Buses are often hailed as the cheaper solution for mass transit. I think there are fundamental flaws in most of the comparisons we see between BRT and rail systems, and that it’s unlikely mainline buses actually make sense in long term planning.

When I say mainline, I mean corridors that will have long term need for transportation. I think Martin’s brought up some great points about what that means – I don’t, for instance, think that we need to build past Redmond at this point, or past Issaquah – we don’t know what is going to happen there in the coming decades, and we don’t have the money to guess. We have some clear centers that are not going to disappear – some are already walkable and dense, like some of Seattle’s neighborhoods, and some are car-centered today, with lots of parking and one-story clusters of development, but ripe for reconstruction to funnel the new growth coming to our region. There’s no “flexibility” argument here, though – urban corridors don’t pack up and move, they never have and they never will. This isn’t a frontier town, this is a major city.

I’m sitting on a bus right now, in stop and go traffic at 9th and Stewart in downtown – so I want to start with the fallacy that building HOV lanes on our freeways is somehow equivalent to building new rail right of way. I think to some, especially to those who use transit already, it’s clear that these are nowhere near the same levels of service. If I head downtown from work, like today, about half of my commute is spent in downtown traffic – a tiny percentage of the overall distance. No matter what we do to SR-520, the 15-20 minutes I spend getting from one end of downtown to the other will not be affected. In order to provide consistent service end to end, we have to build new right of way end to end.

That right of way costs money – lots of money. With a project like University Link, in order to get anything like three minute service from Husky Stadium to the center of Capitol Hill, you’d have to tunnel for buses just as we are for rail. The cost of laying rails in that tunnel is tiny compared to the tunnel itself. You can look at any segment of our light rail system and make similar observations – in the Rainier Valley, we repaved the entire roadway to make space not for trains, per se, but for dedicated right of way. The cost is due mostly to the level of service, not the technology, but that level of service difference is what creates the consistency and reliability that we value in rail systems. When you actually compare the capital cost of a BRT system that provides the same level of service as a light rail system, you find that your right of way costs are exactly the same.

So, you say, you’ve seen capital cost comparisons that meet these requirements, have exactly the same amount of new right of way, but still show BRT being cheaper? Unless they’re in totally different cities or countries with different labor costs and safety requirements, they’re almost always missing one thing – electrification. In this area especially, that’s a big deal. While the cost of oil has doubled in the last couple of years, the cost of our electricity hasn’t. Electrification insulates us from $4.50 per gallon diesel – or $6, or $10. We’re designing a system to last not decades, but hundreds of years – we can’t just shut it down to change over later. But when you electrify, your total cost of construction for rail versus bus is nearly identical – which makes sense, because it’s not any cheaper to lay concrete roadway than to lay rail, and all of your other infrastructure is a product of the level of service, not the technology.

Okay, so what’s the problem? Why are you so hell-bent on building rail if they’re exactly the same, Ben? Two reasons:

First, capacity. Some BRT advocates will tell you that buses can have exactly the same capacity as rail. They’re either uninformed, or they’re lying. Even with double-articulated coaches as in Curitiba, you’re looking at an 85 foot long vehicle with 57 seats. Curitiba claims they can reach 270 passengers – but at the measure of 6 passengers per square meter standing. With half that standing density, 3 passengers per square meter, our light rail cars carry 200 (with 74 seated). If you went by Curitiba standards, we’d carry more than 325 people per car. These cost about the same amount to operate and maintain – for the sake of discussion, about half the operations cost of a vehicle like this is the fuel, and about half the operator, although that now varies a lot more with the high cost of fuel prices, so my comparison gives buses a slight advantage.

But wait – we can tack three more vehicles onto a Link train behind the same operator. If we want to add another bus, that means paying another operator, so Link scales to four car trains at some 5/8 the cost – and a full metro can go much higher, with as many as 12 cars. We can also go down to lower headways than the buses can without affecting service quality – the big limiter is the time taken to board, which is a lot lower for four simultaneous light rail cars than four sequential buses, even when the buses have multiple doors. Rail can also offer a very finely tuned interface between vehicle and platform – on new systems, no ramps or lifts are necessary for wheelchair users.

I’ve already touched on it a bit, but the last reason is long-term cost. A rail vehicle costs more than a bus, but lasts at least proportionally longer – New York City has subway cars well over 50 years old in service today, and recently retired some that were even older. Most buses last ten years, some fifteen. Our Breda coaches in Seattle are now nearly 20 – but that has only been possible after major overhauls. They are nearing the end of their service lives. At the same time, fuel costs for our bus system have doubled, while our electricity prices in the city (I don’t know about you Puget Sound Energy folks) have stayed basically the same.

With any dedicated right of way, ridership is generated largely by the existence of the transportation system. I suspect that this would be the case for true BRT as well, because the factors that generate that ridership have to do more with the pedestrian density generated around stations than with the mode. In the long term I think the immediate space around any system built in any of the Puget Sound urban corridors today will increase in density to the point where the capacity offered by a rail system is absolutely necessary. I think our recent exercise at Reality Check helped make it clear that most of our regional leaders are on the same page in that respect.

Buses are great feeders, but they have no place as a mainline corridor – claims of cost savings are not for equivalent systems and don’t hold up in the long term. If you’re going to build a real transit system, make it rail and do it right.

20 Replies to “Why BRT Doesn’t Make Sense”

  1. i can think of no better post on brt vs light rail than this one. bravo, and congrats on the new site layout as well.

    really, you hit all the high points in one place – electrification vs diesel, boarding speed, operator-to-psgr ratio, dedicated row, equipment capex amortization vs buses, you name it.

  2. I love the open-mindedness and the solution-seeking. I love the flexibility and the desire to seek both short and long-term solutions.

    Yes, this post has it all..


    1. Brad, in 1968, when we were preparing for our first rail vote, when Warren Magnuson had guaranteed us an 80% federal match, the opposition said that they had a BRT system planned that would do just as well as Forward Thrust. They did nothing. Those who fought against Sound Transit in 1995 claimed they had a better BRT solution. They showed nothing, saying only that they “needed more time” to build a package (in fact, they said it for three years before voters got the clue and passed Sound Move in ’96). Now we’ve got commuter rail and we’re about to open light rail, but those buses we had such plans for 40 years ago are still in traffic with everyone else.

      There is no magic bullet for Redmond Ridge. There is no magic bullet for Snohomish. There is no magic bullet for Orting. If you want rapid transit of *any kind* for these places, you’re looking at a lot of tax dollars – more per person the farther apart they live. That’s math – it’s not a matter of open-mindedness or solution-seeking, it’s simple, hard math that when two people live farther apart, it takes more road or more rail to connect them. Unfortunately, the people who tend to live that far from the city are the ones least likely to vote to tax themselves – the per person cost of connecting two next-door apartment buildings is a couple of orders of magnitude lower than the per person cost of connecting two houses on full lots.

      If you have some better math, let’s hear it – but think about fuel prices, think about longevity, and think about cost per passenger mile. It costs well over a hundred dollars an hour to operate a bus – so when we’re looking at adding service hours to a route that commonly has sixty full seats and 30 people standing, and comparing that to a route that gets six riders per trip, where are we supposed to put our service? The trip with six riders doesn’t have six riders because it has low service – it has six riders because it’s in the middle of nowhere.

    2. I think Ben offered a variety of reasons why BRT isn’t the best choice for regional transit compared to a rail solution. Instead of discussing the merits of his points, you say he’s closed-minded and not trying to build a consensus. He seems closed-minded regarding BRT because he understands its faults, and he’s not trying to build a consensus at the sacrifice of the region’s best transit options.

      Where do you want BRT to and from, Brad? I asked you a few days ago for your ideas, and you didn’t even mention BRT.

  3. This is about what is needed. Even with short term solutions you need to start thinking long term. That’s what rail is, a long term solution. If anyone doubts the reasons for rail, read this

  4. Great post. I always chuckle when I hear pro-BRT type stories on the air. I end up screaming at the radio “It’s the right of the way , stupid!” Unless you have dedicated ROW, you might as well run clown cars. You get about 30 clowns into a 11 foot car…hmm.. Clown Rapid Transit, I like it.

  5. one other thing worth mentioning is that BRT doesn’t necessarily come online faster than rail, either. look at the scheduled start dates for rapidride. the silver line brt in boston took many years to get going and still isn’t finished.

  6. Pingback: Schwerzler.com
  7. It’s taking Via Metropolitan Transit in San Antonio, TX 5 years to implement BRT on a single straight-line corridor.

  8. You should really try and get a guest editorial in the Times or PI. Preferably the Times to get this word out to the right crowd. Or maybe talk to the guys over at Crosscut to write a companion piece to Douglas MacDonald’s. This blog just isn’t loud enough.

  9. Ben, what do you think of busways a la Pittsburgh or Ottawa? Basically, as opposed to just running buses along the right-of-way as the Orange Line in LA does, is it better to dedicate a long right-of-way to a buses, then run a set of buses coming from different routes through that right-of-way into a CBD?

    Obviously, from the point of view those along the right-of-way, this won’t be as nice as strict mainline LRT would be (boarding times, noise, diesel fumes, comfort, operations costs, etc. would be worse, though frequencies might be better) but the people beyond the right-of-way do get a quick, one-seat ride while the people in the right-of-way get fast, frequent service, so it seems like it might be an easier political sell than an LRT that seems too far away to use. Then again, people don’t like buses, and for obvious reasons.

    1. Like the DSTT? Or are you speaking more for a highly restricted form of the Portland Transit Mall

      1. AJ –

        What Pittburgh has is sort of like the DSTT, only longer and lower-budget (it’s mostly at-grade, has shelters rather than stations, etc.) I think it was built in an old rail right-of-way.

        I guess you could argue 3rd Avenue is a little like this, only 3rd Avenue kind of sucks (allowing cars on some stretches makes enforcement difficult, it’s already overcrowded with buses at rush hour, and the frequency of intersections makes the whole thing run pretty slowly).

      2. Steve, comparisons like that include a huge amount of existing right of way. Pittsburgh only built some five miles of new busway – they had a $320 million budget, I think, but they had a lot of existing infrastructure. This wasn’t a from-scratch project, it was incremental. My comments are more about building from scratch, like we’d be doing with any new system here.

        If they could have spent the same money on rail, that could have been a better choice, but if that was their only way (and place) to spend their money, it was fine for the short term. Often the money just wouldn’t have been there for anything else.

        With the DSTT, we did a conversion, and it’s better that they have the dedicated right of way to convert in the future than not have anything at all.

  10. Bravo to a a well written post that is certainly applicable to any metro area thinking about how to expand its mass transit service.

  11. Thanks Ben. Excellent discussion of many of the important details in rail v. bus along mainline transit corridors.

    It is obvious to realize the benefit in terms of labor savings by having multiple cars attached to each other and of having multiple wide doors for alight/board. However, I never put the two together to think about the benefit of having multiple cars stopping at a platform simultaneously: huge savings in alight/board times over multiple buses arriving sequentially. I think in a transit class we usually assumed a 15 second station/stop stand time for alight/board. If one train of 3 cars stops at a platform thats 15 seconds for its station stand time. But if 3 sequential buses stop at a platform, that’s 45 second station time.

Comments are closed.