To summarize: 40% of tonnage but 50% of tonnage-kilometres. I thought freight volume would be measured in ton-kilometres in the first place.

To further give context: the article is saying that most of the fuel transported around is done for long distances, so when something removes fuel use in the consumer side it has a double dip effect: less fuel consumed and less fuel used to transport the fuel, since long fuel supplies route diminish. It's a third order of thinking and that's why it's confusing. The article then argues that reducing that consumption in the buyers side is more effective:

> This is the part that fuel-first narratives tend to miss. In a serious energy transition, coal demand falls, oil demand falls, and gas demand falls. That means fewer bulk carriers and tankers moving fossil energy around the world. The maritime sector does not have to find a one-for-one replacement fuel for all of that work, because a material share of the work should disappear.

I would argue that chipping away at all three sides of the equation reducing the amount of fuel used, the amount of fuel used for transport and transporting things using other that fuel are worth pursuing.

I wonder if this actually has the potential to have the opposite effect.

If an oil producer electrifies faster than average, for example Norway, then oil that might have been consumed domestically instead is shipped overseas.

Norway has an abundance of hydro power. They've had an almost-entirely renewable grid even before the energy transition begin. Combine that with their small population and large oil reserves, and they have always exported far more than they have consumed domestically.

In theory the small amount of additional available oil due to domestic electrification might become available for export, but I expect that the global drop in demand due to worldwide electrification will make that unlikely: they'll just slow down production to avoid flooding the market and crashing the oil price.

There are relatively few oil exporting countries so probably not a big effect