Hacker News Reader

Comment by rootusrootus | Hacker News Reader

> Just a reminder that all this car infrastructure is max ~100 years old

100 years ago we had one third the population we have today. Undoing that kind of development in another 100 years is plausible, but I suspect the oil problem will have resolved itself around that time anyway.

And installing the same size network of public transit is far more expensive than paving that many miles of roads. Vastly, incredibly, insurmountably more expensive. The real ask is for changing population patterns to become more dense, and this is a decades-long evolution.

View on HN

> Vastly, incredibly, insurmountably more expensive.

I don't know why you chose to exaggerate here, but rail lines are +/- the same cost per mile as interstate highway, and that's without the economies-of-scale benefit that comes from there being 100x the number of interstate miles. They're also vastly cheaper in terms of materials; the cost is mainly not materials. Concrete production is also extremely CO2 and energy intensive. Unsurprisingly, a boring 19th century technology deployed with vastly less energy and raw materials is more efficient. We're moving backwards in terms of overall system efficiency because of the economics of supporting a individualistic consumer society. And there's a phalanx of people spreading FUD to defend it from disruption.

The best case is that running diesel trains along all the interstate routes would be cheaper than the interstates themselves. I'll give you that, it is probably true (or close enough).

But the vehicle I use on the rails has to stay on rails, and the cost of running rails becomes significantly more expensive as density increases. Politically and economically both. And can you replace every last paved road in America with rails? No, of course not, so you still have some kind of road network. Which brings us back full circle to why people like interstates to begin with.

Systemic energy efficiency is one very simplistic measure of a transportation network. And I would argue the least important.

Though I would love to hear a description of your utopian transportation system. Assuming it doesn't rely too heavily on a wholesale redesign of lifestyle preferences (e.g. "everyone lives in a skyscraper" is kind of a non-starter).

There's no compelling reason why you couldn't have your rails embedded in the roads, most likely for light tram-type vehicles in denser urban areas, although you could go heavier if you like. Then you can have both wheel and rail transport together.

As you're putting up pantographs for the trams, there's no reason to not also have trolleybuses using them. When you have a network density built up to less than 5 minutes walk to the nearest tram/trolleybus stop, with a good timetable for daytime, there is little reason to have any private transport for within-town transport. Add provision of freight carrying by tram (use passing loops and specially designed trams for carrying pallets or delivery cages) and the number of vehicles with good reason to be in the town decreases further.

As you get into lower density areas, it makes more sense to go for links between towns and villages by rail, with ICE or electric buses feeding in. There should be no more than 15 minutes walk for normal users to reach a bus stop (Note: all bus stops should be very nice, with roofs and walls that keep the rain out). In all rural and urban cases, there should be the option of an app or call button which summons a battery minivan (or trolleybus in suitably served areas) to a person's door at anytime, with confirmation after the length of time for estimated collection is received. Automated scheduling systems with 24 hour service for dispatching electric minivans at reasonable cost to the end user to a location would remove the need for a lot of car journeys.

The thing is, none of this requires new technology, and could be marketed as a jobs program alone. Automating parts would be nice, but there's no reason this couldn't have been done anywhere in say, 1920.

It looks like it takes 1500 tons of concrete per mile of highway, and a ton of concrete produces 900 kg of CO2, which I will estimate to be 1 ton of CO2. So, 1500 tons of CO2 / mile of highway. It is said to last 60 years, so 25 tons CO2 / year.

It appears to take somewhere between 115 - 215 tons of steel per mile of highway track, but you're going to want double track if you're comparing to highways, so 200 - 400 tons of steel. Steel produces somewhere between 1 - 3 tons of CO2 per ton of steel depending on how it is produced. So say 300 tons of steel at 2 tons CO2 / 1 ton of steel. So 600 tons of CO2 per mile of railroad track. Track lasts from 2 - 40 years, so if we say 20 years on average, we get 30 tons / year.

Maybe highways don't last 60 years, or track lasts longer than 20. Also, if you are thinking of high speed passenger rail, it uses a lot of concrete so that it can be on an elevated track. But it looks like concrete highways and railroads are comparable as fair as CO2 from construction. Certainly rail isn't so "vastly" better that we need to be concerned about phalanxes of FUD.

Now, operating costs in terms of CO2 probably favor railroads pretty heavily, at least for freight. Since passenger rails is a lot less dense in terms of cargo, the story for passenger rail might be different.

(These figures are sourced from asking DDG "co2 per ton concrete", etc)

Reply on news.ycombinator.com