I developed a matrix for all North American urban rail systems, light rail and heavy rail, including info such as system length, stop spacing, average speed and City population. This post explores the implications from this matrix for Sound Transit’s Link light rail system.

Link light rail, both the initial segment and the Sound Transit 2 (ST2) expansion, has been designed for high capacity and with lots of grade separation. These characteristics have led some to classify Link as metro-light. As shown in the two charts below, its stops are widely spaced and the average speed is high compared to other systems. In fact, only San Francisco’s BART has stop spacing wider than Link with ST2. Based on the advertised travel times for ST2 expansions, Link with ST2 will be the fifth fastest system in North America. The Link system has the characteristics of a system designed to be competitive with automobiles around a large region. Will these characteristics lead to success in terms of high ridership?  More after the jump.

In Chart 1 the average stop spacing of all North American urban rail systems is shown, sorted from left to right. The stop spacing columns are blue for light rail, white for heavy rail, and Sound Transit is highlighted in red (for cities with multiple systems, each is shown separately). The narrow green bars are the ridership for each system, normalized by “ridership per mile.” Note that several high ridership systems are on the right side of the chart, with stop spacing around one half-mile. The correlation between stop spacing and ridership is negative and weak (r2 value of -0.16). This means that the stop spacing and ridership are only loosely connected, and if anything shorter stop spacing trends towards higher ridership.

In Chart 2 the average system speed for the same urban rail systems is shown, sorted fastest to slowest from left to right. The highest ridership systems are scattered across the spectrum from faster to slower speeds. The correlation between average speed and ridership is trivial and negative (r2 value of -0.10).

Among all the characteristics compared in the matrix, the factors most correlated with ridership per mile are City population and population density. Average speed and stop spacing have very limited correlation to ridership. Based on this analysis it appears that adding infill stops to Sound Transit Link (shorter stop spacing) would not hurt ridership, even if it results in lower average speeds, and if anything could slightly help. Shorter stop spacing would allow more people to access more destinations within the same number of system miles, and hence increase the ridership per mile.

90 Replies to “Link Light Rail in the North American Context”

  1. I’m confused by your divisions between light rail and heavy rail. The NYC subway uses electrical multiple units, and the trains aren’t heavy enough to run on FRA tracks, so why are they heavy rail? Likewise with Boston and BART… There is a clear division in Boston, they run both heavy rail (the MTA commuter trains) and light rail (the subway). But, here they’re both clasified as heavy rail… What is your definition of light rail?

    1. the definition that has existed for like ever.

      A heavy rail system is an electric railway with the capacity to handle a heavy volume of traffic. The term is often used to distinguish it from light rail systems, which usually handle a smaller volume of passengers. In North American, heavy rail can also refer to rapid transit, when referring to systems with heavier passenger loadings than light rail systems, but distinct from commuter rail and intercity rail systems. It is characterized by high-speed, passenger rail cars running in separate rights-of-way from which all other vehicular and foot traffic are excluded

      1. The classifications of light and heavy rail are based on those used in American Public Transportation Association’s quarterly ridership reports (the source of the ridership statistics). I have not included commuter rail, i.e. systems on primarly tracks shared with freight. I have also not included small streetcar systems (in mixed traffic), although APTA counts those as light rail.

    2. The major difference between the two is grade sepration and scalability. Heavy rail cannot have any grade crossings (though, Chicago El’s has one odd grade crossing). Light rail can. Heavy rail tends to have an enormious capacity.

      Common misconceptions are that heavy rail must be third-rail powered. Heavy rail, and light rail too, can be powered though any means; including OCS, diesel, and thrid rail. Commuter trains cannot be classified as either since they’re an entirely different system.

      The classification has nothing to do with the physical rail or vehicle. ST’s Link uses RE112, which is a common kind of freight rail. ST’s vehicles are heavier than a NYC or El car, and they’re twice as long.

      1. Light rail was a term coined in the US to describe the advanced tram systems that many German cities were implementing. The German systems basically used tram-like vehicles but improved their speed and reliability by building dedicated right-of-way (usually tunnels) for them in the city centers, while retaining street-running alignments in the suburbs. These systems became known as Stadtbahn (city rail) in Germany, to differentiate them from trams (Strassenbahn) and commuter rail (Stadtschnellbahn). US and British cities started emulating these systems in the early 70’s and called it light rail. Maybe if we had used the literal translation of city rail instead of light rail there’d be less confusion. If you look at a system like Frankfurt’s Stadtbahn you’ll see the similarity to what we’re building in Seattle.

      2. The term “streetcar” also comes from Strassenbahn, so I read. German immigrants applied it to the urban railways in the US, and the term caught on. Whereas other English-speaking countries call them trams.

      3. That’s true. But nowadays the term “tram” is even catching on in Germany to differentiate streetcars from Stadtbahn and Stadtschnellbahn. Too many S words!

    3. Colin,

      The MBTA (hasn’t been the MTA since 1964) runs three types of rail operations:

      1)Diesel Commuter Rail which was taken over by the “T” from the Boston and Maine, and from what was becoming Conrail at the time. These are locomotive-hauled trains that resemble Sounder. Note that even though there is electric wires over the tracks from Boston to Attleboro, MA and Providence, RI, the MBTA choses to operate their trains on these tracks with diesel locomotives.

      2)High-Platform Heavy rail which generally takes its electric power from a Third Rail. These are the Red, Orange and Blue lines. But Boston being Boston, wouldn’t you know that due to icing concerns along the Blue Lines route from the Boston airport to it’s terminus in Wonderland, the Blue Line cars also have pantographs and get their electric power from overhead wires on this stretch.

      3)Streetcars which have evolved into low-platform/low-floor light rail. A number of stations on the Green Line are underground, but the light rail vehicles which serve them are perfectly happy in an outdoor median or even doing some street running, which they still do on the E-branch from Brigham Circle to Heath Street.
      The Boston Green Line is the closest thing to LINK that Boston has.

  2. Note that a negative R^2 value does not mean the correlation is negative, it means the fit you have is worse than just fitting a horizontal line. In other words, I would ignore the discussion of R^2 values since they don’t really have any meaning in this context, only to say that there is no correlation between ridership and spacing, or ridership and speed.

    1. Actually, I’d say a negative R^2, or any other real number squared, means somebody read the wrong number off of their stats software…

  3. I’d guess that the strongest direct relationships are to density. That is, in a dense city you have both closely spaced stops and high ridership simply because of that density. If that’s correct, simply adding stops in a less-dense city wouldn’t increase ridership but only slow down the system. Of course, in the long run closely-spaced stops may encourage denser development, but the stops themselves might not lead to significantly increased ridership unless someplace has been missed by existing stops.

  4. I shouldn’t be surprized by the graphs, but it’s striking how the Canadian systems mostly have very high riderships per mile(7-25k), compared to most U.S. counterparts, with the exception of NY/Phily area. Central Link is at about 1k per mile now, with the second highest stop spacing (1.2mi) compared to the rest of the pack somewhere around 1/2 mile. It just shows how our car oriented culture struggles to layer mass transit systems over the freeway grids.
    Trading some system speed for shorter walks in most urban settings seems to pay off.
    Thanks for the brain food this morning!

    1. Well, obviously my brain needs feeding, as my correction just confused things.
      ST2 at 1.6 miles is the 2nd longest stop spacing on the graph 1. I’m going to breakfast now!

  5. Huh? I was with you until the last paragraph. You basically show data that shows stop spacing is unimportant for ridership, but then end with “Shorter stop spacing would allow more people to access more destinations within the same number of system miles, and hence increase the ridership per mile.”

    The logical end to this post would be something about increasing density around stations (or building stations in dense areas, i.e. Capitol Hill and U-District).

    1. Yeah, this confused me too. If speed and stop spacing seems to not be correlated, then I would tend to say go with the speed (due to lack of stops), it’s cheaper!

    2. Yes, agreed that the current and any new stops need to have increased density surrounding them to support ridership. City population density is a factor, as Martin points out below, that actually does correlate to ridership (r2 of 0.48).

      My logic is this: additional stops could help ridership (more places served by the train) but are unlikely to hurt ridership (since there is no correlation between ridership and speed, which could be slowed by the additional stops).

  6. Can you remake the graph with the distance between TIBS and Rainier Beach cut down? That 5 mile gap kind of skews the results on a 16 mile system.

    1. Median stop spacing is probably a better metric than average stop spacing. The outliers in any system will tend to throw the numbers a bit.

      1. Also, NYC has express trains. That means you need to throw them out as a data point: they have *both* closely-spaced stops and low speeds (on the local trains) *and* widely-spaced stops and high speeds (on the express trains).

    2. I’d be interested to see how this graph would change with station infilling on ST1. For instance, what if Convention Place was to become a Link station someday and the deferred Boeing Access Road station is built? Also putting stations at Graham/MLK Way and S. 133rd St. in Tukwila. Adding these 4 stations to the 16 mile route would bring the number of stations to 17, so the spacing would be less than a mile.

  7. I have one problem with your final hypothesis.

    You should look at ridership for transit in that entire area, not just on that line when figuring out if stop spacing affects ridership. If the spacing of stops in a line if farther away, that may not mean that more people will ride that line, but will they be more inclined to write transit, or give up their cars?

    From what I have seen, when you put a fast arterial rail system in, bus service becomes far more efficient as well since you don’t need to have as many complicated routes.

    Also, link stops less in less dense neighborhoods, and then stops more in Seattle. How much density permits how frequent stopping? That would be a good thing to figure out.

    As for adding stops, we should only do that if we keep separate center tracks in those new stations, and run both local and express trains.

  8. Link needs more stops if it wants to be anything more than a commuter train. Although, if Seattle gets serious about building a robust Streetcar network, maybe that won’t be necessary. Still, I would have liked for Link to have at least a few more stops between Capitol Hill and UW, NIMBYs be damned.

    1. Additional ULink stations would have been difficult to construct.

      After CHS, the tracks take a healthy 4.11% plunge down to get below the Mountlake Cut and then pop up to get into UWS. A station has to be on fairly level ground for ADA reasons, and the transitions take serious distance, so for a 450′ station (4-car train is 400′), you’d end up needing a thousand feet to make the grade-station-grade transition, like we see with the track profile at CHS. The only station that migth have been able to fit the grade profile would be around 20th and McGraw, but that’s in the corner shoved against a hill. I can’t find my ULink track chart so I don’t recall how deep it is around there. But deep!

      First Hill was not possible for many reasons and it would have pushed CHS north east while increasing the grade down to the Mountlake Cut.

      There really wasn’t any NIMBYism involved. It just ended up not being feasible.

  9. U-Link certainly doesn’t have the density for any extra stations but North Link definitely does. A station around 55th, an area that is nearly entirely students, would get HUGE ridership just to the Brooklyn/University stations, not to mention Cap Hill, Downtown, and the Airport. Also, I really think a short jog to Greenlake at around 75 would have been smart. You may even be able to do it with the current route if you have a pedestrian tunnel that comes out on the other side of the freeway. Combine that with a Greenlake Loop Streetcar and you would have a really awesome urban area.

    1. I’m hoping that ST may, in the future, add more stations along the route. Hasn’t this been done in other areas? The new stations don’t need all the bells and whistles of the current stations, but at least need to be able to accomodate the future 4 car trains.

      1. MAX did it. However, they were existing, concrete, above-ground platforms. All TriMet had to do was add station finishes and BAM, light rail station.

        It would be hard and expensive to add an underground station to an existing, operating rail line. Sadly, I doubt Sound Transit will do this. There hasn’t been a peep about Boeing Access Road since it was axed.

      2. Has a new station ever been added to a *bored* tunnel?

        NYC added them to tunnels every so often, but those were cut-and-cover (so, lift off the cover, widen, put the cover back on).

      3. Is the reason the new stations don’t need all the bells and whistles is because the Link operators bring the bells with them?

    2. I don’t think a 55th Station would get much ridership. It’s really close to 45th, and the area around it isn’t all that dense, and doesn’t have much potential for higher density. A jog over to Green Lake would take too much time. However, a station between Roosevelt and Northgate might be good, although I can’t think of anywhere along there that would be a good location. I think a 125th stop (Haller Lake Station I suppose) would be great.

    3. JoshMahar, at 55th a streetcar might be nice but it’s only half a mile to 45th/Brooklyn or 65th/Roosevelt. Also, the north part of the U-District isn’t “nearly entirely students.” I used to live in an apt (converted house) at 52nd and none of our neighbors were students (though 3 of us were UW staff). One of my favorite lowrise developments in the last decade, the 12-unit Park Modern, is on the Ave at 56th–none of the residents are students as far as I can tell. There are students in the area too, of course.

      1. I agree 55th is close enough to both Brooklyn and Roosevelt stations that it really doesn’t justify putting in an additional station, especially an underground one.

        A better solution would be a streetcar between East Greenlake, Roosevelt, 15th/University Way down to at least Campus Parkway. This could hook up with the North end of either the SLUT or the First Hill Streetcar.

  10. Link ridership is below expectations. ST2 will increase ridership, but overall Link is likely to continue as the nation’s best example of how NOT to build light rail. I suspect Sound Transit is rigging their studies.

    1. A number of things have conspired to result in Central link’s ridership coming in below expectations. First is many of the current transit riders along the line are rather resistant to change and it will take time for them to warm up to link. Second is metro has done a poor job of re-aligning bus service in SE Seattle to feed link. For example the 7 should continue up Boren and Broadway rather than going downtown, the 39 should go to the West Seattle Junction rather than downtown, and the 106 should go to Georgetown and White Center rather than downtown. Third the recession has reduced both transit and traffic volumes as fewer people are commuting to work.

      As for ST2 I suspect the ridership projections will prove rather low, especially the Westlake to Northgate segment (though the ridership looks strong all the way to Lynnwood). Even if first year ridership projections for U link are off as much as the have been for central link that still means U link has around 50000 riders per day and the system as a whole (assuming no growth in central link’s ridership) has 65000 riders per day or so which is pretty damn successful by any measure.

      1. But, Link is still 7,000 below projected numbers. Thats 1/3 of its projected ridership! Even though the economy is in the toilet, driving hasn’t decreased by 33%, nor has transit ridership. This kind of makes me wonder how honest that 50,000 number is for ULink.

      2. Sound Transit projected around 16,000 daily boardings for the first 6 months of service, isn’t that about what we’ve been seeing?

      3. oops, I made a mistake, I was looking at total north link boardings for the full Westlake to Northgate line, not U link by itself. So the projected 2015 boardings for U-Link alone are 47,500 which means that if ridership is off by the same factor as Central Link the Westlake to UW segment will add about 30000 boardings to the system for a total of 46000 daily boardings system-wide (assuming no additional ridership growth on Central Link and no net ridership gain if S. 200th is built before U Link opens).

        However I suspect U Link will see higher than projected ridership given the transit demand in this corridor and population growth since the SEIS was prepared.

      4. the numbers for U-Link make prefect sense because Downtown Seattle, the U-District, and Capitol hill are the densest areas in the state in that order. Currently, the 71,72,73,and 66 provide express service between the two areas all day, and still they leave people stranded many times.

        Also, one of the reasons ridership is so low is because the current bus planning makes it faster to get many places by bus than by link. That is a huge waste of bus hours and money. Link ridership would be higher if it were viewed as an arterial transit corridor (like I-5 is an arterial road corridor) rather than a fast connection between a few communities. (I-5 wasn’t simply built to provide fast travel between Walingford and Takwila, but as an artery for the whole region)

      5. Chetan, I have wondered about this for a while. Why IS Link not being seen as the backbone of the Transit System? Is it b/c it isn’t big enough yet? So this will change in the future? Or is it a result of the ST/Metro disconnect?

      6. Chris,

        You can not mean what you said about the #7; you’re normally too level-headed a poster. That is a very high usage route and forcing all those people to transfer at Mt. Baker for a two and a half mile ride would be unconscionable.

        The reason that Link is not meeting its projections is that it’s in the middle of a well-served transit area and it’s a pain to transfer back and forth. The different fare structures and frankly the bureaucratic harassment that is the rats nest of different rules mean that many people just get on the Metro bus and say “screw Link!”

        The thing should have been built and operated by Metro and if Sno County wanted to continue it on to Lynnwood it could have built it and bought a few trains. The hours operated on Sno county could be paid to Metro on some kind of cost and slight plus basis with the costs audited by the State.

      7. A better idea would be to have one transit agency running everything in the region.

        The same fare structure is also extremely necessary. (if we can’t get one agency running all transit)

      8. I’m dead serious about the 7. Not having the 7 go downtown would allow other transit markets to be served and would greatly improve it’s reliability. I’d also have the route turn on S. Henderson to serve the Rainer Beach Link Station rather than continuing South to S. Prentice.

        I’m not the first one to have thought of this. Take a look at the Rapid Trolley Network plan prepared by KC Metro.

        Frankly most places around the country remove redundant bus routes and force all riders onto rail when a new rail line opens.

        The service provided by the 7 wouldn’t go away entirely, it would still provide frequent local service with a one-seat ride to other high-density and high ridership corridors such as either 23rd or Broadway. Essentially the route would be the 9 with the frequency and local stop service, the route could go further North to follow the route of the 49 from Aloha, an alternative as seen in the Rapid Trolley Network would be to follow the route of the 48 from Mt. Baker to the U-District.

        I won’t disagree that Link service and Metro service should have been made much more seamless, at the very least no more hassle than going between ST Express buses and Metro service.

        On the other hand I disagree that the services Sound Transit provides should have been kept in the hands of the local transit agencies. For one the additional tax authority would likely never have been granted by the legislature. For another Metro would likely be siphoning off revenue intended for capitol projects or additional service to fill its budget shortfall if everything was in the same agency. For an example look no further than what has happened with the TransitNow! money.

      9. Metro actually tossed around (at the low levels) turning the 7s into 9s, but it ran in to two practical problems:

        1) There’s actually nowhere to park the large volume of buses staging to run this route on Capitol Hill.
        2) As soon as you announce “we are eliminating the 7,” you have a riot. It really doesn’t matter what you say after that.

      10. I believe the S.200th station is not far enough south, plus it’s only a parking garage, which means it will attract few riders in the reverse-commute direction. And I don’t see how the station area has development potential to guide growth. Stopping at S.200th is just not good enough. The terminus should reach at least the community college and some commercial center where development can occur.

        As for Capitol Hill and Husky Stadium Stations – they too are not good enough. Link does not handle all transit users coming to those districts. ST2 will take Link to Northgate after 12 years of construction, but Northgate too is not a major destination for a terminus. For the money, it would be better to build south and east before the U-Link tunnel. Thus, Link will continue to be uninspiring for many years.

      11. I believe the S.200th station is not far enough south, plus it’s only a parking garage, which means it will attract few riders in the reverse-commute direction. And I don’t see how the station area has development potential to guide growth. Stopping at S.200th is just not good enough. The terminus should reach at least the community college and some commercial center where development can occur.

        S. 200th isn’t the ultimate Southern terminus it is just further along in engineering and environmental review than the segment to Highline CC and Star Lake. I believe it is fairly likely that a way will be found to get to Federal Way TC by 2023 or soon thereafter. The reason S. 200th may happen fairly soon is ST would save money by speeding up construction and there may be some Federal stimulus money to do it with.

        I’m not real familiar with the area down there but it looks like it has as much development potential as any of the rest of the mess along 99/International Boulevard.

        As for Capitol Hill and Husky Stadium Stations – they too are not good enough. Link does not handle all transit users coming to those districts. ST2 will take Link to Northgate after 12 years of construction, but Northgate too is not a major destination for a terminus. For the money, it would be better to build south and east before the U-Link tunnel. Thus, Link will continue to be uninspiring for many years.

        Huh? Capitol Hill and UW stations aren’t “good enough?”. Two stations and they’ll have more riders 12 months after opening than all of either East Link or Federal Way to Downtown Seattle will in 2030. Link isn’t intended to handle all of the transit users coming to any of the areas it serves. Heck even the NYC subway doesn’t handle all of the transit users going to Lower Manhattan.

        You are quite wrong, Northgate is a major destination for a terminus. It is already a very busy transit hub, it is an urban center zoned for high density commercial/office and residential. Furthermore it is a generator of all-day two-way ridership.

        You seem to be contradicting yourself, first you complain about it taking 12 years of construction to reach Northgate, then you advocate delaying construction even further, which is it? Delay will mean it takes even longer to serve one the highest ridership corridors in the region and it will increase the cost even more.

        Besides U-Link is unlikely to be stopped now that major construction is underway and North Link is likely to proceed as fast as is possible with the available funding. Furthermore because of sub-area equity there is no way the money being used to build U Link or North Link could be spent going East or South.

        You also seem to ignore that the North corridor from downtown has the strongest ridership of any part of the ST2 Link system. Upon system completion there will be more riders crossing the King/Snohomish County line on Link every day (~50,000) than riding either the East Link or between Stadium Station and Federal Way.

      12. There are two reasons Northgate was made the terminus in earlier plans. One, the mall is a major pedestrian destination. A basic tenet of rapid transit is it should serve all downtowns, malls, and stadiums if possible. Look at Metrotown station in Vancouver, one of the highest-ridership suburban station.

        Two, Snohomish County and Shoreline expresses could meet the train there rather than going all the way downtown. That would save a huge amount of fuel and get those buses off downtown streets.

        Now that Link is going all the way to Lynnwood (hooray!), the situation is even better. Buses can meet the train at Lynnwood, 175th, 145th, or Northgate, whichever is closest. That is, if the Everett-Seattle express can be truncated, which I’m not sure but North Link might be fast enough for it.

      13. Below projected ridership levels are alarming, given that:
        1. FTA would not allow ridership extimates to include new ‘expected’ development, and the public was assured by ST and the Expert Review Panel that these are ‘very conservative estimates’. Significantly lower ridership levels suggest the model used is not calibrated very well. (ST2 uses the same model)
        2. Operating costs and depreciation are spread over fewer riders, significantly raising the cost per boarding, as compared to other modes.
        Why would you force #7 riders off a $4.00 bus ride to inflate the ridership numbers on Link where the net cost is coming in at over $14.00 per boarding. What’s in it for the #7 rider you just kicked off a nice warm bus?

      14. Mike – on the one hand, if you got a few thousand new riders on Link, you wouldn’t increase costs by $14 per rider. In fact there is plenty of capacity so you would have no net costs at all, and the cost per rider would go down dramatically.

        Anandakos – you are dead on that the different fare structures between Metro and Link create wasteful and ridiculous aversion to use of Link. All the riders know how to use Metro and what it costs. Link is confusing to them, it charges different fares, it will no longer accept their paper transfers. The incremental revenue (if there is any) is not worth the obstacles it creates for these low income riders to use Link.

        In areas where Metro and ST service overlap, the fares should be consistent and riders free to choose whichever service comes next.

        If ORCA will be forced on people who must transfer, there need to be machines to let people add value to their ORCA cards in high traffic locations that are off the rail network, especially in low-income high-transit areas where people may not have credit cards they can use to refill their cards.

      15. Carl, that was my point in #2. Spreading fixed costs over fewer riders just inflates the cost/rider for the service. Yes, it cost no more to make #7 riders get on link, nor does it save link any money either, except now Metro must share the revenue with another agency based on ORCA data for the day.
        My question was, What’s in it for the #7 rider to switch modes? At this point little or nothing.

      16. Now that the latest round of Metro fare increases have made it cost MORE THAN LINK (in all places at rush hour and some places at all hours), people living its corridor are not “resistant to change,” but rather INSANE to keep taking the horrific buses.

      17. I’ll take the bait – what the heck.

        That might be interesting – if it were true. While the cost to take the link to some places outside the downtown core (i.e. Tukwila, Seatac) is .25 cheaper than the Metro peak fare, Metro remains fare-free in the downtown area (while you get charged $1.75 to ride the Link), Senior/Disabled fares are .25 more; youth fares TWICE AS MUCH as Metro’s .75, and you can still take an off-peak bus for a lower fare than link to/from similar locations along the Link line.

        So – problem with your comment “Metro fare increases have made it cost more than link. . .in all places at Rush hour” is flat out untrue, and “some places at all hours” only partially so.

        Speaking of the word “so” – um – – SO?

    2. I believe San Jose’s VTA is still the nation’s best example of how NOT to build light rail. Seattle isn’t even in the running.

  11. Have you tried grouping the analysis you did above by population density? I think it might be more interesting to see, for cities around Seattle’s size and rail systems of approximately the same length, what the correlation is. There may *still* be none. :-)

    1. I’ll take the hit for this. Chad’s original draft had a chart comparing core city density with ridership and found a really strong correlation. I thought that was obvious and, as editor, left it out.

      But take my word for it: it was a strong correlation.

      1. While there is a strong correlation between density and transit ridership, I would also posit that the destination or employment draw is an even higher component of predicting transit ridership. I’ve written about this before but in Gary Barnes’ papers and other research it has been shown that higher transit ridership results from increases in residential density that are tied to employment centers by transit. This means that if you look at the jobs around link, it’s likely that you’ll find the comparison to other systems similar. You can have all the dense neighborhoods and stop spacing in the world, but if it doesn’t connect people with where they want to go it doesn’t really matter.

  12. Oh yeah, exciting story i wanted to add:

    Yesterday I was wandering around SLU and at the new SLU park I decided to see how far I could get on rail in a half hour. I got to the Mercer and Westlake stop at 3:16, waited about 4 minutes and then rode the SLUT to Westlake. I walked to the Pine entrance instead of using the 5th Ave elevator and literally touched the doors of a leaving Link train. Waited about 6 min, then hopped on the next Link. I was at the Mt. Baker station by 3:47!

    To me that is amazing. Once things are built out (Amazon completed, more housing around South stations) I think we will consider our rail system a pretty awesome success.

    1. Great story! It’s a fun race. Link has been open 5 months now and I’m still shocked Seattle pulled it off. My dad has lived here for 30 years and he thought he’d die before he ever saw something like the Link in Seattle.

  13. Closer stops may not affect ridership but it can lead to frustrated passengers. Seattle’s 1.2 mile separation may be wide for the US but it’s small for Russia, where stations are two miles apart (Moscow metro and St Petersburg metro). This allows the metro to extend twenty miles outside the city center while still providing decent travel time. NJ PATH is the closest equivalent, and BART is perhaps second. But both those serve only a corner of the city while the Moscow/St Pete metros cover the whole city.

    So, it can take 40 minutes or an hour to walk between stations, but infill is provided by streetcars, trolleybuses, and (in the outer parts) diesel buses. All running every five minutes, because most people don’t have a car and can’t get around any other way.

    What is needed is a fast, medium, and slow system all running in parallel. North and East Link will be medium, and South Link is a slower medium. Good enough to get to Northgate, Bellevue, and SeaTac within a 30-60 minute timeframe, and maybe Lynnwood and Redmond too, but they’ll leave you irritated if you’re going to Tacoma or Everett (assuming Link is extended there eventually). You’ll still ride it because there’s nothing faster (except maybe the ST Express buses), but it won’t make you as eager to get out of your car as an express train would.

    1. Actually North Link is “Fast”. It’s only 13 minutes between Westlake and Northgate, which is dang hard to do in a car, even at 3 AM. East Link is medium to medium-fast (depending on final alignment and if Downtown Bellevue is grade separated). I grant you that the line South through Rainier Valley is a slower medium, though the segment from Boeing access road to Federal Way will be as fast as North Link.

      1. Well, I measure things as “within a 30-minute commute” and “within a 60-minute commute”. If you can get from Westlake to Lynnwood in 30 minutes, that would be fast. I think you should also be able to go Westlake-Everett and Westlake-Tacoma in 30 minutes (substitute Intl Dist if you prefer). They’re both 30 miles away, and a car traveling 60 mph can do it, so transit should at least match it. Link will clearly not be able to do that; that’s why I call it a medium-speed service.

        Just like there’s no high-speed service between San Francisco and San Jose, which I’ve never understood either because there’s a lot of people that travel between those places. It’s 60 miles and takes an hour by car, but 1:30 on Caltrain and 1:45 on BART (1 hr to Fremont and 45 min on the Fremont-San Jose bus).

      2. I’m sure that North Link from Lynnwood will beat driving times at most times of the day. WSDOT’s 95% reliable travel time from Lynnwood to downtown Seattle is 59 minutes at 9:00am and 30 minutes at noon. Lynnwood is only twice the distance as Northgate, so I’m guessing travel time on Link would be near 30 minutes.

    2. Sounder is our “fast” transit service, and the only rail service that can have competitive travel times from Seattle to Everett or Tacoma. Hopefully capacity improvements for HSR will allow Sounder service to increase in span, frequency and speed.

      1. What are you basing this hope on (not trying to be snarky rereading my post it kinda looks like it)? What improvements are needed to get Sounder faster and more frequent, and are these the same needed for HSR?

      2. Sounder takes a clean hour to get from Everett to Seattle. That makes 25mph. I imagine it would take billions to get it down to the 20min-30min range that could reasonably be called fast.

      3. john,
        It’s 34 miles between Seattle and Everett (reference Amtrak Cascades or Empire Builder timetable). That makes it between 34 and 35 miles per hour. By comparison, Amtrak Cascades is 40 mph and the Empire Builder has the same travel time as Sounder, so it too is 34 to 35 miles per hour.

      4. Sounder to the Tacoma Dome is an hour. Bus 594 is 50 minutes, runs much more frequently, and continues to downtown Tacoma. South Sounder could probably make it in 30 minutes if the tracks didn’t make a huge detour to Auburn and Puyallup, but they were laid a hundred years ago so there’s nothing we can do about it now except wonder if Sounder (South and North) is worth it.

        Thanks to aw about the baby bullet. It didn’t exist last time I rode Caltrain.

      5. “Bus 594 is 50 minutes…”

        Ever try that on I-5 southbound on a Friday afternoon? Trust me, take Sounder. You’ll actually get to Tacoma

      6. Ha ha, what a way to spend a Friday afternoon, stuck on a bus stuck in traffic on one of the most boring freeway segments in the region. Thank goodness for the MCIs but I’ll take the Sounder with tables, leg room, free Wi-Fi, pastoral views, water fountain, toilets, and no traffic.

      7. @Oran, you write “Thank goodness for the MCIs”

        Remember: it takes 10 minutes to load me + another 10 minutes to unload me. That’s if the operator knows what they’re doing when it comes to the lift.

        I’ve hated the MCIs ever since I first laid eyes on them

      8. Doesn’t sound like the MCI coaches are terribly ADA friendly. I’ll say the Sound Transit Gilligs with reclining seats are about as comfortable as the MCI coaches without the access issues for passengers with mobility challenges.

      9. Not terribly, as most of them require a pretty clunky procedure for boarding wheelchairs from the side of the coach, rather than from the front loading doors like everyone else.

        For that reason alone – I don’t believe that they – or the double-deckers – should be run at all, but again that’s my 20 year prior career as a nonprofit social worker speaking.

      10. The double-deckers don’t have the same issues as the MCI coaches AFAIK. The double-deckers are low-floor and allow wheelchair loading through the front doors using a flip-out ramp.

        True the mobility limited can’t really access the upper deck, but the same could be said for the rear seating area of a low floor coach.

        I’ll agree with you about the MCI coaches, especially since they have a number of other problems that make them not really suitable for use as transit coaches.

        At this point the best replacement would likely be some New Flyer 60′ coaches. I’m not sure if paying extra for the hybrid option makes sense for long routes on the freeway. Similarly I’m not sure if the BRT models as ordered for SWIFT or RapidRide have any real advantages over the standard LFA models.

        Let me also say I don’t think 40′ LF coaches work well at all for ST Express service. The ones I’ve been on seem very cramped, uncomfortable and awkwardly laid out.

      11. I’ll agree with you about the MCI coaches, especially since they have a number of other problems that make them not really suitable for use as transit coaches.

        All that wasted space where luggage is supposed to go comes to mind.

      12. Community Transit received very positive customer and operator feedback about the double-deck buses so they decided to order 23 of them, to arrive in the middle of this year and replace their older 60′ coaches.

      13. Those Double Deckers are a nice ride, but ever watch what happens to trees at Westlake Park when they roll by? They whack every branch. I noticed a few weeks back that they have added a safety bar on the door side second level to hit the branches before the window does… but still has some “striking issues”…

        As to the MCI’s… those are charter/parlor style coaches built for longer runs… The are designed for long haul use and built with longevity in mind (the earlier model, the MC 9, was touted as the first million mile coach). The air ride beats even the rail ride in my book.

        Most all charter/parlor coaches are challenged by the ADA issue of getting wheeled device to the level above the bins. Lifts generally are mid coach, or aft, and are stowed in the bin area below, which means fold out time to open the lift out to the sidewalk, then use, then refold. Even the old SF transit style (built in elevator midship) has limits (widths and weight). As wheeled devices get bigger, heavier and wider, its a real challenge for lift builders.

      14. Well, Westlake Park is hardly the only place in the city where buses and trucks regularly strike tree branches hanging over the road.

    3. Yes, you’re quite right — in the long run, you need multiple systems running at different speeds and stop spacings.

      NYC has this in profusion, with local buses, limited-stop buses, local subways, express subway trains, commuter railroads, express commuter rail runs, Amtrak “locals”, and Amtrak expresses…..

  14. All of this is pointing out to me the wisdom of the Nickels approach- let the city focus on streetcars and let ST build the transit spines. Considering how many Link riders will be coming a considerable distance, you want to avoid making their trip interminable.

    The figures are muddied somewhat, for example, while the average speed of the subway is low and the stations are many, when you want to someplace far away, you get on an express. Which you would wish you were riding if you ever went to Coney Isle.

    I’m beginning to think the city should build cross-town streetcars that intersect with Link at Link stations. At the bottom line, new development is more likely to take place around new streetcars if that option is available, and new development seems to be the only way to increase density in Seattle.

Comments are closed.