Necessary, not sufficient
Jan Rosenow shows that electricity wastes less. That removes one objection to replacing fossil fuels, not the rest.

REPLACING AN ENERGY system built on stocks of dense fuel with one built on intermittent, diffuse flows is plainly problematic, and some policymakers think it impossible. In a new paper in npj Clean Energy, Jan Rosenow and his colleagues warn that policymakers may hesitate to support clean energy technologies ‘if they incorrectly believe replacing fossil fuels is impossible’. And the view that it’s unrealistic, they say, ‘rests on a systematic misinterpretation of energy system size’.
It doesn’t. What they show is that the view rests on a systematic misinterpretation of energy system size as long as everything else it rests on is excluded. Nothing is impossible once you set aside what makes it impossible.
In the energy transition, electrification is necessary. But necessary is not sufficient. A transition also has to replace the fuel that electricity can’t easily reach, keep the lights on when the weather fails, and get built. What interests me is how the paper deals with those other necessities, as Rosenow describes it in his Substack essay, Electrofficiency and the Primary Energy Fallacy.
Both the paper and the essay open with Smil, who doubts that a system supplying more than 80% of the world’s energy can be replaced in a few decades, and Yergin, who says the fossil share has barely moved. The analysis quantifies what, as Rosenow says, is not new to energy researchers: electricity wastes less. Then a section of the essay headed ‘What the analysis doesn’t say’ sets aside hydrogen for ‘hard-to-abate sectors’ and whether any of it can be built in time: ‘We don’t address that here.’ Then, having set them aside, the close says a major problem is that people overlook the efficiency gain, and so think the transition is bigger than it is.
So the concerns are dismissed after the things they concern have been set aside. It’s a fascinating rhetorical style, and it travels.
Take the sailing ship. It burns no fuel at all, so it’s infinitely more efficient than a steamer, and anyone who worries whether its cargo will arrive is misreading the size of the energy system. That holds, provided we set aside the doldrums.
Or take a house in Edinburgh with solar panels on the roof. At noon in June they make more electricity than the house can use, so worries about heating it in January rest on a misinterpretation. That holds too, provided we set aside winter.
Here’s what was set aside.
The fuel electricity can’t easily reach. The paper’s electrified system still uses 101 EJ of gas, coal and oil a year, a third of its primary energy: chemical feedstock, steel, cement, chemical plants, most flying and shipping. Replacing most of it with electricity needs 25% to 64% more electricity than the paper shows. The efficiency saving survives, but the build grows.
The weather. The paper’s one figure for it is the energy lost in storage, and the sum is wrong: 56 TWh, not 30. That barely moves its headline, but it’s the only number the paper puts on the weather. The table it took the figure from, in a Royal Society study, assumes a quarter of demand is met by steady baseload, which the paper’s system doesn’t have. The study concludes that wind and solar with storage could meet Britain’s demand, but it also finds that wind ‘can vary over time scales of decades’. Its 37 years of British weather include persistently low winds from 2009 to 2011 that make ‘a very large call on storage’, and it finds tens of TWh of very long-duration storage will be needed, ‘over 1000 times that currently provided by pumped hydro in the UK’. It warns that ‘Studies of single years cannot cast light directly on the need for storage lasting over 12 months’. The paper takes that study’s yearly storage losses, but not the store it says the doldrums need.
The build. The paper’s system gets 181 EJ a year from wind and solar. That’s 11 times what they produced worldwide in 2024. Replace the rest of the fossil fuel with electricity too, and it’s 14 to 19 times. Neither the paper nor the essay says so.
None of this is hidden. The paper names each of these necessities, and of weather and reliability says: ‘We treat these factors as constraints to be addressed in deployment rather than as reasons that the structural efficiency story is wrong.’ Earlier it says the system-size argument ‘collapses’. The fact that the size argument collapses doesn’t collapse the objections. And the efficiency gain that ‘removes one of the most commonly cited objections’ doesn’t rescue the paper’s opening claim from all the other objections.
I’m setting out the detail in a comment I’m submitting to npj Clean Energy, and the full working in a technical paper that I’ll be publishing soon on this site.
Rosenow’s essay ends: ‘The transition is big. But it is not as big as the current fossil fuel system makes it look.’ Agreed. It’s also bigger than the efficiency gain makes it look. The sailing ship burned no fuel and still lost to steam, which burned a great deal of it and arrived when it said it would.