Negative prices change which industrial heating technology wins
Serguey Asael Shinder2 min readResearchers in China have compared the life-cycle economic performance of thermal energy storage (TES) against heat pumps (HP) for industrial heating under negative electricity price conditions, pv magazine reported on 23 September 2026.
The model varies three things: how often negative prices occur, how far below zero they go, and what the equipment costs. Its finding is that the preferred option depends strongly on the temperature the process requires, with storage gaining the advantage for steam production.
The authors set out what was missing before: prior techno-economic comparisons between heat pumps and thermal storage for industrial heating mostly assumed conventional time-of-use tariffs, without systematically quantifying full life-cycle performance under negative prices.
What it means
The two technologies answer different questions, and negative prices sharpen the difference. A heat pump is an efficiency machine: it multiplies each unit of electricity into several units of heat, and that multiplier is what justifies its capital cost. Thermal storage is a timing machine: it converts electricity to heat at one-to-one at best, and earns its keep entirely by choosing when to buy.
When power is merely cheap, efficiency wins — the heat pump needs less of it. When power is priced below zero, efficiency stops being the point, because using more of something you are paid to consume is not a penalty. What remains is whether you can absorb energy in the hours it is free, which is what storage does and a heat pump sized to a process cannot.
The temperature finding follows from the same physics rather than from economics. Heat pump coefficient of performance falls as the required output temperature rises, so at steam temperatures the multiplier that justified the heat pump has largely gone — and with it the argument against a simpler device that stores heat.
⚠️ This is a model, not an installation. It is sensitive to assumed equipment costs and to how often negative prices are assumed to occur, and neither of those is a constant across markets. The useful part is the shape of the answer: the more hours a grid spends below zero, the more the question stops being about efficiency at all.