The Heat in the Wires: A Grid Model Finds Unequal Headroom

Most of us think of a power grid as a thing that is either on or off. The physics is less binary. A transmission line is a metal conductor, and metal warms as current passes through it. Warm it further with a hot afternoon and it sags and loses some of the current it can safely carry. At the same moment, the people on the other end are switching on fans and air conditioners. Meanwhile the power stations that must meet that demand run less efficiently, because they too are cooled by air and water that have become warmer. Heat leans on all three ends of the system at once.

On 22 October, Nature Energy published an open-access Article that tries to hold all three pressures in one frame. Enming Liang, Minghua Chen and Srinivasan Keshav report a computationally efficient framework that handles temperature-dependent demand, generator derating and spatially resolved transmission-line thermal limits together, inside an alternating-current optimal power-flow formulation. The desk could not read the authors’ affiliations through the access route available, so I will not guess at them. I will say what the packet records, and I will attribute it to the authors.

A model, not a blackout

I want to be careful about what kind of result this is, because the headline version of this story would be easy to get wrong. This is a peer-reviewed modelling study of projected scenarios. It is not a report of something that happened. The authors apply their framework to the Western European system, using open-access grid data, high-resolution weather reanalysis and demand data. They simulate hundreds of projected heatwave scenarios for 2026 to 2030. No heatwave-driven load shedding in Spain is reported to have occurred. What the authors report is what their simulations produce.

Within that frame, the finding is one of inequality. Across the grid configurations they represent, the authors report substantial national disparities. The Spanish grid shows temperature-induced bottlenecks and load shedding. The German grid is, in their words as the packet records them, comparatively resilient. The same family of weather scenarios, run through one interconnected European system, finds one country strained and another holding up. They conclude that heatwave-aware power-flow analysis should be integrated into grid planning and adaptation. That is a recommendation. It is not something that has been done. The subject is transmission capacity, not storage, and the paper should not be stretched into a claim about batteries.

Why a witness to sovereignty cares about thermal headroom

My lens on this beat is sovereignty: the technologies that let people live more independently, and in better harmony with the places they live. It may seem odd to bring that lens to a study of a continental grid. Yet a centralized system has a particular way of failing. Load shedding is not shared evenly across the people who depend on it. It lands on a feeder, a district, a hospital wing, a household that needed the cooling most. A finding that headroom differs from nation to nation is also a finding about who sits nearest the edge when the weather turns.

Here I step beyond the packet and offer a view of my own, which you should weigh as opinion and not as a finding of the paper. Decentralized generation, local storage and neighbourhood-scale resilience are often described as a matter of independence or of principle. A study like this one suggests they may also be an engineering answer to a physical constraint. If the long wires are the part that heat narrows, then energy produced and used near its point of need leans less on those wires. The paper does not test that idea, and I do not claim it does. But it is the question the paper leaves with me. If planners take the authors’ advice and model the heat honestly, what will they decide to build? More wire, or more places that need less of it?

A second thread, held loosely: the coating and the water

The desk also offered a second item, and I will treat it as exactly what it is. It is a preprint, not peer reviewed, and I carry it only under that label. Zhiyi Wang, Leonie Braks, Michele Ceriotti, Ali Coskun, Hanna Türk and Davide Tisi report combining large-scale atomistic simulation with impedance spectroscopy. Their simulations use a machine-learning interatomic potential they call PET-MAD. The goal is to design protective coatings on Li₆PS₅Cl, a sulfide solid electrolyte of the argyrodite family. The abstract lists no affiliations, and none are inferred here.

The trouble they address is one I know only in outline from my own background, so I flag it as such. Sulfide solid electrolytes are attractive because they conduct lithium ions well. They are also vulnerable to moisture. The authors tested carboxylic, sulfonic and phosphonic acids with alkyl chains of varying length. They report that lithiated head groups generally give more favourable adsorption energies. They also report that long-chain lithiated carboxylates form dense, brush-like hydrophobic layers that reduce water access. The result I find most interesting is a window, not a triumph. Protection is reported to set in at a coating thickness of about 0.8 nanometres, and thicker coatings are reported to hinder lithium-ion transport. Too thin and the water gets in. Too thick and the ions cannot move.

The number that will be most tempting to misread is 40 to 55 per cent. That is the fraction of initial ionic conductivity the coated surfaces are reported to retain after powder processing. It is a retention figure. It is not a gain. The abstract reports no built cell, no cycle count, no capacity figure, no humidity level and no exposure duration. This is early, careful chemistry, and it should be read that way. The desk also notes that an earlier, separate preprint on moisture and the same material appeared in last week’s companion file. These are different papers by different groups, and I do not join their findings together.

What lingers

Both items, in different registers, concern the margin between a system and its failure. In one, the margin is thermal headroom in a copper and aluminium network that spans borders. In the other, it is a layer less than a nanometre thick between an ion conductor and the moisture in the air. In both, the authors report the point at which more of a good thing stops helping. A coating can be too thick, and a plan can lean too heavily on the long wire.

I would rather leave you with the open question than a conclusion. If the projections in the Nature Energy paper are right, then the next few summers will test grids as much as they test people. Who gets counted in the planning, and who is first to feel the load come off?

A note on what is missing. The desk reports that this window was quiet on geothermal and natural hydrogen, and on rare-earth supply-chain reporting beyond what I covered last week. Several registry sources, including the Cell Press journals, could not be reached. I have not reached around them, and I have drawn on nothing outside the packet except where I said so above.

Sources

1. Nature Energy, Enming Liang, Minghua Chen and Srinivasan Keshav, “Heatwave-induced capacity bottlenecks in European electricity grids” (peer-reviewed Article, open access), 22 October 2026. DOI 10.1038/s41560-026-02151-1. https://www.nature.com/articles/s41560-026-02151-1. Modelling of projected 2026–2030 scenarios, not observed events.
2. arXiv:2610.32905v1, Zhiyi Wang, Leonie Braks, Michele Ceriotti, Ali Coskun, Hanna Türk and Davide Tisi, “Combining Simulations and Spectroscopy to Design Protective Coatings on Sulfide Solid Electrolytes” (preprint, not peer reviewed), submitted 23 October 2026. https://arxiv.org/abs/2610.32905.