Post-CRC Transit: Time for (Real) Commuter Rail?


Now that the Columbia River Crossing is (probably) dead, is Clark County commuter rail a viable alternative?
Now that the Columbia River Crossing is (probably) dead, is Clark County commuter rail a viable alternative?

If the Columbia River Crossing mega-highway and light rail project is actually dead, a question arises: what are we to do to address mobility across the river, especially when it comes to high(er) capacity public transportation?

Without the CRC, opportunity to improve transit to Clark County, Washington remains limited at best. The likelihood of a light-rail only bridge is slim, and busses will continue to be stuck in I-5 traffic. But there is also the cultural divide, for it seems that north of the river, TriMet in general and light rail in specific is often perceived as something suspect. MAX is a crime train. MAX is the long arm of Portland’s urban culture. MAX is government interventionism and social engineering. And whatnot. (For an excellent and balanced take on this, see this article from last June in the Columbian.)

So is there a form of improved public transit that is achievable with lower cost and less exotic engineering, is far less expensive than the CRC project (and therefore more affordable), and also that is socially acceptable to Clark County? The answer may be yes: commuter rail.

Now hang with me. I don’t mean anything like the growing but heavily flawed TriMet WES line. As commuter rail, WES was built to mimic light rail as much as possible, and as a result sacrificed pretty much everything that typifies commuter rail.

To understand true commuter rail, look north at the Sound Transit’s Seattle-centric Sounder. It exhibits typical commuter rail characteristics not found in WES, but found in dozens of highly successful commuter rail systems across the nation:

  • (Mostly) unidirectional service. Much as TriMet’s express busses do today – mornings towards downtown Seattle, evenings away from it.
  • Lengthy. The southern line extends to Lakewood, south of Tacoma, approximately 45 miles. The north line to Everett is about 35 miles.
  • Very high capacity. These are locomotive hauled trains with multiple cars. Each car is two-level and can carry about 160 seated passengers, and almost double this counting standees. Sounder often runs six to nine car trains. A six-car commuter train thus can handle about 1000 people seated, and 2000 standing. (That’s more than ride all of WES in a day with far lower operating employee costs.)
  • Wide stop spacing. Sounder stations are typically 5-10 miles apart, so that trains can actually reach their 60 miles-per-hour maximum speed and hold it for longer periods.

But using Sounder as a possible model for a Portland to Clark County high(er) capacity transit, it is not only the particulars of the mode that are important. Rather, two other characteristics are vital: achievability, and cultural acceptance. That last one is the most important one, but I want to dispense with some of the numbers first.
On the achievability front, Sounder operates over the freight rail tracks of the BNSF Railway, alongside Amtrak. Sound Transit made significant capital investments in BNSF’s rail infrastructure in order to support the operation, and jointly worked with Amtrak on improvements that would benefit both Sounder and Amtrak Cascades services. Building off of existing infrastructure and partnering with other passenger rail projects resulted in shared gains and a more efficient use of infrastructure.

I should note, Sounder did weigh in at a pretty high construction cost. According to Sound Transit numbers obtained via email from ST spokesperson Kimberly Reason, the original segment from Seattle to Tacoma, inclusive of track and signal upgrades, stations, and equipment, cost $554.8 million. That’s just shy of $14 million per mile in capital and startup costs. Compared to the Portland-Milwaukee Light Rail project’s approximately $205 million per mile, however, Sounder was a bargain.

(Caveat: the numbers Sound Transit provided are about ten years old, and are not expressed in 2013 dollars.)

As with Sounder, the tracks that might host a Clark (and Cowlitz) County commuter rail project are owned by the BNSF Railway. Similarly, the northern line to Longview also hosts Amtrak Cascades. This means that a Clark (and Cowlitz) County commuter rail operation would likely have a host railroad experienced with and friendly towards commuter operations [PDF], and a potential infrastructure partner in Washington DOT / Amtrak Cascades.

Commuter rail, with controlled access, cute stations, and car parking is a mode that appeals far more to suburban voters.
Commuter rail, with controlled access, cute stations, and car parking is a mode that appeals far more to suburban voters.

But enough about numbers and statistics. The bigger question is, is commuter rail the type of high(er) capacity transit that Clark County would be willing to accept? Consider:

  • Unlike light rail, access on commuter rail is controlled. This reduces the perception of criminal elements moving about. Whether that is a legitimate concern is beside the point – in fact I think it’s a bogus one (see also here [PDF]), as is beginning to get out – but politically it is extremely important, especially in Clark County.
  • Commuter rail has a “Better Homes and Gardens” personality that makes it appeal to suburban residents. Locomotives and cars are seen as cute or quaint or friendly. Stations are often small-town depot buildings complete with rose beds and cast-iron street lamps. There is no smack of urban uncertainty to a commuter rail station.
  • Commuter rail does not challenge auto-oriented suburban culture. A significant number of riders drive to a park-and-ride facility and then board the train. This means riders can keep their car, continue to live in low density suburbs, and not feel that transit is challenging their lifestyle and self-image.

For those of us who are dedicated transit riders and advocates, the above points may seem ludicrous, frustrating, or like caving in to what Gordon Price calls “Motordom”. But ask yourself this: is it better to have thousands upon thousands of cars parked at park-and-rides in Clark County, or parked in the city center? Which would be better for the air quality, for commerce, for general quality of life on our side of the river?

Commuter rail – the real thing, not WES – may be the one higher capacity transit mode that would satisfy Clark County voters while also being achievable, buildable, and functional. As we contemplate what to do now that the Columbia River Crossing – all twelve lanes and two light-rail tracks of it – is (probably) consigned to the scrap heap, commuter rail to Clark County may be an idea whose time has come.


38 responses to “Post-CRC Transit: Time for (Real) Commuter Rail?”

  1. I’ve thought this for years. Why would anyone prefer to take light rail from Vancouver? It takes 23 minutes to Union Station from the Expo Center. You can figure at least 30 from downtown Vancouver. Amtrak does it in 15 minutes.

    But for selfish reasons, I’d add a stop by the Fred Meyer in St Johns. Service to NoPo is crap and a stop there would cut commute times from 40 minutes to 10. It would add 3 minutes to the Vancouver train, but it’s well worth it.

    I know there are people that take Amtrak from Oregon City, even. The main downside is the once a day in each direction.

    But count me as a fan.

  2. Either existing potential route in Clark County would share North Sounder’s fatal flaw: both run along the edges of the populated area they serve, rather than through the middle of it. Even in downtown Vancouver the BNSF station is quite a distance from the activity center. The Camas-Washougal route would require double tracking east of about Andresen.

    There is a potential for commuter rail to Battle Ground via the county-owned line, but it would require a half-billion dollars in track upgrades to achieve.

    Then there are the high labor costs of commuter rail. It’s what makes WES so expensive to operate per ride, not the vehicles themselves, though they are rather small. Since any commuter rail to Clark County, even a Battle Ground service, would require shared use of BNSF tracks to cross the river and access central Portland or the West Side, it would need the usual crew of at least four people required for passenger trains: two in the cab and two or more in the cars.

    These problems are not reason to reject the idea out of hand, but this would be very expensive to operate and serve relatively few people easily, at least, without the Battle Ground option.

    • Anandakos, can you cite your source for your crew comments? Amtrak routinely operates passenger trains with only one engineer in the cab. I suspect these trains could be operated with only two employees – an engineer and conductor, with fare inspectors used to lift tickets.

      • On their own trackage or shared? They may have some waiver on their own trackage or where there’s some form of automatic train stop, but since the METRA train ran the red signal west of Sylmar I though they had to have two people in the cab on shared trackage.

        What does Sounder so?

      • OK. As I said below to a different poster, if Sounder can fun with a lone engineer, so could this system. I would bet that both the NorthStar and Sounder run over some sort of Automatic Train Stop.

        However, lowering the crew costs does nothing to help the absolute uselessness of a line through Ridgefield, Woodland, Kalama and Kelso.

        And even though Camas is a growing city, it has one and only one bus per day with a commute orientation. Everyone who wants to use transit to commute drives to Fisher’s Landing and takes the bus from there. Such a commuter train wouldn’t even serve the Park ‘N’ Ride, just downtown Camas, which would be a double back for folks from a large part of the settled uplands.

        The only place commuter rail might be useful in Clark County is the county owned line to Battle Ground. And that would be a political cat fight because of the cost and NIMBY freak outs.

        • The Battle Ground route connects to the Cascades route facing north, making it… less than quick. But just for kicks, here’s my proposed list of stations:
          Portland Union Station
          Lombard Street (North Portland)
          Vancouver Amtrak
          78th St/ Vancouver Lake
          78th St/ St Johns Rd
          119th St
          Battle Ground
          Moulton Falls Park
          Yacoult

          Frankly, it doesn’t seem plausible. Commuter rail on the *Oregon* side, to Oregon City, would make more sense.

          Running commuter rail down the north shore of the Columbia River where Amtrak-from-Spokane runs doesn’t work either. There are reasonable station locations and reasonable speeds — but because of the backtracking involved in going west to the BNSF rail bridge, crossing into Oregon, and then going east again to get to Portland Union Station, I-205 will be consistently faster.

          Running commuter rail up the Cascades line is a possibility, I suppose; stops could be
          Portland Union
          Lombard St
          Vancouver
          Ridgefield
          Woodland
          Kalama
          Kelso/Longview

          But thinking about it — a simple shuttle between Vancouver Station and Portland Union Station might be the most effective commuter rail you could build.

  3. Two other important attributes of effective commuter rail systems:

    * Strong anchors on both ends. When one anchor is a big-city downtown, that criteria is met, but good commuter lines often connect two big cities, or one big city and several significant suburbs. Vancouver itself qualifies, though I’m not sure anything else in Clark County does. Longview/Kelso don’t function much as bedroom communities for Portland, and nothing else between there and the ‘Couv would drive enough demand.
    * Appropriate fares. Despite all the other issues with WES, it’s too cheap to ride. Most commuter rail systems have distance-based fares, and are a separate ticket from local transit. Riding Sounder from Lakewood to Seattle costs $5.25, and you still have to pay extra if you want to ride King County Metro or Pierce Transit on either end. Kent to Seattle, a journey of similar distance to WES, is $3.50. A ride on WES is only $2.50 (less if you have a pass) and the price includes a transfer to the rest of the TriMet system.

  4. Another complication is that the “Portland Triangle” is a notorious bottleneck in the rail system. Both BNSF and UP conduct overlapping operations, long distance tracks go north, south, and west (up the gorge), plus you have the assorted spurs associated with the Port of Portland. It’s a huge knot, and it’s projected to take several hundred million dollars to untie it. I’m not sure any of the incumbent parties would welcome significant further operations in the area without a lot of progress on sorting out the mess.

  5. I wouldn’t say that commuter lines need strong anchors on both ends, looking at rail maps of Chicago and Boston a lot of those lines go nowhere in particular. Rail lines have to serve lots of housing to make sense, but as Alexander notes that housing might well be low density sprawl.

    I think whether such sprawl in Clark County is desirable depends on how willing Portland is to allow more housing close in. I’d rather people working downtown take a train than drive, but I’d even more prefer they be able to live where they can walk, bike or take Trimet.

  6. Is the potential ridership there for a commuter line? How many people commute from points in Clark County and Cowlitz County along the BNSF railway to Union Station?

    And what’s the potential construction cost, ridership and operating subsidy for a commuter train vs. a light rail bridge to downtown Vancouver?

    • All of the past studies I’ve read on this topic indicate the ridership is not there to support the cost of the required infrastructure upgrades. The question in my mind is whether the ridership might be there to support operations if improvements are made through another project, such as the Cascadia corridor passenger rail improvements for which planning is underway. Commuter rail in Clark county may be feasible if they can spin up operations by buying a trainset or two and publishing a schedule. Not sure if intermediate term ridership would justify much more of a capital expenditure.

  7. Anandakos, I’d like to point out Minneapolis’ Northstar and Seattle’s Sounder commuter rail operations (both operated by BNSF crews) are operated with a TWO person crew. ONE engineer and ONE conductor. Older “legacy” commuter rail operations do operate with 3-5 person crews as stipulated in their older crew consist agreements. But the trend in newer operations is with a reduced crew to curb the manpower portion of operating costs.

      • The citizenry in general has fallen for the neo-con garbage that has been spewed for so long. That is “people cost to much”.

        That is such a bogus red herring yet so many people have had their minds distorted by years of neo-con propaganda.

        The only other ‘government’ function that is being held up to this neo-con standard is the postal service, which is actually running in the black.

        Every penny that the neo-cons take from the citizenry ends up in the pockets of the power elite.

        Do people actually think that by removing my health benefits the rest of the citizens will benefit?

        The brainwashing is so total that most ‘citizens’ have lost long ago any sense of the reality that is our government today.

  8. Everyone likes to use Portland to Milwaukie MAX as how expensive light rail is.

    It has an entirely new bridge over a major river in a downtown corridor! Yeah. That’s expensive!

    It also will be usable by busses, streetcar, bicycles and pedestrians, and emergency vehicles – not only MAX.

    How about we compare light rail lines costs where there isn’t a major multi-use bridge included?

  9. I feel like I keep repeating myself I guess that’s cause I keep repeating myself.

    This is not complicated folks, high capacity transit can be available tomorrow.

    All that’s needed is to buy buses,lots of them, high capacity buses, that have nice interiors with nice comfortable seats and simple amenities like high speed internet, run them frequently (as in every 10 minutes all day long) and people will use it.

    No turmoil or construction needed. One 60 ft bus can haul 200 people comfortably that’s 200 cars off I5. If you had a whole fleet of those you can remove thousands of cars from the existing roadway.

    Planners like to plan and providing reliable and frequent service in big buses doesn’t employ any planners and makes no money for construction outfits nor does it make profit for developers.

    Why are people attached to the complicated. The solution lies in the obvious.

    Commuter rail goals are easily achievable today, with buses.

    • “Commuter rail goals are easily achievable today, with buses.”

      Only if the Feds, Washington State DOT and ODOT all agree to convert a lane of I-5 each way to buses only (or buses and 3+ person carpools only, a least), for the equivalent length of route. Adding tolls would also work. Currently, this isn’t possible, without a new vote by congress and the state legislatures.

      We would also need to build new stations and bus stops. It would be cheaper than a whole new commuter rail system, but not negligible.

      If capacity is an issue, 60 foot buses can’t really hold 200 people comfortable; most have 60 seats. Even 80 foot buses only have 80 seats, and these would be hard to fit into the Transit Mall. We could buy double-tall 40 foot buses (like Community Transit north of Seattle) which seat 77 people and take up a similar curb space. And 85 foot commuter rail car, like the M7s used by Metro North, seats 100 to 110 people.

      So, It would take almost 14 double-tall buses to equal the seats in a 10-car commuter train. Does Vancouver need 1000 person trains? I’m not sure. But a 1000-person train with a crew of 3 is going to have lower operating costs per passenger than 14 buses with 14 drivers. That’s an ongoing cost.

      BRT would still be cheaper, especially if the demand is not there for, say, 2000 people per hour to commute downtown at the peak (2 trains vs 28 buses). But it has real capital and operating costs and a large political issue that needs to be solved, if buses are not going to be stuck in traffic.

      • Only if the Feds, Washington State DOT and ODOT all agree to convert a lane of I-5 each way to buses only (or buses and 3+ person carpools only, a least), for the equivalent length of route

        ~~~>I don’t buy that. By your figures only 60 seats=120 people= 120 private cars. 6 of these buses would travel in one hour which would be 720 cars. Multiply that over the day and you will have a more open freeway.

        Furthermore you are making the assumption that it has to be fast for people to use it. I disagree. If riders are provided with RELIABLE FREQUENT COMFORTABLE service with minimal amenities (internet) they will happily wait out the congestion or traffic jams.

        And those operating costs you keep spouting, you have allowed the neo-cons to invade your brain.

        This government waste billions right in front of our noses so don’t talk to me about costs of operating transit.

        • “This government waste billions right in front of our noses so don’t talk to me about costs of operating transit.”

          Now who is spouting Neocon talking points? :-)

          I would, personally, support raising the gas tax, adding tolls to freeways, and increasing milage-based fees and parking fees to pay for more transit. I would also support higher property taxes on areas near transit. I would support reducing money spent on new highways.

          But right now, Trimet and C-Tran don’t have enough money to run their existing bus routes at a reasonable frequency, and they don’t have the ability to raise taxes, or take money from highways.

          If you want BRT or Commuter trains or more frequent buses in the city, it will take political will in the State capitols to change funding, or at least locally to pay for more service.

          So, let’s assume we stabilize finances at Trimet and C-Tran thru new legislation, reform the administration and get a new contract with the union. Now we can add 1000 person-hours a day of bus drivers, mechanics, conductors, fare-checkers, etc (~$250 million a year), and we have $5 billion to spend on capital projects over the next 20 years (~250 million, half from the feds). We both agree we should run buses more frequently in the city, and buy a couple hundred larger buses (double-decker or 60 foot articulated).

          Now we have a few hundred hours a day for operations, and a few billion on capital costs to spend. If we build lots of BRT, we can get good value out of the capital dollars – but only if ODOT lets us take away 2 lanes from cars on I-5, I-84, Powell, Barbur, etc. And if our new BRT project is really successful, we might need to run a bus every 5 minutes all day; that’s 200 driver hours a day, and then there are bus maintenance costs and gas and right-of-way maintenance.

          With rail, we probably would only be able to build 1/2 as many miles of new service. But if the route is really popular, as in the example above, 2 trains an hour would take care of the demand, and need 1/2 as many people to operate. So we use up all our capital dollars, but have more operating costs left over to spend on extra bus service.

          I’m trying to say that operating costs matter, even if we get more money, and it makes sense to consider if we might need high capacity on certain routes. I don’t know that Vancouver-Portland will have enough demand in the future to need high capacity. But it might, if habits about driving continue to change, gas prices go up, and the downtown core continues to grow. In that case, we might need high capacity transit, with lower operating costs per ride, if we are going to afford to continue to provide subsided transit service to all the people who may need it in the future.

          • In theory- 5-minute-headway buses would be a more desirable product than big trains that only came every 30 minutes. That way, scheduling around the bus would be totally supurfluous. Bigger vehicles means less frequency which means more planning required

        • Al,

          No, you won’t have a more open freeway. If 2,000 people switched to the bus, 1,500-1,800 new houses would sprout in the I-5 corridor with owners who would drive.

          People will drive if there is capacity for them on the road. The only way not to keep building roads until the central cities are coked is to stop building roads.

        • “Furthermore you are making the assumption that it has to be fast for people to use it. I disagree.”

          It has to be RELIABLE for people to use it. If it’s caught in traffic, it won’t be reliable.

          I realize that Amtrak is setting ridership records despite being unreliable — but the major alternatives are also unreliable (airplanes and intercity buses) and long-distance driving is a particularly awful experience.

          For a local service to compete with cars, it needs to run on schedule. You cannot do that without bus lanes.

      • And since when did it become a crime to provide good paying jobs for your fellow citizens?

        Since Reagan the great liar came into office spouting all that garbage that he himself never believed for one second.

        He played his role like the oscar winning actor he never was.

  10. “Commuter rail goals are easily achievable today, with buses.”

    >>>> ALL transit goals here in Portland are better achievable with buses.

    • Hey Nick and Al (and anyone else who thinks buses solve every problem), are you ready to take lanes away from SOVs to provide for buses traveling on I-5 between Portland and Salem? Otherwise the buses would get stuck in northbound congestion every evening between Wilsonville and Portland.

      Unless, of course, you want to widen the freeway for the exclusive ROW?

  11. I agree that commuter rail may be the best shorter-term solution for transit to Clark County, specifically to Vancouver. What I’d like to throw into the mix is the idea that it should be electrified. The only difference is that you use an electric locomotive to pull the coaches (you don’t need to spring for self-propelled cars). Electric trains are faster to accelerate than diesel ones; this reduces the penalty for adding additional stations (such as one in St. Johns), and also allows them to accelerate out of the twisties faster. All of this adds up to eventually allowing the trains to be run with tighter headways than is possible with diesel rolling stock. Oh, and of course they don’t emit carbon or any other kind of emissions (except, I suppose, for perhaps a small amount of particulate matter from brake pads). This helps with community relations; nobody who lives next to the tracks wants to hear about the possibility of even more dirty diesel trains going by.

    As Joseph Edge pointed out, this really makes sense if combined with the electrification and improvement of the Cascadia corridor (which, again, could be done with existing rolling stock simply by swapping in electric locomotives and electrifying the line).

    I’ve seen estimates of the cost of electrification at $2 million/mile, which really isn’t bad compared to the total cost of these proposals (much less the total cost of the CRC).

    • Warren Buffet is not going to let you hang wire over his railroad. At some time in the future he might decide it makes sense to hang wire over his railroad and power it with his coal. And then you can have your electrically hauled commuter rail.

      Not before.

      • Yes he will. BNSF will allow governments to pay for electrification — we know this from Denver.

        BNSF will demand enormous overhead and side clearances (particularly on this route, due to the airplane parts business). BNSF will charge twice the amount which seems reasonable. But BNSF has made it clear that it’s entirely a matter of money — they’ll do it for the right price. (Unlike UP which has refused outright.)

        Unforunately, the BNSF rail bridge over the Columbia has overhead girders, so meeting BNSF’s enormous overhead clearance requests would probably mean buying BNSF a new bridge.

        • Well, I guess you’re right about the electrification. But commuter rail across the river still a silly idea because both the Camas-Washougal line and Ridgefield-Woodland-Kalama run along the edges of the catchment area they’re supposed to serve.

          The only decent commuter rail opportunity is the county-owned railroad along St. Johns to Battle Ground.

          And that would cost a cool billion to upgrade.

          Clark County is not the Peninsula or western Cook County.

  12. “(Mostly) unidirectional service. Much as TriMet’s express busses do today – mornings towards downtown Seattle, evenings away from it.”

    This is a bad policy by Sounder, not a “positive” feature of commuter rail. It is true that sounder doesn’t have the money or the track rights to run in both directions, but it would be much more useful as a transportation option if it did. Have you seen the meager ridership figures for sounder?

    Almost all regional rail trains run in both directions in the rest of the world, including the best systems in Europe (Switzerland, Germany) and Japan. That’s because it is cheap to run trains away from downtown if you are already planning to run trains TO downtown. Unless you have a big train yard to park trains near downtown, you have to do something with the vehicles.

    Running service both ways during the day, every 30 minutes or at least every 60 minutes, makes the service useful as a form transit. If the train only runs to work in the morning and to home in the evening, you might get stuck. What if you need to get back home unexpectedly, or you only need to work 1/2 the day, one day? What if you need to stay late, or want to eat dinner after work? Commute rail works best if it runs during the day and at night, in both directions, just like transit.

    Stop spacing is another risky trade-off. If you have stops only every 5 to 10 miles, the train will have a high average speed, but it may take a long time to get to the nearest station from home, and you may have to take another transit service to get anywhere except downtown. Because of this, most German S-Bahn systems have stops every 1 to 3 miles in the suburbs, and every 1/2 to 1 mile in the city; Japan’s regional trains often stop every 1 mile. This makes the service useful for getting to dozens of destinations from the suburbs, not just to one central station. Of course, these systems in German and Japan are almost always electrified, so less time is lost with each station stop, due to fast acceleration (like MAX rather than WES).

    I would support a regional-rail system on the existing BNSF/Amtrak route, but we would do better to imitate successful systems in other countries, with at least 2 tracks each way, service in both directions every 30 minutes or better for 18 hours a day, and stops at least every 2 to 4 miles in the suburbs, and several stops 1 mile apart in the city. Service like this could also be successful from Downtown to Troutdale along the tracks next to I-84, if Union Pacific ever changes their attitude.

    Perhaps what many people want is not regional rail, but better intercity rail. There is a place for trains that stop every 10 to 20 miles and run every hour – that’s what the Amtrak Cascades line should be offering, along with express trains that serve only the some of the existing stops, every hour – like Bolt Bus.

  13. Talk to UP about commuter service on their mainline south to Salem. They won’t have any of it (without some per$ua$ion). Mainline railroads are generally loath to accommodate commuter rail service because the tightly packed schedule of trains during the peak periods impacts their freight shipments. And wouldn’t a commuter rail require a countywide vote in ClackCo?

    But I do think BIDIRECTIONAL service would be a boon to people who live in Portland and work in Salem and currently don’t have a way to take a train to work without spending the night in Salem. Apparently that will change in the next few months as ODOT will add a new Cascades trip going southbound in the mornings and northbound in the evenings.

    Consider stops at: Woodland, Ridgefield, Vancouver, St. Johns, Union Station, OMSI, Milwaukie, Clackamas, Oregon City, Canby, Woodburn, Keizer, Salem.

Leave a Reply to Anandakos Cancel reply

Your email address will not be published. Required fields are marked *