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.
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.
> 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.
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.
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:
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
That's comparing new cars (all EVs are fairly new) to the full range of passenger rail, including 40-year-old diesel trains trundling around Wales.
Look at "Eurostar", which is a not-even-that-new high speed electric train: 6g/km. Though the calculation probably also takes account of the number of people on the train, and Eurostar will have better utilisation than average.
(NB coal and ore is moved at about 60mph in the UK, to avoid slowing other trains.)
the problem is that cars have a much worse form of overhead: parking. A rail line that transfers 1M people's commutes per day doesn't use up any space in the city that people commute to. If the same 1M people commute by car, you end up needing roughly 15 square kilometers just for the parking.
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.
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.
> 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.
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.
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.
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.
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-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.
> Most train platforms are maxed out at well below 10 cars (Caltrain is 6 cars)
> a perfectly run commuter train system is _just_ barely comparable with a regular 6-lane freeway
Also look at any major route in Britain, like Manchester to London, no 6-lane highways anywhere in sight.
You are comparing some weak-ass train with a giant highway. It's easy to extend a rail platform to accomodate more train cars - you just need to knock down a few buildings in a local area. Now try widening a 3-lane highway into a 6-lane, that could be hundreds of building and NYMBY's across three cities.
Allegedly the Eurostar, which links Paris to London, does:
"Inter-Capital" sets consisting of two power cars and eighteen passenger carriages. These trains are 394 metres (1,293 ft) long and can carry 750 passengers: 206 in first class, 544 in standard class
I can't find info on the upcoming HS2. However I do testify that extremely long trains are common in Eastern Europe, Russia, India, etc. they might not be high-speed but they do move a lot of people at once.
20 car trains are a distraction here the important correction is 150 passengers per train is way low. 250 is common and going higher (e.g. double decker) is pretty easy if you need to
> 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.
So you'd need 6 of those highways converging on the middle of a city to compare to Tokyo's single train station. To make it really fun to solve, no rational city would allow that disastrous use of above-ground land, the parking for another 2m cars would be an engineering feat of note and you're dragging millions of tons of steel around for no reason. We haven't begun on emissions or need for fuel stops for those 2m cars.
Oh, and that's not the entirety of the Tokyo metro population, which is ~37m. The entire of Texas is ~29m-ish and that's spread out so far and wide they can afford to fuck around with 14 lane highways. The scale and solution are incomparable.
that's not 24 tracks. you need more tracks at the station since the trains are stopped, but you can probably do that with 4 to 6 tracks for the portions where the trains are up to speed.
BART has a capacity of 200 passengers per car in their legacy fleet, 241-256 for their new fleet, and regularly has 10 cars per train during peak hours [1] and travels 80mph.
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.
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.