You know your blog has made it when Randall O'Toole writes an incomprehensible critique. Which he did after reading, apparently, a couple of sentences of this blog. I commented on his blog, so read that over there.
The other thing that gets lost in the ITDP and O'Toole pro-bus arguments is that you need way more labor in order to provide the same service. This is great in South America where labor is relatively cheaper and maintenance of complex systems relatively more expensive than in the US/Canada/Europe. But here, you get more bang for your buck if you invest upfront in the complex system (rail infrastructure) with lower operating costs per trip over time.
Those 350 buses per hour require 350 operators per hour driving them. That's to provide the same capacity as a heavy rail line with 30 trains per hour and 30 operators driving them!
First, Curitiba's BRT does have 40 buses per hour, and they're at capacity, so much so that Curitiba is now planning to dig subways to increase capacity.
Second, the GAO's report is only valid if you consider that "buses painted a different color" counts as "BRT", which is pure lunacy. In Gatineau, Québec, an actual BRT line was built for 21 million dollars per km, so around 34 million dollars per mile. Cheaper than an LRT line? Maybe, then again, Besançon in France built its tramway line for 16 million Euros per km, so about... 34 million dollars per mile. And they actually were under-budget... and that included rebuilding a bridge in the center of town. Québec City also made a study of various BRT options versus a tram option recently, the BRT line was found to be less expensive, but only by 40%, for a capacity around a third of the tramway alternative.
In fact, the reality is that building a road made to support frequent bus passage or laying tracks is roughly the same price per mile. LRV are more expensive than buses to buy, but their capacity is much higher and their life expectancy is twice as long as buses, minimum (hell, Hiroshima was still running trams that survived the atom bomb in 2010...in regular service!). So when you normalize for capacity and life expectancy, LRV aren't actually more expensive than buses, the initial cost is higher, but you have to renew the fleet only once every 30-40 years rather than every 15 years. Finally, most LRV vehicles are electric, so you need to build power substations and power lines, which you don't need on diesel buses, which will cost more in fuel.
So building a BRT with trolley buses on a new right-of-way would costs about the same as building an LRT line. Buses aren't magically cheaper like O'Toole claims. However, in the North American context, where we have massive roads everywhere and few tracks, BRT can be much cheaper if they can reuse existing roads. One of the rare cases of the opposite, where an LRT line was built on existing tracks with diesel-powered vehicles is Ottawa's O-train, which cost 2,5 million dollars per km to put in service, 4 million dollars per mile.
Speaking of Ottawa, which has a full-fledged BRT system, the cost of building their "Transitway" was 440 million dollars... at the same time, Calgary built its first LRT line, of a similar length, for 543 million dollars. Again, true BRT on a new ROW revealed itself to not be substantially less expensive to build than LRT, and I'm not sure that cost for Ottawa included vehicles. And today, to confirm the lie that buses are cheaper to run, Calgary's transit authority spends, overall (including maintenance and overhead) just 3,10$ per passenger on their transit network, Ottawa's transit authority spends 3,90$ per passenger. These cities have similar populations and transit use, Calgary is more sprawled out, yet Calgary's C-train proves itself to be significantly cheaper than Ottawa's BRT to operate. There is no surprise that Ottawa is now correcting their mistake and replacing their BRT with an LRT line, the Confederation Line, something they expect will yield dozens of million dollars in savings every year for the transit authority.
For high-frequency uses, rail is more cost-efficient than buses, anyone who denies that is either ignorant or in bad faith. However, it has higher initial costs, especially when buses can simply reuse existing roads, which makes its use for low-ridership routes uneconomical.
The only reasonable point that is made is that, as it stands, there are few transit corridors in the United States not currently served by rail that have the ridership to justify building rail in them. That is a valid point. And indeed, many US cities could look at mid-sized Swedish cities for inspiration for transit, with trunk bus lines benefiting from articulated buses, stops at every quarter-mile, POP, etc... All cheap ways to better service for buses to be actually useful.
But in a way, that is also a self-fulfilling prophecy... how can a transit corridor have the ridership to justify rail when CURRENT services have a lower capacity than that which would justify rail? Not only does the lack of current capacity constrain ridership, it also constrains urban developments as residents and jobs flee the congested areas since there is no additional capacity to use in the transport network. Rail in such a circumstance can lead to a great influx of development downtown, and also serves to expand urban areas, in a way that downtown express buses absolutely cannot do, as they bring people only to one bus terminal downtown and do not give rapid access to other neighborhoods. One thing that is frequently seen when rail is built is that the rail network increases use significantly off-peak and during the week-end, whereas downtown express buses have next to no off-peak use. That's because it's not expensive to maintain adequate (10-15 min headway) service in rail due to the low operating costs per vehicle, whereas maintaining a vast quantity of express buses running would be extremely expensive, so they run few of them, once every 30 minutes or hour, and since they require transfers, that makes every trip very slow, so users are repulsed.
A good example of this is Phoenix's LRT. Phoenix's transit network was very poorly used, and Phoenix itself is very sprawling, yet they built this one line, and the ridership on it exploded, and now it is above 40 000 riders per day and still climbing, at a level that actually justifies building it. It has also caused a flurry of development in the areas it went through, at least, where development is actually allowed. If they don't use planning power to restrict TOD and keep up with demand, who knows how high it will go? Calgary's LRT started that way and it now carries half of the transit riders of the local transit authority.
Calgary is huge, yes. And while there are a lot of cities that don't need rail based on current projections (and places like Euclid where demand will probably never be exceeded by buses especially since there is an underutilized rail line between the endpoints), in the context of Boston, good lord there needs to be an investment in high capacity rail lines. This report revives the Urban Ring concept, which is DOA as a bus route because there is no way to do it in a straight line and no feasible way to do it without a lot of mixed-traffic, which defeats the purpose. No one takes the CT2 in the afternoon because it sits in 20 minutes of traffic to get across one bridge! There needs to be a real investment, not a few bus lanes where the corridor is wide enough. And this Boston BRT study is basically lines on a map (and some pretty pictures!) that really goes nowhere. Also, it's late and I'm ranty.
I think he's just angry at how bad Portland's rail system is, and I actually agree with him on that, though I still think a competently implemented rail system would be preferable to BRT even in that case. But Portland's rail system really is pretty terrible! The streetcar goes all the way up to a top speed of 20 mph, when it's not stuck in traffic or stopping at every stoplight, and the frequency is so abysmal that it's almost always faster to walk. And while the light rail has some fast and efficient sections, that's cancelled out by the incredible slowness of Downtown, which has way too many stops and a very low average speed as a result. Plus there are a few truly baffling route choices with sharp 180 degree turns to fit around freeway infrastructure. He's probably right in thinking that a bus system wouldn't have that sort of nonsense. But he's very, very wrong in thinking that every other light rail system is as incompetent as Portland's (though some definitely are, like San Jose).
The next post looks at some cities where light rail has been implemented well. Particularly Calgary. 300,000 people with service every two minutes on surface-level streets downtown. Why do we look south to cities very much unlike Boston (is anyone going to mistake downtown Belo Horizonte for Beacon Street?) when maybe we should be looking north instead?
The other thing that gets lost in the ITDP and O'Toole pro-bus arguments is that you need way more labor in order to provide the same service. This is great in South America where labor is relatively cheaper and maintenance of complex systems relatively more expensive than in the US/Canada/Europe. But here, you get more bang for your buck if you invest upfront in the complex system (rail infrastructure) with lower operating costs per trip over time.
ReplyDeleteThose 350 buses per hour require 350 operators per hour driving them. That's to provide the same capacity as a heavy rail line with 30 trains per hour and 30 operators driving them!
In case it doesn't appear on his site...
ReplyDeleteFirst, Curitiba's BRT does have 40 buses per hour, and they're at capacity, so much so that Curitiba is now planning to dig subways to increase capacity.
Second, the GAO's report is only valid if you consider that "buses painted a different color" counts as "BRT", which is pure lunacy. In Gatineau, Québec, an actual BRT line was built for 21 million dollars per km, so around 34 million dollars per mile. Cheaper than an LRT line? Maybe, then again, Besançon in France built its tramway line for 16 million Euros per km, so about... 34 million dollars per mile. And they actually were under-budget... and that included rebuilding a bridge in the center of town. Québec City also made a study of various BRT options versus a tram option recently, the BRT line was found to be less expensive, but only by 40%, for a capacity around a third of the tramway alternative.
In fact, the reality is that building a road made to support frequent bus passage or laying tracks is roughly the same price per mile. LRV are more expensive than buses to buy, but their capacity is much higher and their life expectancy is twice as long as buses, minimum (hell, Hiroshima was still running trams that survived the atom bomb in 2010...in regular service!). So when you normalize for capacity and life expectancy, LRV aren't actually more expensive than buses, the initial cost is higher, but you have to renew the fleet only once every 30-40 years rather than every 15 years. Finally, most LRV vehicles are electric, so you need to build power substations and power lines, which you don't need on diesel buses, which will cost more in fuel.
So building a BRT with trolley buses on a new right-of-way would costs about the same as building an LRT line. Buses aren't magically cheaper like O'Toole claims. However, in the North American context, where we have massive roads everywhere and few tracks, BRT can be much cheaper if they can reuse existing roads. One of the rare cases of the opposite, where an LRT line was built on existing tracks with diesel-powered vehicles is Ottawa's O-train, which cost 2,5 million dollars per km to put in service, 4 million dollars per mile.
Speaking of Ottawa, which has a full-fledged BRT system, the cost of building their "Transitway" was 440 million dollars... at the same time, Calgary built its first LRT line, of a similar length, for 543 million dollars. Again, true BRT on a new ROW revealed itself to not be substantially less expensive to build than LRT, and I'm not sure that cost for Ottawa included vehicles. And today, to confirm the lie that buses are cheaper to run, Calgary's transit authority spends, overall (including maintenance and overhead) just 3,10$ per passenger on their transit network, Ottawa's transit authority spends 3,90$ per passenger. These cities have similar populations and transit use, Calgary is more sprawled out, yet Calgary's C-train proves itself to be significantly cheaper than Ottawa's BRT to operate. There is no surprise that Ottawa is now correcting their mistake and replacing their BRT with an LRT line, the Confederation Line, something they expect will yield dozens of million dollars in savings every year for the transit authority.
For high-frequency uses, rail is more cost-efficient than buses, anyone who denies that is either ignorant or in bad faith. However, it has higher initial costs, especially when buses can simply reuse existing roads, which makes its use for low-ridership routes uneconomical.
DeleteThe only reasonable point that is made is that, as it stands, there are few transit corridors in the United States not currently served by rail that have the ridership to justify building rail in them. That is a valid point. And indeed, many US cities could look at mid-sized Swedish cities for inspiration for transit, with trunk bus lines benefiting from articulated buses, stops at every quarter-mile, POP, etc... All cheap ways to better service for buses to be actually useful.
But in a way, that is also a self-fulfilling prophecy... how can a transit corridor have the ridership to justify rail when CURRENT services have a lower capacity than that which would justify rail? Not only does the lack of current capacity constrain ridership, it also constrains urban developments as residents and jobs flee the congested areas since there is no additional capacity to use in the transport network. Rail in such a circumstance can lead to a great influx of development downtown, and also serves to expand urban areas, in a way that downtown express buses absolutely cannot do, as they bring people only to one bus terminal downtown and do not give rapid access to other neighborhoods. One thing that is frequently seen when rail is built is that the rail network increases use significantly off-peak and during the week-end, whereas downtown express buses have next to no off-peak use. That's because it's not expensive to maintain adequate (10-15 min headway) service in rail due to the low operating costs per vehicle, whereas maintaining a vast quantity of express buses running would be extremely expensive, so they run few of them, once every 30 minutes or hour, and since they require transfers, that makes every trip very slow, so users are repulsed.
A good example of this is Phoenix's LRT. Phoenix's transit network was very poorly used, and Phoenix itself is very sprawling, yet they built this one line, and the ridership on it exploded, and now it is above 40 000 riders per day and still climbing, at a level that actually justifies building it. It has also caused a flurry of development in the areas it went through, at least, where development is actually allowed. If they don't use planning power to restrict TOD and keep up with demand, who knows how high it will go? Calgary's LRT started that way and it now carries half of the transit riders of the local transit authority.
Calgary is huge, yes. And while there are a lot of cities that don't need rail based on current projections (and places like Euclid where demand will probably never be exceeded by buses especially since there is an underutilized rail line between the endpoints), in the context of Boston, good lord there needs to be an investment in high capacity rail lines. This report revives the Urban Ring concept, which is DOA as a bus route because there is no way to do it in a straight line and no feasible way to do it without a lot of mixed-traffic, which defeats the purpose. No one takes the CT2 in the afternoon because it sits in 20 minutes of traffic to get across one bridge! There needs to be a real investment, not a few bus lanes where the corridor is wide enough. And this Boston BRT study is basically lines on a map (and some pretty pictures!) that really goes nowhere. Also, it's late and I'm ranty.
DeleteI stopped feeding the troll long ago... I do enjoy these interactions though as its humorous how logic gets overlooked.
ReplyDeleteI think he's just angry at how bad Portland's rail system is, and I actually agree with him on that, though I still think a competently implemented rail system would be preferable to BRT even in that case. But Portland's rail system really is pretty terrible! The streetcar goes all the way up to a top speed of 20 mph, when it's not stuck in traffic or stopping at every stoplight, and the frequency is so abysmal that it's almost always faster to walk. And while the light rail has some fast and efficient sections, that's cancelled out by the incredible slowness of Downtown, which has way too many stops and a very low average speed as a result. Plus there are a few truly baffling route choices with sharp 180 degree turns to fit around freeway infrastructure. He's probably right in thinking that a bus system wouldn't have that sort of nonsense. But he's very, very wrong in thinking that every other light rail system is as incompetent as Portland's (though some definitely are, like San Jose).
DeleteThe next post looks at some cities where light rail has been implemented well. Particularly Calgary. 300,000 people with service every two minutes on surface-level streets downtown. Why do we look south to cities very much unlike Boston (is anyone going to mistake downtown Belo Horizonte for Beacon Street?) when maybe we should be looking north instead?
Delete