Here's the question batteries don't answer well: how do you store energy for many hours, or days, cheaply? Lithium cells are excellent for short bursts but get expensive as you add duration, because every hour of storage means more cells. Storing energy as heat flips that economics — heat is cheap to hold, and you can add hours by adding inexpensive thermal mass.
Rondo Energy's grant US12644396B2, "Thermal energy storage system coupled with thermal power cycle systems" (issued June 2, 2026; inventors John Setel O'Donnell and Yusef Desjardins Ferhani), describes the mechanism. The CPC tags map it cleanly: F28D 20/00 is heat-storage apparatus, and F01K 3/02 is steam-plant arrangements for heat accumulation. Surplus electricity heats a storage medium to high temperature; later, that stored heat drives a thermal power cycle to make electricity (or process heat) on demand.
The way this actually works is a deliberate round-trip through heat. You take cheap, intermittent solar or wind electricity, convert it to high-temperature heat, and bank it in a low-cost medium. When the grid needs power — or a factory needs heat — you discharge through a power cycle. You lose some energy in the conversions, but you gain duration at a fraction of the per-hour cost of chemistry-based storage.
Map it to the grid problem and the niche is specific. The interconnection and deployment stories on the energy desk keep hitting the same wall: short-duration batteries can't cover multi-hour or overnight gaps when renewables go quiet. Thermal storage targets exactly that gap. It is not competing with lithium on fast frequency response; it is competing on the long, cheap hours where batteries get costly — and coupling the store to a power cycle is what makes the discharge dispatchable.
The sober qualifier: round-trip efficiency for heat-to-electricity is lower than for batteries, so the economics depend on cheap input energy and applications that value duration or direct heat. A patent on the coupling mechanism is not a claim that thermal storage wins everywhere. But it documents a real architecture for the part of the duration curve batteries serve worst — which is why this corner of the storage landscape is filling with IP that looks nothing like a battery patent.