Trying to build on what Adam said in his great post about station spacing, I want to look at the transit system with the farthest spaced stations, BART, and compare it to its “sister system”, the DC Metro, and talk a bit at what that might tell us about Link.

DC Metro
Crystal City Station, photo by mischiru

Christof Spieler over at Citizens’ Transportation Coalition, a transit advocacy organization based out of Houston, had a blog post comparing BART and DC metro, and he first noted the similarities between the systems:

San Francisco is in the 5th largest metropolitan area in the country. Washington DC is in the 4th largest. Both cities have old, urban cores with major employment centers surrounded by extensive post-war suburban development. In the 1960s, both decided to build a heavy rail system. And not only do those two systems use very similar technology, they are nearly the same length (104 miles vs. 106 miles).

There is one significant difference between San Francisco’s BART and Washington’s Metrorail, though: Washington has 2 1/2 times the ridership (902,100 average weekday boardings compared to 338,100.)

Mr Spieler sees three main reasons for the massive difference in ridership:

  • BART serves the suburbs. Metrorail serves the suburbs and the urban core.
  • BART saves money by using existing rights of way; Metrorail maximizes ridership by puting lines where the transit demand is.
  • BART stations are where the cars are; Metrorail stations are where the people are.

The big result is that BART has ended up being a big commuter rail system, and DC Metro is both a commuter rail system and an urban Metro. What this means for the core city is that the DC Metro can be used to get to different areas around town, while inside San Francisco BART acts more like a single line on the Muni Metro, itself a sort of small-scale commuter rail system. Commuters who use BART have to transfer to local transit if the BART station isn’t within walking distance of their destination.

Obviously there are historical, geographical and political reasons for these differences. BART had to be passed through many levels of local government while the DC Metro was first created under Federal Law, and what the Feds want, they get. The DC Metro also received much more Federal funding than BART did. With more money, most of the DC Metro is underground, where as BART is underground only in San Francisco, in parts of Oakland, and in Berkeley. In the last case, the city of Berkeley had to pay for its tunnel. The city of San Francisco is also relatively isolated on a pennisula, with large bodies of deep water on the west, north and east, and hills throughout the greater metro area. DC is in the middle of a valley with rivers but not much else in the way of natural obstacles. Finally, BART had a system of sub-area equity, where all money raised in a sub-sections of the metro area is spent there, and because so much of the DC Metro’s initial money came from the feds, sub-area equity was not an issue.

Comparing the two, Seattle’s situation resembles BART a lot more than it does the DC Metro. We’ve been getting a significant federal contribution, but we still need to come up with most of the money ourselves. We have a lot of local governments involved in the process, and Sound Transit has even had to fend-off repeated attacks from the state legislature and initiative process. We have deep and large bodies of water throughout our metro area, and both the city and the region have hills throughout.  And obviously Sound Transit has sub-area equity as well.

Taking an even deeper look, it seems that Seattle’s extensive bus system and burgeoning street car system are  a direct analogy to San Francisco’s Muni buses and Metro. The Muni Metro is a European-style light rail-streetcar hybrid, with a nine-station subway (with more on the way), twenty four surface stations – which resemble Link’s surface stations – and dozens of surface stops that look about like bus stops or SLUT stations. The First Hill and South Lake Union streetcars are the first steps toward a system that requires an inter-modal transfer from light rail toward “surface” transportation. There are more streetcar lines in the works.

Sound Transit's Long Range Plan

The bright spot in all of this is that Link already looks a bit better than BART in terms of its urban railway characteristics, and there are still opportunities to improve this further. There are just eight BART stations inside the San Francisco city limits and thus it only serves a small portion of the city, while Link after ST2 will have 17.5* stations that serve a much larger portion of the city.

In addition, if you take a look at the Long Range Plan Sound Transit’s board adopted in 2005, you can see a dashed-blue line from Downtown to Ballard and from Ballard to the University District (click on the image to get a better look). That blue line represents a potential rail expansion within the city limits that could be included in a future Sound Transit expansion package. The ST2 package that passed last November included money to conduct a study of high capacity transit in that corridor. Whether this would be Link light rail or a streetcar, and whether it could be built affordably and in a reasonable time-frame it’s too early to tell. But due to sub-area equity, any future ballot measure would have to include something in Seattle, and if it is more light rail, that might be the difference between Link resembling the DC Metro more than BART.

Even having said all of that, San Francisco’s transit system isn’t a bad one to emulate. BART may not be the second busiest “heavy rail” system, but it’s fifth. And the Muni Metro is the nation’s second busiest light rail system. San Francisco is also one of only a handful of cities, along with New York, Boston and DC, where most commuters do not travel by car. It’d be amazing if Link and our Streetcar network could combine the best qualities of both systems, but only time will tell.

* Central Link has 11 stations in Seattle: Rainier Beach, Othello, Columbia City, Mt Baker, Beacon Hill, Sodo, Stadium, International District, Pioneer Square, University Street, Westlake. U-Link has two: Capitol Hill, UW. North Link has 3.5 : Brooklyn, Roosevelt, Northgate and 145th st, which straddles the city limits with Shoreline and East Link has one: Rainier

51 Replies to “Lessons from DC Metro and BART”

  1. I’ve been on both systems and personally, I like DC better, I think for the reasons stated (you can use Metrorail to get around the city – and believe me it’s quite important being in a wheelchair). It’s nice being able to also check out the ‘burbs (shopping in Crystal City) then heading into the city for the Smithsonian (which has it’s own dedicated station on the blue/orange line)

    Best part? They honor out of town disability ID cards (like our Regional Reduced Fare Permit) if you head to their office :)

    1. That’s pretty cool. DC Metro is mostly underground, too, which somehow makes it seem faster.

  2. BART is kind of a mess. In addition to what you already mentioned: the fare system is not integrated with MUNI, the trains are much louder than DC’s for some reason, and the airport line only comes like once every 1/2 hour. I was really unimpressed after being used to Metro.

    That said, at least it’s there! More than you can say for us!

  3. I was just in that station two days ago, I love DC metro, it just rocks! It was also very full for a Saturday. Tourists also use the METRO just as much as commuters. They have built many lines around the tourist HOT SPOTS which also happens to be where a lot of commuters work. I also do not see trains running as frequent as DC Metro where they only post the last train of the day. I really have a feeling that ST will have schedules for their trains

  4. Both operate at around 60% fare box recovery. I think only the NY Subway does better in this country (London has shown greater than 80% and Japan operates at a profit). If fare box recovery is roughly equal it tells me Metro isn’t operating any more efficiently than BART despite it’s high ridership. Could that be the difference between the percentage user funded vs federally funded?

    BART only got to SFO in what, 2001-2002? I think it’s still 1/2 hour headways at the airport. It wasn’t able to achieve late night 15 minute headways until just a few years ago. I expect that to be the case for East Link but Central Link and University should be able to get there much sooner.

    Average speed for BART is around 45mph? What’s the average speed for Metro? Projections for Link? I feel there is too much emphasis on station spacing for East Link rather than extending the line (no ridership from Bel-Red vs a huge demand from Redmond. I think the lesson to be learned regarding station spacing is more stations are good to the point that they start to decrease average speed to what is achieved by “surface” transportion.

    One transfer isn’t bad. Lack of good transfers is the problem (i.e. long waits and reliance on screwball point to point routes rather than circulators).

    1. BART has very high fares compared to almost every other US heavy or light rail system. BART also requires payment for every ride and doesn’t offer much in the way of a discount for buying a large number of trips at once (a $64 ticket costs $60). BART has no transfers and does not honor transfers from any other agencies. BART also relies on unspent balances on so called low-value tickets as a source of revenue. I believe these factors help increase the farebox recovery of BART.

      With East Link I don’t think the Bel-Red stations are a waste. While they may not have much ridership with current conditions in the area, there is scheduled to be significant development in the area in the next 20 years.

      1. Redmond and north east King County have the demand now. 20 years from now that increased demand will dwarf Bel-Red. As it sits the Redmond extension is on the same time schedule as the Port’s purchase of the BNSF property; postponed indefinitely. Zig zagging back and forth through here is nuts. It’s a tiny triangle that starts out about a 1/2 mile wide and pinches down to a point. What’s more there are stations planned at each end which cover most of the area already. The whole thing is only 1.5 square miles (and Bellevue wants a big swath of that to be maintenance facility, the part closest to downtown). 7.5 stations for Bellevue population 120,000. 17.5 stations for Seattle population 500,000. 1.5 stations in Redmond population 50,000. Where’s the equity in that? Especially when you consider Renton, Kirkland, Bothell, Issaquah, and all off east King county are left out entirely.

        What’s going to benefit King County residents more over the next 20 years, a station at Marymoor Park or 130th & Bel-Red?

      2. First thing, deleting the Bel-Red stations won’t free up enough money to build segment E. Not even close. Second, according to the DEIS the ridership in segment E really isn’t all that huge, even using the 2030 numbers. Third it isn’t clear if the maintenance facility will be in Bellevue or Redmond. Based on the DEIS I’d guess ST probably favors the MF5 location in Redmond. Building the maintenance facility isn’t part of ST2 anyway.

    2. Sound Transit’s projected travel times put the average speed around 30 MPH. Adding stations isn’t particularly costly compared to the cost of building the track, but I do think they should focus more on travel time, especially for people like me who are used to a quick and direct 545 trip. ;-)

      1. Japanese trains are fairly expensive

        Not really. The base fare for the Tokyo Metro subway is 160 yen, and that will take you as far as 6 km; compare that to New York’s $2 fare regardless of distance or BART’s $1.50 minimum fare. Intracity transit by rail is comparatively cheap in Japan; intercity transit is more expensive per kilometer,

        extraordinarily crowded

        Only in Tokyo will you find trains pushing 200% of capacity, and even then only at certain times and only on certain segements. Most other major cities have 100-130% crowding on inbound trains on major lines during the morning peak.

    3. I believe that Philly’s light and commuter rail systems (PATCO and SEPTA) also do around 60%, but those (together with the aforementioned NYC, BART, and DC Metro) are the best in the country by far for municipal systems. Obviously the Northeast Corridor does better, as intercity rail.

  5. Good analysis. I think the topography provides similar constraints between Seattle and the Bay Area. But, the lessons that we can take advantage of are priceless.

  6. I was just in SF this weekend. The lack of a common pass between BART and Muni is almost ridiculous. The fact that there are no TVMs for Muni was frustrating for my friends (I bought a three-day pass from a vendor). I’ve never heard of light rail system that has people pay at the front door, is it common?

    1. I’ve seen that in Boston and Pittsburgh. Boston’s green line is pay at the door everywhere except in the downtown subway…Pittsburgh’s “T” light rail system has fareboxes on board, but fares are collected by attendants at larger stops during peak times.

    2. Muni Metro is a mixture of mode characteristics in one system. On old street trackage works like a streetcar, with stops every few blocks and people would have to watch for car traffic to cross from the sidewalk and step up from the street pavement. Its proof-of-payment so you can board at any door or the 2nd car, but need to be at the front to pay the operator if you have no pass or transfers. In the tunnel, the stations and the trains operate like a normal metro, passengers have to go through turnstyles and board any door. On the new trackage at 3rd Street, it operates like a modern LRT, with TVM’s at the stations, however you can still pay the operator if you don’t like to fiddle with the TVM. The Muni Metro is slow as molasses on the surface routes, can go fast in the tunnels (if the computer lets it), its automated under Market Street, and manual mode above ground. The LRVs are high floor with steps that mechanically move up to create a flat floor when in the tunnel stations or surface stations. Most stops at the old trackage is non-ADA compliant but select stops have a mini-high platforms that the trains line up their front door with steps in the up position to load passengers. Its a neat system to check out San Francisco’s dense first ring suburbs. If you are getting off BART, you can get a Muni discount on dispensers near the exit turnstyles.

    3. There is a common pass, but only throughout the city. A city-fare muni and bart pass exists (I used to get one every month).

      1. What is it? I needed a three day pass and shared transfers and stuff like that. Is the shared pass only for monthly use?

    4. There are TVMs in the all of the subway stations, and most of the bigger stations, but not in the bus-stop like places.

      I think on-board payment is a feature of very old systems. Muni is about 97 years old now.

      1. TVMs for Muni? All the turnstiles took change, but I didn’t see any TVMs for Muni. I can’t find any references to TVMs on Muni’s website about fares, either!

  7. What happens if you add BART and MUNI ridership? It is a forced transfer, but the only way to get from an East Bay suburb to a SF Neighborhood (Say Orinda to the Sunset). On WMATA all such trips may be taken on a single integrated system/single ticket – even with required transfer.

    Demographics of users are pretty different between SF and DC as well.

    All that said on farebox recovery (60% is VERY good BTW). Skytrain, if taken out of the rest of the translink, does have 100% recovery; however, the supporting feeder bus service is no place near.

    1. Muni Ridership is pretty big, but combined they still fall short of the DC Metro’s ridership.

  8. The problem I see is that connecting Belltown and Ballard to Downtown is really vital, more vital than anything not yet being worked on and some things that are being worked on, but the cost will be huge. Because of the SAE, it will take forever for the North King area to have the funds to build this–it will have to wait for lots of less valuable outlaying projects to be ready in order to justify the tax rates in those areas. What is the most optimistic outlook on a Ballard line right now? Early 2030’s? With Northgate being finished in 2020, or even 2018, that means Seattle will go at least 15 years without any new service, even though it will vitally need it. Absurd.

    There needs to be a change in the system so that Seattle residents can tax themselves more for the projects that they needs. People want this and are willing to do so, at least in good economic times, as seen by the continual support that was shown to even the half-baked monorail plan. I don’t know if this is something the city could speed up by doing some tunneling work itself. If so, they really need to focus on that. I loved the streetcar network plans and hope it gets done, but we need to have the right priorities here. A dedicated line to Ballard (through Belltown) and eventually continuing to West Seattle, should be at the top.

    1. While I love streetcars, and I ride the South Lake Union line all the time, I wish that Seattle would instead focus on extending Link. I think that the money that will be spent on new streetcars would be better spent on extending Link to Belltown, Queen Anne, Fremont and Ballard. Seattle needs a high-speed transit system and that’s what Link will provide. The city should focus on expanding Link into a city-wide system and use busses for local service. Is there any way for the city to funnel money into Sound Transit to accelerate the expansion of Link? How about extending the Downtown Transit Tunnel through Belltown?

      1. Maybe when they have the ST3 vote, it will pay for most of the U District-Ballard-Downtown line, and then the remainder would be picked up by a special Seattle-only levy.

      2. It depends on how big a package is proposed for ST3.
        A Ballard-U-District line could cost as much as all the projects in the North King subarea for ST2 if it’s put underground, less if it were above ground, but then it would be very politically sensitive.

        A Ballard-Downtown line could be extraordinarily expensive, since a new downtown tunnel would likely be needed.

      3. The good news is due to sub-area equity something is going to need to be built in North King if Link is extended to Everett, Redmond, Issaquah, Tacoma, etc.

        I don’t believe all of the Sound Move and ST2 taxing authority will be needed to pay for O&M and bond servicing once the ST2 projects are complete. So even without additional revenue sources there still will be money for adding to the system. Though I suspect getting voter approval for further expansion is going to get easier not harder (there is the small problem of getting new tax authority though).

      4. Well, it has been argued on this blog before that any new line to Ballard/West Seattle would need a new tunnel through Downtown, as the capacity of the DSTT will be used up when Eastlink opens. This makes it a particularly costly venture. You probably would have to tunnel under Belltown and Seattle Center as well. Getting through Interbay is easy, but you have to tunnel again under the ship canal. So it looks pretty expensive.

        Seattle could go ahead and build the new tunnel themselves, like they did with the DSTT, but that is certainly not going to happen in the current climate. Dream scenario would be that they find someway to integrate transit into the 99 tunnel and thus piggyback on state funding, but that almost certainly won’t happen.

      5. A new Downtown tunnel could be cut-and-cover to Belltown, though it’d have to be bored under the battery street tunnel, so that could be cheaper than it seems at first.

      6. Maybe the most realistic plan, ultimately, would be to build as many streetcar lines as possible, and then later build the dedicated ROW and longer platforms for the most popular.

      7. No, the people at ST are pretty good at estimating ridership, so doing that would just add extra costs. And yes, the new downtown tunnel will be expensive, but it will definitely have great long-term benefits, like an ST4 or 5 line along Aurora. I would have the Ballard line be underground from U Dist-Ballard-across the ship canal, then elevated along 15th until before Uptown, where it would go underground, and stop at Uptown, Seattle Center, and Belltown. Then it would go in a tunnel either under 2nd or under the DSTT, before going over the West Seattle Bridge and going in a tunnel with stops at Avalon and Alaska then popping up and continuing down California until Burien or so.

      8. I totally agree that’s what we want and need, but I’m just not sure if it is going to happen, or if so how long.

      9. I don’t know a tunnel would be necessary for crossing the Ship Canal. Assuming an at-grade or elevated alignment a bridge would work too.

        On the other hand I suspect Ballard would demand a tunnel for any alignments West of 15th. Similarly a tunnel is probably the only politically viable option on 45th East of Stone Way.

        Building in Seattle isn’t going to be cheap, but it will have good ridership compared to the rest of the system. Again with sub-area equity it means if the other sub-areas are getting something expensive in ST3 there will be money to build something expensive in North King.

      10. A bridge would not work, because ships need to pass through the canal. Trains getting stopped at a draw bridge would defeat the point, and a a bridge high enough to avoid this would be way too big.

      11. I have no real idea of the feasibility of a bridge. Certainly one could be built to clear ship traffic, similarly a drawspan could be built. I know the SMP was planning to use a bridge across the canal. I’m not sure how much delay a drawspan would introduce into Link operations. I assume that would be one of the things that would need to be studied.

  9. I’m actually a regular Seattle Transit Blog reader, so I was happily surprised to see my nearly 3-year-old post resurface.

    I don’t know Seattle nearly as well as I know SF and DC, but I do think that BART is worth looking at as a cautionary story. The Central Link system seems to be starting out pretty well, and the next round of extensions to Northgate and Redmond seem pretty sensible, too: they link employment centers and universities and serve fairly dense places.

    After that, though, the plan seems to consist of just going further outwards. But does it really make sense to connect to Everett and Tacoma with both light rail and commuter rail? Wouldn’t it make more sense to try to connect the dense, walkable neighborhoods in Seattle? What about light rail on the proposed monorail routes? The streetcar system can serve that need — but I’d push hard for transit that doesn’t share lanes with cars, even if it is harder and more costly to implement.

    San Francisco, like Seattle, does have really high transit use. But the majority of those passengers are carried on slow, crowded buses stuck in city traffic. In San Francisco, the Geary corridor carries 100,000 bus riders a day (on Geary itself and nearby streets.) BART’s most recent extension (Milbrae/SFO), by contrast, carries only 28,000. A subway on Geary has been needed for 50 years, but providing better (or even acceptable) service for existing transit riders is not a regional political priority.

    Transit that serves higher density areas is nearly always more cost-effective than transit that serves lower density areas; transit whose riders walk to the stations is nearly always more cost-effective than transit whose riders drive to the stations. If it were not for politics, serving the urban core would nearly always make more sense than extending out to the suburbs.

    1. Christof,

      The current round of funding is laying all the track from University to the city border.

      As for actually getting to Everett and Tacoma, state law requires that a similar amount of money be spent in the Seattle city limits, so we’re going to see some effort to connect more Seattle neighborhoods as you describe. Unfortunately, that round of funding exceeds our state-authorized revenue, so it may have to wait a few decades for bonds from the current projects to be paid off.

      Furthermore, the district is constructed so that plans must have many suburban votes to pass, so to do so you really have to deliver rail service out to the suburbs.

    2. Link is not exactly going to the suburbs, it is going through the suburbs to employment centers in Tacoma and Everett, where it will actually terminate. The Sounder and Link will not be redundant, because they will serve seperate functions. Most commuters going from Tacoma to Seattle will probably use the Sounder, but commuters from the suburbs will use Link to get to Tacoma.

      Furthermore, this spine is going through the oldest, centrally-located suburbs, giving the most potential for sprawl-reducing TOD. There are plenty more less dense suburbs further East, South and North that Link is wisely not going to go to. Of course, the plan for much of North Link so far seems to be building on I-5, which would be a big mistake from this perspective.

      Of course, these outside-of-Seattle lines are still probably not as valuable as expanding within the city, but, as Martin pointed out, there are a lot of political problems with prioritizing the city lines. I don’t think anything can be done about that unless either the city or a new tax district are willing to put up a lot of funding themselves, or lots of federal funding is thrown at it.

      1. Well Northgate is right along I-5, downtown Mountlake Terrace is about 2 blocks from I-5, the downtown Lynnwood plan has 20 story buildings right around Lynnwood TC next to I-5, Alderwood is next to I-5, and many P&Rs are next to I-5 between there and Everett. Once it gets to Everett I would like to see it go underground through downtown Everett after Everett station, and then continue either on the surface MLK-style or elevated up to Everett CC.

    3. That is one of the big problems with regional planning — political pragmatism ends up driving a lot of decisions. That happens everywhere. But good transit planning, with a clear vision clearly articulated, can help. And sometimes (often) the public have to prod for good planning and articulate a vision themselves.

      1. You’re totally right about this. One of the things I’ve learned writing this blog is that the decision makers on transportation generally want to take the public’s input, and are interested in what the people have to say.

        However, our state legislature on the other hand…

    4. BRT is being put on Geary, I don’t know how exactly they are going to improve service on top of the 38 route, but anytihng is an improvement.

  10. ST has a very large district. Each of its three counties are regional in size. Together, the three counties of TriMet are one-half the size of King County alone. the governance has a lot to do with ST’s system becoming BART-like. They intended wide stop spacing and chose rare 1500 volt overhead for that purpose. ST Link LRT will a lite-Metro between South McClellan Street and Northgate; that is the key segment. It has less grade separation along MLK Jr. Way South. The East alignment is yet to be determined. ST2 will probably not reach Tacoma and Everett. Other modes already serve the intercity function. Will there be an ST3? ST2 is still a long ways off. How will transit mobility be improved within Tacoma and Everett? Such centers will grow again with global warming and higher petro prices.

    1. You’re definitely right about Tacoma and Everett. Without some additional taxing authority, I don’t think ST3 could be on the ballot maybe 2024 or 2028.

    1. Born in 1958 I expect to be free of transit concerns by 2080. Maybe I should be less pessimistic and more forward thinking. If only my grandfather had foreseen the need for carbon fiber buggy whips we could have energy independence today ;-)

  11. As a visitor to both Washington and San Francisco – one right after the other on the same trip, I can say that the DC Metro captured my heart. It is convenient for getting around town on, has cohesive and restful stations – my only beef was the sparse weekend frequencies. I think that the DC metro supports TOD lifestyles much better than most regional suburban-oriented lines which often must be paired with park’n’rides (ie MAX in Portland). BART was fine – definitely convenient for getting in from the airport on (much appreciated), but that was about it. As a visitor the San Francisco area, other than my exploratory trip to Berkley, I only used it to go to the airport. Since my visit to San Francisco was to the central city I used the articulated trolley buses to get around.

    I think ridership potential and the benefits of mass transit investment are higher in more densely urbanized areas.

Comments are closed.