Link to Run One-Car Trains

Photo by Oran

As a cost-saving measure, Sound Transit will begin running one-car trains on evenings and weekends when no event makes high ridership likely. The one-car runs will begin as early as 7:30pm on weekdays and in some cases all day on weekends.

Spokesman Geoff Patrick explained:

It is a cost-reduction measure in response to the recession’s significant impacts on agency revenues. Demand during those late-evening hours can be met with one-car trains, and the agency will save about $460,000 annually. We will operate two-car trains when there are sports events or other happenings that will increase rider demand.

The change does introduce some operational complexity in that in-service trains must be split in the O&M Facility or the Pine Street Stub Tunnel, but there is a net savings in doing things this way.


  1. Mike Skehan says

    Another way to reduce cost would be to turnback every other train, during certain times of the day, at Henderson Stn. It’s my understanding that has been the plan all along, after N. Link comes on-line. Tough times call for hard choices.

    • Mark Dublin says

      More likely: some Friday pm rush with games at all stadiums and crashes on all freeways at once, trapped drivers start spontaneously start painting white letters on their hoods and roofs in miles-long messages visible from outer space saying:”HELP! SOMEBODY GET US A TRANSIT SYSTEM!”

      Just a thought.

      Mark Dublin

  2. Shawn says

    Forgive my ignorance but how does this save money? There is still only one driver regardless of the # of cars he’s driving…

      • Adam B. Parast says

        Yeah must be. If anything it increases driver costs, because you have to shuttle the trains around a bit.

      • Charles says

        Hmmm. While Sound Transit uses “sinking funds” to create reserves for future replacement costs, the concept of depreciation in a public agency isn’t very relevant. Usually a for profit corporation uses depreciation to recapture “cash” from tax payments.

        I think the sinking funds payments are based on a fixed time schedule rather than actual usage of the vehicles. It would be ironic if they tried to slow such payments because it would only hurt them when it came time to purchase replacements.

    • Norman says

      Driver cost is a very small percentage of total operating cost for Link. This is the fallacy of the contention that running 4-car trains in the future will result in great cost savings (per boarding) for Link. It won’t.

      How much do Link operators get paid?

      • Stacy says


        You write: “Driver cost is a very small percentage of total operating cost for Link.”

        It is no smaller than everywhere else. It is just one of the costs of doing business. The agency’s overall costs are better than expected, and better than what other agencies are incurring. The 4-car trains will be needed as we’ve just about maxed out road capacity. What exactly are you whining about?

      • Norman says

        “The agency’s overall costs are better than expected, and better than what other agencies are incurring.”

        You have figures of operating cost per vehicle hour for Link, and for “other agencies”? You have a link to that data?

      • Mike Skehan says

        I posted all the peer system cost here some months ago, so just do a search for the data. At any rate, it was not less.

  3. striatic says

    they’d save a lot of electricty costs, i’d imagine, from not pulling all that dead weight from SEATAC and back.

    • Nathanael says

      Yep. That and maintenance costs are the two main savings. Fuel and maintenance.

      As long as they correctly estimate when the system will have peaks (games, for instance?) this is just sensible.

  4. Sam S. says

    Hopefully they have a good understanding when the system will receive ridership shocks. Otherwise, it seems like a reasonable idea.

    • Andy Walker says

      Actually, a ‘train’ does not have to be more than a single car/engine/cow, at least technically, in railroad terms. Whether or not one considers Link to be a railroad is another question, of course (and it wouldn’t be under the same technical terms I am thinking of. :) )

      The Bovine Pedant

      P.S. Cor! Cor!

      • Brent says

        So, if single-car trains are not really trains, can they then be at a tunnel platform at the same time as a bus?

      • kwakiudl says

        A train can have one to four cars in it’s consist. Sorry, the signaling system prevents trains and buses from sharing the platform.

      • Mike Orr says

        Since the buses are “trains” inside the tunnel, we could have one Link train at one bay, and a Metro train or ST Express train at the other bay. :)

  5. Gary says

    Seems totally reasonable. That’s why they have this sort of car configuration. Now when they finish the tunnel so that they can add a car for peak loads that will show the real flexibility.

    But yeah, who rides at 11pm except on Mariners game days?

    • lazarus says

      Exactly. They went with LRV’s of this length/configuration to give themselves operational flexibility. They were predominately looking towards the capability of running 3 car trains (temporarily limited during U-Link construction), but it also gives them the capability of running 1 car trains during low demand periods.

      ST never predicted they would need 2 car trains at off-peak ours, but the fact that they can save money by doing so is a good thing. Kudos to ST.

    • Norman says

      “But yeah, who rides at 11pm except on Mariners game days?”

      There are poeple on this blog who think Link trains should run all night, so they can go out drinkng and ride Link home at 2am, 3am, or 4am.

      • Alex says

        As someone who stays super late at my college, I am VERY Thankful the CTA Red Line runs 24 hrs. or I’d be SOL 6 miles from home many days. I project this gratefulness in my wish that link ran 24 hours because when I come back to Seattle, I’m still going to be working till 2 AM (architects work a LOT) on projects and need a way home.

      • Nathanael says

        Oh, it’s a good idea, but the thing is, two-track systems generally need a daily downtime for track cleaning and maintenance. Of at least two hours.

        The CTA in Chicago gets around this on the two-track portions of the Blue and Red by various strange and complicated methods, but it’s actually rather difficult. Single-tracking can manage it, as can longer weekend outages.

        Once you have a four-track system, or two parallel two-track systems, it starts getting really easy to run 24-hour service.

      • Nathanael says

        That said, one could probably do something like running the trains 24 hours on Friday and Saturday nights without massive disruption — the nighttime track maintenance window would still exist 5 days a week.

      • Kyle S. says

        Yes, they should. Given the number of people I have encountered out drinking in Seattle on weekend nights who are coming from the Eastside, it is obvious that we need all-night regional public transportation.

      • lazarus says

        Oh man, that would be wonderful. And it would really cut down on drunk driving — a true win-win.

        And if McGinn succeeds in extending bar closing times, then it would certainly be advisable to extend Link operating hours too (probably in one car trains so you can hose them out easier).

      • Brent says

        But couldn’t night-owl buses from Belltown also serve the same purpose, without the costs involved to keep the tunnel open?

      • Brent says

        The night-owl bus doesn’t have to run down 2nd. Just pick a street close by that doesn’t get party-locked. Schedule extra runs overnight Friday and Saturday.

      • Bernie says

        Because virtually all drunk driving can be traced to bars within walking distance of Link stations? And of course irresponsible drunks that get behind the wheel of a car are naturally going to be the demographic that gravitates to living in TOD. Yeah, I can see that.

      • Kyle S. says

        “Anna’s Ride Home is not meant to be used as a replacement for responsible decision making when you go out drinking, so please plan ahead to get home safely by using a designated driver or taking your own cab to and from your destination.”

      • Mike Orr says

        In Europe this is not even an issue. Night buses cover the whole city and suburbs, because people are out for whatever reason and need transit, just like they do in the daytime. Plus it provides an alternative to drunk driving, as others mentioned. In large cities the night buses run every day; in smaller cities they run just on Friday and Saturday till 3am. But they run every HALF HOUR, not with 1- or 2- or 3-hour gaps. There are parallel lines a mile or so apart, not one line for a 5-mile area.

        I hadn’t realized it, but their night service is like Seattle’s day service. :)

  6. groan says

    Maybe they could just run a “shadow bus” service for the off peak hours, contract it out to Grey Line or something.

  7. Norman says

    I was going to mention this on Sunday’s open thread, but I think it is related to this story.

    For the first time this year, I saw a decrease in month-to-month ridership on Link from July to August, from my weekday trips on Link. Since every time my counts showed ridership increased month-to-month this year, ST’s ridership estimates also showed month-to-month increases, I am expecting ST’s August ridership estimate to be below their estimate for July, as my counts were.

    This would indicate that my theory that Link ridership probably peaks in the summer and will decline in the fall and winter could be correct.

    The fact that Link is reducing train size in evenings and on weekends starting in October certainly indicates that Mariner games give a boost to Link ridership, and those games will end after Oct. 3 this year. It also could indicate that Link does not expect ridership to keep increasing after September, since they seem confident that single cars will have sufficient capacity in evenings and weekends starting in October. If ST was expecting significant ridership increases over the rest of this year, they probably would not be doing this.

    • Zach Shaner says

      Please, enough already with the manual counts. If you enjoy them, fine, and like you I often count when I’m on Link too…but I do it just for fun and I harbor no pretensions that my anecdotes in any way provide useful/generalizable information. It’s not possible for a single person, even if he/she rode nonstop 20/7/365, to have a statistically robust sample from which to extrapolate. Your counts (or anyone’s) don’t mean anything at all as far as trends are concerned. Even if your expected trends are ‘confirmed’, it would definitely be correlative not causative.

      • Norman says

        lol Polls use samples in the hundreds to predict elections where millions of people vote. And those polls are often pretty accurate, particularly at spotting trends. I am not claiming to be able to give a precise ridership count (and neither should ST — they always call them “estimates”). But, my trends in riderhsip from month-to-month have been the same as ST’s, so far.

        I see a lot of other Link cars on my trips, so it’s not just the counts I do in the cars I ride in. I can get a pretty good sense of how full or empty a lot of other Link trains are during my trips. It was pretty obvious just from looking at Link trains in stations, as well as those I was on, that ridership took a noticeable dip in August compared to July.

        Are you disputing this? Or just taking a shot at my “methodology”?

        And, don’t you think that the fact ST is reducing train size in October is an indication that they think ridership is not going to show any large gains after September?

      • Zach Shaner says

        I don’t dispute your experiences, not at all. I think you’re honest. It was a purely methodological ‘shot’.

    • Yorik says

      I don’t think they’re doing this because of slow growth in ridership. Even one-car trains (sorry, railcars) have a listed capacity of 200 people, and 10-minute evening headways give plenty of people-moving capacity and room for growth (2-3 times more capacity per hour than any existing bus route by my count). My guess is that the tradeoff will be comfort and bicycle/luggage capacity.

      • Bernie says

        My guess is that the tradeoff will be comfort and bicycle/luggage capacity.

        Yep, that’s what you get for double triple the cost. But starting in Oct I’m not seeing a big bicycle demand after 7:30PM. And there’s more than adequate luggage capacity with a single car. Looking at the budget numbers I have a hard time seeing this as any thing but an attempt to try and “spin” links over capacity problems off peak.

      • Brent says

        “Spin” is your word, Bernie.

        Now that ST has a full year of ridership data, it knows when one car will probably be sufficient. This is a common-sense maneuver with or without a recession.

        There is no cooking of the books here, only efficient provision of service. Do you have any reasons why the trains shouldn’t be cut down to one car (except for predicted times of crushload) on evenings and weekends?

      • Nathanael says

        It actually makes perfect sense to overprovision service a little bit when you start up, until you figure out what the service pattern is, and then cutting back like this (from 2-car to 1-car trains at low-volume times). It’s better than underprovisioning it at first, because then people redirect their trips in strange ways and distort your statistics, making planning difficult.

  8. P DeMan says

    There are a couple of people on this blog who go out of their way to make only negative comments, and I don’t mean the constructive type. Tear down, tear down is their only message. We are building light rail. thank God. Let’s make it work.
    actually it is as we can see from the increasing ridership.

    • Sean says

      Don’t worry about the nay-sayers. They lost, and they’re just bitter. It’s the guard of failed white guys – they’ll move to Arizona and pass away soon enough! In contrast, Link is here to stay!

      • groan says

        Wow, I guess the concern for government accountability is a function of gender and race in your worldview.

        Transit for chicks of color!

      • Kyle S. says

        “Accountability” is not synonymous with “stopping construction of things I don’t want built.”

      • Nathanael says

        The demographics of aggressively-hostile-to-rail are remarkably consistent. A few are female, but most are male, white, middle-class, and born in a particular time period.

    • AndrewN says

      It’s amazing how those commenters come out. Minneapolis, which has nearly 50% higher average weekday ridership while operating at similar headways for most of the day, runs single-car trains on non-event nights/weekends. Therefore, this change in ST policy doesn’t demonstrate failure of light rail in Seattle, so I would hope the trolls would stop trying to make mountains out of molehills.

      • Oran Viriyincy says

        Right. Phoenix does a similar thing as well and they have higher ridership than Seattle. Last summer I had a short layover there and went to check the Airport station. They ran a single car during midday weekday. It looked crowded but there was enough room.

      • MSPdude says

        Minneapolis originally ran one-car trains on weekends (never regularly in the evenings, to my knowledge) for a year or two after the Hiawatha Line opened, but they almost never do any more. You’ll occasionally see one, but they are unusual, never the whole line run with only one-car trains. This month we will finally see some limited 3-car train use outside special events. Unfortunately there won’t be enough LRVs to run a more robust 3-car schedule until 2012 or so.

        But I agree, this really isn’t a big deal as long as they change to 2-car trains when needed.

      • Nathanael says

        I wonder how things will change in MSP when the Central Corridor opens.

        There’s pretty much 24/7 demand for trips to the Airport and Mall of America, but downtown St. Paul kind of closes at night, so the Central Corridor may well end up with one-car trains in the evenings….

    • Shawn says

      When University Link goes online I immagine ridership could actually decline in the summer since the # of stdents drops drastically due to not taking classes. We will find out soon enough… It all evens out eventually.

      • says

        Kind of, but not really. Staff will ride too, and they’re relatively constant year-round.

        And the 70 series are still PACKED in the afternoons during the summer. The only time they’re not is during quarter breaks.

      • philw says

        UW Medical doesn’t close for the summer. UW also seems to have plenty of other things going on during the summer session.

    • Bernie says

      My bet is on an Oct. peak. But, if you’re a betting man bet against me since I’m usually wrong.

  9. Mike Orr says

    It makes sense to save money now, when Link doesn’t cover most of the highest-ridership corridors yet, and maybe have some extra money to spend later when U-Link and North Link open. Not only are these areas higher-ridership, but they have a concentration of evening activities people can ride transit to.

  10. Brett says

    This is not a big deal and is a prudent move by ST. I’m eager to see 3-4 car trains running 10 years from now. There can’t be that many light rail systems that run 4 car trains – I guess that is why people are describing Link as “light metro”.

    To certain comments that suggest labor costs aren’t that much of a factor, I would argue they are one of the things that makes light rail a good investment from an operational cost perspective. While Link might represent a large up front capital expenditure, operationally it will be the lowest cost per passenger to move people around the region. Given it has nearly dedicated ROW, it should also be reliable and fairly speedy.

    Looking at the construction cost of Link and comparing it to moving the same number of people with buses is akin to looking at the cost of building a new house versus renting for a few years – of course building a house isn’t cost effective on a short time horizon. It only makes sense if the house is going to last for a long time. By building the first all-new North-South ROW through downtown Seattle in 40+ years, this project has to be looked at from the long view.

  11. Brent says

    Given the choice between increasing headway and reducing capacity, I also agree that ST is doing the right thing.

    As more cars arrive and are available for service, the amount of splitting and reconnecting will go down, but they have to start testing their capacity to do this at some point.

    I wouldn’t be too disappointed if Link were shut down on Sunday evenings as well, and maybe earlier in the evenings the rest of the week, as long as shadow service is available. If shorter Link hours can buy 24-hour service on the route, and give Norman some 194 trips to look forward to late at night, I say: Go for it.

    One-car trains also make it easier to have the fare inspectors to hit a higher percentage of the passengers, and guard better against fare evasion.

    On the reverse side of the coin, if the crushloads overwhelm the trains after games, put some 3- or 4-car trains into service if that is cheaper than doing a short spurt of decreasing the headway. We may as well test ST’s ability to go up in train size as well.

    • Kyle S. says

      You want to shut down the train to the airport on Sunday evenings?! If anything, that’s the time during the weekend in which we should run two-car trains to make sure airport passengers have enough room for their luggage.

      • Brent says

        BTW, Does anyone have an idea why ST seems to put luggage- and wheelchair-unfriendly Greyhound-style buses on the 574 route? That one baffles me.

      • Bernie says

        Because of a cost driven desire to “standardize” the fleet they don’t have anything better? Just a guess. If they were able to contract the service to a private company that provides it’s own equipment would that help?

      • says

        Near term, I don’t think ST would contract to an agency requiring them to provide their own equipment. Back when ST Express started, they could have done that, but instead chose to invest millions in their own fleet.
        And they’re planning on building their own maintenance base, so the likelihood of them outsourcing their fleet is even less likely.
        See my post below for the real reasons.

      • says

        From the owner:

        The buses are part of a pilot project to evaluate the operation of high-capacity single-door buses in long-haul transit service. The buses are in Regional Express bus service originating from Pierce County. … In 2003, an MCI 45-foot long-haul bus was tested on Seattle Express routes between DuPont, Lakewood, Tacoma and Seattle. During the month-long test, customer surveys showed overwhelming support for the smooth, quiet and comfortable bus.

        Not sure why you call them “luggage-unfriendly” when there are four luggage bays the width of the bus on the outside plus the overhead racks on the inside.

      • aw says

        Do they actually use the exterior luggage bays on these buses? I would think that opening them at a downtown bus stop would be a fine way to clog up the street for three or four minutes.

      • says

        Who are “they”? The operators? Probably not. The passengers? Yes, I’ve seen luggage being loaded. The 574 doesn’t go downtown. The only MCIs in downtown Seattle run the 590 series and the occasional 578.

      • Z says

        The MCI’s are there for passenger capasity as the peak trips to the airport draws a lot of commuter traffic. And why there are no artics was because some facilitis in pierce county cannot handle them well, plus they proved less than optimal on the long freeway hauls.

      • Mike Orr says

        Because ST thinks those cars look more luxurious to passengers? One of ST’s marketing points for ST Express was that they were more like charter-bus coaches than your typical city bus.

    • Jason Rogers says

      ST can’t operate 3- or 4-car trains through the DSTT at the moment because U-Link construction has shortened the usable length of the Pine Street stub tunnel (which can only accommodate 3-car trains anyway). You’d have to short-turn them at Stadium, which isn’t the worst idea in the world since that is where the crowds would largely be, but that introduces some operational oddities that ST might just as soon avoid. I’m not sure if ST has the rolling stock available either to run a couple of “special” long trains for sporting events. I know they physically possess enough cars (they only need 18-20 to run service with 10 min headways), but how the maintenance requirements fit into car availability I don’t know.

      • kwakiudl says

        The stub has no construction at this time so 3 car trains could still operate to Pine St. if anyone in Sound Transit wanted it.

      • Jason Rogers says

        Thanks I didn’t know that. I seem to recall that a few months ago ST made a big deal about moving the end-of-line bumper forward in the stub tunnel (which would preclude 3-car trains) a few months ago, but I could be wrong.

      • Nathanael says

        I think they did move it, but I think they finished the part of the construction which caused them to move it.. I don’t know whether they moved the end-of-line point back or not, but they *could* now.

  12. Brent says

    Does anyone want to take bets on whether Metro will follow ST’s common sense by truncating the 101 and 150 at Rainier Beach Station on evenings and weekends in order to maintain service and headway on these routes, while significantly reducing operating costs?

    • Adam B. Parast says

      Actually if they do that it should be during peak periods, not off peak. Peak service is much more expensive, plus the high frequency of LINK makes transfers much easier.

  13. Bernie says

    Number of trips with a single car:
    60 trips per weekday,
    128 Saturday,
    112 Sunday,
    300 trip per week = 225 service hours. So about $40/hr cheaper to run a single vs a double car configuration.

    In the ST 2010 budget (pg 19 of 125), Central Link Light Rail has an O&M budget of $48,170,000. A $460,000 saving amounts to 1%. The move to single car trains is largely symbolic to reduce the appearance of waste. It looks a lot better if you’ve got 20-30 people on one car than a train with only a dozen riders in each car.

    Cost per revenue vehicle hour:
    Central Link Light Rail = $ 348.33
    ST Express Bus = $ 140.57

    Remember that a two car train is two “vehicles” so a two car train costs $700/hr to operate. While saving $40/hr is nice and all the cost for that single vehicle is running ~$660/hr. Two car trains are the smallest viable consist for Link. Page 37 reports 92,192 trips for Central Link. That’s $522.50 cost per trip. A trip is 45 minutes. That works out to $700/hr for a two car train. Working it backwards, 138,288 revenue hours operated comes out to 90 minutes per trip. Oh yeah, two vehicles… 90 minutes divided by 2 is 45 minutes per trip. Yep, $700/hr running two car trains.

    At 2.5x the cost per hour of a bus (350/140) you need 150 boardings per car per hour to break even with a bus. Not a crush load by any means but certainly standing room only for much of the route. That’s 225 boardings per trip. 2010 projection is <90 per trip. A single car will be almost 5x the cost of a bus (cost per "train" goes down $40 but cost per revenue vehicle soars). You'd have to have 280 boardings per hour or 200+ people per trip. If you had that sort of demand you'd be back to the break even point with buses running two car trains. Even if they could get Central Link cost down to Tacoma Link levels it's still 3X the cost of using a bus. So, for the same money you could provide 3X the headway with buses.

    The only way to economically scale rail service is to eliminate runs that can be handled more cost effectively with a bus. Really want to save money; no Sunday Link Service. 112 Sunday trips is 84 service hours. Shutting the system down for the day would save over $3 million a year. Even if you provide the same headways with buses (overkill) the total savings would be $2.4M, almost 5X the "savings" of leaving one car in the barn. And that's not counting the additional overhead savings of completely closing for the day. As Mike said, "Tough times call for hard choices."

    As an aside, the numbers I recall for the SLUT were about $180/hr. How can the Tacoma streetcar cost $424.47/hr? I think I know the reason but curious as to what some of the gurus have to say.

    • Mike Orr says

      “The only way to economically scale rail service is to eliminate runs that can be handled more cost effectively with a bus. Really want to save money; no Sunday Link Service.”

      We can make adjustments like this, but we need to keep reasonably short headways and fast travel times. The 550 is 15-minute daytime and 30-minute after 7pm and all day Sunday. The DSTT used to be closed at 7pm and weekends. Those made a measurable difference in terms of how nice it is to ride transit, whether it’s Chicago-like or Tacoma-like. We need to build up our transit service in designated transit corridors, not make it skimpy like the rest of the system. Otherwise people will say, “Look, Link is as undependable and infrequent as the existing buses, I guess I’ll still need to drive.” Link’s headways make a good model for the rest of the system, so if we cut it that’ll just set us back another decade.

      A full-time shadow bus is a good idea, serving the local stops as well as the Link stations. It would simplify things for riders if one bus went the whole way rather than “the 8 goes only between Mt Baker and Rainier Beach”. Then we could MAYBE reduce Link runs on Sundays in favor of the shadow bus, if the travel time is not 50 or 60 minutes.

      I’m not sure how it would work on Beacon Hill though. A bus would either have to come up Columbian Way and backtrack, or come up on Holgate and skip SODO stn. The latter would confuse riders but the former would add to travel time. But it might be good for Beacon Hill riders.

      • Charles says

        Transit should be frequent enough that you can depend on it. The Seattle area is a long ways from being thought of as dependable transit.

        In the past 2 months, I’ve been car-less in Chicago and have gotten around almost exclusively on CTA buses and trains. I’ve been very impressed with connections usually being not more than 10 minutes. Buses in many corridors are frequent – from as little as 6 minutes to a max of 20 minutes. It works because people can reasonably rely on it to get them where they need to be without long waits in the heat or cold.

        Contrast this to my current visit this week here in Seattle where I planned a trip using the 107 to connect with Link to get downtown. The 107, a nice little connector (using a single door “shorter” bus) that goes over hill and dale between the Renton Transit Center and MLK only runs approximately twice an hour has a stop very close to where I’m staying. The next closest bus is little under 3/4ths of a mile a way also only twice an hour service. While my overall trip time was less than metro calculated, the 107 was nearly 15 minutes late in arriving. We met a Link train at Rainier Beach station whose driver chose to wait for the gaggle of about 10 of us passengers scampering across the road to make the train. (thanks train driver) Otherwise, an additional 10 minute wait. Oh, btw, at 10:15 am, the 2 car train was S.R.O.

        I recognize that funding is tight right now, but the single most valuable thing we can do to get people out of their cars is to create a transit system so convenient that driving becomes thought of as expensive and more hassle than hopping on the “next” bus or train or trolley and if you just miss one, the next one is not long in coming. That is what it is going to take.

      • Alex says

        I fully Agree, I too live carless in Chicago and I never ever worry about trying to get anywhere because between the two train systems and the busses I can go anywhere in a reasonable amount of time. I in fact do not even look up what bus to take half the time because the routes are so well planned that I can assume, and be correct that if I go to a major street it has a bus running its length, and if I go to a major street going perpendicular it also has a bus running its length. Effectively making me able to get anywhere in the city on two busses Max, 1 E-W, 1 N-S. Major streets are 1/2 miles apart-ish here so its never ever a long walk to where you are going, and its never a long transfer, and it is just sooooo EASY to get around.

    • Nathanael says

      Seriously, there’s no reason not to run one-car trains in the off-peak if it saves money. Cancelling service, on the other hand, causes you to *lose passengers*.

      Even on trips where you didn’t cancel service.

      You didn’t figure that into your “savings”, did you? Closing for the day could cost well over $2.4 million in lost patronage if you’re not careful. You have to analyze both sides of the equation.

  14. EvergreenRailfan says

    This might be off-topic, but I have observed something on SOUNDER-North recently, they have varrying train lengths. The first train out of King Street Station in the Afternoon is only 2 cars long right now, plus the locomotive. The others are between 3-4 cars in length.

    • aw says

      Makes sense, if they can serve more people with a better allocation of resources. Ignoring for a moment the length of platforms, I wonder if it would make sense for some of the south trains to be eight cars, with others having six. I’m assuming they are all seven car trains now.

  15. says

    Exactly HOW will not running the trains save $460,000? I doubt this number is accurate. This doesn’t pass the smell test with me. They still will have to pay the mechanics, etc. And with such new trains, I doubt any expensive parts on the train would have to be replaced after one year of service. I wonder if the whole cost saving aspect is a red herring? With well over 70 employees at ST making over $100,000 per year in salary (a company that, btw, subcontracts out most of its work), I think of better ways of saying a half million dollars a year.

    • Mike Skehan says

      Just a guess, but PM (preventive maintenance) is done on routine schedules, based on service hours or miles, or inspection results. That accounts for most of the Mr. Goodwrench stuff, not fixing things that are broken.
      Fewer miles = fewer shop hours.

    • Bernie says

      Well, if you look at the budget there’s some $2.5M for “utilities”. I’m assuming most of that is the electricity to move around a 105,000 pound “light” rail car. That divides out to $28 per trip or $37.44 per hour. That’s darn close to the $40/hr figure you come up with dividing the $460,000 by the service hours saved. Not all the “utilities” go to powering the Link cars but I’m sure there’s more than a couple of bucks an hour in maintenance savings. FWIW, that’s actually a fairly high cost for energy. ETBs consume ~$4 and hour and Diesels eat the better part of a Hamilton. Of course the trolley buses spend a huge amount of time each hour standing still and you can carry 4x the people with a light rail car but if you’re not then yeah, pretty expensive train set to be running around empty.

      • Bernie says

        Oops, tripped up by “trips” again; forgot to divide by 2 to get “per vehicle” energy cost. Simply dividing the cost of utilities by the total vehicle revenue hours comes out to $18.73 per hour. That sounds a lot more reasonable based on energy costs for buses and trolleys. Multiply that out by the service hours saved and you only get $220,000 in energy savings. It is sort of hard to come up with that extra $20 to $25 dollars an hour in savings. Far and away the big ticket item in the budget is “PURCHASED TRANSPORTATION SERVICES”. Since there is nothing broken out for maintenance I’m assuming that falls under this line item. But, unless they are cutting back mechanics hours there’s very little savings here. If a car gets used part of the day the regular stuff like cleaning needs to get done anyway. Depreciation and amortization is a huge part of the total Central Link Budget. All I can think is they are “accounting” for some savings by jiggering these numbers but that doesn’t really do anything for cash flow does it?

      • Mike Skehan says

        I’m glad someone is paying attention to details here.
        I wasn’t going to call you on it, but your Link cost do not include depreciation and amortization, while the bus comparison costs from Metro do.
        It would more than double the Link costs per ‘whatever’, and I’m tired of debating whether those cost are relevant for comparisons. Certainly some are, vehicles wear out, tracks and electrics are replaced, etc. Not everything lasts 100 years, as some would argue in dismissing that rather large gorilla in the room.

      • Bernie says

        The bus numbers I was using were ST cost per hour but I’m pretty sure Metro costs of ~$130/hr for buses do not include depreciation and amortization either. I didn’t included depreciation or cost of capital in the O&M cost for bus or rail to try and make it apples to apples. Total transit operations (depreciation & amortization) cost is given as $876.73 per vehicle revenue hour for Link. Total cost for ST buses, $166.70 per hour. Now, cost per vehicle hour would be slashed almost in half if four car trains are the norm but that doesn’t make them any more cost effective unless that level of capacity is required.

        It’s pretty hard to make a direct comparison between bus and rail on the capital side. ST pays to maintain the rails. Metro doesn’t pay to maintain the streets but tax payers certainly do and there’s no doubt buses put a lot of wear and tear on the roads. ST has to negotiate a lease or buy ROW and station property. Bus stops are for the most part just placed on public property (the tunnel and P&R lots being the exception) and Metro doesn’t give WSDOT a dime for use of the HOV lanes. I also think the depreciation & amortization cost is overstated for rail. Yes, train cars have to be replaced eventually but I think the schedule for “writing them off” is exaggerated. For buses it’s pretty cut and dry as the fleet is sold off and replaced on a fixed schedule. Plus, the residual value of buses is next to nothing but the future value of ROW, stations, etc. will likely be much higher than the purchase price. I guess what I’m saying is that a good deal of the “cost” of light rail is truly an “investment”. How you “account” for that is difficult.

      • Nathanael says

        Depreciation is very real on rail vehicles. It’s also very hard to estimate. Some vehicles lasted 100 years in hard use with little maintenance; others collapsed much faster.

        Modern practice is to depreciate the individual components of the train separately. The component lifetimes are somewhat better understood. For some of the expensive ones — motors and wheels — wear-and-tear really is per-mile, the lifetimes are well-understood, and the savings is clear and genuine. For others (structural steel, furnishings), who knows?

      • Bernie says

        Very true. Which is why I didn’t try to add those costs into the cost of rail. As far as ST’s claimed savings though I’m hard pressed to see how such savings on depreciation come anywhere close to what they’re claiming. Surely depreciation must be less than the cost of energy to run the trains. The other aspect with depreciation is simple accounting. For example on a rental home you take the maximum depreciation schedule you can for tax purposes. Of course if you sell the rental and it’s worth more than it’s depreciated value you end up paying back the tax savings.

    • Bernie says

      “Downtown Seattle Transit Tunnel” is a line item in both the ST Express and Central Link budget; $5M charged to light rail and $4M to buses. How do they divvy this up? I wonder if it’s on a per vehicle basis and part of the “savings” is a lower charge for tunnel usage. If so then that really is an accounting slight of hand.

  16. Brendan M. says

    Other issues aside, the evening/weekend use of one-car trains might be of interest to photographers looking to get a picture of such a train.

  17. William R. says


    Excellent analysis! Thank you for all your effort digging into this important subject. You are pointing at the “devil in the details”, and the social engineers have no response. Keep on it, you’ve got your finger on the “understory”, the one the MSM never will touch. Without somebody cutting through the underbrush we’d never see the forest OR the trees.

You may want to read our comment policy.