Photo by Slack Action

At the start of my Transit Master Plan series, I stated that the study used the current Seattle streetcar operating cost of $220 per hour, a figure that would likely decline if the system achieved economies of scale. In fact, although that was true in an earlier draft, Nelson/Nygaard engineer Tom Brennan tells me the data I posted in the series already incorporated a newer figure of $187 — still a substantial premium over buses, but more in keeping with other North American systems with larger rail operations.

The original post has been corrected.

51 Replies to “Correction: Streetcar Costs”

  1. We should also keep in mind that since streetcars use electricity instead of diesel fuel, they will be insulated from any future rise in the cost of gas. Our electricity is also very clean and cheap. Also if the Ballard streetcar was really able to use double-size streetcars and fill them up consistently, the cost per rider will drop significantly.

    1. Our current (no pun intended) electricity is clean and cheap, but any new electricity we need dirtier and more expensive? I thought the hydroelectric potential was pretty tapped out and any new electricity needs would have to be dirtier (coal) or more expensive (wind from Central Washington). Am I missing something?

      1. The Ross Dam on the Skagit was designed to be expanded, unfortunatly that expansion would inundate some Canadian land, so we can no longer do that (Besides the greens would kill us ;)
        Is is not fair to say that the Hydro potential in the cascades is Taped out, it is probably more acurate to say that politically we can not expand that resource any further than it already is.

        Lor Scara

      2. Even with hydro, dirtier and more expensive isn’t the only option. There’s also clean and more expensive ;-). We still have wind, sun, and even the potential of geothermal. Not to mention the cleanest and cheapest fuel – efficiency.

      3. We only use fossil fuels on the hottest days in the summer when the hydro capacity is maxed out. Adding wired vehicles would increase our electricity use, but some of that may be displaced by route reorganizations, and of course the amount of power Seattle mass transit uses is a tiny fraction of the total power consumed in the state. If a thousand people charge their electric cars every night, it would require more electricity than a thousand people riding light rail. So if electric cars take off, it could increase Washington’s use of electricity dramatically, but that may not necessarily be a bad thing if they’re indirectly using natural gas rather than gasoline.

      4. Most new electricity use is offset by efficiency gains, especially if a whole new power plant is needed. Most utilities will do whatever they can to get people to use less power if it means avoiding the capital costs of a new power plant. Coal is a non-starter in Washington. People are doing whatever they can to phase out the only coal plant in the state. Any new power is going to come from natural gas (much cleaner and less CO2 than coal) or renewables, so it’s all good. Energy may get a little less expensive, but keep in mind that electric rates in Seattle are around 3.5 cents/kwh, while back in Iowa I was paying about 9 cents/kwh. It’s cheap in Washington.

      5. Don’t forget that we buy electricity from California during the peaks, and who knows whether that comes from coal.

      6. Is that completely true? Without raising the height of the dam, could we not add additional generating capacity to the existing dams? Such as when the third generating house was added to Grand Coulee in the 1970’s? Also, the existing generators at Ross Dam are probably long past their design usefulness and could probably be replaced with one’s that were far more efficient.

        Right now, Seattle City Light facilities only generate 1/2 of the total power needs of the service area. The remaining energy is purchased or exchanged with BPA and others.

        We also have tremendous potential to gain power capacity from wind. And to get a little creative, there is the idea of wind in combination with catchement. That is, use wind power to pump water into a catchement for later release and power generation.

      7. Something tells me nuclear would go over less well than coal (too bad because coal is dirtier). I also doubt much can be re-used as the plant has been abandoned for nearly 20 years.

      8. Good grief y’all have cheap electricity. Out here in upstate NY it’s 11 cents a kWH — or more.

      9. Definitely supplement at least one of your hydro plants to pump-generating, so you can store production from wind for when it is needed….

      10. Good grief y’all have cheap electricity.

        Seattle has cheap electricity (Seattle City Light, publicly owned by the City);4.61¢ per kWh (first 10-16 kWh per day depending on season). All additional kWh per day at 9.56¢. If you’re a City Light customer in Tukwilla you pay 5.27¢ per kWh and 10.50¢ per kWh respectively but Shoreline gets just about the same deal as Seattle, go figure. Over the course of a year City Light is able to wholesale power it produces to the grid and rebates that money to City residents. Most of the rest of us are served by the privately owned and non-publicly traded PSE (Puget Sound Energy) and residential rates are 8.5 cents for the first 600 kWh and 10.3 cents for every kWh over that. BTW, City Light doesn’t extend it’s best large customer rate to Metro for the ETB network; they pocket a handsome profit that takes County funds and rebates the profit to City residents.

      11. @Bernie, having participated in Seattle City Light’s public planning process I came to understand that they are spending about $250Million a year on capital maintenance items and financing a good chunk of that with bond borrowing. I and other citizens raised our concern over using bond borrowing authority for expenditures that didn’t effectively increase output or distribution capacity. The borrowing will result in approximately $100Million in interest costs in a few years. That’s not good. SCL only generates 1/2 of the total energy demand of it’s service area.

        Also, I don’t know what arrangement Metro has with SCL but it is almost certainly cheaper than buying power from PSE eh? But they should negotiate some favorable agreement. I’m curious though from where does Link get its power?

      12. The Grand Coulie complex has some pumped storage. I don’t know if there is any other pumped storage in the area. Of course the other thing that can be done is to generate less hydro power when the wind is blowing.

      1. Yes, just like electric trolleybuses. Your point? I was not commenting on mode choice, just that the operating costs of streetcars compared to regular diesel buses might not seem so much higher when gas prices go up in the future. Trolley buses are great for many applications, but clearly Ballard and Fremont have the demand to warrant higher-capacity vehicles.

  2. Martin, I must be mission something here.
    Slut and Tacoma Link are about as similar as you can get. In-street running, short distance, etc.
    Yet, Tacoma Link costs nearly double what the report says Seattle can do it for, or is currently doing it for.
    Tacoma Link ~ $326/hr excluding depreciation of about $100-$200 per hour depending on years used and assets depreciated)
    How can Seattle do it for half of what ST is doing it for? Either one is excluding too many costs, or the other is getting tagged with too many – maybe both.

    1. I think it’s because ST can subcontract the operations on the SLUT to Metro but actually runs Tacoma Link. Since it’s the only service they actually operate the overhead kills them.

      1. I sometimes wish I could see the books of various transit agencies without going through a public disclosure request because the costs just seem astronomical for some of the services that are being run. Just look at Central Link at $420/hr per car, excluding depriciation or ammortization of the dept.
        Or, why is Metro bus service so much higher than some of the peer agencies like San Diego Transit at $66/hr. (
        That report from 2009 shows San Francisco the highest in the nation at $147/hr, however there are 6 major cities all below $100/hr.
        Labor? Fuel? Admin? Maint?
        I’m sure our transit managers are privy to the reasons why, but maybe some public disclosure would set the stage for getting more bang for the buck.
        Isn’t that what good transit is all about? Moving lots of people from A to B or Z, at affordable costs.

    2. Aside from all the typos above (sorry), I did some digging in the ST 2011 budget for Tacoma Link.
      For a two train operation each day shift they have 2 operators, 1 supervisor, 1 maint supervisor, 2 technicians and 1 manager. To cover all the shifts, 7 days a week, they have 18 FTE’s. That’s to put one or two single unit streetcars out on the 1.6 mile road. Plus on top of that, they sub-contract out all the track, signal, and janitorial services.
      So the $420 an hour starts to make sense strictly from a bean counters perspective, but is that a wise way to provide transit service for 2800 riders per day?
      Should Seattle jump on the band wagon, or is this just a streetcar named desire.

      1. Central Link at $420/hr per car, excluding depriciation or ammortization of the dept.

        Ah, now you’re realizing that the cost presented is per revenue vehicle and a two car train is two revenue vehicles. Running two car trains is a big/b> loser (headways should have changed). Three car trains running a bit more than half full is break even with buses and four car trains (if you have the demand is a clear winner). East Link???

      2. Tacoma Link needs all that to keep the bums from riding around in circles. Seriously, I used to take TLink every day and its little more than a parking lot shuttle. And the security is beyond normal levels, but the thing runs through downtown Tacoma so…

      3. Sound Transit always knew that Link would not really start being cost-effective until University Link. That’s why they built 4-car platforms so that in the future they can run longer trains full of more people. It’s gonna work out, people, it’s just going to take another 5 years.

      4. But this is about Tacoma Link. Its one thing as a starter line that is being expanded, but there are no real plans in place for a TLink expansion. This short line is going to serve as a very expensive parking lot shuttle for years to come.

        That said, the line really has done a lot to improve the aesthetic of downtown. Between it, UW Tacoma and the museums, the south end of downtown isn’t so bad.

      5. I’d like to see a breakdown of the O&M costs for central link. In particular I’m interested in what the fixed costs are and what the marginal costs are. I assume some costs will scale as link gets built out and the overall cost per hour will go down. At least I hope the cost per revenue hour, per car goes down once U-Link opens.

        The current cost per service hour for Link is way out of line compared to peer agencies. To my mind it is worth asking why that is and if there is any way to bring the numbers more in line with other agencies.

        Similarly the cost for Tacoma link is crazy. There it is obvious part of the problem is a lack of economy of scale. Perhaps it is worth seeing if contracting with PT for O&M or trucking the cars to the Link O&M base for heavy maintenance could bring the costs down to the same range as the SLUT (a similarly small streetcar operation).

        With future streetcar expansion in Seattle I’m not worried so much because economy of scale starts kicking in as the system gets bigger. Presumably costs would begin to match other surface rail systems in North America. BTW another reason to build the connector line down 4th/5th is to save on operating costs, both for the motormen and on maintenance. Presumably the current Harrison St. base for SLUT could be turned into just streetcar parking and storage and all of the maintenance/supervisor/management activities for the streetcar system consolidated at the Charles St. base. Furthermore I assume some benefit is gained by the proximity of Charles St. to the Atlantic/Central/Ryerson complex. I’d also assume OCS maintenance could be consolidated with the current ETB OCS maintenance department.

      6. What would it take to put Central Link trains on Tacoma Link track? Is it just changing the voltage and transformers on the vehicle? Or are the vehicles themselves incapable of running on the same track?

      7. I wonder if ST, Pierce County, and Metro could work out a deal that would let Metro operate Tacoma Link. With a few scattered streetcars, it makes sense for one agency to operate all of them.

      8. @MikeOrr, my guess is that C-Link trains couldn’t make the turn from 25th to Pacific Avenue. Besides, it’s all single-track anyway except for the passing track at UW-T/Union Station.

      9. TLink was originally supposed to end up as part of Central Link… but this seems to have been a badly thought out idea and has been pretty much abandoned. The result is that it’s a bit of a “mode orphan”. I also wonder if there’s some way to get its maintenance etc. merged with that for the Seattle streetcar system, though this sounds difficult due to the total lack of track connection and distance apart.

      10. How’s the ridership on Tacoma Link? People always make fun of the Portland Streetcar and how it is essentially a moving sidewalk since it’s so short and slow, but it is always completely packed every time I’ve ridden it or even glanced at one going past. Clearly it’s enhancing people’s mobility or enhancing their quality of life, otherwise it wouldn’t be SRO all the time. If Tacoma Link can get that kind of ridership, then it’s probably worth it.

  3. Martin, since many readers are new to transit discussions, can you give some quick perspective on the cost of various modes of transportation, including automobiles? Also, since a balance sheet has two sides- how does the “credit” column look as well?


    Mark Dublin

    1. New York recently tried replacing some buses with a shared taxi stand (from the Upper East Side to Wall Street — a very wealthy demographic and one with no realistic competing subway service). But it’s been an abject failure — very few people are taking the cabs and the buses a few avenues over are more full than ever.

      1. My mistake. I confused the taxi stand with the jitney service that was intended to replace bus service with privatized vans and has since been canceled. You’re right; the taxi stand is doing fine.

  4. I recently took a trip to the University of California Santa Barbara and observed a wonderful demonstration of how to provide cheap and effective mobility for a ~1 1/2 sq mile area, about the combined size of downtown and south lake union. While their 30-60 minute headway bus service was nothing extraordinary, they had a wonderful system of bike-friendly streets and trails that crisscrossed the area allowing anyone to safely and effortlessly get from anywhere to anywhere on their own muscle power. Five things that really stood out about their bike network were that:
    1) It really went everywhere, not just to a handful of destinations.
    2) The bike routes were separate from not only cars, but also pedestrians, so no one got into each others way.
    3) There was lots of bike parking besides every building in a tiny fraction of the land that the equivalent amount of car parking would have cost.
    4) The bike network contained roundabout intersections when perpendicular bike paths intersected. During the day, the traffic volumes were high enough so the roundabouts were actually necessary.
    5) The bike network had direct connections to regional trails allowing people from all over the area to bike to campus safely. There is even a bike path to the Santa Barbara airport 1 1/2 miles away, which I got to try!

    In addition to the bike network, the university did a lot in the design of their campus to discourage car travel for short, local trips. While the entire campus was accessible by road, the road network was full of dead ends, making lots of trips considerably shorter in distance by bike compared to any motorized vehicle, including a private car. Furthermore, most of the car parking was in garages near the edges of campus, with the smaller lots in the middle of campus restricted to faculty and staff only during the day.

    The result was an environment in which the typical person’s choice for local trips was to walk, bike, or skateboard, with the car or the bus being relegated to longer distances or heavy loads.

    So, moving back to the topic of streetcars, one doesn’t need to be a transportation export to conclude that the capitol cost of a 10-foot wide bike trail is a tiny fraction of the cost of a streetcar line for the same distance. And once its built, the only recurring cost is fixing potholes. Mobility-wise, you can get from anywhere to anywhere, within the 1.5 sq mile university area by bike, door-to-door in under 10 minutes. With streetcars, by contrast, a 2 1/2 minute walk plus a 5 minute wait plus a 2 1/2 minute walk at the other end already adds up to 10 minutes even if the streetcar warps you from one stop to the other instantaneously. When the actual running time of the streetcar, which is no faster than a bike, anyway, is added, the streetcar looks like the real loser in mobility, in spite of its vastly higher cost. And that’s before you factor in that streetcars are only really useful for trips along the line where it’s built while most real trips involve some amount of travel in the perpendicular direction (in spite of the fact that many urbanists use wishful thinking to convince themselves otherwise).

    So, how does this apply to Seattle? Well, when we compare ourselves to UCSB, our biking infrastructure in the downtown-south-lake-union area is woefully inadaquate. There, you get from anywhere to anywhere along trails, or occasionally low-traffic, low-speed streets, usually with bike lanes. This means almost anyone can feel safe doing it and parents can feel safe allowing their kids to do it. Here, going almost anywhere useful requires some amount of riding in mixed traffic and, even when their are bike lanes, you still have to be really careful for things like car doors, right turns, and streetcar tracks. While an experienced cyclist can certainly do it, downtown Seattle fails the test of whether a typical parent would allow a 10-year-old child to ride alone there, while UCSB passes it.

    If we really had a comprehensive UCSB-style bike network, our transit network would become a whole lot simpler. We would be able to get away with far fewer routes than we have today, allowing more resources to be put into the routes that remain allowing them to run faster and more frequently, making them worth riding to more people. We would also be focusing our transit dollars on longer trips where a powered vehicle actually has the potential to save a non-trivial amount of time over a human-powered vehicle.

    1. Are you really comparing a college campus to the downtown of a major city?

    2. That doesn’t really work when you have normal city blocks, unless you want to build bike tunnels or something. I’m afraid cycling will always be pretty difficult in Seattle due to topography.

      1. You ought to try riding more before you make the “hill” claim. I ride 32 miles round trip daily and come in over Capital hill. It’s not that bad. I’m a middle aged guy like John Bailo and with low gears I just grind up the few steep sections.

        My friend who just started riding and is probably 50lbs overweight takes the bus up the hill to the top going home and rides the rest of the way. If you are clever, know how to read a map, you can find the old streets that were laid out for horse & carts and ride those hills. I’ve even been up Queen Anne but of course I didn’t ride the counter balance hill.

        Besides cities like San Fransisco which didn’t sluice off the top of their hills have a large bicycling community. It’s not the hills that keep people from riding. It’s the traffic lanes.

      2. South Lake Union is like a pancake and downtown isn’t that hilly. Political will is a much stronger impediment to biking than topography in the central city.

      3. While I don’t think you can get a college campus like bike network in a built up urban area I do think it is possible to have better infrastructure than we do currently. For example take a look at what has been done in other urban areas, particularly in Europe. Another method of making walking and cycling safer is to both cut down on the number of cars in the central city and slow the remaining ones down. Of course this won’t win you any friends among the “war on cars” crowd.

    3. If the city installed a “bicycle sharing” system they could have eliminated the trolley. Yes you would still need a bus or access van occasionally for wheelchair and disabled riders but no where near the service levels or cost of the current SLUT.

      BUT the key is to big or not at all, ie install enough bike stations so that you don’t have to bring your own bike from home on the bus to make it work.

      SLU is nearly flat. No need to worry about hills.

    4. I was in Santa Barbara last February, although I was on the opposite side of town on the beach. What impressed me was the bus service: it was pretty frequent and ran late into the evening, in spite of the city’s small size. I never rode the bus because I could walk to the center in 30 minutes, but I did see them and took a look at the schedule. Besides the regular buses there’s a tourist-centric bus going along the whole beach and making a T on the main street to the center. I think the tourist bus had a cheap fare like 50c.

      What I was not impressed with were the rail/bus connections to LA. My friend said it’s “only” a 90-minute drive from downtown LA, but both the Pacific Surfliner and the buses take three hours, and run peak hours weekdays only. I wanted to meet my friend in LA in the evening after my conference, but I’d have to leave before 5pm (skipping part of it) and there was no way back until the next morning (meaning I’d have to get a hostel room and also miss the morning talks).

      1. There aren’t that many places that serve trips that take 90 minutes by car with frequent transit service late into the night. 90 minutes is, for example, about how long it takes to drive from Seattle to Bellingham; San Francisco to Santa Cruz; Chicago to Milwaukee; Denver to Colorado Springs. For a long-distance trip like that you’d need pretty high fares and full loads to make it work. I bet LA-SB service stacks up pretty well with those other trips.

      2. I would expect at least an hourly local bus from Santa Barbara to Ventura, and from there to LA. That’s what the New York region and Europe have.

      3. Or to put it another way, Seattle-Bellingham has more transit than Santa Barbara-LA, even though southern California has four times the population.

  5. Even in NY, driving say from Midtown out into Connecticut can be faster than the trains in the evening. During the day and commute times – never.

    Still, folks use the trains for the good reason that they don’t need their cars in the city, and parking can be expensive (though I must say for a fair amount of Manhattan, and in the Boroughs, street parking remains free and accessible.

Comments are closed.