The International Energy Agency put a number on industrial heat that deserves more attention than it got: thermal energy storage built from sand, cement or bricks costs about USD 15–20 per kWh and stores heat up to 1,000 °C, with no critical-mineral supply chain behind it. That is roughly a tenth of what a grid-scale battery system costs per kilowatt-hour. The comparison is not like-for-like, and that is exactly why the economics work: industrial heat never has to be turned back into electricity. The same report puts the scale of the prize at 35 billion cubic metres of EU gas a year and 48 bcm in China — and prices the electricity bill that comes with it at 600 TWh and 1,700 TWh respectively.
The source is the IEA’s Renewables for Industry: Electrification of low-temperature heat and steam, a 95-page study that was first published in late 2025 and reissued in a revised version marked February 2026. It was written before China’s 15th Five-Year Plan was released, and it is not a new document. It is, however, the most complete public accounting of a market that the battery and storage industries have largely left alone — and three of its findings are worth carrying around.
What the IEA Actually Counted
Industry consumes about 30% of global final energy, and most of that is still fossil. But the report deliberately steps away from the sectors everyone writes about. Steel and cement get the attention; the IEA looks at the less glamorous ones — food and beverages, textiles, chemicals, paper, wood products, transport equipment — which depend mainly on low-temperature heat and steam.
Together they account for roughly 70% of global industrial energy consumption. In 2023 they emitted close to 3 gigatonnes of direct energy-related CO₂ — half of all direct industrial emissions — although that is down about 8% since 2013. And despite very different industrial structures, every major economy has converged on the same number: electricity supplies only about 4–5% of industrial heat.
That last figure is the one to hold on to. This is not a technology gap. Industrial heat pumps are commercially established to 150 °C; electric boilers can make steam to 350 °C at around 70 bar; induction furnaces reach 3,000 °C at up to 90% efficiency. All of it is available today. The reason the share is stuck at 4–5% is the price of electricity relative to the price of the fuel it replaces.
Thermal Energy Storage: Why $15–20 per kWh Is Not a Battery Price
The IEA calls thermal storage “the enabling technology” that connects cheap, variable renewable electricity to a factory that needs heat continuously — and the reason is the cost structure.
A thermal store holds heat, not charge. Sand, cement or bricks at up to 1,000 °C, charged during the cheapest hours of the day and discharged as process heat. There is no cathode, no electrolyte, no inverter, and no round trip back to electricity, which is where most of a battery’s losses and most of a battery’s cost sit. Compare it with the last grid-storage price this publication worked out: the three Saudi battery sites we calculated at about USD 144 per kWh of system capacity. Thermal storage lands at roughly a tenth of that — on a thermal kilowatt-hour, which is a different commodity and the reason the gap is real.
The IEA’s modelling backs the practical case. Pairing electric boilers with thermal storage produces steam at a lower cost than an electric boiler alone in every EU member state it examined except Germany — where grid charges that scale with connection size eat the saving. Storage also turns an industrial site into a flexible load: the IEA estimates that if the full technical potential were paired with storage, Germany’s demand-response resource would reach up to 25% of its 2024 peak load, with Ireland similar and the rest of the countries examined at 8–11%. For scale, that is a bigger flexible resource than most national battery fleets.
The projects are small but real. A 4 MWh thermal battery went onto the steam grid at Yara’s Porsgrunn fertiliser plant in Norway in 2020. A Finnish food producer installed 10 MWh in 2023, delivering up to 2,000 MWh of steam a year and cutting energy costs by USD 140,000. The largest is under construction in Vantaa, Finland: Varanto, a seasonal cavern store with 90 GWh of thermal capacity, due in 2028 and described by the IEA as enough to heat a medium-sized Finnish town for a year.
The Two Numbers That Land on the Grid
Here is where an industrial-heat story becomes a power-system story.
| European Union | China | |
|---|---|---|
| Industrial fossil fuel reduction potential | ~3,000 PJ | ~9,000 PJ (35% of industrial heat fuel use) |
| Natural gas avoided | 35 bcm/yr | 48 bcm/yr |
| Additional electricity demand | ~600 TWh/yr | 1,700 TWh (4,800 → 6,500 TWh, +36%) |
| IEA’s own comparison | Germany + the Netherlands, combined | China’s forecast solar PV generation growth to 2030 |
The Chinese figure has a second framing that matters more. The IEA puts the additional 1,700 TWh at 63% of the projected growth in China’s renewable generation between 2025 and 2030 — and says that if it were met from new renewables, Chinese industry could cut emissions by 600 Mt CO₂ a year. In other words, industrial heat electrification is not a competitor to China’s renewables build-out. It is the load that build-out has been waiting for.
The EU number has a harder edge. Six hundred terawatt-hours a year averages about 68 GW of continuous demand, and the IEA’s spatial analysis finds that only around 30% of the new electricity demand sits within 3 km of an existing high-voltage substation. Close to the transmission network is not the same as connected to it. That is why the agency’s second priority action is to anticipate heat electrification in long-term grid planning rather than treat it as a queue problem — the same conclusion its earlier grid work reached, which we covered in our piece on grid-enhancing technologies and which sits alongside the flexibility questions raised by assets like Australia’s 850 MW Waratah battery.
The Barrier Is the Price Ratio, Not the Technology
The report is unusually blunt about what is actually blocking this, and the answer is arithmetic.
Levelised cost of heat is dominated by operating expenditure, so the electricity-to-fuel price ratio decides everything. For an electric boiler, that ratio needs to be close to 1. For a heat pump it can be higher, because a coefficient of performance around 3 buys three units of heat per unit of electricity. The IEA finds industrial heat pumps are already cheaper than gas boilers in several EU states — but the range across countries is 41–74 EUR/MWh, a spread driven mostly by national differences in electricity prices, energy taxation and how network costs are allocated.
That spread is a policy artefact, not a physical one. The energy component alone was about 61% of the industrial electricity price in the EU in 2024, up from 46% in 2019. Finland illustrates what happens when the ratio is favourable: in 2024 electricity was cheaper than natural gas for 6,000 hours of the year.
China’s version of the same problem is starker. The IEA puts the levelised cost of heat for heat pumps in selected Chinese provinces at USD 40–50/MWh, and notes that natural gas boilers — about 22% of industrial heat supply, running largely on imported LNG — already cost around USD 57/MWh in operating terms. Heat pumps beat imported gas today. But against domestic coal, they need a carbon price of roughly USD 20–30 per tonne of CO₂ to break even. The report says it plainly: grid-connected heat pumps are attractive compared with gas, but “struggle while cheap domestic coal is available as a heat source.”
The one lever that changes the maths without policy is self-supply. Connecting industrial users directly to captive solar, wind or hybrid generation could almost halve the cost of steam in the provinces the IEA examined — from USD 70–100/MWh grid-connected to around USD 50/MWh.
What the Report Leaves Out
Two caveats belong next to all of this.
The first is that the numbers are technical potentials, not forecasts. The IEA is explicit that they assume complete or near-complete electrification of the heat supply in the sectors examined, using commercially available equipment. Nothing in the report says this will happen; it says it could, and it prices the inputs. The capex assumptions it publishes are worth reading in that light — an industrial heat pump at USD 1,500 per kW of thermal capacity against an electric boiler at USD 150, plus a grid connection costed at USD 1 million per MW. A 3 MW heat load is a grid project before it is a heating project.
The second is timing. This report predates China’s 15th Five-Year Plan, which explicitly targets long-duration storage and lists batteries for intelligent robots, ships and aircraft among its categories; the industrial-heat policy stack the IEA describes is built on the 14th plan, provincial renewable-consumption mandates requiring industry to source 25–70% renewable power in 2025–2026, and an emissions trading scheme that moves from intensity-based allocation to absolute caps from 2027. China’s policy direction is not in doubt. Its industrial coal price is still the binding constraint, and no report written last year can settle that.
The Bottom Line: The IEA’s Renewables for Industry study prices thermal energy storage at USD 15–20 per kWh — roughly a tenth of the USD 144 per kWh we calculated for a grid battery system — because industrial heat is stored and used as heat and never converted back to electricity. The prize is 35 bcm of EU gas and 48 bcm of Chinese gas a year, and the bill is 600 TWh and 1,700 TWh of new electricity demand respectively. China’s increase equals 63% of its projected renewable generation growth to 2030. Nothing here is blocked by technology; electric boilers reach 350 °C and heat pumps 150 °C today. It is blocked by the electricity-to-gas price ratio, by grid connection cost, and in China by the price of domestic coal.
Notes: All figures in this article are from the IEA’s Renewables for Industry: Electrification of low-temperature heat and steam, first published in late 2025 and reissued in a revised version marked February 2026; it is not a new report and the numbers are the agency’s analysis rather than observed outcomes. The technical potentials are modelled scenarios assuming near-complete electrification of the sectors examined, not forecasts, and the cost figures are levelised costs under the specific assumptions published in the report’s annex. The USD 144 per kWh comparison is our own arithmetic from a separately reported storage contract and is a system-level electrical figure, not directly comparable with a thermal kilowatt-hour; the relationship between the two is explained in the text. The 68 GW continuous-demand figure, the 1-in-10 cost ratio and the project-capacity comparisons are our calculations from published figures. EVsays has not visited any of the projects described, has not tested any equipment and has no relationship with the IEA or any company named. See our editorial policy and correction policy.
Sources & Further Reading
- IEA — Renewables for Industry: Electrification of low-temperature heat and steam (95 pages, CC BY 4.0) — the primary source for every figure in this article, including the global industrial energy shares, the EU and China technical potentials, the electricity-to-fuel price ratio discussion, the thermal storage cost range and the six priority actions.
- IEA — executive summary of Renewables for Industry — the condensed statement of the 35 bcm EU figure, the 600 TWh electricity implication, the 48 bcm Chinese figure, and the 4–5% industrial electricity share that all major economies have converged on.
- IEA — full report PDF — the source for the EU cost analysis (LCOH, the 41–74 EUR/MWh range, the 61% energy component and the Finland 6,000-hour finding), the electro-thermal storage analysis including the German exception, the Chinese provincial cost assumptions, the 30%-within-3km spatial finding, and the project examples in Norway and Finland.
- Energy-Storage.news — “IEA: Thermal energy storage key to industrial heat electrification that could cut EU gas demand by 35 bcm annually” (9 October 2026) — the trade report that surfaced the thermal storage figures. The outlet returned an access error for EVsays at the time of publication, so no link is given.
- EVsays — Saudi Arabia’s First Grid Batteries Broke Ground — the USD 866 million for 6 GWh calculation used here as the electrical-storage comparison.
- EVsays — Grid-Scale Battery Storage: An 850 MW Shock Absorber and a 30-Year Cell — on the two kinds of grid storage asset and what each is paid for.
- EVsays — China’s Battery Industry Five-Year Plan — the 15th Five-Year Plan framework that postdates this report and reframes its Chinese policy chapter.
