Potential new light rail lines for the upcoming Sound Transit 3 ballot measure were clarified further after a recent Sound Transit board workshop. From 4 options for the Ballard to Downtown line, only one provides a stop in Belltown, while the rest cover South Lake Union. This dichotomy need not be true. It is possible to combine service to the fastest growing job center (South Lake Union) with a stop that serves both the second densest residential neighborhood (Belltown), which is also a major employment center and some of our most popular tourist attractions (Space Needle and Olympic Sculpture Park).

This article builds upon the grade-separated (tunnel + elevated) alignment proposals only. Grade-separated lines can carry as much as three times as many people per hour compared to at-grade lines and do so with significantly higher reliability. Given that we are spending billions of dollars for a system that is to support strong population growth over the next 50 years, we get the highest return on our investment by building grade-separated lines.

The Seattle Department of Transportation has proposed two stops to Sound Transit: one at the heart of South Lake Union (Denny & Westlake) and one in the northwest corner of the neighborhood (Aurora & Harrison). Sound Transit has accepted the first proposal as part of its alignment options, but is treating the latter as a separate project. The proposal below moves that latter station to Denny & Broad and goes over the implications of that in terms of mobility, future extendibility, cost and performance.

Let’s start by looking at the 10-minute walksheds (in blue) of the two stations (courtesy of IsoScope):

Belltown alignment
Belltown alignment
Harrison alignment
Harrison alignment

Before we look at the benefits, it’s important to understand that the lines are roughly the same cost because:

  • The Belltown alignment is only around 600 feet or 0.11 miles longer. Based on University Link costs of around $200 million per mile for tunneling and track work, this would make it $20 million or around 0.5% more expensive.
  • The Belltown station is likely less complex and cheaper as it can be shallower than the SR-99/Harrison station, which has to allow for the light rail tunnel to go under the SR-99 tunnel approach. As stations run $180 million each (University Link again), the savings from the Belltown station can offset the cost increase from the slightly longer tunnel.

Travel time is also roughly the same. While there are sharper curves in the Belltown alignment, they are all next to a station where the train has to slow down to a stop regardless.

To understand what we stand to gain or lose with each station, we can compare the difference in walk sheds. The area in red below is the unique non-overlapping part of the walkshed added by each station:

Belltown station walkshedHarrison station walkshed

  • High density residential area (6-25 floor buildings) with some office buildings (3-14 floors)

  • First Avenue and Second Avenue Belltown retail/entertainment corridor

  • Major employers: Cisco, Zulily, Kiro7, KPLU, KOMO, UWMC; 30+ smaller office-based employers

  • Tourist attractions: 2 minutes to Space Needle; Olympic Sculpture Park, Clipper Ferry Terminal

  • One-third of additional area is zoned to allow 8-floor buildings with a section less than 1/5 zoned to allow up to 25 floor buildings

  • Most of the additional area is north of Mercer and is currently zoned for 4-6 floor buildings, although a potential Uptown rezone can change that

  • Major employers: Gates Foundation, UW, Allen Institute

  • Tourist attractions: Lake Union Park

Note: New development shaded in.
Note: New development shaded in.

From a visual perspective, the Belltown stop could serve more residents, visitors, and likely a comparable number of workers to the SR-99/Harrison stop.

It also serves tourists better as a direct line from the airport covering all of Downtown, Pike Place, the Space Needle, WSF and the Victoria Clipper.

The 8 Subway

A tunnel under Denny also lays the foundation for a potential “8 Subway”, as proposed by the grassroots organization Seattle Subway. The Denny corridor has similar ridership to the Ballard-Downtown corridor (14,700 vs 17,200 projected in 2030, for each line respectively) and it is a great candidate for light rail. While a South Lake Union stop does provide a connection to Capitol Hill already, there are more destinations on the corridor that are outside of its effective walkshed — Summit, Hilltop, Madison Valley to name a few.

Regardless of whether an 8 Subway is considered, the Belltown stop can still serve far more people than a SR-99/Harrison stop in terms of residents, workers, and visitors. Even just the stretch on Denny from Westlake to Broad is very congested, yet a common corridor for South Lake Union workers living in Belltown.

If you want Sound Transit to consider a Denny & Broad station, e-mail them, tweet referencing @SoundTransit, or submit a comment online.

We hope you loved this article. If so, please consider subscribing or donating. The Urbanist is a 501(c)(4) nonprofit that depends on donations from readers like you.

Anton has been living in the Pacific Northwest since 2005 and in Seattle since 2011. While building technology products during the day, his passion for urban planning and transportation is no less and stems from a childhood of growing up in the urban core of a small European city.

Inline Feedbacks
View all comments
Ken Haufle

You made a great case that a light rail station at Denny and Broad is superior to Aurora and Harrison. It would also enable a great opportunity to introduce a new mode of high capacity transit, the urban gondola. , Broad street is a prime corridor for an urban gondola stretching from the waterfront at Alaskan Way all the way to the streetcar station Lake Union Park.

A combined light rail and gondola station at Broad and Denny would expand the throughput and usefulness of that station offering fast, frequent connections to many top business, residential, and tourist locations.

Heading northeast from Denny a station at Harrison would have a walk shed including the Gates Foundation, massive Seattle Center parking garage (gasp!), and presumably some areas of South Lake Union east of 99 once the tunnel is completed that would have been in the Harrison/Aurora station walk shed.

The next station would be in the vicinity of Mercer and Dexter opening up frequent access to a lot of residents along Dexter north of Mercer. This station would be a turning station with the routing now heading through the small lonely block of Broad Street that connects Roy and Valley with a final station at Lake Union Park next to the streetcar station. This station in addition to increasing it’s usefulness by adding another frequent high capacity transit connection also offers great access to tourist locations and eventually dense business or residential developments across Valley Street.

Heading the other direction, a terminus at Alaskan Way would be within walking distance of the Olympic Sculpture Park, the waterfront and many businesses and residences. Instead of a terminus it could actually be a turning station creating frequent access along the waterfront down to the aquarium or beyond.

Some things to consider are high power lines in the vicinity of Harrison and the landing path for sea planes onto Lake Union.


Excellent article and interesting proposal. There are a few things to considers:

When the SR 99 work is done (when Bertha finishes digging the tunnel) you will be able to cross Aurora north of Denny, up to Harrison. Thus a stop at Harrison would enable access to a bus that goes east-west on a street that has at least one lane reserved for buses. This would provide much better access to South Lake Union than the one or two stops that barely skirt the edge of it.

A stop at Denny and Westlake as well as a stop at Denny and 3rd Ave have some similarities. Both are on the edges of the area they purport to serve. They both have different weaknesses. A stop at Denny and Westlake is very close to the main Westlake station. This cuts significantly into its value (as shown by your excellent map). I doubt anyone will take a train from Westlake to Denny. I think a lot of people will avoid the transfer as well, and simply get off at Westlake and walk. The station is a good one assuming you are making a line to Ballard, but we shouldn’t overvalue its impact as serving South Lake Union when it barely does so.

The Belltown station has a different problem. It is far enough away from Westlake to warrant a trip on the train (although you could say the same thing about Harrison). Meanwhile, the high density areas are all to the south. The highest density census block in Seattle (by a wide margin) is on Western, to the southeast of the sculpture park. This station is not very far from there, but it isn’t centered around it either. The biggest problem with the station placement is that it is adjacent to the Seattle Center, an area already served by the monorail. The Seattle Center employs relatively few people, and doesn’t house anyone. It can be thought of a giant park. It really doesn’t matter which side you serve, and the Uptown site serves it adequately. The Seattle Center eats into the potential ridership quite a bit. A stop at say, 2nd and Vine would have much higher ridership, and more importantly, provide a much greater value.

Which is not to say that it is possible to serve both South Lake Union and Belltown with sufficient stops. This comes closest to serving both areas and is is an intriguing idea. It might even be possible to do some of this with cut and cover, which would might save money (although I doubt folks would allow Denny to be shut down for a few years).


Yeah, I think it would make the most sense to make Thomas the main east-west bus route. That would open up the possibility of it running through the Seattle Center. As far as streetcars go, ours don’t make any sense, because they have a smaller capacity than our buses (and cost more to operate, have limited routes, etc.). But if that is what it took to get transit through the Seattle Center, then I would support it. But until then, our buses will continue to carry the bulk of the transit riders, and our streetcars will continue to be irrelevant.

But your suggestion makes sense with bus service. The city is already working on getting various improvements in the area (bus only lanes, a BRT with off board payment and signal priority, etc.).

I do think, though, that the best overall solution is still the WSTT. It would connect all those places with more frequent service (since buses would run more often than trains) and make the connection to Aurora much better. You would have a stop on (roughly) Aurora and Denny*, but also a one seat ride along Greenwood/Phinney/Fremont/East Queen Anne/Belltown and the rest of downtown.

* The WSTT map shows a stop at Aurora and Denny, but I assume that could be moved to Aurora and Thomas. I could also see dual entrances depending on the size and depth of the station.


Oh, no, not the streetcar debate again. OK, I made a mistake I should have said “our streetcars don’t have a greater capacity than our buses”, rather than “our buses have a greater capacity than our streetcars”. Sorry about

As should be obvious, the main reason the streetcar numbers sound better is because of the seat configuration. It would be misleading if I said that our buses seat way more people, because the seats can be arranged any number of ways. But seating capacity is an obvious, indisputable fact. One person per seat. But standing room is not, and the numbers you claim (or the company claims) are debatable, to say the least: http://seattletransitblog.com/2014/07/29/streetcars-a-momentary-lapse-of-reason/#comment-511430

In other words, just change the seats on the buses and you get the same capacity. The buses are a bit shorter, but a bit wider. It’s a wash. The two vehicles are roughly the same and have roughly the same capacity, depending on seat configuration.

Not that it matters. Even if the streetcar can carry 5 more passengers, what difference does it make? The money spent on them could be put into more service. You only get an operational advantage if you have very big streetcars (which we don’t have) and have maxed out the headways (the buses are backed up behind each other). We are nowhere near the latter.

Smoother ride? Who cares? I honestly have never met anyone who would turn down a bus ride for a streetcar (assuming they are going to the same place). Why would you? So that you can test the capacity limits of the streetcar (my, so smooth when we are all packed in Tokyo style). It is also laughable to suggest that the city won’t pay for pavement improvements but will pay for a rail line used by a very small percentage of transit riders. I’m sure that decision goes over really well with bikers.

Streetcars only make sense in certain areas, and this isn’t one of them. They are still being used by Toronto because it is easier to just maintain their current fleet, rather than switch over. Yet Vancouver doesn’t have any. The third highest transit ridership in North America per capita (about three times our ridership) and they don’t have any streetcars. They have a single bus line that carries 55,000 people a day (which is more than our light rail line) and yet not one streetcar. New York has a huge bus system, some fairly flat areas, yet no streetcars. I guess they don’t need the extra capacity. Maybe when New York gets to be as big as Seattle (or as dense) they can think about streetcars.

Streetcars are inflexible. You are absolutely right — our current line is bad. So why not move it? Just the other day the route for the 10 was changed. So why not move the streetcar route? The answer should be obvious.

What happens when an accident occurs on the street? The streetcar is stuck. How about construction that requires the vehicles (with the help of a
flagger) to move to the other side of the road? Can’t do it with a streetcar. You can do that with a bus, you can do that with a trolley (to a certain extent), but you can’t do that with a streetcar. Some of the newer trolleys are even capable of traveling off wire to avoid a street (if the street gets closed down for some reason). They are just more flexible.

To say nothing of being able to handle our hills. Madison won’t work for a streetcar. Neither will Yesler. Thus you get the screwy routes that we have for streetcars, and the sensible ones we have for buses.

The city is in the process of improving various surface transit areas. They only have so much money. Right now there is the possibility of having what they call the “Full BRT” option for Roosevelt to downtown transit. The buses would average over 25 MPH south of Denny, and over 21 MPH north of it. This means that a trip from 45th and Roosevelt to Westlake Center would take ten minutes. From Westlake Center to anywhere in the South Lake Union area would be less than three minutes. A trip from Eastlake … well, you get the idea. This is at rush hour! This is competitive with light rail. These speeds are shocking, really, yet with the proper treatment, quite possible. Building them won’t cost billions (like like rail) either.

But it isn’t cheap. It isn’t clear at all whether we will have this, because we only have so much money. Do you really want to sacrifice that line so that some riders can have a smoother ride?


You make a lot of good points. Let me respond to them by number:

1 — I agree completely.
2 — Fair enough. High capacity surface rail certainly has its place.
3 — I’m not trying to say that our streetcar lines are bad, I’m saying that just about any surface only rail line in Seattle is a bad idea. We just have too many hills. Surface running is fine, but there are very few areas where you can do that for the distances necessary to make it work. There is nothing wrong with Link running down Rainier Valley (quite the opposite). If light rail runs to Ballard, then running down Elliot makes a lot of sense. But those lines require tunnels to avoid the major obstacles. The lines integrate surface running with tunnel running. Most of the plans for what the city is calling streetcars don’t do that.
4 — Madison BRT has center running where it is needed most. There will be BAT lanes downtown. I am concerned about that as well, which is why I wrote this (http://seattletransitblog.com/2015/11/16/faster-madison-brt-through-downtown/). I also think BAT lanes (or center running) should be extended farther east. But the studies suggest neither the contraflow lanes downtown or similar improvements to the east are necessary. If the studies are wrong, then some of the changes (such as adding BAT lanes to the east) are trivial and very cheap. Other changes (such as moving the entire thing to allow for contraflow or adding additional center running) is not. But it is nothing like moving a rail line.

As far as seeing real BRT (gold level or even silver) I think it is quite possible. I think it is way more likely than actually getting a similarly rated light rail line. The lack of urban spacing through our most urban areas, the lack of any consideration of bus to rail interaction, the focus on areas extremely ill suited for rail (while the handful of places where it would make sense are left with nothing) means that the whole system would probably get a grade of “Not Light Rail” if a rating agency made such a judgement. We are very fortunate that a rating agency exists for BRT — we need one for rail.

Oh, and I think the chances of anything called a streetcar in this town making as much difference in the lives of as many people as BRT is very slim. Like you said, it really is about the route — and a Madison route would simply be impossible with surface running rail.

5– Yes, the Madison BRT costs more, and it will carry more people. Six minute headways all day and night (better than Link, by the way) along with very fast speed along a very congested and urban landscape. It is likely that more people will ride that than will use the Link station at Capitol Hill. That is worth pondering. There is only one station between the UW and downtown, but unless ridership greatly exceeds expectations (and so far ST’s estimates have fallen way short) then more people in the area will ride a system costing less than 10%. It will allow for a fundamental change in the bus service in the area, which is something that was impossible with Link (because again, there was only one station). You can finally get rid of the 43, because now the transfer that is required to get downtown is simple (did I mention six minute headways and very fast speeds?). You move the 11 onto John, which means the 10 can go back to where it was.

In a few months Link will get to the UW and lots of riders in the Capitol Hill/C. D. area will suddenly hate their new bus route. Madison BRT will be the opposite of that.

Oh, and the high cost has a lot to do with center running (which our streetcars lack). Apples to apples service is just cheaper with rubber tires.

6 — Yes, everyone (including me) thinks that the 99 B-Line should be converted to rail. Well, everyone but the voters in B. C. I guess. But my point is that despite the wait, despite the overflowing number of riders, it carries a huge number of people. But would you rather have that, or a train that runs every ten minutes? That is the top end, the most egregious example of a line that should have been converted to high capacity rail a long time ago, yet it is still a better experience (i. e. get you where it wants to go faster) than many of our proposed light rail lines (e. g. West Seattle). That is because it is frequent. None of our streetcar lines will come close to that number (even our light rail line has fewer riders).

7 — OK, so buses make the riders sick and trains don’t. I’m sorry to hear that. But I’m claustrophobic and sometimes all of these tunnels freak me out — does that mean we should avoid subways? Of course not. Besides,

I think it is unrealistic to assume that the bulk of your trips will involve only rail. Skytrain carries ten times the number of riders that Link does, yet the bulk of the ridership is still on buses. The two work together, which is key.

Conclusion — I agree, it is about alignment. But let me propose an alternative:

a) Ballard to UW light rail.

b) The WSTT

That is probably cheaper, but even if it isn’t, it will carry way more people, despite the fact that some of the vehicles with your proposal have more capacity. Do you really think we should spend the city’s obviously limited funding (so limited they are willing to run buses on BAT lanes downtown or throw away 23 MPH averages through South Lake Union) so that we can lay some rail?


4: OK, and what happens if, as people fear, Madison encounters slow downs? What happens if they have to move from BAT lanes to contraflow as I suggested. Again, this is a trivial change with buses, but not with trains.

Besides, as you have pointed out, the advantage of trains only come into place when they are longer. Being longer means having much bigger stops. That adds to the expense (of course).

5. If you are going to repave a street, you might as well put in rail Yeah, tell that to the next biker being carried off by Medic One.

6. I wasn’t comparing SkyTrain to their BRT. I was comparing their BRT with our light rail. West Seattle light rail, for example, is slated at 10 minutes peak headways. Peak! Run buses down several corridors and even if by some magical happenstance the buses are too full, it will still be a much better, much faster experience.

As far as what the voters voted for, I am well aware of the reasons given, but the fact remains that they voted against it. To make an obviously symbolic, probably pointless vote against a very good set of proposals suggests to me that the current system is not that bad. People in Seattle hate Metro. They are one of the least respected, most insulted agencies. But when push came to shove they voted for more bus service because we need that big time. Vancouver, apparently, does not feel that way about the latest proposal (as good as it was).

7. Again, claustrophobia is real, too. To quote Wikipedia, extreme claustrophobia effects somewhere between 5 to 7% of the population. Are you suggesting we give up subways?

Yes, there are people who prefer rail over buses. There are people who prefer taking a ferry or taking the monorail. I get that. But in the grand scheme of things, there are very few places in the world where rail ridership comes close to bus ridership, and essentially none the size of Seattle. Almost all of those cities had subways built a long time ago. The small number of extra riders that you gain by choosing one mode over the other is minimal. What matters is to pick the right tool for the job, as Vancouver has done.


Just to be clear, the planned headways on West Seattle are not a freak accident. Running a big train is simply more expensive than running a bus. To justify running it you need to pick up a lot of people. If you run in an area with fewer people than the Madison corridor (and West Seattle rail will be one of those corridors) then you can’t afford to run a train very often. BART runs every 15-20 minutes, despite serving an area way more populated (with way more densely populated segments) than us.


Trade-offs and facts. It is important that we have these right.

4. There are things a bus can do that a train can not. There are things a train can do that a bus can not. A bus can go up a very steep hill. Part of the reason our streetcar lines are so bad is because of that limitation. A bus can avoid an obstacle (temporary or long term). The biggest trains have much higher capacity than the biggest buses. It makes sense to use them where such capacity is needed.

5. Bike an rail interaction is a problem no matter where you put the rail. Imagine Madison as a streetcar line.Now imagine every intersection, and every bike that trying to cross Madison. They aren’t trying to ride on Madison, but trying to cross it. They have to swerve to avoid a highly acute angle and going down. Swerving equals confusion equals danger. The city can do things to reduce the risk, but those usually require taking more space for bikes. Since the city has a limited budget, those sorts of safety improvements mean that other safety improvements are made. Again, trade-offs. A smoother ride for transit users versus a few injuries a year (or worse) for bikers.

You actually don’t have to imagine that scenario, you can just look at Jackson, where there is center running, and accidents have already occurred (even though no steetcar has run there): http://www.seattlebikeblog.com/2015/06/17/the-new-jackson-street-isnt-safe-heres-a-way-to-fix-it/

7. If I call a cab, I would rather be picked up by a limo than a Prius. But mostly I just want to take the first vehicle that arrives. Most people are like this. Most people will cram themselves into a crowded bus and curse the bus driver for not letting them on instead of waiting a few minutes to get a much more comfortable ride. This is why, for example, the streetcar has very few riders, despite being in a very high demand area (it has way fewer riders than the 70 or the monorail).

8. Sorry, you simply don’t have your facts right. It is more expensive to run one train than it is to run one bus. Do I really have to cite a source for that?

You are claiming that the high capacity of a train will make up for that. That is only true if the trains actually carry a lot of people. That is my point. Even now, even years after our first rail line, the farebox recovery rate of our light rail is about the same as our buses (both ST and Metro). They have decent fare box recovery because they don’t run that often. A line out to the airport every couple minutes would have horrible farebox recovery, which is why they don’t do that.

SkyTrain saves money by automating the operations. But most of the savings occur because the trains are full. Toronto has better farebox recovery. This makes sense, because Toronto is simply a much more densely populated city. Of course their trains are more crowded. But Vancouver has great numbers: http://www.comparetranslink.ca/#sthash.JNtGfh8u.dpbs. We do not. Vancouver has (as you mentioned) very frequent trains. We do not. This is because Vancouver puts their rail where it makes sense to put it. We do not.

Look at the numbers for farebox recovery: https://en.wikipedia.org/wiki/Farebox_recovery_ratio. Find Toronto. Find Dallas. It is not that the trains in Toronto are cheaper to run, it is that they simply pick up a lot fewer passengers. They pick up a lot fewer per trip, despite running very infrequently.

I think we both agree that Vancouver has a very good transit system, and that the voters there were fools for voting against the expansion. But we aren’t building what Vancouver has. The only thing that is even remotely similar is the tunnel, which Sound Transit didn’t build. We are building low capacity surface streetcars (which Vancouver doesn’t have) along with distant, DART type rail. It will run infrequently (despite all the claims) and will represent a very small portion of the area’s trips (even ST doesn’t deny that). About the only thing we (the city, not ST) is doing right around here is building BRT, which will soon do for the Central Area what Link failed to despite spending ten times as much: serve First Hill with fast, frequent transit and enable a much improved bus network. Roosevelt BRT could do the same — deliver speeds similar to light rail — but only if we spend the money. The city, unlike Sound Transit, doesn’t have a lot of money, which is why spending it on rail when it isn’t needed (and has the limitations mentioned) is a really bad idea.

High capacity surface rail is fine. But it needs to be on a line that makes sense. Rainier Valley makes sense because it can connect to downtown via a tunnel. 15th West makes sense because it could connect Ballard to downtown via a tunnel. But low capacity rail (what we call streetcars) just doesn’t make sense for us because of geographic limitations and higher capital costs.


My larger point is that the Seattle Center doesn’t need three stops. If you are headed to the Center, it doesn’t much matter which stop you take. The Uptown one would be fine for most riders. If you are headed to the Space Needle, or the glass museum (e. g. a tourist) then nothing will get you closer than the monorail. If you are headed to a Storm game, then Uptown makes sense. If you are headed to Folklife or Bumbershoot, then any stop will do.

This makes it different than a stop geared towards day to day travel. Spending an extra five minutes walking to work each day (and back) makes a big difference.


Oh, I agree. I just think that without that stop those same people just get off at Uptown (or takes the monorail). This is different than the person who lives (or works) in Belltown, who simply won’t take Link unless it has a stop like the one you suggest. There are plenty of people like that, but there would be more if the stop was farther to the southeast.


Excellent. Thank you very much for that. It would also be interesting to see a map like that with the WSTT stops (which are the same as this, but with the addition of another stop at Aurora and around Thomas. There would be a lot more overlap, but would still have more coverage than any other option (largely because of the addition of another line).


Would that type of stop in Belltown really be beneficial for going southbound? During most hours 3rd and Cedar and 3rd and Bell have buses going through to Westlake station every 2 minutes or so. It seems like the two extra stops between the Belltown stop and Westlake would take longer then any bus down 3rd.

With that, those in Queen Anne and north Belltown would benefit but I feel like the highest density areas of Belltown wouldn’t.


Your proposed alignment would be awesome, Anton, as it would hit the densest and most walkable parts of the north-downtown neighborhoods. Belltown definitely should not be left out, and SLU is important because of the job density there. Let’s make it happen!

On your map you have a Denny/Aurora station – I think that could be combined with the SLU station if that tight curve is anything close to what the actual alignment would be, as a quarter-mile spacing is too close for light rail service.

Chris Corr

>>as a quarter-mile spacing is too close for light rail service<<

No it's not. It's not even too close for heavy rail service. It may be relatively rare with both light and heavy rail to have stops that close together, but it's not unheard of.


Good article! I think that the Gates Foundation, which is on the corner of Mercer and 5th, is covered by the Belltown DSTT2 alignment according to your walkshed calculation! You should include that in the “both” category.