Earlier quoted context omitted.
Thanks. That was my guess. I mean, I didn't think that aluminum production was big there. I wonder what the energy efficiency of petroleum refining is. From the US DOE,[0] I get: TBtu 3542 primary energy 2082 applied energy 1460 energy lost So ~59%. That's a lot of energy to blow off. 0) https://www.energy.gov/eere/amo/dynamic-manufacturing-energy...
This illustrates something that is curiously lacking from most ICE vs BEV energy efficiency calculations, namely: More electricity is spent just refining the gasoline used by an ICE than is used in total by a BEV to go the same distance.
The relevant comparison is arguably greenhouse gas emissions per km. For ICE, you clearly must include emissions during refining. But that could be ~zero, if all the process energy were wind, solar or nuclear electricity. Or it could be an additional 70%, if all the process energy were petroleum. So the range is 1.0-1.7 times nominal fuel emissions.
For BEV, total emissions could be ~zero, if all the electricity were wind, solar or nuclear. But if the electricity were generated from fossil fuels, you'd have nominal fuel emissions plus both refining process energy and generation loss.
Electricity from natural gas could be the clear winner, because you get additional energy from two H2 for C converted to CO2. Unless leakage were too great. And there's ~no conversion loss in the vehicle. So that's maybe 0.4-0.7 relative to gasoline/diesel fuel emissions.
Electricity from coal and petroleum would clearly be worse. For coal, because there's ~no H2, and because generation efficiency is lower than for natural gas. And for petroleum, because you have the 0%-70% process energy, less H2 per C, and somewhat lower generation efficiency. But there's ~no conversion loss in the vehicle for either. So overall that's maybe 0.8-1.4 relative to gasoline/diesel fuel emissions.
Bottom line, relative to nominal ICE fuel emissions:
ICE ~1 to ~1.7
BEV ~0 to ~1.4
But hey, those are totally thumbnail estimates.