In its Capital Committee meeting on Thursday, Sound Transit announced its intent to exercise a contract option for an additional 30 light rail vehicles (LRVs) to be purchased from Siemens. This option comes on top of the 122 vehicles Sound Transit ordered last September. The 30 new vehicles will arrive no later than 18 months after the original order has been fulfilled, and the total order will more than triple Link’s fleet to nearly 220 vehicles. Ordering now locks in a lower unit price from Siemens, with price escalation if the Board waits to order beyond May 2017.

Once Northgate is open in 2021, Link will move to all 4-car train operations. The 122 vehicles originally ordered are required to operate the Lynnwood and Overlake extensions opening in 2023, and the additional 30 are required for the Downtown Redmond and Federal Way extensions opening in 2024. The current operating plan calls for 8 minute peak headways on each line and 10 minutes off peak, with combined 4 minute peak and 5 minute off peak headways between Lynnwood and International District Station. Current riders from International District to Angle Lake will see their peak headways decline a bit, from 6 minutes to 8, but will also see capacity boosted by 25%, from 24 LRVs/peak hour to 30.

The Siemens S70 vehicles will be roomier, quieter, and have more bike storage than the current fleet. The aisles will be wider, especially in the center articulated section, improving passenger flow. Otherwise they will function much like the current Kinkisharyo vehicles, with 4-car trains and 8 cab cars running in fully interchangeable push-pull trainsets. This maximizes operational flexibility but also diminishes capacity slightly.

Sound Transit anticipates the new vehicles will begin arriving in mid-2019, with a steady but slow drip of 1-3 vehicle deliveries per month between 2019-2024. This delivery schedule means that though ST will be able to boost capacity on current operations between 2019-2021, they will also not be able to fully backfill the loss of tunnel buses within the One Center City timeframe. The vehicles take roughly 100 days for commissioning and testing, some of which can be done in revenue service, somewhat minimizing the need to do overnight testing. By purchasing off-the-shelf, the S70 LRVs are a known quantity and should be able to expedite burn-in processes. This model already successfully runs in places such as Portland, Minneapolis, San Diego, Salt Lake City, and suburban Paris.

As new vehicles arrive and are certified, the current Forest Street Operations and Maintenance Facility (OMF) will max out its capacity. Once the new East Link OMF is complete in 2020, a mix of old and new vehicles will be trucked there for storage, where ST promises to “keep the trains warm and move them around.”

Once fully operational, Link will have two nearly-maxed out OMF facilities, and additional cars will be stored on the lines overnight. Lynnwood and Redmond will be able to store two trainsets overnight, much like Angle Lake and UW can today. But in the south, Sound Transit is planning much more substantial storage. An extended tail track south of Federal Way will hold additional vehicles, and ST is also planning a pocket track between Federal Way and Star Lake that can hold 1-2 additional trainsets. Having vehicle storage at all 3 termini will allow quicker start of service each day, and minimize vehicle deadheading.

The full Board will likely approve the $132m option at its regular meeting on April 27.

61 Replies to “Sound Transit Boosts Railcar Order by 25%”

  1. This is awesome news. I’ve been on Trimet and the only praise I have for Trimet Light Rail is around their Siemens S70s. They’re spacy and look great (if used inappropriately IMHO) as per my photos:

    I’m hoping we’ll see four car trains sooner rather than later. Real easy to find a train standing room only.

    One last thing – journalism like this is why I donate $15/month here. Hope others will do the same.

  2. Beyond our obvious issues with the trains that were selected, this is a bit of a surprise.

    The 30 cars and hour in a 3 car, every 6 minute (every three minutes where they overlap) uses the same number of cars, is a bit more expensive to operate, but is MUCH better for riders, particularly riders who are transferring.

    We’ll add that to our list of things.

    1. I agree, this is a degradation of service right when we are more dependent on Link. Worth mentioning is that Madison BRT runs every six minutes. This means the replacement for the 7 (the most popular non-RapidRide bus) might also run every six minutes. In a few years folks in Rainier Valley will have two fairly fast ways to get downtown and the one that would be more frequent would be a bus.

  3. >>Once Northgate is open in 2021, Link will move to all 4-car train operations.<<

    That's good to hear. Variable-length trains is stupid.

  4. When did they change the plan from having peak service every 6 minutes on each line to every 8 minutes on each line? That sucks.

    1. At 6 minutes, trains appear to be braking lots more on MLK. The added time between stations (because of stopping at signals between stations) seems to make the 2 minute benefit awash if one is traveling through on MLK.

      1. If that’s the case they should fix the signal priority so trains don’t stop between stations. The solution is not to increase headways.

  5. The Siemens S70 that ST is ordering is not an off-the-shelf specification.

    Kinky’s are 1.5kvdc and Siemens uses the standard 650-750vdc. The Siemens cannot use a standard that Kinky claims is proprietary. So they will use a step-down transformer. My further estimation is that in actual use, the Kinkys won’t cross the lake, instead relegated to I5 service.

    The sets won’t interoperate, and marketing types are too stupid to understand the why’s and how’s. And ST’s engineering staff aren’t traditionally known for PR acumen.

    1. There is nothing proprietary about 1,500 volts. It’s what Chicago suburban lines use as well as significant main lines in Europe, where these are allowed on the main line.

      1. Pie in my face. I’m not completely wrong though. Seattle is unique. DC is going to be using 1500vdc with CAF equipment.

        The Seattle Siemens cars will be custom and not off-the-shelf. Will they be capable of 70mph? Passengers are going to think that they are weak sauce performance-wise if they do the step-down transformer method.

        Kinky likes to do things in a different way. Dallas runs variable voltage based on load. And in a way, they actually are locked into what they have. I know just enough to dig a trench in sand on the particular details. I can inquire about it on Tuesday with an acquaintance who probably still know about it (Dallas).

      2. 1,500V isn’t a Kinky decision. ST picked it because it allowed them to build fewer traction power substations and to meet the demands of running fully loaded 4-car trains at tight headways.

    2. Different power configurations are the least of their concerns. It’s highly common in Europe to have even one train that supports multiple power systems and communication systems for international use. The Netherlands use 1500V DC nationwide, while Germany use 15kV 16.7 Hz AC and France uses 25kV 50 Hz AC. A train that runs through those countries (e.g. those from Thalys) has to support all of them. It’s really mature technology at this point and it’s nothing unusual.

    3. The KI cars are going to be stored at the East OMF while the new Siemens cars are burned in on the Central Link alignment. No word yet on whether or not they will be run when East Link opens.

  6. Since the operating plan is for IDS Angle Lake at 8 minutes peak with 4 minute peak from IDS north, could ST start operating at these headways now but with three-car trains?

    MLK is a mess lately because the signal priority is no longer working right at 6 minute peaks. Trains are braking between stations much more frequently than before. I’m wondering why!

  7. It would be really cool if they allowed feedback about the interior design of these trains, such as seating config, signage, etc.

  8. But will they be getting a new livery scheme to replace the goofy-looking waves??

  9. Consider if ST moved to the 4/8 peak headway in September 2018, when the buses get kicked out of the tunnel. Interline a full loop with a turn-back loop between SODO and UW, for 4-minute peak headway on the SODO-UW portion of the line. Use 3-car trains on the full loop, and 2-car trains on the turn-back loop. My math says this can be done with 13 trains on the full loop and 5 on the short loop, for a total of 49 LRVs.

    The turn-back trains would hopefully absorb enough boardings that the 22.5 LRVs per hour on the full loop would be more than sufficient for the riders continuing further south. The key metric would be riders on the full-length trains at their fullest points: travelling from Pioneer Square to ID/C.

    That seems like as good a time as any to switch to the 8-minute headway for each line, counting the turn-back loop as its own line.

    1. 2021 will almost certainly need 6-minute trains between downtown and Northgate, even though East Link will not have opened yet.

      1. Capacity. The combined load of the 41, 64, 76, the existing U-link load, plus, maybe the 255/545 is a lot of people.

    2. Thanks for doing the math, Brent.

      As I mentioned above, trains on MLK seem to be stopping at signals between stations more often. That defeats the advantage of having 6-minute headways for longer-distance riders south of SODO and riders north of SODO would have wildly more frequent service.

      Finally, ST could begin to roll out two-line branding and signage – something that will be needed anyway once East Link opens in 2023.

      1. I’m not sure how much extra stoppage was created on MLK due to reduced peak headway, and how much already existed. Perhaps ST has some data?

      2. My experience is anecdotal. I am experiencing more delays than usual though. I’ve even been on MLK and watched the train get priority and then lose it before it arrived – because the train was stopped by an earlier signal. Someone has really messed with the light rail signal priority!

        I believe that ST reports on-time performance as within 4 or 5 minutes. Unless ST is drilling deeper in the data, additional 1-2 minutes of signal delay won’t show up.

    3. Brent: do not the ST SIPs state that the Angle Lake UW cycle time is 114 minutes or 49+49+8+8 = 114? 114/8=15. Today, 114/6=19. Three-car trains on Angle Lake would take 15×3 or 45 LRV. That would leave 17. ST needs gap trains; is that 2×2 or 4? The 13 remaining would not be enough for the SODO-UW line, so ST does not have enough LRV for what you suggest?

      1. The 114 minutes is inflated by a minute each way of forced over-dwell at Airport Station. When the buses get kicked out of the tunnel, that should save about three minutes each way. Lengthening trains will reduce dwell time. 104/4=13.

  10. Waste of money. Fixed-line transit will be obsolete before the cost of these cars is amortized. Same with ST3 light rail. Huge waste.

    Point-to-point transit service on micro BRT is the answer.

    1. Use a different name, or pick an avatar.

      I do NOT appreciate the impugning of my character in this way.

      1. Well that’s offensive in like three separate ways:

        – you think you’re the only person on Earth with ‘Bradford’ in his name
        – you think anyone would bother to troll you
        – you think my comment is a troll, when it is a sensible and forward-looking position concerning Seattle transit

        Rename this blog, or else let’s get real about the future of city transit. Fixed line transit will be obsolete for local transportation. Rail is the most expensive, least flexible, and least adaptable form of it. Why commit $60B to a fixed line system just as we are entering a period of innovation in automation, data-driven dynamic routing, and possibly even low cost flight? This is the time to keep our options open, and BRT is the best compromise. Technology will reduce the cost of operations significantly in the next 20 years. P2P micro bus transit will effectively triple road capacity but it cannot run on rail lines. Rail investments will be nearly worthless – nobody will take the time to walk to/from a train stop when a micro bus will arrive at your door in five minutes.

        Anyway BRIAN, I will use a different handle hereafter so that others don’t mistake your comments for mine.


      2. I said the troll comment (below), not Brian.

        ” Fixed line transit will be obsolete for local transportation. Rail is the most expensive, least flexible, and least adaptable form of it. Why commit $60B to a fixed line system just as we are entering a period of innovation in automation, data-driven dynamic routing, and possibly even low cost flight? This is the time to keep our options open, and BRT is the best compromise. Technology will reduce the cost of operations significantly in the next 20 years. P2P micro bus transit will effectively triple road capacity but it cannot run on rail lines. Rail investments will be nearly worthless – nobody will take the time to walk to/from a train stop when a micro bus will arrive at your door in five minutes.”

        Now we’re getting somewhere. If you had said this in the first place it wouldn’t have sounded as trollistic, because you give reasons for your assertions. However, I still disagree with them. You’re betting on technologies which are still in research, testing, or used by only a few people. Whether it will scale to millions of people in a city remains to be seen. Keeping our options open means preparing a conventional fallback in case we need it. Putting all our eggs in a technology which may be in production in twenty years gives a good chance it won’t work and we’ll fall flat on our face. If as you say P2P microbuses will be much cheaper, they can replace some of our expensive coverage routes and extend service to blocks that don’t have it in lower-density areas. But we’ll still need trains for the bulk of movement in the major corridors because we’ve run out of freeway space. Technology can also improve in trains. Our activity centers will not move because the cities/counties are concentrating jobs/housing there, so the rail corridors will not become obsolete anytime soon.

    2. That does not scale in cities. We need a fixed high-capacity transit backbone to carry the bulk of people going in the major directions simultaneously, and some combination of fixed-route bus and flexible-route service for last-mile connections and areas the trunk doesn’t reach. See for extensive evidence.

      Brian Bradford: there are others who have the same one-word, two-word names. You can’t expect no other Bradford to sign up and not use his name, or to know that you exist. But you can say ‘Brian Bradford (not “Bradford” the troll)’ when it’s necessary to distinguish yourself from him. Some others do something similar.

    3. Any sort of high capacity transit without its own ROW is pretty useless. If you want to avoid fixed routings, how do you intend to keep your micro buses out of traffic congestion? Separated bus lanes on every street? LOL, good luck with getting cagers to give up their precious SOV lanes.

      1. You really only need BRT lanes on spines to make all the difference. That’s achievable and many are already in place and underutilized. Dynamic routing is the only way to make a dent in converting car drivers to transit. ST mostly just moves transit riders around from one mode to another which is zero sum, because all the transit modes are similarly flawed from a drivers perspective.

        Transit ridership is down across the country except for a handful of cities. Seattle is up, but thats more driven by a rise in population than a change in behavior. Light rail is a regressive form of transportation – fewer stops and less flexibility, and incredibly expensive. It’s the worst option on the table.

      2. >You really only need BRT lanes on spines to make all the difference. That’s achievable and many are already in place and underutilized.
        HOV/HOT/BAT lanes are *not* BRT. Please look up “BRT Standard.” I don’t believe there is a single true “BRT lane” in Seattle. We can’t even get 100% dedicated lanes for RR/RR+ projects, so it is most definitely not “achievable.” Imagine taking one lane in each direction on every major thoroughfare from Renton to Northgate away from cars. There would be riots.

        >all the transit modes are similarly flawed from a drivers perspective.
        Okay, then make it harder to drive. We subsidize cagers enough as it is.

        >Light rail is a regressive form of transportation – fewer stops and less flexibility, and incredibly expensive. It’s the worst option on the table.
        It’s also got throughput that you will never be able to match with bus service unless you build out a Curitiba-style BRT system, at which point most of the cost savings of BRT vs rail evaporate. BRT is only cheap if you do a bad job of it.

      3. “Dynamic routing” is Silicon Valley speak for “I live in the suburbs and can’t be bothered to walk or bike to a transit stop.” SDOT’s goal is “72% of Seattle households are within walking distance of 10-minute transit service by 2025.” If you have that sort of high frequency coverage, dynamic routing is completely unnecessary.

        Further, most cities in the region are trying to drive development adjacent to existing or planned HCT stations, specifically to encourage easy access to transit and to avoid expensive transportation solutions like dynamic routing.

        Dynamic routing may make sense if you’re trying to figure out, for example, how to get people in Mercer Island or Sammamish to and from the closest Link station. Driverless vans may very well be a game-changer for those low density neighborhoods But it’s simply not relevant for most of Seattle or other urban areas in the region.

        Also, Bradford-not-Brian is right that the tail of bus lines don’t need bus lanes, but that’s relevant for a “local” bus route, not HCT. For a BRT line to be actual HCT, the goal is exclusive lanes for the vast majority of the route. Different requirements for different transit modes.

      4. Your thinking is too static, Pat. Your only game plan is to punish people until they accept a worse alternative to driving. It makes no sense, unless you are a sadist.

        Dynamic routing for micro buses creates a situation that is better than driving. No need to walk in the rain, stand on a crowded bus or train, and add 20 minutes to your door to door commute time. When it’s implemented, families may not buy that second car. They may not drive to work anymore.

        And this will happen, and ST ridership will plummet. Fantasies of bus lines that feed static rail lines will evaporate. Either ST needs to embrace this, or die from it.

      5. No, in a lot of cases, dynamic routing is worse than driving or fixed routing. I work at Microsoft, and for several months, I used the dynamically-routed shuttles a lot to travel between buildings on campus. It was horrible. I needed to schedule a pickup, but once I did schedule it, the shuttle could be coming any time in the next ten or fifteen minutes. Then, once I got on board, it could detour a quarter of the way across campus – in heavy traffic – to drop off or pick up a couple other people, before finally taking me to my destination (with a couple other people still aboard waiting to go somewhere else.) Thanks to all these delays, if I had a meeting I really needed to get to, I’d need to call a shuttle at least twenty or thirty minutes in advance even though a direct drive might take only ten minutes. I frequently wished for fixed-route shuttles, even if they wouldn’t take me door-to-door.

        Things won’t be quite as bad in a situation where just about everyone’s going to the same train station, but there’ll still be a lot of detours to pick people up. Times will be unreliable, people won’t be able to plan ahead, and they won’t want to ride it.

  11. It is our comment policy to not allow use of names already used by regular commenters. In this case, it’s likely accidental and not quite a duplicate. Also a common name.

    But still confusing. “Bradford” didn’t leave a valid email. If he intends to comment again, a more unique handle would be wise.

  12. Something I’ve always wondered about: why does it need to take 100 days to “test” these new train cars? Train cars aren’t exactly a new technology and the tracks they’re running on have been in use for years. What could go wrong in day 99 of testing that is at all likely to go unnoticed in the first week or two?

    1. Products tend to have a higher failure rate at the beginning of their life cycle, which can’t always be caught in a manufacturer’s testing. I assume the idea here is to identify and repair any problematic cars before they’re run in revenue service with passengers on board.

      Bathtub Curve

      1. Yes, I’m aware of the bathtub curve. I’m just kind of surprised that train cars are considered to require such a long period of regular use to get past the initial peak of the curve.

        When people buy a new automobile they start relying on it right away. Maybe it does break down a bit more often at first, but not often enough for the average person to feel it’s worthwhile to only use it for nonessential trips for 100 days to break it in before trusting it. I’ve never heard of airlines running new planes empty for such a long time before trusting them either.

        What makes a transit vehicle so different?

      2. Great gobs of specification items, each of which means that few transit agencies get an “off the shelf” car. Instead, they each wind up getting tweaked to fit the needs of each individual agency.

        This means different control software and hardware details.

        For automobiles, they are typically making tens or hundreds of thousands of them in a batch. Testing a few dozen pre-production autos isn’t a big deal, and some of those tests are destructive. Can’t afford to do that if you are only making 30 of something.

      3. It’s usually more than about the cars. BART Warm Springs finally opened a few weeks ago — after a long delay because of the communication system, for example. Any time there is new track, a number of problems can arise besides the cars.

        I would even speculate that the cars are usually the least of the worries because the cars can be broken in in other ways.

    2. Well, for one, new software issues that don’t always close doors when they should be closed.

    3. I assumed it was regulations, the way ST has to run trains on new extensions for a month or longer without passengers, which seems way longer than necessary. There’s over 200 runs in the first two days alone. How many runs does it take to discover a problem?

    4. By the way, does a “run” mean a one-way trip, or all the trips in a driver shift? I’m assuming it means a one-way trip.

  13. I like the idea to use standard train sets. But if we use 4 car setup, isn’t it a waste to have 4 heads? Portland already eliminated one of two with a few seats, what about if we could turn that area into efficient bike storage?

  14. Pingback: ST3 Kicks Off

Comments are closed.