Joe Wolf (Flickr)

Last Wednesday, CH2M submitted the I-90 and Mercer Island Mobility Study prepared on behalf of Sound Transit.

If you’ve not been following the drama, a quick recap: Mercer Island has been making considerable noise since 2015 about perceived loss of mobility due to East Link construction. Whereas prior complaints were more generic and white-hot, of late the complaints have been much more focused on mitigation for the closing of the express lanes for East Link construction and the coming SOV-to-HOV conversion of the westbound ramp from Island Crest Way to I-90. Islanders maintain that they are owed mitigation for this by the 1976 agreement granting their solo vehicles special access to the I-90 express lanes, and also by a 2004 amendment to the agreement holds that:

[t]o the extent of any loss of mobility to and from Mercer Island based on the outcome of studies, additional transit facilities and services such as additional bus service, parking available for Mercer Island residents, and other measures shall be identified and satisfactorily addressed by the Commission, in consultation with the affected jurisdictions.”

To date, WSDOT and Sound Transit have held that the agreement clearly and permanently committed the center lanes for transit and that Mercer Island SOV access was a temporary and conditional use. The agencies noted in a letter to Mercer Island that SOV use of HOV lanes, even temporarily, would be a violation of federal law and a trigger to repay grant funds to the federal government. In response, Mercer Island announced its intent to sue Sound Transit, threatening both the schedule and budget for East Link.

Things have softened a bit since then. Mercer Island reneged on its threat to revoke shoreline permits for Sound Transit, and the ST Board committed in March that the city and agencies meet regularly to negotiate issues around Mercer Island Station beyond traffic. The current study was an olive branch between the agencies and the city, seeking data to quantify the scale of any lost mobility, in the spirit of the 2004 amendment.

Well, the study results are out. In a memo to Sound Transit board members last week, Secretary Millar summarized that:

the Mobility Study concludes that the overall mobility for people traveling to or from Mercer Island will be similar to or slightly improved compared to existing conditions during the six-year East Link construction period, and will be improved once East Link light rail service begins in 2023.   A short summary of the study is attached. [emphasis mine]

The study examined five options. The No-Build Alternative would cancel East Link. Option 1 would maintain the current construction schedule and roadway plan, but allow Mercer Island SOVs to use the HOV lane, whereas Option 2 (the current plan) would not. Option 3 would cancel the HOV lane plan and all lanes of westbound I-90 would be general purpose from Island Crest Way to I-5. Option 4 would allow temporary use of the HOV lanes for SOVs using the westbound ramp, but would require an immediate merge into the general lanes. The No-Build Alternative isn’t a serious threat, and WSDOT notes that only Options 2 and 3 comply with federal law.

The study notes that Island Crest Way only serves 20% of Mercer Island’s SOV traffic onto westbound I-90, with just 2,000 vehicles daily. The other two primary access points, West Mercer Way and 76th Ave SE, carry 4,000 vehicles each. During peak periods, the combined peak hour volume for Island Crest Way is approximately 300 vehicles, with less than 100 vehicles typically entering via the ramp during the morning commute peak hour.

Crash risk would be increased with Option 4, as higher differential speeds between the GP lanes and the HOV lane would increase the risk of required merging. And general non-compliance would be a fatal blow for Option 1, as other area commuters would quickly learn that there would be no realistic way to enforce the HOV requirement west of Mercer Island.

A 50 percent HOV non-compliance rate was used for modeling the operation of this option in order to reflect a realistic expectation that with limited means for enforcement, many SOV motorists will choose to stay in the HOV lane traveling across the bridge. This non-compliance may also encourage other mainline travelers to do the same. The required left to right merge from a higher speed HOV lane to a congested GP lane increases risk of crash frequency.

The current plan (Option 2) scores the best in terms of speed and throughput:

Construction Option 2 is the only option that meets the HOV speed and reliability policy requirement of 45 mph or greater, at least 90 percent of the time, for travel between Bellevue and Seattle.

In terms of total travel time, Option 2 greatly improves transit and HOV speeds during the 6-year construction period, saving 2.3 minutes per trip. SOVs would see a 2 minute increase in AM westbound travel time. But the increased inefficiency and merging pressure of Option 4 (where SOVs would be allowed but required to merge) would be even worse for Mercer Island SOVs, adding an additional 24 seconds per trip compared with Option 2. Option 3 (canceling the HOV lane entirely) would save SOVs a mere 6 seconds over Option 2.

You can read the full study here, or WSDOT’s Summary Memo here. But the results are clear: there is no reason to delay East Link on account of a mere 300 vehicles per AM/PM peak period. Travel times for Islanders will not be significantly worsened, and 76th Ave SE and West Mercer Way are able to take up much of the slack. The current plan is the only one that complies with federal law. And six years later, Islanders will have all-day, traffic free, high capacity transit the likes of which they’ve never seen.

Any mitigation that seeks to increase access to the Island Crest Way ramp should focus on increasing the HOV rate. And if budgetary resources allow, agencies should shower the island with additional peak transit service, whether in the form of more hours on Routes 204, 550, or 630, or to a resurrected Route 202. But East Link construction should proceed on schedule, and I-90’s thousands of current transit commuters deserve to have complete HOV/Transit priority across Lake Washington.

35 Replies to “Study Shows No Significant Impact to Mercer Island”

  1. Great work, Sound Transit and WSDOT!

    So about that parking for Mercer Islanders … Wouldn’t banning a bus transfer facility at Mercer Island Station and forcing more off-islanders to drive there in order to take Link be a violation of the agreement to provide more parking for islanders?

    It would also, ipso facto, be a violation of the agreement to provide more bus service, since islanders traveling to jobs in Eastgate would have to take Link to Bellevue and then transfer to a bus, rather than just catching that bus at Mercer Island Station.

  2. The study included some interested diagrams of roundabouts around the light rail station for bus turnaround, for a faster turnback for buses coming from the East on I90. Are those new to this EIS? I don’t remember seeing them in prior studies, and I think they include some clever way to improve bus throughput around MI station.

    I’m looking at 2.2.2 on page 2-6, Exhibits 2-3 to 2-7. Very end of the Executive summary pdf.

    1. Yes they are new to the EIS, although both of those options have been out in public for a couple of years now as considerations.

      The 80th Ave option grew out of earlier feedback to avoid driving most buses around MI local streets (especially SE 27th and North Mercer Way) and avoid private property takings, so it is the “newest” option. The roundabout option at 77th/N Mercer is the only surviving roundabout option from earlier proposals at 76th/N Mercer or 77th/Sunset Hwy; the 76th/N Mercer option had somewhat severe impacts on local traffic, and the 77th/Sunset option had impacts to parks and nonmotorized facilities as well as impacting the 77th station entrance, while not being substantially better operationally than the 77th/N Mercer option (except for the lack of property takings).

  3. Humor me for a second. :-) Was a serious study ever done for a tunnel under Lake Washington to Bellevue? I know that it would have challenges due to depth, but ST is experienced with tunneling now. The only arguments against this my searching could find seemed pretty hand wavy, like it was just a known fact that you couldn’t tunnel under Lake Washington but no real explanation why you couldn’t if it was deep enough. And a direct tunnel option would kill some other birds with the same stone. A line running from Downtown would have stops in First Hill and the expanding Central District before heading straight to Bellevue where it could have stops on both the westside and eastside of Bellevue supporting much better walkability to destinations. It would be faster and more reliable over all, not requiring new technology or slower speeds on floating bridges. And the Rainier Ave/Mercer Is stops are pretty low performing over what a First Hill/CD stops would be. It seems the better long term investment to me. Especially if Bellvue continues to expand and intensify. And as a stand alone line – with a little rejigging of the existing plans for East Link beyond Bellevue it could be built as a true light-metro instead, with fully automated running for continuous service all the way out to Redmond.

    1. Actually depending on how you route it such a line might also be routed via Madison/23rd, interchange with CapHill station, Denny/Westlake, Belltown, and then south along 1st before heading to West Seattle.

    2. Whether it’s doable or not, what we do know about the cost of tunneling in general and tunneling under deep bodies of water more specifically suggests it would have been considerably more expensive. What do you cancel to pay for it?

      1. True on cost. Just doing a crude back of envelope now. It looks like the roughly 7 miles of tunnel and 4 or 5 stations required would be between $3.5-4.5 billion. Couldn’t (quickly) find a breakdown of Eastlink segments to see what the the 10 mile section from International District to Bellevue is predicted to cost, but if it’s roughly proportional to the overall line length and budget this segment would be around $2.5 billion. So if this very rough calculation held up, there would be a premium of $1-2 billion for the tunnel route. Would that have been worth it? (If that was the real difference, then I’d suspect so.)

    3. There’s been talk of a Sand Point – Kirkland tunnel. I don’t remember if it was in the Ballard-Redmond corridor study in 2014. At one point ST disfavored it, but after protest it restored it for consideration. ST is considering a Seattle-Eastside crossing somewhere between 520 and 522, which could either be on the highways or a separate bridge or tunnel. The most likely separate location is Sand Point – Kirkland, because the lake is narrowest there, it would conveniently serve downtown Kirkland, and it could serve U-Village and Children’s along the way.

      The lake is extremely deep and has no hard bottom: it’s a U shape that goes down 200 feet and then gets gradually solider with sediment. That’s why traditional arch or suspension bridges aren’t feasible, and a tunnel would have to be in a tube in the middle of the water or slushy sediment. That would be extremely expensive. Plus it would have to descend gradually from UW and Kirkland, requiring long tunnels on land. Note how UW Station is deep because it’s right next to the Ship Canal, and TIB Station is high because the track goes over nearby highways.

      The corridor report said a northern cross-lake line would be very expensive and have mediocre ridership; that’s why it wasn’t pursued in ST3. It did consider segments of the corridor, to allow Ballard-UW first, UW-Kirkland second, and Kirkland-Redmond maybe never. ST3 funded a further study to look at options more closely. That could result in anything from a 520 line, a Northgate-Bothell-Kirkland line, a UW-Sand Point-Kirkland tunnel, or anything vaguely similar.

      A northern lake crossing would connect north Seattle and the northeast Eastside in a way they’ve never been connected before, not even with a car. (Imagine driving from Kirkland/Redmond to Ballard to see how revolutionary it would be. There’s no way to drive without going through the I-5 bridge traffic or the U-District traffic.) So it could have major, major potential. But we’d have to weigh the cost and the Eastside’s adversion to density.

      The report said a second Kirkland-Redmond track would be redundant with East Link. But reading between the lines, a Seattle-Kirkland line could turn southeast to join the East Link track at 120th and continue to Redmond, or it could join the Kirkland-Issaquah route.

      A tunnel around I-90 wasn’t considered because the highway was rebuilt in the 1980s to be convertible to center rail someday, and we’d be throwing that investment down the drain.

      1. Thanks for the detailed reply. On the aversion to density, that is one good aspect of such a hypothetical (revisionist history at this point) Lake Washington Subway line. It would connect right through with two stops in Bellevue where density is growing up, and the Bell Red Corridor where density is coming with new developments, Microsoft’s huge campus, and the moderately dense Redmond. Such an investment for Kirkland and Laurelhurst/Sand Point seems questionable by comparison to that.

      2. I should say, the study addressed only the 520 and separate bridge/tunnel options, not the 522 options. Those come directly from ST’s long-term plan, which has a Northgate-Woodinville corridor and a Bothell-Kirkland corridor. But the study in ST3 will presumably include some of these north lake options.

      3. Didn’t McGinn insist that convertibility for future rail be included in the design? I thought that made it in. ST’s default alternative is 520.

      4. Late in the game someone (Seattle Mayor?) pitched a fit and demanded that the new 520 bridge be light rail compatible. WSDOT reran some numbers and said, “Good to Go”. There’s probably less than 0% chance that GP vehicle lanes would ever be converted. Loosing the HOV lanes for exclusive use of light rail would cause almost as much of a pitchfork rebellion and be a huge lose for buses.

      5. “As shown in the graphics below, light rail could be accommodated either by converting the bridge’s transit/HOV lanes to light rail, or by adding more width to accommodate light rail in each direction along with the four general-purpose and two transit/HOV lanes.”

    4. Might want to check this out, erentz. BART could lay its tubes on top of the bottom of San Francisco Bay.

      Lake Washington is narrow, but extremely deep. Creating either unclimb-able grades in both directions. Or all operations kept underground about the depth of the lake bottom. Every sky-scraper in Seattle is equipped with tall enough elevator shafts.

      Though remember, ST and all its contractors: Sandwich signs and stairs aren’t gonna “get it.” Maybe just to be safe we can find some of those mining pit-head lifts, you know, with steam engines and those big wheels. Okay, fire the boilers with used Krispy Creme oil. It think some of them could lift railroad cars full of coal.

      Theoretically possible to create tunnels moored to the lake bottom, that float far enough below the surface they don’t interfere with watercraft. Check it out.

      Says the Chinese are already working on it.


      1. Thanks Mark. I’ve been doing a bit more research. Details of the geology under the lake seem to be hard to come by. This is the best info I could find so far [1]. Here’s the cross section from where they took the samples (along the 520 bridge):

        The Qlg layer is described as “Lacustrine deposits consisting of very soft diatomaceous ooze with some organic silt layers.” While the Qvrl layer is “lacustrine deposits… of very soft silty clay to clayey silt.” And glacially overridden material is “compact gravelly silt SAND to gravelly sandy SILT… [and]… “compact intercalated gravelly sandy SILT and gravelly SAND.”

        I have no idea if you can tunnel through any of that. But assuming you can’t tunnel through the Qvrl and can somehow tunnel through the glacially overridden material, that puts the depth required at ~375 feet. Coming from the west side assuming a station at ~23rd and Union that’s at +315′ to -375′ in approximately 2 miles. Or from MLK and Union it would be less at +235′ to -375′. Assuming we can build a deep station at 250′ below surface (Washington Park in Portland is 260′ below ground), that’s a slope of around 1 in 24′ or 1′ in 29′ respectively. So best case 3.45 degree slope. Is that too steep for a metro over a longish 2 mile run? (I’m not sure.) Either way, yes it would be deep.

        Slope would seem to be much less of an concern anyway if you could somehow tunnel through the Qvrl layer. Still – leave it to someone who knows geology and tunneling to comment on the feasibility of those aspects.


    5. Self-reply (see longer response to Mark for links and stuff for the interested): the more I look into this it seems tunneling through the Qvrl may be possible, and the “glacially overridden” layer even more so. This (again I’m a very lay person) is apparently similar to the stuff we’re tunneling through with the existing University Link and Big Bertha tunnels. Just have to hope there are no hidden steel pipes under Lake Washington.

    6. Is there any TBM in the world which can withstand the pressure of 240 feet of water and fifty feet of silty muck below it? I doubt it very much. Trackhead depth would be -320 at least. That’s like climbing Queen Anne Hill I believe.

      1. The Channel Tunnel (Chunnel) is deeper and longer; they must have used TBMs, right? Of course, that would probably cost the equivalent of the WA state yearly budget to build today ;).

      2. Tunneling below a body of water has been done, so it’s feasible. The grades (getting from surface on-shore to some hundreds of feet below and then back to the surface) would be punishing. Maybe more for the trains than for a TBM.

        But let’s suppose all that is fixable. It’s ludicrously cost-prohibitive. We have plenty of cross-lake rail capacity on East Link for decades to come. The low-cost option for a second lake crossing is almost certainly adding rail to SR 520. Even that’s about $2 billion in current dollars. And it doesn’t add much value that isn’t cannibalization of East Link.

  4. Oh my, only 20% of the total SOV total flow and a mere 100 SOV’s during the peak hour every day? This is hardly worth any consideration at all for mitigation, but I am sure MI will find a way to whine anyhow. Because, you know, this isn’t so much about reality as it is about perceived exclusivity.

    But hey, I can hear it now, “fake news”, “stacked study”, etc. MI will demand what they think they deserve. Facts be damned.

  5. These current flow numbers for the 3 entrances can’t possibly be correct. I drive it daily and have reason to use different routes. I think Island Crest has 10 times the SOV traffic onto I90 westbound than the other two combined, but the article says each of the other entrances have higher SOV volume.

    The only exception is when drivers are detouring off I90 and then getting back on at West Mercer. Is that what we’re working to optimize?

    1. lol I love the hurried I 90 detour drivers when I’m riding my bike across the island. The freeway wasn’t fast enough for them so blame the bike rider for getting in the way

  6. I love this blog.

    A rebuttal post to a mad scientist rant to add more lanes and get into boats, over 100 comments!
    A post on a WSDOT study that puts Eastlink back on track, merely a meh at 20 comments.

    I love this blog.

    1. We’ve already talked over the Mercer Island issue in multiple other posts with dozens if not hundreds of comments; what is there to add here? “Yay! Knew it! Go East Link”? What’s more, unfortunately this isn’t even final until Mercer Island decides whether to go ahead and sue anyway.

    2. You should see the articles about Link project decisions. 529 comments in 2015 when ST suggested an ST3 that was all spine and no Ballard (or West Seattle Link and a Ballard streetcar). 296 comments in2013 on David Lawson’s Metro network proposal..

  7. Goes to show, if you have money you can get an expert to say whatever you want them to say.

    1. BAP,

      Insinuate much?

      No one can argue that Mercer Island has money. There is still hope that Mercer Island will choose to use that money for something useful, instead of suing to keep the HOV lanes overfull of SOV cars.

      I wish the study addressed improved travel times for actual HOVs on Mercer Island.

      That said, this part of the document remains laughable.

      WSDOT also has an operational policy for managing its HOV facilities, which specifies vehicle operating speeds to be 45 mph or higher 90 percent of the peak period. In both 2020 and 2035 Build conditions, the HOV lanes in both directions would meet this policy during both peak periods…

      That 45 mph condition isn’t going to be met without both HOV3 and tolling equivalent to 520.
      WSDOT has its head in the sand – presumably until no one from Mercer Island is on the State Transportation Committee.

      1. “…presumably until no one from Mercer Island is on the State Transportation Committee.”

        Judy Clibborn, a Mercer Island state representative, is the *chair* of the House Transportation Committee.

  8. “In response, Mercer Island announced its intent to sue Sound Transit, threatening both the schedule and budget for East Link.”

    Have any WA state bar members who live in North King looked into the possibility of “intervening” in the Mercer Island suit? The intervenors* could argue that if East King lawsuit prevails, East King forfeits the right to sub-area equity funds and North King intervenors ask the court that ST funds be sent to a North King project.

    *have MI fight a two front war.

    1. You can’t just put a gun to somebody’s head and say, “We’re taking your subarea equity funds.” What would be the legal justification for making this lawsuit able to change subarea equity? It’s East King taxpayers’ money, and subarea equity simply says it should be spent to benefit East King. What does that have to do with controversial lanes on a bridge?

      If you’re concerned about the impact on North King, the only one I foresee is Judkins Park Station. That was so unimportant to North King that it was originally in East King’s budget until East King begged North King to take it on so that East King could afford Bellevue’s downtown tunnel. The blue line trains can run on Lynnwood-Intl Dist jf the Eastside isn’t ready to open yet. It would just depend on how much East Link’s money is needed for their operating funds. But since East Link funds were intended to operate them anyway, ST could argue that the stub line is still fulfilling the original plan (and it’s East King’s fault its segment isn’t ready yet).

Comments are closed.