Finland · Thermal storage
The same hot stone can last four days or a month
A storage capacity becomes a duration only when you say how quickly the heat is being used.
- Capacity and output answer different questions100 MWh describes the heat store. The 1 MW rating describes its heat-delivery power.
- Summer demand changes the arithmeticA lighter load stretches the same stored energy. It does not make the vessel larger.
- Cheap charging is one part of the billBuying electricity at favourable hours does not remove heat losses, equipment costs or network charges.
A Finnish town has a heating asset that sounds almost wilfully unglamorous: a large quantity of hot, crushed stone. Yet the interesting part of Pornainen’s sand battery is what the stone allows the heating operator to postpone. Electricity can be purchased at one time; warmth can be delivered at another.
The project also presents a small numerical puzzle. Its advertised storage capacity is 100 megawatt-hours, and its heat output is one megawatt. Divide one by the other and you get roughly four days. So how can the same installation be described as holding a winter week or a summer month of heat?
The answer makes this particular project more interesting, rather than less credible. Those descriptions use different rates of consumption. Understanding that difference reveals what the installation actually buys a district-heating system—and what its first-year results still leave unanswered.
The sand battery is full of crushed soapstone
The name describes a storage concept, not a requirement to fill every installation with beach sand. Polar Night Energy says the Pornainen store uses about 2,000 tonnes of crushed soapstone, including a by-product of Tulikivi’s fireplace production. Its material explanation is a useful correction to the image the name first puts in your head.
There are no lithium-ion cells hidden between the stones. Electricity produces heat in resistive elements; circulating air carries that heat through the store. During discharge, heat passes through an exchanger into the district-heating system. The air circuit and the district water circuit remain separate. Polar Night Energy describes the operating principle in its September 2025 efficiency explainer.
This is a different category from the solid-state batteries discussed in our electrochemical-storage guide. Here, the useful output is already warmth. Comparing the two solely by how much electrical energy comes back out would overlook the service the thermal store was built to provide.
A week of heat needs a stated rate of consumption
The January 2025 announcement naming Elisa as the reserve-market operator specifies a 1 MW heat output and 100 MWh capacity. The distinction is simple: megawatts measure a rate; megawatt-hours measure an amount. The number of hours cannot be read from either one alone.
Duration in hours = stored heat in MWh ÷ average heat demand in MW.
Start with a full, idealised 100 MWh store. Assume no recharging, no heat loss and no energy held in reserve. Now change only the constant heat load:
| Heat load | Calculation | Duration |
|---|---|---|
| 1.00 MW | 100 ÷ 1.00 | 100 hours · 4.2 days |
| 0.60 MW | 100 ÷ 0.60 | 167 hours · 6.9 days |
| 0.14 MW | 100 ÷ 0.14 | 714 hours · 29.8 days |
Author’s calculations, rounded. The loads are illustrative, not measured Pornainen demand. Actual duration depends on usable energy, changing demand, losses and operating decisions.
Nothing in the third row requires better stone, a bigger tank or a surprising law of physics. Less heat leaves each hour. Conversely, the first row does not disprove the longer-duration descriptions: it asks the store to deliver continuously at its rated output.
Capacity and power also impose different constraints. A store may have plenty of energy remaining while being unable to meet a momentary demand above its output rating on its own. To assess a real heating system, you need its demand profile and the role of the other heat sources—not just a headline number of days.
The first year changed when electricity was bought
In its 11 June 2026 first-year report, Polar Night Energy describes the store as covering approximately a winter week or summer month of local heat demand. It reports efficiency above 85%, a 70% emissions reduction and uninterrupted district-heating supply. These are supplier/operator-reported results; the article does not provide an independently audited dispatch dataset.
The commercially revealing claim concerns charging. A representative of the operator’s owner says the electricity was purchased at prices 70–80% below the average spot-market price. The report attributes the result to charging at favourable hours and describes participation in Fingrid’s reserve markets.
That comparison concerns electricity procurement, not a 70–80% reduction in residents’ heating bills. The distinction matters because a heat bill pays for a system, not just the energy entering one component.
- Conversion and storage losses affect how much useful heat reaches the customer for each unit of purchased electricity.
- Connections, equipment and operation still have to be funded, even when some charging hours are exceptionally cheap.
- Reserve-market income is a separate contribution to the business case, not a reason to treat every stored unit of heat as free.
The first-year account is encouraging evidence of useful operation. A full economic comparison would additionally need the complete cost boundary, the alternative heat source, the amount of useful heat delivered and the period over which equipment costs are recovered. A favourable spot-price comparison cannot supply those missing quantities.
A town can use storage without emptying it once a month
The duration table imagines one charge followed by uninterrupted discharge. An operator does not have to run that experiment every time. A partly charged store can receive more heat when suitable electricity hours arrive, then continue serving demand between charging opportunities.
That creates a more useful question than “How long until it is empty?” Ask how much freedom the stored heat gives the operator to avoid buying electricity at an inconvenient moment. A larger energy buffer can provide more freedom; a higher heating load spends it faster. Charging power and the available cheap hours constrain how quickly that freedom can be restored.
The manufacturer’s technical explanation also distinguishes active-use losses from losses during storage. Holding warmth for longer is not costless. Its scenarios are not evidence that this installation captures a whole summer’s energy and preserves it unchanged through winter.
This energy pathway runs in the opposite direction to many ambient-energy harvesting systems, which seek usable electricity from small surrounding energy flows. Pornainen deliberately consumes electricity to deliver a larger-scale heat service. There is no need to turn the warmth back into electricity to make that service valuable.
Heat delivery and electricity recovery remain different projects
Polar Night Energy’s separate Sand to Power pilot in Valkeakoski runs across 2025–2027 and investigates converting stored heat back into electricity. That development programme should not be silently added to the capabilities of the Pornainen district-heating installation.
Likewise, reported heat-supply reliability does not establish stand-alone operation during a power cut. Fans and controls use electricity; assessing outage resilience requires information about backup power and the wider heating network. The published results do not justify treating the installation as an electricity supply for the town.
The habit of asking exactly what a battery-related number describes is useful elsewhere too. Our guide to battery-passport and used-EV health data separates recorded information from a current condition assessment. Here, the crucial distinctions are between stored heat, delivery power and the demand being served.
Pornainen’s achievement does not need an exaggerated promise of a winter powered by sand. A mundane material is giving a heating system more choice over when to purchase energy. The hot stone is the buffer. The town’s changing demand determines how much time that buffer buys.
Questions about the Pornainen sand battery
Does the 100 MWh figure mean 100 MW of output?
No. MWh is an amount of energy; MW is a delivery rate. A 100 MWh store supplying 1 MW continuously lasts 100 hours in a loss-free, no-recharge calculation.
Is this a way to store electricity for household sockets?
The installed Pornainen system delivers heat. Electricity recovery is the subject of a separate pilot, not a demonstrated function of this district-heating store.
Are the three durations in the table measured performance?
No. They are transparent calculations using three assumed constant loads. They explain how consumption changes duration; they do not reconstruct the site’s operating history.
Primary sources and calculation notes
Sources checked 6 September 2026
- Elisa selection announcement, January 2025 — Capacity, heat output and planned reserve-market operation.
- First-year report, June 2026 — Supplier/operator account, not an independent audit.
- Efficiency explanation, September 2025 — Heat transfer, auxiliary electricity and loss boundaries.
- Storage-material explanation — Pornainen’s crushed-soapstone feedstock.
- Sand to Power pilot — Separate electricity-recovery development programme.
- The duration table divides 100 MWh by each stated load, then divides hours by 24. It ignores losses, recharging and reserve margins; rounded values are illustrative.
