Two systems can store the same amount of energy and still be completely different engineering propositions. One may fit in a small tank. Another may fill a room. One may be light enough for an aircraft. Another may be perfectly acceptable in a stationary building. The missing quantity is energy density.
Energy density measures how much energy is stored per unit mass or per unit volume. Gravimetric energy density answers “how much energy per kilogram?” Volumetric energy density answers “how much energy per litre or cubic metre?” The distinction matters because some systems are constrained by weight, others by space, and many by both.
Wait, what? The highest energy density is not automatically the best technology
Nuclear fuel has enormous energy density compared with chemical fuels. Liquid hydrocarbon fuels can store far more usable energy per kilogram than present batteries. Yet batteries dominate phones and increasingly road vehicles, while nuclear fuel does not power laptops. Energy density is powerful, but technology choice also depends on conversion efficiency, power, safety, controllability, infrastructure, cost, emissions, rechargeability and scale.
A high-density store is valuable only if the complete system can use it well.
The direct answer
Energy density works as a compactness metric. Divide stored or available energy by mass to obtain gravimetric energy density. Divide by volume to obtain volumetric energy density. These numbers help engineers compare how much storage mass or space is required for a target service.
But energy density should almost never be used alone. A complete comparison also asks how much of the stored energy reaches the useful output, how fast it can be delivered, how easily the system can be replenished, and what containment or conversion hardware is required.
Gravimetric energy density
Gravimetric energy density is usually expressed in joules per kilogram, watt-hours per kilogram or related units. It matters strongly in anything that must carry its own energy store: aircraft, vehicles, portable electronics, robots and human-carried equipment.
If a vehicle needs a fixed amount of onboard energy, lower storage mass leaves more payload capacity. That is one reason liquid fuels have remained attractive for aviation: their energy per unit mass is high and the aircraft consumes the fuel during flight, reducing carried mass over time.
Volumetric energy density
Volumetric energy density measures energy per unit volume. It matters when space is scarce: fuel tanks, battery packs, ships, buildings, industrial sites and dense cities. A lightweight fuel can still be difficult to store if it occupies too much volume.
Hydrogen illustrates this difference. Per kilogram, hydrogen contains substantial chemical energy. At ordinary conditions its gas density is extremely low, so volumetric energy density is poor unless it is compressed, liquefied or chemically bound. Those storage methods require equipment and energy.
Energy density versus power density
Energy density answers how much can be stored. Power density answers how quickly energy can be delivered relative to mass or volume. A storage technology can have high energy density but limited power capability, or excellent power density but modest capacity.
Supercapacitors can deliver high power and cycle rapidly but generally store less energy per kilogram than batteries. Fuels can carry large total energy, but the engine or fuel cell determines how fast that energy can become useful power.
Battery energy density
Batteries store chemical free energy in electrode materials and return part as electrical work. Cell-level energy density depends on chemistry, voltage, active-material capacity and cell construction. Pack-level energy density is lower because a usable system also needs casing, cooling, wiring, sensors, structural support and battery-management electronics.
This distinction matters in comparisons. Quoting the best laboratory active-material value and comparing it with a complete fuel tank or installed battery pack creates an unfair boundary.
Why batteries do not need fuel-like energy density for every job
Electric motors can convert electrical energy into mechanical work with high efficiency. Combustion engines lose a large fraction of fuel energy as heat. A battery vehicle therefore may need much less stored energy for the same wheel service than a combustion vehicle needs in fuel.
This is why comparing battery watt-hours directly with fuel chemical energy can exaggerate the practical gap. The correct comparison follows the chain to the final service.
Liquid hydrocarbon fuels
Petrol, diesel and aviation fuels combine high gravimetric and volumetric energy density with ambient-temperature liquid handling. That combination helped make them dominant in transport. Tanks can be filled quickly, fuel can be moved through pipelines or trucks, and engines can deliver high power.
The disadvantage is the complete chemical pathway: combustion produces carbon dioxide and other pollutants, engines reject substantial heat, and fuels depend on extraction, refining and distribution infrastructure.
Hydrogen
Hydrogen has high gravimetric chemical energy density but poor volumetric density as a gas at ordinary conditions. Compression improves volumetric storage but requires strong pressure vessels and compression work. Liquefaction increases density further but requires extremely low temperature and continuous management of boil-off and insulation.
The storage system therefore includes more than the hydrogen molecule. Tank mass, insulation, pressure hardware, valves and safety systems reduce practical system-level energy density.
Synthetic fuels and ammonia
Hydrogen can be converted into other chemical carriers such as ammonia or synthetic hydrocarbons. These may be easier to store or transport than pure hydrogen, but every additional conversion stage consumes energy and requires infrastructure.
The question becomes a trade-off: sacrifice some round-trip efficiency in exchange for better storage density, compatibility or transportability.
Nuclear energy density
Nuclear reactions involve changes in nuclear binding energy rather than electron-level chemical bonds. The energy released per unit mass of fuel can therefore be vastly larger than for chemical fuels.
But a nuclear power system also requires reactor structures, shielding, cooling, turbines, generators, containment, safety systems and long-term fuel and waste management. Fuel energy density is spectacular; complete plant power density and system complexity are separate engineering quantities.
Food energy density
Biology has its own energy-density trade-offs. Fat stores more chemical energy per gram than carbohydrate and does not bind as much water as glycogen storage. This makes fat efficient for long-term biological energy storage.
Glycogen, however, can be mobilised rapidly and supports different metabolic needs. Organisms therefore optimise across accessibility, power, water, mass and metabolic control rather than simply maximising energy density.
Elastic storage
Springs store mechanical strain energy. They usually cannot compete with chemical fuels for total energy per kilogram, but they can accept and release energy quickly and directly as mechanical work. That makes them useful in watches, launch mechanisms, suspension systems, toys and biological locomotion.
The correct metric depends on the task. A spring can be excellent for a millisecond-scale energy exchange even if it is poor for powering a vehicle for hours.
Flywheels
Flywheels store kinetic energy in rotating mass. Increasing rotation speed can raise stored energy strongly, but material stress also increases. High-performance flywheels therefore use strong materials, careful balancing, low-friction bearings and robust containment.
They excel at high power and frequent cycling but usually have greater standing losses than chemical stores over long durations.
Gravitational storage
Gravitational energy per kilogram is modest under ordinary terrestrial height differences. A kilogram lifted 100 metres stores only about one kilojoule. This low gravimetric energy density seems unattractive—until enormous masses of water are available.
Pumped hydro succeeds because reservoirs can move millions or billions of kilograms. Low energy density can be overcome by scale when geography and infrastructure permit.
Thermal energy density
Thermal storage can hold energy by raising temperature or changing phase. Water stores sensible heat through temperature change. Ice stores latent heat during freezing and melting. Molten salts can hold high-temperature heat for power or industrial systems.
Thermal stores may have modest energy density compared with fuels, but they can be highly effective when the final service is heating or cooling because no round trip through electricity is needed.
Compressed air
Compressed-air systems store mechanical and thermodynamic energy in pressurised gas. Practical energy density is limited by pressure vessel or cavern constraints and by thermal effects during compression and expansion.
Large underground spaces can compensate for low volumetric energy density, making compressed air more plausible at grid scale than in small mobile devices.
Energy density in aviation
Aviation is unusually sensitive to gravimetric energy density because the aircraft must lift its energy store. Extra storage mass increases required lift and often increases energy consumption, creating a feedback loop.
This is why battery-electric aviation is much harder for long-range large aircraft than for small aircraft or short routes. Improvements in battery specific energy help, but aircraft design, reserve requirements, payload and charging infrastructure also matter.
Energy density in road transport
Cars tolerate more storage mass than aircraft because they do not need to lift it continuously against gravity. Regenerative braking also allows some kinetic energy to be recovered. This makes batteries practical despite lower gravimetric energy density than liquid fuels.
Heavy trucks face stronger payload and range constraints, so duty cycle becomes important. Vehicles that return to a depot can accept different storage trade-offs from those requiring very long continuous range.
Shipping
Ships are less sensitive to storage mass than aircraft but can be constrained by tank volume and cargo space. Fuel choice therefore involves volumetric energy density, storage temperature, safety, engine compatibility and emissions.
A lower-density fuel may reduce cargo capacity or increase vessel size even if its gravimetric performance is acceptable.
Portable electronics
Phones, laptops and wearables need high enough energy density to provide long runtime without becoming heavy or bulky. Battery improvements therefore translate directly into thinner devices, longer runtime or more computational capability.
Safety becomes especially important because a compact device places substantial chemical energy close to the user. Thermal runaway control, separators and battery-management systems are part of the density problem.
Energy density and safety
Concentrated energy can produce concentrated consequences if containment fails. A high-energy battery, pressurised gas cylinder, fuel tank or flywheel can release energy rapidly under fault conditions.
High energy density therefore often increases the importance of controlled release, thermal management, structural containment and fault isolation. Density is an advantage and a hazard parameter at the same time.
Energy density and refuelling speed
A liquid-fuel pump can transfer a large amount of chemical energy in a few minutes because the fuel is dense and the flow rate is high. Fast-charging a battery requires equally impressive electrical power if the same energy is to be transferred in the same time.
This reveals another trade-off: the stored energy density may be only part of the problem. The infrastructure must also deliver the required power safely.
System-level versus material-level numbers
Material-level energy density can be much higher than installed-system energy density. A battery active material needs packaging and electronics. Hydrogen needs tanks. Nuclear fuel needs a reactor. Thermal storage needs insulation and heat exchangers.
Good comparisons therefore use equivalent boundaries. Compare complete usable systems to complete usable systems, not an active molecule to an installed machine.
Usable energy density
Not all nominal stored energy is available to the user. Batteries preserve reserve margins to avoid damage. Fuel cells and engines have conversion losses. Thermal stores may not be able to cool to ambient without losing useful temperature quality. Mechanical systems may have operating limits.
Usable energy density should therefore reflect the actual operating window and conversion chain for the service being compared.
Three worked reasoning examples
1. Battery versus petrol car
Petrol carries much more chemical energy per kilogram than a battery stores electrically. But the combustion engine loses much of that energy as heat, while an electric drivetrain converts a larger fraction to wheel motion. The practical vehicle comparison must include tank or pack mass, drivetrain efficiency, regenerative braking, range and refuelling or charging infrastructure.
2. Hydrogen aircraft concept
Hydrogen performs well gravimetrically, but storage volume is difficult. Cryogenic tanks need insulation and have shapes unlike thin conventional wing tanks. The aircraft may save fuel mass but lose internal volume or require a different fuselage architecture. Gravimetric and volumetric density pull in different directions.
3. Pumped-hydro reservoir
Energy stored per kilogram of water is low. Yet an enormous water mass and substantial height difference create very large total capacity. The site can then deliver grid-scale power through turbines. Low density does not prevent usefulness when stationary scale is available.
Common misconceptions
- Energy density and power density are not the same.
- Gravimetric and volumetric energy density can rank technologies differently.
- Fuel energy density should not be compared directly with battery stored energy without accounting for conversion efficiency.
- Material-level numbers are not the same as installed-system numbers.
- The highest energy density is not automatically the safest or cheapest option.
- Hydrogen’s high energy per kilogram does not imply high energy per litre.
- Low-density storage can still work at very large stationary scale.
A universal energy-density audit
- Define the final service and required usable energy.
- Determine whether mass, volume or both are constrained.
- Compare gravimetric and volumetric energy density.
- Use system-level rather than material-only values.
- Include conversion efficiency to useful output.
- Check power and refuelling or charging rate.
- Include containment, cooling and safety hardware.
- Measure usable operating window, not only nominal capacity.
- Consider lifecycle, cost and infrastructure.
- Choose the technology whose complete trade-off fits the service.
How energy density fits the wider Energy series
Energy density connects storage, chemical energy, nuclear energy, elastic storage and gravitational storage.
The deeper lesson is that the same number of joules can be physically easy or extremely difficult to carry depending on how concentrated they are. Energy density tells us how tightly energy fits into matter and space; engineering decides whether that compactness is actually useful.
How Energy Works | Main Series
How Energy Storage Works · How Power Works · How the Energy Transition Works