but probably would be faster than slogging in that one lane.

      1. Unfortunately getting from Eastgate to South Bellevue without using I-90 is a royal pain.

      2. @William Aitken, I disagree. I almost never take I-90 from Eastgate to South Bellevue. You just take Eastgate Way to Richards, Lake Hills, SE 8th, 112th, Bellevue Way. It’s actually pretty straightforward. Add in a few light timings that favor busses and you’ve got yourself an express route.

        It’s not the most convenient route but by gum it’s gotta be better than sitting on I-90.

      3. Except of course Bellevue way will be backed up all the way back to downtown Bellevue just like it is whenever there’s above average I-90 traffic. [And Eastgate Way is likely to backed up past the Humane Society just like it is almost every day at 5:30, although you can probably get around that mess by taking Camber road.]

      4. Remember, the I-90 closures are only on the East Channel Bridge, out of range for the I-90 transit lanes, so they won’t be affected.

    2. There needs to be some incentive for people to not bring their own cars unless necessary. Since it doesn’t seem feasible to implement even-odd license plate days, the speed afforded by HOV lanes would be a useful push.

      I am bemused / relieved that I actually _have_ chosen to take that week off.

    3. At the expense of the 550 not being able to serve Mercer Island — the HOV ramps no longer serve the center lanes — or I suppose running it wrong way all the way to Mercer Island.

  4. In 2000, a barge crashed into a 520 support pillar and WSDOT closed the bridge for several days to inspect the structure and ensure that the bridge wouldn’t collapse due to the damage. Suddenly, 100,000 vehicles had to find a new way to commute between Seattle and the Eastside. The first day was absolute chaos, but people quickly figured out that the NASCAR mentality of highway commuting wasn’t going to be successful. Many businesses closed or modified their work hours, but I was stuck in the middle of a critical project and I had to make the Seattle to Redmond commute every day during the closure. Drivers understood that 10 mph was going to be as fast as traffic would allow, but people started using their turn signals, stopped changing lanes and weaving through traffic just to get one car ahead, and let cars merge when needed. The commute was somewhat slower than normal, but the stress level during the commute went way down because the NASCAR aspect of highway commuting disappeared. Hopefully, drivers will handle this coming closure in the same manner.

    1. Exactly. The pathology of the SOV so comes to the fore during these times when repairs and updates are being done. “I want more lanes and no potholes but I don’t want those projects to interrupt MY commute/trip to the ballpark/whatever.” STFH (Stay TheF Home) and do some yard work!

  5. What a great Living Experiment!

    I have always contended that the two floating bridges make traffic worse, not better, for the general public because they disrupt the laminar flow of our highway system at exactly the worst places.

    I keep imagining a simple belt system, high capacity, around the circumference of Lake Washington…and no bridges. What would traffic in downtown Seattle be like without those bridges?

    1. I have noticed anecdotally that traffic on I-5 is often better through Seattle when 520 is closed. 520 is just in the worst possible location, with the worst possible design. Most access points to it are either left-hand on ramps to right-hand exits, or right-hand on-ramps to left-hand exits, and of course there is no express lane access, so buses are stuck in it too. It’s no wonder why the Ship Canal Bridge to Downtown is the worst traffic we experience! Consider how many 4+ lane merges 520 causes in the space of just 2 miles:
      1. University/I-5 to 520
      2. Mercer/I-5 to 520
      3. 520 to Mercer/Stewart/Union
      4. 45th/I-5 to 520

      1. I agree with respect to the part of Seattle near 520.

        What happens when 520 is closed is that the I-90 collector/distributors in both directions get overwhelmed, which leads to backups on the mainline in both directions leading to them. During the second-to-last 520 closure I waited in traffic for 30+ minutes approaching the SB collector-distributor. That’s worse than normal for the SB side, but the NB side is very fragile because it narrows to one lane at the merge point with I-5, and it routinely spills into backups on both NB I-5 and WB I-90.

      2. I suspect the main effect of connecting 520 to the I-5 express lanes will be spreading the traffic jams to the express lanes.

  6. My thought is that the bridge should be HOV-only during this shutdown. Want to get to Seattle without paying a toll or going all the way around? Carpool, bike, or take the bus. Period.

      1. Contractors with vans full of tools will find little time advantage in squeezing through one lane here instead of going around, except for Mercer Island-eastside trips.

        If we managed to keep people moving pretty well except for contractors with vans full of tools going west from the eastside to Mercer Island, we’d be doing amazingly well.

  7. It’s time for all our road agencies to really think about traffic management with these closures. Times of reduced vehicle capacity are times when it’s most critical that there be a strong incentive to use transit and HOV. In this particular, extreme case I think that incentive should be limiting use of I-90 to buses and emergency vehicles between 7am and 7pm. Maybe allow something like 5+ HOV for anyone that can put together a serious carpool. Most drivers under the current plan will find I-90 so slow it’s barely better than going around anyway. So instead keep it useful for vehicles carrying lots of people!

  8. Maybe I’m missing something, but will things really be that bad for HOV users? For the first few days, a bus goes all the way to the first merge, then switches lanes. Then a very short stretch, and does that again. Those merges will be very slow, but that is a very short distance, too. In some ways, it reminds me of the merge on 520, right before the bridge. It will be that times two, but still not that bad. Once you are past the merging, you have two lanes without any further constriction. I think two lanes is faster than three (in this case) since there is less of a chance of cars passing buses (directly or indirectly).

    For the other days, it will be worse. The HOV lane will be slowed down considerably. But depending on when the other cars will be allowed to move into the HOV lane, HOVs will be able to get very close to the merge without still being exclusive. For example, assume that the entire road comes to a standstill for a minute. For that minute, SOVs pile up in one lane, and HOVs pile up in the other. Unless there is a dramatic change in the way we drive (wouldn’t that be nice) the HOV backup won’t be that long. Again, this reminds of 520, or even the reverse commute on I-90. In general I think things will be hellish for SOV drivers, but not that bad for everyone else (no worse than your average reverse bus commuter).

    1. Sorry, never mind — “no HOV/transit priority will be maintained”. Damn. That is nuts. Really, really nuts. How did that ever get decided?

      1. Wait — I’m still confused. I should have lead with these questions:

        What exactly does “no HOV/transit priority will be maintained” mean? On the first few days, can a SOV drive in the HOV lane? If so, why? I really see no justification for this (if the answer is yes).

        For the next change, how soon can a car get into the HOV lane? For safety reasons, I can see why they don’t want to limit this lane change too much; but a half mile ought to do it. Just put up a sign saying “HOV lane ends” right where the sign says you need to merge left.

      2. The HOV lanes are always the first to go in a situation like this.

        During phase one the buses should do OK, depending on how far back WSDOT forces the rightward merge from the HOV lanes. Something tells me that it’s going to be much further back than the 4-car-lengths shown in the “not to scale” diagram. If I were in charge, I’d cone off a special path for HOV/Transit to bypass the 3-into-1 GP traffic merge and join GP traffic after the 405 onramp.

        Phase two is what worries me. Knowing WSDOT, they’ll probably remove the HOV restriction a full mile ahead of the merge, bringing buses to a standstill from Eastgate to Bellevue Way.

  9. WSDOT should look at how it can protect the HOV lanes as much as possible by bring the merge point between GP lanes and HOV lanes to the very front. Stage 1 is horrible with HOV lanes merging first rather than last… With that said the HOV lanes will probably be oversubscribed anyways. It might also be worth looking at making the HOV lanes temporarily 3+ leading up to the construction on I-90.

    1. The problem in these situations with HOV “priority” is enforcement. How do you enforce it? Now those of you in Canyon Park on 405 have witnessed HOV enforcement on a single ramp in the last week or two.

      The HOV violation rate with the “all but one” lane closure would be obscene. It would tie up WSP resources that would be needed elsewhere (enforcement/incident response). A temp 3+ condition wouldn’t work, since it would be unenforceable.

      One of the things accelerating this project was the recent failure of one of the east channel bridge expansion joints. It was repaired enough to limp it along until this project can fix it.

      What Zach fails to mention is that some of the closures WILL impact PEDs and bicyclists on the bridge as the expansion joint will be replaced under the trail, as well.

      1. They say that the only trail impacts will be at night, and pedestrians/bikers will be escorted through after a max-five-minute wait. That seems somewhat reasonable to me, since they’re a lot lighter than cars.

  10. BTW, this type of negative reinforcement is good and healthy. It teaches people there are consequences to bad choices. Perhaps this suffering will be a teachable moment, as Oprah says. Perhaps this will teach people to model themselves after people, like myself, who have chosen to live within walking distance of work.

    Be the change you want to see in the world. Be a hero and save the earth, all at the same time. Live within walking or biking distance of work.

    1. See, I knew you were actually Kemper Freeman. He lives within walking distance of work.

  11. Surely too late to implement this, but…

    From the South Bellevue P&R to the Mercer Island P&R is only 3 miles. A temporary bikeshare installation at the two P&R lots, with a truck to rebalance the bikes, would make this a 15-minute trip for anyone in good enough shape to ride a bike 3 reasonably flat (by Seattle standards at least) miles, most of that on a separated MUP.

    1. So… how does this not solve the problem?

      – Starting with platitudes and generalities. Bikeshare and carshare (like taxis), in cities with excellent mass transit systems, complement those systems by filling in for trips the transit system serves poorly. They rarely have the capacity to actually replace the excellent transit systems. That’s why when the NYC subway went down in the flood they quickly marked off bus lanes.

      – Slightly less platitudinal: A full 550 toward Seattle arrives at South Bellevue P&R, as many do every day during peak hours. One hundred people are on the bus. Are one hundred bikes waiting for them? Are those hundred bikes back at South Bellevue by the time the next 550 shows up? We’re probably on the order of a thousand bikes to replace just the 550’s capacity, and infrastructure to support a massive increase in bike traffic and docking/storage around the P&Rs.

      – This capacity issue is even greater if you hope to extend the benefit to riders from Eastgate and Issaquah. Seattle-Bellevue demand is quite peaky and somewhat directional; service from farther east on I-90 has super-peaky, super-directional demand. Good luck re-balancing right through the I-90 bottleneck!

      – The bike path between MI P&R and South Bellevue P&R includes some pretty big hills, and it’s really not wide enough to support large volumes of traffic (especially when half of it is screaming down a hill and half the rest is trying to pass the remaining quarter going up).

      – Two possibilities: first, you split routes at the bottleneck and force most people to ride. That’s not going to happen. Second, you let them run through as we’re apparently going to do. If the buses run through the bottleneck reliability will be a joke and effective headways will be multiplied at both ends (this is mathematically necessary unless you run far more buses, as both in-service runs and deadheads will take more time) — maybe someone biking can catch up to the next bus ahead, but general operational problems for transit remain.

      The only way to even remotely address the problems involved here is to realize that one G.P. lane on I-90, against the size of the rest of the highway network, is little better than none, and give up on it. One (quite restrictive) HOV lane that’s actually flowing could actually provide meaningful transportation.

Comments are closed.