At 32 feet, it does seem like they have plenty of width for 3 lanes each way with no shoulder. 12 foot lanes are the standard for 60-70 mph interstates, but Aurora's top speed limit is like 45 mph or something and I think it's even lower in the current tunnel. So 10.6 foot lanes seem pretty reasonable.

Take a look at the sides of the I90 E tunnels. Tunnel walls tend to jump out at Seattle drivers. Gotta give people some breathing room and I'm sure there is some consideration for leaving space for emergency vehicles to squeeze by. I am also hoping they increase the speed limit inside the tunnel since there are no ramps. I think there is probably consideration for making it easy for buses to navigate as well.

I think you have a good point about emergency vehicle access.

I-90 is a much higher speed road, though. (Assuming they don't take your suggestion of increasing the speed limit in the tunnel.) Freeway standard is 12 feet per lane (and even the 1940-era Mt Baker tunnels have 12 foot lanes[0]), but again, Aurora isn't a freeway.

Meanwhile, buses are about 8.5 feet wide[1] and easily fit in a 10 foot lane. Most Seattle streets are narrower than that.

[0]: http://www.seattleasce.org/committees/hh/LaceyBridgeReport.p...

[1]: https://cptdb.ca/wiki/index.php/New_Flyer_Industries_XDE60

Maybe interesting datum: the existing Battery Street Tunnel has two lanes with a total width of 25 feet[0].

> The SR 99 roadway width narrows to 25 feet with two lanes in each direction through the BST

> The existing lane width varies from 10.5 feet to 11 feet, and there are no existing shoulders only a gore stripe adjacent to the curb... The existing posted speed is 40 mph.

So there's no real reason we couldn't keep the 10.5 foot lane size, 40 mph speed limit, and do three lanes. The new tunnel has better sight lines and curve radius than the BST. And there's separation between directions, which is better than the Aurora bridge (which also has pretty narrow lanes):

> Many aspects of the existing horizontal and vertical curves do not meet today’s roadway design standards for the posted speed limit. When compared to current design standards for stopping sight distance, horizontal curve radius, and vertical curve length, about two-thirds of the horizontal and vertical curves would coincide with a design speed of less than 40 mph.

[0]: http://www.scatnow.com/DesignDev3Grade.pdf

Is three lanes at 40mph higher throughput than two at 60?

Trick question! "mph" is not a direct measure of throughput. Cars per second (passing a given measurement point) is.

If the spacing of the cars is the same between the two, then a 60 mph lane is moving more cars per second than a 40 mph lane. The spacing never is the same, though, unless these are autonomous cars.

Also, throughput is limited by bottlenecks. If an upstream bottleneck has throttled things to one car every ten seconds, it doesn't matter whether the downstream lane is 30 mph or 60 mph.

Fine points but in this case posted MPH isn't an output, it's an input. So is the number of lanes.

Spacing is a function of speed. Even with autonomous cars spacing is not constant because there are on and off ramps so some will be accelerating or decelerating. We don't even know how those cars will work, maybe you can pay more to drive faster!

The question here is which configuration of the tunnel will maximize throughput.

I can't find any explicit statement of the planned speed limit but I did find a document [1] that implies there will be variable speed limits much like other highways in the area.

[1]: http://www.wsdot.wa.gov/Projects/Viaduct/Media/Default/-NewD...

Really good points. I was suggesting the need for a shoulder so people don’t just drive into the walls like they do on I90.

And what, put in a car elevator to get from the tunnel to first avenue? The tunnel is 200 feet underground at the location of the current ramps.

It's my understanding the first tunnel plans were for a six-lane tunnel. I have not seen the elevation/depth info on the tunnel, but even at 200f, you just need a longer ramp to ascend.

But you have to build that subterranean ramp. The tunnel cost billions already without the ramp, how much more would it cost with that added complexity? There will still be surface streets where the viaduct is located.

I'm not a transit engineer but I do think this is one of the drawbacks in the kind of radical democracy we push in Washington. Everything goes up to a public vote, even things the public can't really fully be experts on.