7 ms·

It doesn't have to. How much do our freeways cost us? esp with externalities?

cyberax 4y ago
Freeways are amazingly cheap. 1 mile of new 6-lane freeway on level terrain is about 3 million: https://www.strongtowns.org/journal/2020/1/27/how-much-does-a-mile-of-road-actually-cost https://www.strongtowns.org/journal/2020/1/27/how-much-does-... Amazingly, railroads are not much cheaper. The current costs of tracks are estimated at about $2 million per mile, and this is without taking into account all other necessary rail infrastructure (such as sorting yards, maintenance facilities, etc.). And CO2 emissions are being fixed by switching from gas cars to EVs.
- adgjlsfhk1 4y ago
  The problem is that 1 mile of rail can move a lot more passengers than 1 mile of road.
  - sixstringtheory 4y ago
    What is the route capacity for a lane of freeway vs a rail line? I assumed it’s the opposite, since a freeway has continuous throughout while a rail line is discrete, but I don’t have a good intuition for comparing the scale of each mode.
    - dragonwriter 4y ago
      > What is the route capacity for a lane of freeway vs a rail line? I assumed it’s the opposite, since a freeway has continuous throughout while a rail line is discrete According to this book (which provides assumptions and calculations supporting) [0], 10:1 in favor of rail, as a conservative estimate. https://eng.libretexts.org/Bookshelves/Industrial_and_Systems_Engineering/Book%3A_The_Art_of_Insight_in_Science_and_Engineering_(Mahajan)/01%3A_Part_I-_Organizing_Complexity/01%3A_Divide_and_Conquer/1.02%3A_Rails_vs_Roads#:~:text=How%20does%20the%20capacity%20of,capacity%20of%20a%20train%20line%3F&text=This%20capacity%20is%20ten%20times,single%20fast%2Dflowing%20highway%20lane https://eng.libretexts.org/Bookshelves/Industrial_and_System....
      - cyberax 4y ago
        > A rush-hour train may consist of 20 cars The author is smoking some hard crack.
        AnimalMuppet 4y ago
        Well, BART has no problem running 9-car trains. I don't know if that's rush hour or not, but I've seen them. If the rest of the math is right, that still gets you 4.5-to-1 in favor of rail.
        dragonwriter 4y ago
        BART (mainline) runs a maximum of 10 car trains, with a “crush” (maximum) capacity of 200+ per car.
        cyberax 4y ago
        Have you ever walked past a 20-car train? It's way too long for commuting. There will be no way to build platforms for it unless you're doing a full green-field build.
        snovv_crash 4y ago
        Europe has plenty of double-story trains. I agree 10 is about the limit otherwise the station sizes aren't people friendly.
        Symbiote 4y ago
        The Eurostar trains between London and Brussels/Paris/Amsterdam have 16-18 coaches, although they are noticeably long. (Other than night trains, I'd expect these to be the longest in Europe.)
    - bluGill 4y ago
      How do you calculate capacity? Safety engineers keep yelling that drivers need to keep 3 seconds between cars, but in reality they most drive about .3 seconds from the car in front. There is nearly a 10x difference in freeway capacity between just those two. Trains tend to maintain longer distances, but if you want to ignore safety we can follow a lot closer.
    - adgjlsfhk1 4y ago
      The big difference is that in a train, you fit 4 people wide and roughly 2-3 ft front to back per person while in cars which have average capacity of ~1.5 per car, you have 1.5 people wide and 50-200 ft front to back. Trains are farther apart than cars, but fit a ton of people per train.
  - cyberax 4y ago
    The calculations on that page are not correct. First, the average car occupancy is not 1, it's 1.5 on average. Second, the lane capacity is 1900 cars per hour (that's maximum at around 45 mph, btw). So this works out to 8550 people per hour. A realistic scenario for commuter trains (that would replace a freeway) is 1 train every 10 minutes, and even this is pretty tough. So you have 6 trains per hour, and to match the throughput you'd need 1425 people per train. Most train platforms are maxed out at well below 10 cars (Caltrain is 6 cars), 20 car trains are just pure nonsense for commuting. So for 10 car trains it'll be around 150 people per car. Caltrain cars are 130 seats per car ( https://www.greencaltrain.com/2014/05/keeping-up-with-caltrain-capacity-needs/ https://www.greencaltrain.com/2014/05/keeping-up-with-caltra... ), with another 40 standing places. Basically, a perfectly run commuter train system is _just_ barely comparable with a regular 6-lane freeway. Sorry train fans, but trains are not that great for commuting.
    - judge2020 4y ago
      Correction, trains specifically. Subways and light rail are immensely higher ROI (as in, benefit vs cost) within a city since it effectively forces people close together, while a 6-lane-on-each-side highway isn't economically feasible for all of the high-traffic areas.
    - richardw 4y ago
      > Sorry train fans, but trains are not that great for commuting. Quick comparison: more people (3.6m) go through Shinjuku Station in Tokyo than the daytime population of Manhattan, at 3.1m. Only half of those travel into Manhattan, using all modes of transport. When things get extreme it’s hard to just double the road network and parking into a single location. https://en.m.wikipedia.org/wiki/Shinjuku_Station https://en.m.wikipedia.org/wiki/Shinjuku_Station
      - cyberax 4y ago
        This station is served by 12 lines. So presumably that's 24 tracks. This means that one track would need to carry a bit more than 150000 people a day on average. A 6-lane freeway can carry around 100000 people a day. As a sanity check, the infamous Katy Freeway carries 350000 cars a day, for about 600000 people. It's 14 main travel lanes wide. So yep, trains are not that efficient compared to freeways.
- ClumsyPilot 4y ago
  > And CO2 emissions are being fixed by switching from gas cars to EVs. Not reall, rail is 9 times more energy efficient than road vechicles. Thats why its cheaper to have a diesel locomotove move freight than to pour the same diesel into trucks. The whole reason rail exists is thsa its the most efficient form of tranportation on land.
  - bluGill 4y ago
    There are a lot of assumptions in the idea that trains are more efficient than road vehicles. Trains tend to carry more heavy bulk goods like coal, if they had more light things the numbers would change. Trains get a lot of efficiency from running very long trains, but that only works out when you have a lot of things going the same way, if you had smaller trains from each warehouse (which now is done by truck) that would reduce th efficiency. Yes trains have some efficiency advantages, but in similar service the difference is small. You only get those advantages when you use trains for things that trucks cannot do at all.
    - mikepurvis 4y ago
      The very long trains of mostly goods like coal are not an inevitability, though— it's a result of rail companies implementing PSR in response to some pretty specific incentives, see: https://www.nytimes.com/2022/10/09/opinion/business-economics/freight-train-mismanagement.html https://www.nytimes.com/2022/10/09/opinion/business-economic... A number of YouTube video essays argue the sides of this as well, here's one based around Sen. Sanders confronting a rail CEO in the wake of the recent Ohio derailment: https://youtu.be/e4w0q5NzCwA https://youtu.be/e4w0q5NzCwA
  - cyberax 4y ago
    > Not reall, rail is 9 times more energy efficient than road vechicles. If you're hauling coal at 35 mph. For people-moving application it's not that more effective: https://ourworldindata.org/travel-carbon-footprint https://ourworldindata.org/travel-carbon-footprint
    - 4y ago
- babypuncher 4y ago
  You're comparing dollars-per-mile of infrastructure when the more important metric is dollars-per-passenger-mile. You can build 1 mile of freeway for not much more than 1 mile of rail, but that mile of rail can serve considerably more passengers.
  - cyberax 4y ago
    That works only in cities (where you can't build freeways anyway), but not for commuter trains.
    - adgjlsfhk1 4y ago
      Why? What makes commuter trains different?
- mikepurvis 4y ago
  Two other factors that aren't coming up in this analysis are long term maintenance costs (highways in Ontario are constantly being resurfaced) and land use opportunity costs (it's simply not viable to run six lanes of freeway into most CBDs, and doing it with grade separation leads to raised highway eyesores or insanity like Boston's Big Dig). In any case, as others have pointed out, we don't have to argue hypotheticals here— China, Japan, France, Germany, etc have all shown that frequent-service electrified passenger rail is perfectly possible and an incredible public good.
waterheater 4y ago
>How much do our freeways cost us? esp with externalities? Rail also has cost, including externalities. For example, as a wise person on the Internet once said, "Trains take you from a place you don't live to a place you're not going to." The "last-mile problem" is a non-trivial factor to building train networks which by definition is not solvable by more rail. You're considering a cost analysis. A more complete analysis is the benefit-cost analysis (BCA). If the benefits outweigh the costs, the project is a net positive for society. Asking only how much our freeways cost us avoids consideration of the known benefits. Here's a BCA from 2021 which investigates a highway project, so you can see exactly how much freeways benefit and cost us: https://www.mdot.maryland.gov/OPCP/I-81BCA_Report.pdf https://www.mdot.maryland.gov/OPCP/I-81BCA_Report.pdf >It doesn't have to. That's a nice sentiment, but sometimes the energy required to reform is higher than expected losses. The entire Amtrak network is kept afloat because of a handful of lines in New England between DC, New York, and Boston. Amtrak basically runs a loss everywhere else.
- babypuncher 4y ago
  > Trains take you from a place you don't live to a place you're not going to. Airplanes have the exact same problem, but I don't see people saying we should stop investing in airports. For long distance travel, I don't think it's a huge problem that you might need multiple modes of transit to get all the way from A to B.
  - bluGill 4y ago
    Trains are too slow for long distance travel. They are useful for short and medium distance travel, but for such trips travel time to the station becomes more important. If this isn't made convenient people will quickly decide their car is better: it goes when they want to go, and goes directly to where they want to be.
  - waterheater 4y ago
    >Airplanes have the exact same problem, but I don't see people saying we should stop investing in airports. Fair point. We should then see if air travel holds a key advantage over rail travel in the USA. As I see it, the answer is in both space and time savings, both of which minimize cost and maximize benefit. The time savings are particularly pronounced, especially over distances greater than, say, a few hundred miles. Happy to elaborate on the savings in more detail, if you desire such. >For long distance travel, I don't think it's a huge problem that you might need multiple modes of transit to get all the way from A to B. That is, of course, your opinion. I'm sure there are tens, if not hundreds, of millions of Americans who will strongly disagree with you because they are, in no particular order: feeble, disabled, terrified of a particular mode of transit, hurried, cost-conscious, traveling with multiple young children, etc.

Freeways are amazingly cheap. 1 mile of new 6-lane freeway on level terrain is about 3 million: https://www.strongtowns.org/journal/2020/1/27/how-much-does-a-mile-of-road-actually-cost https://www.strongtowns.org/journal/2020/1/27/how-much-does-... Amazingly, railroads are not much cheaper. The current costs of tracks are estimated at about $2 million per mile, and this is without taking into account all other necessary rail infrastructure (such as sorting yards, maintenance facilities, etc.). And CO2 emissions are being fixed by switching from gas cars to EVs.

adgjlsfhk1 4y ago
The problem is that 1 mile of rail can move a lot more passengers than 1 mile of road.
- sixstringtheory 4y ago
  What is the route capacity for a lane of freeway vs a rail line? I assumed it’s the opposite, since a freeway has continuous throughout while a rail line is discrete, but I don’t have a good intuition for comparing the scale of each mode.
  - dragonwriter 4y ago
    > What is the route capacity for a lane of freeway vs a rail line? I assumed it’s the opposite, since a freeway has continuous throughout while a rail line is discrete According to this book (which provides assumptions and calculations supporting) [0], 10:1 in favor of rail, as a conservative estimate. https://eng.libretexts.org/Bookshelves/Industrial_and_Systems_Engineering/Book%3A_The_Art_of_Insight_in_Science_and_Engineering_(Mahajan)/01%3A_Part_I-_Organizing_Complexity/01%3A_Divide_and_Conquer/1.02%3A_Rails_vs_Roads#:~:text=How%20does%20the%20capacity%20of,capacity%20of%20a%20train%20line%3F&text=This%20capacity%20is%20ten%20times,single%20fast%2Dflowing%20highway%20lane https://eng.libretexts.org/Bookshelves/Industrial_and_System....
    - cyberax 4y ago
      > A rush-hour train may consist of 20 cars The author is smoking some hard crack.
      - AnimalMuppet 4y ago
        Well, BART has no problem running 9-car trains. I don't know if that's rush hour or not, but I've seen them. If the rest of the math is right, that still gets you 4.5-to-1 in favor of rail.
        dragonwriter 4y ago
        BART (mainline) runs a maximum of 10 car trains, with a “crush” (maximum) capacity of 200+ per car.
        cyberax 4y ago
        Have you ever walked past a 20-car train? It's way too long for commuting. There will be no way to build platforms for it unless you're doing a full green-field build.
        snovv_crash 4y ago
        Europe has plenty of double-story trains. I agree 10 is about the limit otherwise the station sizes aren't people friendly.
        Symbiote 4y ago
        The Eurostar trains between London and Brussels/Paris/Amsterdam have 16-18 coaches, although they are noticeably long. (Other than night trains, I'd expect these to be the longest in Europe.)
  - bluGill 4y ago
    How do you calculate capacity? Safety engineers keep yelling that drivers need to keep 3 seconds between cars, but in reality they most drive about .3 seconds from the car in front. There is nearly a 10x difference in freeway capacity between just those two. Trains tend to maintain longer distances, but if you want to ignore safety we can follow a lot closer.
  - adgjlsfhk1 4y ago
    The big difference is that in a train, you fit 4 people wide and roughly 2-3 ft front to back per person while in cars which have average capacity of ~1.5 per car, you have 1.5 people wide and 50-200 ft front to back. Trains are farther apart than cars, but fit a ton of people per train.
- cyberax 4y ago
  The calculations on that page are not correct. First, the average car occupancy is not 1, it's 1.5 on average. Second, the lane capacity is 1900 cars per hour (that's maximum at around 45 mph, btw). So this works out to 8550 people per hour. A realistic scenario for commuter trains (that would replace a freeway) is 1 train every 10 minutes, and even this is pretty tough. So you have 6 trains per hour, and to match the throughput you'd need 1425 people per train. Most train platforms are maxed out at well below 10 cars (Caltrain is 6 cars), 20 car trains are just pure nonsense for commuting. So for 10 car trains it'll be around 150 people per car. Caltrain cars are 130 seats per car ( https://www.greencaltrain.com/2014/05/keeping-up-with-caltrain-capacity-needs/ https://www.greencaltrain.com/2014/05/keeping-up-with-caltra... ), with another 40 standing places. Basically, a perfectly run commuter train system is _just_ barely comparable with a regular 6-lane freeway. Sorry train fans, but trains are not that great for commuting.
  - judge2020 4y ago
    Correction, trains specifically. Subways and light rail are immensely higher ROI (as in, benefit vs cost) within a city since it effectively forces people close together, while a 6-lane-on-each-side highway isn't economically feasible for all of the high-traffic areas.
  - richardw 4y ago
    > Sorry train fans, but trains are not that great for commuting. Quick comparison: more people (3.6m) go through Shinjuku Station in Tokyo than the daytime population of Manhattan, at 3.1m. Only half of those travel into Manhattan, using all modes of transport. When things get extreme it’s hard to just double the road network and parking into a single location. https://en.m.wikipedia.org/wiki/Shinjuku_Station https://en.m.wikipedia.org/wiki/Shinjuku_Station
    - cyberax 4y ago
      This station is served by 12 lines. So presumably that's 24 tracks. This means that one track would need to carry a bit more than 150000 people a day on average. A 6-lane freeway can carry around 100000 people a day. As a sanity check, the infamous Katy Freeway carries 350000 cars a day, for about 600000 people. It's 14 main travel lanes wide. So yep, trains are not that efficient compared to freeways.
ClumsyPilot 4y ago
> And CO2 emissions are being fixed by switching from gas cars to EVs. Not reall, rail is 9 times more energy efficient than road vechicles. Thats why its cheaper to have a diesel locomotove move freight than to pour the same diesel into trucks. The whole reason rail exists is thsa its the most efficient form of tranportation on land.
- bluGill 4y ago
  There are a lot of assumptions in the idea that trains are more efficient than road vehicles. Trains tend to carry more heavy bulk goods like coal, if they had more light things the numbers would change. Trains get a lot of efficiency from running very long trains, but that only works out when you have a lot of things going the same way, if you had smaller trains from each warehouse (which now is done by truck) that would reduce th efficiency. Yes trains have some efficiency advantages, but in similar service the difference is small. You only get those advantages when you use trains for things that trucks cannot do at all.
  - mikepurvis 4y ago
    The very long trains of mostly goods like coal are not an inevitability, though— it's a result of rail companies implementing PSR in response to some pretty specific incentives, see: https://www.nytimes.com/2022/10/09/opinion/business-economics/freight-train-mismanagement.html https://www.nytimes.com/2022/10/09/opinion/business-economic... A number of YouTube video essays argue the sides of this as well, here's one based around Sen. Sanders confronting a rail CEO in the wake of the recent Ohio derailment: https://youtu.be/e4w0q5NzCwA https://youtu.be/e4w0q5NzCwA
- cyberax 4y ago
  > Not reall, rail is 9 times more energy efficient than road vechicles. If you're hauling coal at 35 mph. For people-moving application it's not that more effective: https://ourworldindata.org/travel-carbon-footprint https://ourworldindata.org/travel-carbon-footprint
  - 4y ago
babypuncher 4y ago
You're comparing dollars-per-mile of infrastructure when the more important metric is dollars-per-passenger-mile. You can build 1 mile of freeway for not much more than 1 mile of rail, but that mile of rail can serve considerably more passengers.
- cyberax 4y ago
  That works only in cities (where you can't build freeways anyway), but not for commuter trains.
  - adgjlsfhk1 4y ago
    Why? What makes commuter trains different?
mikepurvis 4y ago
Two other factors that aren't coming up in this analysis are long term maintenance costs (highways in Ontario are constantly being resurfaced) and land use opportunity costs (it's simply not viable to run six lanes of freeway into most CBDs, and doing it with grade separation leads to raised highway eyesores or insanity like Boston's Big Dig). In any case, as others have pointed out, we don't have to argue hypotheticals here— China, Japan, France, Germany, etc have all shown that frequent-service electrified passenger rail is perfectly possible and an incredible public good.

waterheater 4y ago

>How much do our freeways cost us? esp with externalities? Rail also has cost, including externalities. For example, as a wise person on the Internet once said, "Trains take you from a place you don't live to a place you're not going to." The "last-mile problem" is a non-trivial factor to building train networks which by definition is not solvable by more rail. You're considering a cost analysis. A more complete analysis is the benefit-cost analysis (BCA). If the benefits outweigh the costs, the project is a net positive for society. Asking only how much our freeways cost us avoids consideration of the known benefits. Here's a BCA from 2021 which investigates a highway project, so you can see exactly how much freeways benefit and cost us: https://www.mdot.maryland.gov/OPCP/I-81BCA_Report.pdf https://www.mdot.maryland.gov/OPCP/I-81BCA_Report.pdf >It doesn't have to. That's a nice sentiment, but sometimes the energy required to reform is higher than expected losses. The entire Amtrak network is kept afloat because of a handful of lines in New England between DC, New York, and Boston. Amtrak basically runs a loss everywhere else.

babypuncher 4y ago
> Trains take you from a place you don't live to a place you're not going to. Airplanes have the exact same problem, but I don't see people saying we should stop investing in airports. For long distance travel, I don't think it's a huge problem that you might need multiple modes of transit to get all the way from A to B.
- bluGill 4y ago
  Trains are too slow for long distance travel. They are useful for short and medium distance travel, but for such trips travel time to the station becomes more important. If this isn't made convenient people will quickly decide their car is better: it goes when they want to go, and goes directly to where they want to be.
- waterheater 4y ago
  >Airplanes have the exact same problem, but I don't see people saying we should stop investing in airports. Fair point. We should then see if air travel holds a key advantage over rail travel in the USA. As I see it, the answer is in both space and time savings, both of which minimize cost and maximize benefit. The time savings are particularly pronounced, especially over distances greater than, say, a few hundred miles. Happy to elaborate on the savings in more detail, if you desire such. >For long distance travel, I don't think it's a huge problem that you might need multiple modes of transit to get all the way from A to B. That is, of course, your opinion. I'm sure there are tens, if not hundreds, of millions of Americans who will strongly disagree with you because they are, in no particular order: feeble, disabled, terrified of a particular mode of transit, hurried, cost-conscious, traveling with multiple young children, etc.