4 ms·
> But obviously they are dwarved by the single commuters, or else the average wouldn't be 1.4 persons per car. Well, 1.4 persons per car is already quite a bit
by rklaehn 7y ago
> But obviously they are dwarved by the single commuters, or else the average wouldn't be 1.4 persons per car.
Well, 1.4 persons per car is already quite a bit higher than your original claim of "at least 2500kg of material to move 1 person".
> Let's look at the numbers again. To not disadvantage the car, we assume 20% capacity, which is the minimum possible for a car.
Yes, OK, let's look at some actual numbers about kWh per 100 passenger km, averaged over all times and not just rush hour. Here is something I could find: https://www.forschungsinformationssystem.de/servlet/is/342234/ https://www.forschungsinformationssystem.de/servlet/is/34223...
https://www.forschungsinformationssystem.de/servlet/is/342234/Stadtverkehr_gross.jpg https://www.forschungsinformationssystem.de/servlet/is/34223...
So a subway car or a commuter rail car gets 12.5 kWh per 100 person kilometers. That is still better than a tesla model 3 of 160 Wh / km or 16 kWh / 100km. But not by much.
https://insideevs.com/news/347916/tesla-model-3-epa-energy-consumption/ https://insideevs.com/news/347916/tesla-model-3-epa-energy-c...
If you assume 1.4 people on average, you get 11.4 kWh / 100km / person, so basically the same.
And if you have 3 or 4 people, the model 3 is significantly more efficient than the average commuter rail.
> In my area the actual train during communiting hours is at least 70% occupied, BTW.
But trains also go outside commuting hours. That is the whole problem with a system that has large units and is not on demand.
> I'm not trying to win internet points here; there just is no way around the fact that electic cars are not in any way a meaningful solution to our climate problem.
Neither am I. I was doing back of the envelope calculations about what it would take to do a battery powered commuter rail for the relation I frequently use, and was genuinely surprised how inefficient commuter rail is on average.
- Symbiote 7y agoIt would be good to have figures for trains that are up to 10 years old (rather than everything up to 40 years old or whatever), which should be more efficient, and the manufacturing energy cost for the vehicles -- I suspect the cars will fare quite badly here. We can also take into account the cost of accidents, which we could probably assign an energy value to avoiding.
- HPsquared 7y agoMaybe an easy win could be to downsize the trains outside peak times - send just a single carriage?
- bildung 7y agoFrom my casual obervation this is exactly what is done in my town: The rush hour trains usually are longer.