2020-05-15 01:49:43 +03:00
|
|
|
# West Seattle mitigations
|
|
|
|
|
|
|
|
_Draft, updated May 14, 2020 by Dustin Carlino (<dabreegster@gmail.com>)_
|
|
|
|
|
|
|
|
In March 2020, the West Seattle bridge was closed due to cracks forming. As of
|
|
|
|
May, COVID-19's impact on commuting means the area still hasn't seen how the
|
|
|
|
area will handle losing the main route to the rest of Seattle. A local group,
|
|
|
|
HPAC, published a list of
|
|
|
|
[requests to SDOT](https://www.westsideseattle.com/robinson-papers/2020/05/04/highland-park-action-coalition-calls-seattle-officials-traffic)
|
|
|
|
to prepare the area for these changes.
|
|
|
|
|
|
|
|
This page will try to explore some of the problems and solutions from HPAC's
|
|
|
|
document using [A/B Street](https://abstreet.org), a traffic simulator designed
|
|
|
|
to explore the impacts of changes like this on different modes of
|
|
|
|
transportation.
|
|
|
|
|
|
|
|
> **_NOTE:_** You might want to read [how A/B Street works](../how_it_works.md)
|
|
|
|
> first.
|
|
|
|
|
|
|
|
## 16th Ave SW and SW Holden St
|
|
|
|
|
|
|
|
HPAC has been asking for a protected left-turn phase at this intersection. I'm
|
|
|
|
unfamiliar with this intersection and currently unable to scout in-person, so
|
|
|
|
I'm blindly guessing the traffic signal currently has just two phases:
|
|
|
|
|
|
|
|
![existing_diagram](existing_diagram.gif)
|
|
|
|
|
|
|
|
From watching the traffic, it seems like the east/west direction is busier, with
|
|
|
|
lots of eastbound traffic headed towards WA-509. Holden St has no turn lanes, so
|
|
|
|
a protected left turn phase makes sense. Let's make the change and see what
|
|
|
|
happens:
|
|
|
|
|
|
|
|
<iframe width="560" height="315" src="https://www.youtube.com/embed/6tooJaZLa0Q" frameborder="0" allow="autoplay; encrypted-media" allowfullscreen></iframe>
|
|
|
|
|
|
|
|
Unfortuately, we can't evaluate the change yet, because the simulation gets
|
|
|
|
stuck with unrealistic traffic jams in other parts of the map. This is mostly
|
|
|
|
due to data quality issues in OpenStreetMap and incorrectly guessed traffic
|
|
|
|
signal timings. These problems can be fixed with the help of somebody familiar
|
|
|
|
with the area.
|
|
|
|
|
|
|
|
## Re-evaluate arterials
|
|
|
|
|
|
|
|
The 9th item from HPAC's list asks for measuring the amount of east-west traffic
|
|
|
|
to figure out what streets people are using as arterials. That's an easy
|
|
|
|
analysis, using the _throughput_ layer.
|
|
|
|
|
|
|
|
<iframe width="560" height="315" src="https://www.youtube.com/embed/yzp9c7gHhOI" frameborder="0" allow="autoplay; encrypted-media" allowfullscreen></iframe>
|
|
|
|
|
|
|
|
By 6am, the busiest streets include Admiral Way, S Charlestown, SW Genesee, SW
|
|
|
|
Alaska, SW Holden, and SW Roxbury St. Again, it's necessary to first fix data
|
|
|
|
quality problems and run a full day before doing more analysis.
|
|
|
|
|
|
|
|
Once the simulation is running smoothly, A/B Street can be used to make changes
|
|
|
|
-- like lowering speed limits, adding a protected left turn phase, or converting
|
|
|
|
part of the road into a bus lane -- and evaluate the effects on individual trips
|
|
|
|
and aggregate groups.
|
2020-05-28 02:21:15 +03:00
|
|
|
|
|
|
|
## Next steps
|
|
|
|
|
2020-05-28 02:52:39 +03:00
|
|
|
It'd be useful to first get a baseline with the high bridge restored. Should now
|
|
|
|
be possible using edits.
|