Comments

1

I enjoy being told how urgently we need rapid transit and then when we get to vote on something I learn that "urgent" means we can wait 20 years to have light rail out to Everett.

In 20 years time we had laid down a huge portion of the entire interstate system and yet we can barely choke out a couple dozen miles of light rail in that same time. So yes, our expectations are in an endless downward spiral.

2

"Metro’s parabolic barrel vaults" -- picture looks more like cycloid arch profile to me.

3

Fantastic reading, thanks David.

4

20 years between JFK and Nixon's Southern Strategy? I think you need to check your history timelines, Charles. 20 years after JFK's election was when Reagan took office, 6 years after Nixon resigned.

5

@1- well, if we had accepted the Feds' offer to build light rail in the 70's, we'd be fine. short sighted has always been the watchword in the american west, and our current lightweight, fragile infrastructure for all things is a result. Remember winter of 07/08, when we lost both interstates for extended periods of time? Total economic loss of $75M. http://www.wsdot.wa.gov/NR/rdonlyres/8FCFF2CF-4ACC-461A-96A6-AA310CCF6050/0/WSDOT_I5_90ClosuresFinalReport.pdf

6

@4: A math error on my part that slipped through the editing process. Thanks for the heads-up.

7

Japanese cities do very nice mass transit.

8

@4,
Freudian slip there... Charles didn't write this, David Cole did. I can totally get the mistake though, while reading it I assumed it was a Mudede piece as well.

9

The suburban ideal of the single-family house with a Mustang in the garage inspires Jonestown levels of religious fervor, with cheap gasoline standing in for the Kool-Aid

This critique of car-oriented cities, along with the quote above, really reminds me of James Kunstler. Great article, I share your frustration that we can't seem to get things done anymore.

10

Great article, thank you.

11

I seem to recall something about the voters repeatedly asking for a monorail system. I'm recalling the word "elections". Lots of them. And a much lower price tag then ST3. Then there was this jackass - i think his name might have been "Greg" - who did his best to kill it.

So here we are with Link instead. Let's make it the best system we can. Link's travel times will be fine in the sections that don't run on streets. NO MORE grade level tracks, and let's get working on elevating the Rainier Valley section.

12

@11 Monorail? That sounds like a Shelbyville idea

13

Once Lynnwood Link opens in 2024 (hopefully), a train ride from Westlake to Lynnwood (16.5 miles) will be under that 30-minute threshold. We may have screwed the pooch on the Rainier Valley segment (thanks Eyman, Sims, et al), but the newer suburban segments should be comparable to a decent rail system.

14

Very well written piece. Thank you, David.

15

@12- It's certainly not a Nickelsville idea.

16

Splendid article, Mr. Cole. Inspiring anger and action. I, for one, am ready to make you West Coast Dictator for the next 10 years. People seem to hate the planning and building of rapid transit (including bullet trains), but when it is completed, all is forgiven.

17

Hey David,

Good article!

I do have a few comments on your calculation as I did very similar work for an article 2 years ago.

You said the best we can hope for is 10 trains per hour or a headway of 6 minutes. So on the segment from International District to Lynwood where East Link and North Link interline the planned headway is 3 minutes. We (The Urbanist) had also checked with people from Sound Transit and they said that it is technically possible to achieve a headway of 1.5 minutes / 90 seconds on that segment with upgrades to the ventillation of the various tunnel segments and potentially egress capacity of the stations. These upgrades may seem expensive now, but as the system gets good ridership they will become the cheapest way to boost capacity. So, we could get 4 times higher capacity and actually have link compare to subway systems.
You used an average occupancy of cars of 1.5 and considered it generous. For the Seattle region the average is actually already 1.6 according to PSRC data. Same conclusion, but just wanted to point that out as I had researched it.
This isn't something I covered, but I am optimistic. You mentioned that an automated system will be better and that we can't automate ours. Well, automating a rail system is a much simpler problem than automating cars. In fact, the key technology required is automated braking - something that is already standard on many new models from common brands like Toyota or Volkswagen! I think that automating street running light rail, as a technology problem can be solved earlier than automating cars that can go anywhere. Whether agencies will be willing to make the mental leap (the humanities question) is a whole another story.

Here is the article I wrote - I got good feedback and readership on it (and unlocked my lifetime achievement of Jarret Walker tweeting about it), so I think you'll enjoy it:

https://www.theurbanist.org/2016/05/26/the-supply-and-demand-of-street-space/

Anton

18

Good article, but you seemed to have missed the bigger point. There is nothing inherently wrong with running surface rail in Rainier Valley. Nor are the trains particularly slow. Nor is it a big deal if a vehicle carrying a thousand people has a driver. Those are all reasonable trade-offs when it comes to spending billions (and I literally mean billions) on a system that serves less than a hundred thousand a day.

In your quick summary of mass transit in America, you skipped over the great turning point: BART. The problem with BART is not that it fails to serve Marin County (who gives a fuck) nor that it is slow (it is extremely fast) but that it does such a piss-poor job of serving the city. Look at how it compares with D. C. The DC Metro has a shitload of stations, all throughout the heart of the city. The only part they left out was Georgetown (for political reasons). BART, on the other hand, barely touches the city. BART has only one line through San Fransisco, and a split line in East Bay. The DC Metro covers the fuck out of DC, and the nearby areas. BART focuses on the suburbs.

This was an experiment worth taking back in the day -- BART actually sounds like it would work -- but there is no excuse now. We know now that most of the people who ride BART do so inside the city. The entire system is designed to favor places like Pleasant Hill or Union City, while shortchanging San Fransisco and East Bay -- yet ridership in San Fransisco, Oakland and Berkeley dwarfs that of the other areas. We should have learned from the great BART experiment, but we didn't.

America has "embraced" mass transit, but only on its suburban terms. She has ignored building what actually works (urban mass transit) and focused on what we know won't work (long distance rail).

We went on to build similar, flawed systems, in places like Denver and Dallas. Long, sprawling subways that are extremely expensive to operate, have poor ridership, and thus suffer from horrible headways (15 minutes if you are lucky).

Canada, of course, took a different approach. Rather than focus on every distant suburb, they kept it close. Not that every system is perfect (Vancouver should build the UBC extension before fucking around with more rail in Surrey) but by and large they got it right. Multiple lines in the heart of the city, along with very good bus to rail integration. It wasn't that they spent more, or built more (Toronto has a relatively tiny system) but that they focused on the urban areas.

Unfortunately, Seattle is fucking up, like so many American cities. We are building out, while neglecting our inner core. Holy shit, we will soon have the largest light rail system on the continent. Only a handful of cities (New York, Chicago, Mexico City) will have more miles of rail. We will double every Canadian city in terms of rail distance -- and that includes cities cities like Toronto and Montreal!

Yet despite the massing investment (much larger than any Canadian city) we fucking ignored the Central Area and Belltown. Holy shit -- here is a census map of this area: https://arcg.is/1v5Tfb. Now keep in mind that the city (in both absolute terms, as well percentages) is growing faster than the suburbs. Given all that, in what fucking universe does this make sense: https://tinyurl.com/y6vb8hfb? It's insane. Miles and miles of track to distant, largely vacant areas, while the core of the city waits for the next multi-billion dollar package (ST4? ST5?). It is all just bullshit, designed to please the handful of mayors that will soon have rail to their tiny suburb (when what they could really use is decent bus service) knowing all the while that ignorant folks in the city will vote for any fucking transit system. It is just a repeat of every fucked up transit system in America, while we jealously look across the border and wonder why Canadians do this shit better than us.

19

@17 - Yes, ST has said they can run the trains every couple minutes if they have to. They are worried about some bunching (delays) when they run under three minutes, but theoretically 90 seconds would be fine. Based on the way this system is designed (skipping important stops like First Hill) capacity isn't likely to be a problem.

Speed is largely meaningless with an urban system. Trains typically don't reach top speed. They simply accelerate from one station, then immediately decelerate to the next station. If our trains were BART style (80 MPH) instead of light rail (55 MPH) most people wouldn't notice. What does take a while is the time spent at each station. For whatever reason, Link spends a lot of time at each station. That could change.

As it turns out though, we have built a system with some very long stretches (e. g. Rainier Beach to the airport). On those stretches, a faster train would save some time. But generally speaking, that really doesn't make a difference -- very few people ride it that far. Despite the overall fast speeds from downtown to SeaTac (because of the very few stops) there still aren't tens of thousands of riders south of Rainier Valley. Most ride it in town (as is the case with every transit system in the world).

In terms of automation, the value of such is overblown. Vancouver saves some money by running automated -- and loves to pat itself on the back -- bit it really doesn't make much difference. The cost of a driver is minimal. Unless, of course, you don't have many riders. But then other costs start adding up. It costs money to maintain the trains, as well as the tracks. The more often you run the trains, and the longer they run, the more trains you need to buy. That is why Denver just cut back on service on one of their lines. It wasn't that the driver cost is so expensive, it is that sending a train out miles and miles just doesn't add up unless the train is full.

Which means that if you are really are concerned about cost, you should ignore driver cost, but look at fares per hour. How many people board per hour. For a typical suburban trip, not that many. The train takes a long time, and folks are largely just getting on the train headed into town (and getting off while leaving). For an urban trip, it is the opposite. It doesn't take that long to run a train from downtown to Northgate, but all the while people will getting on an off the train, as they do in every city around the world. They will take trips from Roosevelt to Capitol Hill, or Northgate to the UW -- trips that are popular but slow today, but will be fast in the future. On the other hand, you just won't get that many people taking the train from Tukwila to Fife, or Ash Way to Mountlake Terrace (trips that are both fast and unpopular right now). The design of the system (where you put the stops, how the various lines intersect) has a lot more to do with the cost effectiveness of it than whether you automate it or not.

20

@18- Completely agree that the city needs to be well served (and hope I live long enough to see the Ballard line running), but I have to say that NO urban rail should EVER be run at grade level. Period. No upside to it and it can only slow things down. In our case, it's a giant bottleneck on the system running south and a big reason that it takes longer to get to the airport on Link than it often did on the bus. San Diego had a grade level crossing on their trolley line near the airport and wound up spending bg $$ to separate the two because of traffic pileups. Let's not repeat those mistakes.

21

@20 -- I hope I live long enough to see the Ballard line as well. By that I mean the Ballard to UW line. But since ST hasn't even proposed it, my guess it will take a while. It may never happen, while we focus on transit to places like Fife and Ash Way. Meanwhile, the Central Area and Belltown have nothing.

Anyway, surface rail is really no big deal. The most popular light rail system on the world runs on the surface through downtown (Calgary). In our case, the cost to riders is minimal. There just isn't enough distance between stations to go much faster. There will be even less once Graham Street is added. It isn't why trips to the airport are slower on the train than they were on a bus. It is because the train makes lots of stops. That's what subways do. If you ride the subway from Queens to Manhattan, it will take a while, even though the train is capable of going very fast. You will likely be better taking an express bus. Folks don't really seem to get this. Subways aren't great for long distance travel. The best ones aren't designed for it. The worst ones (like BART and DART) carry few riders.

You also seem to ignore the benefit to riders of surface rail. Consider the difference between taking the train at Capitol Hill versus Othello. With Capitol Hill, you spend a good five minutes just getting from the surface of the street to the platform. At Othello, you just walk right there. That is a lot of savings for a lot of riders. That means folks in Rainier Valley come out way ahead, especially if they are taking trips within Rainier Valley (e. g. Columbia City to Rainier Beach). The only people who are hurt are those who skip Rainier Valley (e. g. those headed to the airport). If you add up the big time savings of those in Rainier Valley and the minimal cost to those who skip it, my guess is surface rail has a big net savings. It would be different if the surface rail was in the middle of the system, but there simply aren't that many people trying to get to the airport (on the train). It is highly likely that as Rainier Valley grows in population, ridership in the valley will continue to increase, and the overall savings will increase.

You also can't ignore cost. I would much rather make a down payment on a UW to Ballard subway, or a Metro 8 subway before burying the track.

22

Here is another way to look at it. Imagine we operated like SkyTrain, but with our current alignment (as built and proposed). So that means the Rainier Valley train would be elevated or underground. The trains would all be automated. It would be less than a minute faster to get to the airport, while getting to a stop at Rainier Valley would take longer.

It would be way more expensive that what exists (or is planned) in Vancouver. We have a lot of deep bore tunnels, and new giant bridges and way more miles of rail. That is before you add in the extra cost to bury (or elevate) the trains in Rainier Valley. Because of the extra rail, it is a lot more expensive to operate and maintain.

But it would still suck compared to what Vancouver has. Look at a SkyTrain map. Better yet, look at the Vancouver Frequent Transit Map (http://infomaps.translink.ca/System_Maps/frequent_transit_network_map.pdf). Now imagine trying to get from one reasonably popular place to another. You might need to take a bus, but for the most part, the bus is fast (no major bottlenecks). You might might even take the ferry. Through some combination of relatively fast and frequent bus, ferry and train, you can get just about anywhere. That is why the system (and not just SkyTrain) is so popular. Ridership on SkyTrain is much higher than ridership on Link, and their buses are even more popular.

Would it be fucked up if the trains had drivers, or the Canada Line ran on the surface? Not really. It would mean that it would take a little longer to get to the airport, but no big deal, really. It isn't what drives ridership, and the airport isn't that far away (which is why the train doesn't run there as often).

Now look at the ST3 map. Imagine trying to get from one place to another, and you can see that there are bottlenecks everywhere. Getting from the Central Area to downtown or C. D. to the U-District (trips that are very close and very popular) remains extremely difficult, despite the fact that you likely cross over the train. First Hill manages to be in the middle of it, yet ignored.

Meanwhile, even places that have stations aren't great. Try getting from Ballard to the UW. Even if you are standing by the station, it isn't worth taking the train. You are better off slogging your way on the very slow 44. What if you are on Phinney Ridge, and want to get to Lower Queen Anne? It is only about a mile to the Ballard train station, but you can't get there -- there is no way a bus is going to head down Phinney Ridge, then take a hairpin turn to serve Ballard. You would have to take two buses just to get to the train. It isn't worth it.

Here is my favorite: Imagine being at Fairview and Harrison, headed to 15th and Harrison. It is only about a mile as the crow files. Right now you have to walk to catch the 8 (which is very slow) and then walk some more on the other end. After spending billions supposedly serving these exact areas, you really don't have a better option. You could walk much farther to the station on Denny (or the one on Aurora), then catch a train, transfer to the other train, then walk much farther from Capitol Hill Station. You are better off slogging your way with the 8, walking, or hiring a cab. This is for a trip that is only a fucking mile, in the heart of the city. But when you fuck up the stations and the lines, the system becomes fucked.

We are spending more than Canadian cities on rail. We are making them "fast", by building big tunnels and bridges. Those Canadian cities will continue to have much better systems not because they are automated (most aren't) or avoid the surface (Calgary doesn't) but because they don't have their head up their ass when it comes to planning. They actually plan for trips that involve the bus and design the thing so that it serves the entire region. We simply fuck around, draw lines on the map and think "maybe a subway to the Tacoma Dome would be cool"

23

The author was quite wrong about why BART was never built to Marin County, ascribing it to "wealthy homeowners". Actually Marin voted to join BART by a wide margin. But when San Mateo County elected to leave the BART District, it reduced the tax base, making it unfeasible to build a line to Marin across the Golden Gate Bridge. The BART board asked Marin to withdraw from the District.

24

MARTA sucks! I lived in the Atlanta area from 2002-2015 and trust me, our transit here is way better than the broken-down, smelly, dangerous MARTA system. And, by the by, the stupid nickname for MARTA isn't that offensive, unless you're a timid snowflake. MARTA officially is: Metropolitan Atlanta Rapid Transit Authority. The "offensive" name? Moving African Americans Rapidly Through Atlanta. Woo! So awful!

Washington D.C.'s public transit is excellent, though, yes, it is starting to really need some major refurbishment.


Please wait...

and remember to be decent to everyone
all of the time.

Comments are closed.

Commenting on this item is available only to members of the site. You can sign in here or create an account here.


Add a comment
Preview

By posting this comment, you are agreeing to our Terms of Use.