Storing heat and cold

Water, salts, rock, ice and buildings as energy stores. Theory, formulas, worked examples and applications.

Sometimes it is absurd to convert twice

Why store heat directly instead of going through electricity

If the final service needed is heat or cold — heating a building, domestic hot water, industrial refrigeration, or steam to turn a turbine —, converting that energy first into electricity, then storing it chemically in a battery, and finally converting it back into heat through an electric heater is, in many cases, an unnecessary detour that loses efficiency at every conversion. When the final service is thermal, it can be much simpler and more efficient to store that heat or cold directly, without electricity as an intermediary. A hot-water tank, a molten-salt tank, a block of rock, concrete, specialized phase-change materials, or even ice, receive thermal energy when it is convenient to generate or capture it, and return it later, when needed, without that double conversion.

Materials matter as much as design

Each thermal storage material has its own profile of advantages and limitations. Water is cheap, abundant and easy to handle, but needs considerable volume to store significant amounts of energy, and requires good insulation to avoid losing heat quickly. Molten salts — used in large-scale concentrated solar power plants — can operate at very high temperatures, allowing much more energy to be stored per unit of volume than water, but can corrode their containing materials or solidify if temperature falls below a critical threshold, complicating operation. Ice allows shifting refrigeration's electricity consumption in time — producing it in the early morning, when electricity is usually cheaper and cleaner, and using it to cool during peak-demand hours. Even a building's own thermal mass — its walls, its structure — can smooth indoor temperature swings throughout the day with no separate tank at all, if the architectural design makes good use of it. Pumps, heat exchangers, pipes and proper insulation form, in any case, the complete system that makes it possible to move that heat from one point to another and preserve it for the time needed.

A large-scale example: molten salts in solar plants

The most advanced concentrated solar power plants, such as Gemasolar in Spain or Crescent Dunes in the United States, use a field of mirrors to concentrate sunlight onto a central tower, heating molten salts to temperatures of several hundred degrees. Those hot salts are then stored in an insulated tank, and can be used hours later — even throughout the whole night — to produce steam and generate electricity exactly as if the sun were still shining. This ability to "store sun for use at night" is precisely what distinguishes concentrated solar power with thermal storage from conventional photovoltaic solar, which stops generating as soon as the sun sets.

Heat losses are inevitable, but manageable

Any thermal storage system loses heat over time, simply because heat always tends to flow from where there is more temperature to where there is less. The greater the temperature difference between the storage material and its surroundings, and the worse the tank's insulation, the faster that stored energy quality is lost. Even so, many of these materials are abundant and cheap, can withstand a great many thermal charge-discharge cycles without significant degradation — unlike a chemical battery, which does degrade with every cycle —, and for the specific thermal service they provide, they entirely avoid the need for a more expensive electrochemical battery with its own chain of critical materials.

The cheapest storage starts by reducing losses, not enlarging the tank

Before considering building or enlarging a thermal storage tank, the first economic and environmental priority is always improving the existing system's insulation and control: reducing avoidable losses is almost always cheaper than storing more energy to compensate for those same losses. When comparing thermal storage options, it is worth looking at the useful temperature needed, the real storage duration required, available space, necessary materials, corrosion risk, long-term maintenance, and the final destination of those materials at end of life. And it is always worth remembering that a thermal kilowatt-hour is not automatically equivalent to an electrical kilowatt-hour: it provides a different service, and comparing the two without that distinction leads to wrong conclusions.

The storage system starts by reducing losses

In favor: avoids the double electricity-heat-electricity conversion when the final service is thermal; many materials are cheap, abundant and withstand a great many cycles without degrading like a chemical battery.

To watch: heat losses are unavoidable and accelerate with worse insulation and a larger temperature difference from the surroundings; materials like molten salts can corrode equipment or solidify outside their design temperature range.

Before enlarging a tank: improve the existing system's insulation and control first — reducing avoidable losses is usually cheaper than storing more energy to compensate for them.

Do not confuse a thermal kWh with an electrical kWh. They provide different services, and comparing them without that distinction leads to wrong conclusions about which technology suits each case.

The numbers, with their source

Ten hours of electricity, stored as heat

The Crescent Dunes plant (Nevada, USA) has a power output of 110 MW and a thermal storage capacity of 10 hours at full power, equivalent to 1.1 GWh of energy stored as molten salts heated above 550°C. That capacity lets it keep generating electricity many hours after sunset, and even operate exclusively at night when the grid requires it — something impossible for a photovoltaic solar plant without separate storage. The Spanish plant Gemasolar, at 50 MW, was a smaller-scale precursor of this same central-tower, molten-salt technology.

Sources: Interconnection.fyi — Crescent Dunes Solar Energy plant data; Inside Climate News — 24-Hour Solar Energy: Molten Salt Makes It Possible.

Why this does not compete with a battery, it competes with the night

The relevant comparison for a concentrated solar plant's thermal storage is not "is this better than a battery?" but "does this allow delivering electricity once the sun has set, at a reasonable cost per additional hour of autonomy?" Ten hours of storage at full power is a timescale far beyond what is economical today for most large-scale chemical batteries, precisely because the storage material — molten salt, relatively cheap — does not degrade with every cycle the way a chemical cell does.

What this table is missing

This figure describes a single reference plant; the real performance of a specific molten-salt system depends on the tower's design, the exact salt formulation used, the storage tank's insulation, and local climate — there is no universal "hours of storage" figure that applies to every concentrated solar power plant.

Sources