We use cookies to store your privacy preferences, understand how you use our site and to improve your experience.

You can at any time manage your cookie choices by clicking the Privacy link at the bottom of each page.

Homepage Energy storage | v1.6 | energy.at-site.be | May 2023

Energy storage - a capacity cost, size and efficiency comparison

Motivation: wanted to quantify how suitable batteries are to store renewable energy for different purposes, and see how this relates to other energy storage techniques.

Solar and wind energy fluctuate, and are often not matched with energy demand. Typically fossil fuel energy sources are used as energy backup when insufficient wind or solar energy is available. Another solution is storing a wind or solar energy surplus for use later on. The main technique to do so in Europe is pumped hydro [13], which provides electrical energy backup for a few hours. The storage need is expected to increase as more solar and wind sources are used. Also other storage options become available at a decreasing cost. This page summarizes the energy storage state of the art, with focus on energy density and capacity cost, as well as storage efficiency and leakage. Power capacity is not considered and can be found in literature [13].

The initial focus of this page was battery energy storage. Later data for comparison of other storage technologies were added. We discuss following topics:


1. Energy quantities

At present fossil fuels dominate the energy mix and when discussing energy storage it is important to speak about energy quantities that matter for the world. Often it is assumed that energy storage will be solved soon since batteries become cheaper and more of them are produced each year. Unfortunately it is observed that production volumes are low for battery energy storage, and most upcoming battery capacity will be used for portable devices and for small transportation applications [1]. For Lithium-Ion batteries a production capacity of 1 TerraWatt-hour (TWh) per year was expected for 2028, over four times as much as the battery production capacity in 2018 [2]. Recent studies indicate a faster ramp up of battery production to about 9 TWh in 2027 [3].

In table 1 we estimate how many minutes of global yearly energy can be stored in 1 TWh of Lithium-Ion batteries gives. We do so for primary and final energy, as well as for electricity for the year 2018, using data from "EU energy in figures, Statistical pocketbook 2020" [4]. Since a single TWh is typically consumed in less than 5 minutes globally, a TWh of battery capacity can only cover a few minutes of global energy consumption before they need to be recharged. Scaling storage capacity up to 10,000 TWh allows to store a month of final energy and several months of electricity.

Table 1: Global energy consumption in 2018, and average storage time for energy storage of 1.0 and 10,000 TerraWatt-hour. Data source - EU [4]
2018MtoeTWh1.0 TWh : minutes of storage10,000 TWh : days of storage
World primary energy14 400 Mtoe168 000 TWh3.1 minutes22 days
World final energy9 940 Mtoe116 000 TWh4.5 minutes32 days
World electricity26 600 TWh20 minutes137 days

For electricity a TWh of battery capacity can on average cover 21 minutes of global electricity needs. Electricity demand is expected to expand significantly though, to replace fossil fuels by electricity for transportation, heating and air conditioning. Contrast this with fossil fuel storage capacity:

To put this in perspective an anticipated production capacity of 1.5 TWh would need over 5 000 years to produce the current storage capacity of oil and gas. And assuming a price point of 100 Euro per kWh this would cost 1 000 000 billion Euros for a storage capacity of 10 000 TWh. One can argue that:

So we conclude that though Lithium-Ion battery production will rise significantly, it can only cover short term storage needs. Still reduced battery prices allow several use cases to reduce fossil fuel consumption, and we discuss these in the next part.

2. Battery use cases

The next table summarizes four applications, assuming a battery storage (capacity) costs of 100 Euro per kiloWatt-hour (kWh).

Table2: battery capacity cost and normalized production volume examples
ApplicationkWhBattery price [100 Euro/kWh]Million units/TWh
E car454 500 Euro22
E bike0.550 Euro2 000
Home battery101 000 Euro100
Home seasonal heating battery2 500250 000 Euro0.40

E car use case: a conventional car uses typically between 50 and 100 kWh fossil fuel for 100 kilometer (km). An electric car (E-car) uses approximately 15 kWh for 100 km. Hence a battery of 45 kWh offers a range of almost 300 km. A production capacity of 1 TWh can sustain production of 22 million such cars yearly, at a capacity cost of 4500 Euro per car battery when the assumption of 100 Euro per kWh holds. If a longer range is desired less cars can be made with this production capacity. Unless significant extra battery comes online, we can expect that E cars will remain a small percentage of the global car fleet the coming decade. A key motivation to electrify transport is to lower carbon emissions. In most parts in Europe electricity generation produces less than 500 gram of CO2 per kWh [8], which boils down to less than 75 gram of CO2 per km. Such levels are not achieved by cars with an internal combustion engine. Of course the size of car has a big influence on the carbon footprint. When electrifying cars we can expect that small to medium cars require less electricity and have a lower carbon footprint than a big sport utility vehicle (SUV).

E bike use case: an argument to motivate the lower range of an E car is that most daily uses need only a limited range. However for short distances an E bike makes even more sense. With a 0.5 kWh battery ranges of 50 up to 100 km can be covered, depending on desired driving assistance and weight. Assuming that electricity generation produces less than 500 gram of CO2 per kWh we can expect in a range of less than 2.5 and 5 gram per km, which significantly outperforms a car. Another advantage is that bikes enable more throughput on a lower amount of road infrastructure and requires less parking space as well. These properties are a significant benefit in cities, such as shown in the city of Copenhagen in Denmark. At present a bike battery is still a costly component, and we can expect that price will become more affordable the coming decade.

Home battery use case: home batteries allow to store solar energy for to spread energy availability from the noon peak for use during most hours of the day, including evening and night, or to compensate for a day with less sunshine. At present the home battery capacity cost is not economical. A size of 10 kWh makes sense, since a yearly consumption of 3600 kWh in a country of is typical, about 10 kWh per day. At a price point of 1000 Euro home batteries become more affordable. With 1 TWh battery capacity 100 million homes can be foreseen with a capacity of 10 kWh. Home batteries are too small however to compensate seasonal fluctuations, which we consider in the next and last use case.

Home seasonal heating battery use case: one thing a home battery cannot do is seasonal storage, for example for heating. Most household energy in Belgium is used for space heating. Most popular energy sources for heating are natural gas or heating oil. Average yearly heating gas consumption is 17 000 kWh in Belgium [7]. Space heating energy consumption is peaking in the winter season, when solar capacity diminishes significantly. For homes using heating oil a small fuel tank typically is 2 000 liter, or about 20 000 kWh. When using electricity with a heat pump more insulation is advisable. Let's assume a heat pump with a coefficient of performance (COP) of 4, and a battery backup of 2 500 kWh, eight times less than an oil tank. Even at a price point of 100 Euro/kWh a seasonal heating battery of 2500 kWh is not affordable: a quarter million Euro. This money is better spent in additional insulation. With 1 TWh of energy storage less than a million homes can be fitted with a seasonal heating battery of 2 500 kWh. Therefore we also consider how batteries compares with other energy storage techniques in the next part.

For the shown use-cases we conclude that from a sustainability and scalability point of view E-bikes or lightweight E-vehicles offer most potential.

3. Storage energy density and capacity cost comparison

Up till now we only considered Lithium ion batteries, but other battery technologies can be used for energy storage, as well as mechanical and thermal storage options. In this paragraph we summarize a few different storage options, focusing on their energy density and storage (capacity) cost. For energy density we use data from Wikipedia [9][10] and cost predictions for 2025 from the HydroWIRES study in 2019 [11]. For thermal storage energy density and capacity cost estimates from the German BVES study in 2016 [12] are used. Table 3 summarizes the different options. For energy density we use the unit Watt-hour (Wh) per kilogram and liter respectively. Compared to crude oil energy density is low. To compare, one kilogram crude oil contains 11 700 Watt-hour of energy and 10 300 Watt-hour per liter, rounded to three significant digits. Another representation is Million cubic meters per TerraWatt-hour (TWh). When this unit is divided by the storage height in meter then we obtain the number of square kilometer to store one TWh. Of the listed storage options lithium-ion battery storage offers the best energy density, second only to flywheels. From a capacity cost perspective we observe that thermal storage offers the cheapest storage, then mechanical storage (excluding flywheels) and then battery power. Water heat storage is the cheapest option, but constrained to space heating and domestic water heating - which have a significant energy footprint in places further away from the equator. Rock and molten salt thermal storage can drive a classical steam based electric power generator due to their elevated heat. Battery and mechanical storage energy can also be converted to electricity.

Table 3: energy storage density and capacity cost comparison
Battery storage [9]Wh/kgWh/literMillion m3 per TWhCapacity cost [11], year 2025
  + Lead-acid batteries47.21566.45319-540 $/kWh
  + Lithium ion batteries100-243250-7314.00-1.36308-419 $/kWh
  + Vanadium redox flow batteries [10]10-2015-2566.7-40.0555-951 $/kWh
Mechanical storage [9]Wh/kgWh/literMillion m3 per TWhCapacity cost [11], year 2025
  + Hydropower/pumped storage: 100m drop0.2730.2733 670106-200 $/kWh
  + Hydropower/pumped storage: 200m drop0.5450.5451 840106-200 $/kWh
  + Hydropower/pumped storage: 500m drop1.361.36734106-200 $/kWh
  + Compressed air at 300 bar13955.618.094-229 $/kWh
  + Flywheel storage1001 4700.674 320-11 520 $/kWh
Thermal storage [12]Wh/kgWh/literMillion m3 per TWhCapacity cost [12]
  + water storage60.0-10060.0-10016.6-10.00.4-10 Euro/kWh
  + stone/rock storage50.0-10070-15014.3-6.6715-40 Euro/kWh
  + molten salt storage40.0-11075-20013.3-5.0025-70 Euro/kWh

To have a better feeling of the area and capacity cost magnitudes of different energy storage options we do a simple back of the envelope computation for one and ten thousand TerraWatt-hour. These are provided in Table 4, keeping in mind that:

While area footprint is significant, most options are doable, with lithium ion batteries offering the smallest footprint. When considering large storage capacities most options are not feasible simply because their capacity cost exceeds the gross domestic product of the world, or total manufacturing capacity. This comparison is a summary, a wider set of energy storage options are considered in [9][11][12]. For storage also efficiency, losses and lifetime are important. These properties are summarized in the next section.

Table 4: Area and capacity cost examples for energy storage capacities of 1 and 10 000 TWh
Storage typeAverage storage height - meter1 TWh area: square km1 TWh capacity cost, billion Euro
Lithium Ion batteries50.272300
Hydropower/pumped storage: 200m drop2092.0100
Compressed air at 300 bar101.80100
Molten salt thermal storage51.040.0
Stone/rock thermal storage51.3325.0
Water thermal storage52.002.00

4. Storage efficiency, losses and lifetime

The previous part looked at density and capacity cost of different storage technologies. Besides this also storage efficiency, losses over time and the lifetime of storage solutions matter also. We use the estimates of BVES [12] for these parameters, which are summarized in table 5. This table lists following properties:

The BVES study dates from 2016, and an estimated lifetime of 10000 charges and discharges of Lithium ion batteries seems optimistic. Typically lithium ion battery capacity halves in less than 1000 full discharges and reloads. Fortunately this issue seems addressed, with modern battery topologies capable of full discharges several thousand of times with a lower degradation [15].

Table 5: Overview of storage efficiency and losses due to energy leakage, data source: BVES [12]. Higher efficiency and lower loss values are better.
Storage typeSystem efficiencyStorage efficiencyLoss/dayLoss/monthLoss/3monthsLoss/6monthsLifetime: cycles (c) or years (y)
Lead-acid batteries--87%-92%0.067%-0.10%2%-3%5.9%-8.7%11%-17%3000c || 20-25y
Lithium ion batteries80%-95%90%-98%0.067%2.0%5.9%11%10000c || >30y
Vanadium redox flow batteries70%-80%--negligiblenegligiblenegligiblenegligibleno cycle limit || 20-25y
pumped storage--75%-80%0.0%0.0%0.0%0.0%>20000c
Compressed air storage60%-70%--0.50%-1.0%14%-26%36%-60%59%-84%30-40y
Flywheel storage80%-95%>80%-95%71%100%100%100%100000c || >15y
Water thermal storage--50%-90%0.50%-2.5%14%-53%36%-90%59%-99%20-4000c
Stone/rock thermal storage--98%2.0%-4.0%45%-71%84%-97%97%-100%>10000c || >20y
Molten salt thermal storage--90%-99%1.0%-5.0%26%-79%60%-99%84%-100%>10000c || >20y

When comparing different technologies we see that few are able to store seasonal energy efficiently. Best suited are pumped hydro and battery storage. These are unfortunately expensive for storing large quantities of energy. Another solution is transforming power to a chemical energy carrier, and in the next part we briefly discuss power to gas.

5. Power to gas

In part one we showed that world wide underground storage capacity of natural gas (or methane) in 2016 was 4000 TWh. The idea of power to gas is to convert electricity first to gas, so that it can be stored affordably for later use. Besides methane also hydrogen is considered for power to gas, since it can be made by electrolysis of water. Gas can be used directly, for example for heating processes, or gas can be transformed to electricity. Power to gas or gas to power conversion incurs conversion losses. In this part we first have a look at storage densities for gasses. Next we consider conversion losses. We do so by using public data of Wikepedia [9][16].

Table 6 shows storage density of uncompressed and compressed hydrogen and methane. Per weight hydrogen stores more than twice as much energy as methane. Still energy storage density is more than a factor three lower than methane, since hydrogen atoms have the lowest weight per volume of all atoms. This implies that over three times underground storage capacity is needed for hydrogen to store an equivalent amount of energy as for methane. For road transport applications uncompressed hydrogen or methane require a too large volume - therefore they are compressed. We show power density of hydrogen compressed to 690 bar and methane compressed to 250 bar. We see that energy density of compressed methane is a factor 2 better than for hydrogen and this at a lower pressure. Hence methane is preferable from an energy density point of view. Also the existing methane (natural gas) infrastructure is much larger than that of hydrogen. Still also hydrogen has a number of advantages, as shown in the next paragraph.

Table 6: Energy storage density for hydrogen and methane, data source: Wikipedia [9]
Gas storage at 25 °C [9]kWh/kgWh/literMillion m3 per TWh
hydrogen uncompressed33.32.8357
methane uncompressed14.910.199.0
hydrogen compressed to 690 bar33.312500.800
methane compressed to 250 bar14.925000.400

Table 7 shows energy conversion efficiencies for power to gas and power to gas to power. From an efficiency point of view hydrogen is preferable over methane. Also methane is a greenhouse gas while hydrogen recombines with oxygen to water when released accidentally. Still in all cases significant losses occur. So the idea is to do the conversion when renewables such as wind or solar produce more electricity than needed, and use the gas when demand is larger than renewable energy production. Price guarantees for wind and solar electricity production can complicate the economics for power to gas. If the price of renewable sources is higher than market price then a power to gas to power cycle will result in expensive power.

Table 7: Energy conversion efficiency of power to gas to power, data source: Wikipedia [16]
PathwayEfficiency [%]Conditions
Electricity to hydrogen64%-77%without compression
Electricity to methane51%-65%without compression
Electricity to hydrogen57%-73%80 bar compression
Electricity to methane50%-64%80 bar compression
Electricity to hydrogen54%-72%200 bar compression
Electricity to methane49%-64%200 bar compression
Electricity to hydrogen to electricity34%-44%80 bar compression, with 60% efficiency to convert hydrogen back to electricity
Electricity to methane to electricity30%-38%80 bar compression, with 60% efficiency to convert methane back to electricity
Electricity to hydrogen to electricity and heat48%-62%80 bar compression and co-generation efficiency of 40% electricity and 45% heat
Electricity to methane to electricity and heat43%-54%80 bar compression and co-generation efficiency of 40% electricity and 45% heat
Electricity to hydrogen to fuel cell27%-36%200 bar compression and fuel cell with 50% efficiency

To summarize we see that methane offers the best storage density, and hydrogen has lower efficiency conversion losses for power to gas. When renewable energy is used to produce these gasses then both are able to significantly reduce greenhouse gas emissions.

6. Conclusion

Lithium-Ion battery production capacity is expected to double from 2021 to 2025. Still these production volumes are estimated to be too small to make a significant dent in world wide energy storage. Using this battery capacity for electric transportation seems to makes most sense. We showed that lightweight transportation means such as E bikes significantly outperform bigger alternatives for short distances. From an energy efficiency point of view these are the fastest pathway to reduce the world wide transport carbon footprint, though it goes against the recent trend of using ever larger SUV's for transportation. For energy storage other technologies outperform batteries from a capacity cost perspective, and most are doable with existing technologies. Still capacity cost is significant when considering thousands of TerraWatt-hour of storage capacity, amounts that are reached easily for storage of conventional fossil fuels. From an energy density perspective lithium ion batteries have an acceptable area footprint. Also conversion efficiency of lithium ion batteries and leakage are favorable compared to other techniques. Emerging technologies are power to gas an gas to power. These allow cheaper storage at the cost of efficiency losses. Still power to gas is relevant for storing energy surpluses from renewable sources for later use, and do so with less carbon emissions than backup power using fossil fuels.


The Battery Series Part 2: Our Energy Problem: Putting the Battery in Context
Battery Megafactory Forecast: 400% Increase in Capacity to 1 TWh by 2028, visual capitalist, October 2018
Visualizing China’s Dominance in Battery Manufacturing, visual capitalist, January 2023
EU energy in figures, Statistical pocketbook 2020, ISBN 978-92-76-19443-9
Global underground gas storage capacity, end 2016
Global oil strategic petroleum reserves, last visited June 2022
VREG, evolution of energy consumption (using Dutch language)
Real-time C02 intensity of electricity generation
Wikipedia, Energy Density, last visited in June 2022
Wikipedia, Vanadium redox battery, last visited in June 2022
K. Mongird et al, Energy Storage Technology and Cost Characterization Report, HydroWIRES, U.S. Department of Energy, July 2019
Bundersverband Energiespeicher (BVES), energy storage technologies - publications, 2016 data - website updated, last visited in August 2023
EU Directorate-General For Energy (2020), Database of the European energy storage technologies and facilities. Publications Office.
World bank global GDP - using current USD
J. E. Harlow et al, A Wide Range of Testing Results on an Excellent Lithium-Ion Cell Chemistry to be used as Benchmarks for New Battery Technologies. Journal of The Electrochemical Society, Volume 166, Number 13, September 2019
Wikipedia, power to gas, last visited in June 2022
Oliver Schmidt, Storage Lab, Projecting the future lifetime cost of electricity storage technologies, last visited in May 2023


May 2023
version 1.6, added storage cost definitions and provided links to LCOS cost literature and latest battery storage data from visual capitalist.
June 2022
version 1.5, simplified table 4, corrected broken links
October 2020
version 1.4, added part 4 on storage efficiency and losses and part 5 on power to gas
July 2020
version 1.3, added Vanadium redox battery and CO2 emission estimates for transportation
version 1.2, added comparison between battery storage, mechanical and thermal storage
May 2020
version 1.1, updated yearly battery production capacity forecast, and added comparison with natural gas and oil energy storage
November 2019
version 1.0, initial version
Ⓒ 2018-2024, energy.at-site.be, all rights reserved - Privacy - Terms of use (v1.5) - Disclaimer - E-mail: energy [a] at-site.be