Hydrogen vs Battery for Stationary Energy Storage

Energy storage is needed to solve the variability of renewable energy. Main contenders are hydrogen and batteries. These can be built in regions that don’t have large lakes or mountains for leveling hydropower.

Which ones should we build? To get a feel for the question, we are comparing two existing systems.
Note that these are rough back of the envelope calculations!

The comparison

The world’s biggest battery energy storage at the time of writing is the Crimson Storage System in Southern California, delivered by Canadian Solar.

We compare it against NASA’s new liquid hydrogen tank at Kennedy space center in Florida.


LabelHydrogen tankBattery farm
Energy content6.5 GWh[1]1.4 GWh
PowerN/A [2]350 MW
Round Trip Efficiency~30% [3]88% [4]
Cost$15 million[5]N/A
Table comparison

[1] Energy content for hydrogen storage assumes 50% fuel cell conversion efficiency to electricity
[2] NASA uses this to fill rockets that burn the propellants, it is not converted to electricity.
[3] https://en.wikipedia.org/wiki/Hydrogen_economy
[4] Canadian Solar (90% at start, 87.5% at 20Y)
[5] This is just for the tank, doesn’t include the electrolyser equipment (and there are no fuel cells)

The hydrogen ball’s energy content is impressive, but the round trip efficiency for hydrogen is bad. Cost comparison is not available.

Other differences


The technology is widely different. The hydrogen storage is a huge ball with a vacuum jacket. There is loose small glass ball insulation in the vacuum. The liquid hydrogen is kept at 22 K, while ambient air on the outside can be 300 K (+27 C). This is done with a helium cryocooler and it eliminates boil off (older systems had boil off).

One huge advantage of the hydrogen system is that because of the potential effects of the square-cube-law, there is proportionally less insulation needed if the tank gets bigger. Therefore it would make sense to create one big tank instead of two medium sized ones.

The technology has some commonalities with LNG storage, though LNG is much easier at 90 Kelvin and no embrittlement and less leakage issues.


The batteries at Crimson storage are provided by Canadian Solar. On the outside, regular looking shipping containers stand on short stilts in the desert. There seem to be mostly energy storage containers with probably a minority of different transformer containers in between.

The energy storage container looks simple. Inside, the battery modules are contained in racks, connected with thick wires. They can be serviced by people.
There don’t seem to be much scale advantages. More storage is gotten by adding more units.

Storage systems don’t take so much space compared to generation systems like wind or solar so it’s not such an important question.


Hydrogen leaks are relatively harmless since it’s so light and rises up, even fires (outdoors) are not that dangerous compared to heavier fuels.
The battery systems are isolated into containers with a clear gap, meaning a fire in one container doesn’t cause a cascading failure into others.


Hydrogen might work for absolutely humongous, stadium-sized tanks, one per industrial or energy production area. It could be coupled with direct industrial usage of hydrogen (Fertilizer and steel, aluminum, nickel or chemical production), avoiding the fuel cell inefficiency (and cost). They could be constructed of relatively ordinary materials like aluminium, polystyrene and even plywood. Maybe some of the innovations from the LNG boom could be adapted. Because of the cost of electrolyzers and fuel cells, the power would be optimized to be low and it would make sense as a large longer term energy storage. Would seasonal energy storage be possible with very large hydrogen storage systems?

Large battery storage systems can be constructed in a modular fashion, and there are multiple levels of hierarchy. Some of the components can be interchangeable. The technology is developing rapidly. One could construct a facility from containers from multiple integrator contractors. A container could usem cells from multiple manufacturers. Current batteries are more suited for short term storage because of the cost and the way the technology is developed (for electronic devices and cars). Every city and solar or wind farm could have these.

Post note on tank size

Square cube law would favor larger tanks. On the other hand, if vacuum insulation is used, that means the tank is a pressure vessel and the force grows with area, and thus thicker material must be used for the same strength. Meaning the pressure vessels would have the same mass proportion no matter what the size. That takes away some of the advantages of the square-cube law for the whole tank. On the other hand, if tanks get larger, one could potentially get away with less insulation and not have to pull vacuum. Complicated questions.

What would a truly humongous tank be like? There’s already a good model, the famous Globe n or Avicii arena in Stockholm.

If we assume 1 m insulation thickness for both, Globen would have 4.7 times the effective tank radius and thus 100 times the volume of the NASA tank, with a whopping capacity of 650 GWh

This would mean two weeks’ output of the Olkiluoto 3 EPR 1.6 gigawatt nuclear power plant. Not enough to solve a small nation’s seasonal energy storage yet, but it would have a sizable effect.

In comparison, Germany’s natural gas storage capacity is 250,000 GWh or about 400 Hydrogen Globens.


This entry was posted in Uncategorized. Bookmark the permalink.

Leave a Reply

Your email address will not be published. Required fields are marked *