# Hydrogen storage **Hydrogen storage** is the problem of keeping [[Hydrogen|hydrogen]] somewhere until it is needed. It is the hardest part of using hydrogen as a fuel, and the reason is a single mismatch: hydrogen has the best specific energy of any chemical fuel, about 33 kWh per kilogram, and one of the worst [[Energy_density|energy densities]] per unit volume of anything that burns.[^murphy-app] A kilogram of it is a wonderful thing to carry; a cubic metre of it, at ordinary [[Pressure|pressure]], is almost nothing at all. Every storage method on this page is an attempt to make hydrogen denser, and every one of them costs either [[Pressure|pressure]], cold, mass, or a share of the fuel's own energy. The microsim on this page is live and reused from the Thury Compendium set. In it the reader switches between the three established routes for the same kilogram of hydrogen — compressed to 700 bar, liquefied, or moved by pipeline — and reads two numbers for each: the volume it occupies and the loss it incurs. The governing relation behind the compressed case is the real-gas equation of state, p·v = Z·R·T, where the compressibility factor Z = p·v/(R·T) tends to 1 only as pressure falls toward zero and rises well above 1 at high reduced pressure — so a 700-bar tank holds appreciably less than the [[Ideal_gas_law|ideal-gas law]] would promise.[^yan-gas] On the [[Energy]] flagship this article sits in Part VII — Energy transfer, section *Hydrogen as a store* (row E65), the third child of the C42 storage group after the root [[Energy_storage]] and its short-duration counterpart [[Flywheel_energy_storage]]. Hydrogen holds the opposite corner of that group's Ragone plane: seasonal duration, very high specific energy, poor round trip. The chain it belongs to is set out on [[Hydrogen_economy]]; the element itself is the [[Hydrogen]] page and is not restated here. ## Established technologies Two methods carry essentially all the hydrogen stored today, and the sim compares them directly. Both take the same molecule and attack the volume problem from opposite ends — one with force, one with cold. ### Compressed hydrogen Compression is the incumbent because it needs no new physics, only a strong tank. At 700 bar and 288 K the ideal-gas law gives a density of 58.9 kg/m³, using the specific gas constant R = 4.124 kJ/(kg·K) that follows from hydrogen's molar mass of 2.016 g/mol (derived). The real figure is lower: at that reduced pressure the compressibility factor is well above 1, because the molecules are close enough for repulsion to dominate, and Z rises rather than falls.[^yan-gas] Ignoring that correction is, in the textbook's phrase about the ideal-gas model generally, a common mistake.[^yan-gas] The volumetric penalty is best seen against a liquid fuel. Taking 58.9 kg/m³ as an optimistic ceiling, 700-bar hydrogen carries 1.94 kWh per litre, while a gallon of propane holds about 91,500 Btu, which works out to 7.08 kWh per litre — some 3.6 times more (derived).[^murphy-therms][^murphy-app] A car's five-kilogram tank is therefore 165 kWh of fuel in at least 85 litres of pressure vessel, before the vessel's own walls are counted. The tank is the real object of the engineering. A 700-bar vessel is a filament-wound [[Carbon_fiber_reinforced_polymer|carbon-fibre]] [[Composite_material|composite]] over a gas-tight liner, and it is sized by [[Fatigue_(material)|fatigue]] over thousands of fill cycles rather than by a single burst pressure. Filling also heats the gas, so a fast fill must be pre-cooled or the tank will be short-filled when it settles. ### Liquefied hydrogen Liquefaction attacks the same problem with [[Cryogenics|cryogenics]]. Hydrogen cannot be liquefied by pressure alone at any ordinary temperature: it must first be brought below its [[Critical_point_(thermodynamics)|critical temperature]], which lies only a few tens of kelvin above absolute zero, and that requires a refrigeration plant rather than a compressor.[^yan-crit] The reward is smaller than intuition suggests — liquid hydrogen is denser than 700-bar gas, but not by the order of magnitude that liquefying a normal gas would give. The costs are three. Liquefaction consumes a large fraction of the fuel's own energy content. The vessel must be a vacuum-jacketed cryogenic dewar, not a pressure tank. And the store boils off continuously, because no insulation is perfect and the [[Latent_heat|latent heat]] of vaporisation of hydrogen is small, so any heat that leaks in removes fuel. Liquid hydrogen therefore suits things that are used soon after filling and tolerate the plant — launch vehicles above all — and suits long-term storage badly. ## Chemical storage Chemical storage abandons the idea of holding hydrogen as hydrogen. Instead the gas is reacted into a compound that is liquid or solid at ordinary conditions, stored as that compound, and released again by heating it or reacting it. The gain is volumetric density and easy handling; the cost is that both directions of the reaction have an enthalpy to pay and a [[Catalysis|catalyst]] to find, and the release step is usually the slow one.[^kerlin-hydrogen] ### Hydrogenation of CO₂ Reacting hydrogen with carbon dioxide gives liquid carriers — formic acid, methanol, or on to methane — storable in existing tanks and pipelines. What is stored is the hydrogen; the carbon is a reusable handle. The thermodynamics is ordinary [[Gibbs_free_energy|Gibbs]] accounting: hydrogenation releases energy, so the release step must be driven, and the handling advantage is paid for in round-trip efficiency. ### Metal hydrides Many metals and [[Alloy|alloys]] dissolve hydrogen into their [[Crystal_structure|lattices]] and form hydrides, holding hydrogen atoms — not molecules — between the metal atoms. The volumetric density that results can exceed liquid hydrogen's, because the lattice packs the atoms more tightly than the liquid does, and the store operates at modest pressure, which removes the tank problem entirely. The obstacles are mass and heat. The metal is heavy, so the *gravimetric* fraction is small even where the volumetric density is excellent, which inverts hydrogen's one great advantage. And because absorption is exothermic, desorption is endothermic: the store heats when filled and must be heated to discharge, so a hydride tank is a [[Heat_exchanger|heat exchanger]] as much as a vessel, with its release temperature set by the hydride's formation [[Enthalpy|enthalpy]]. [[Le_Chatelier's_principle|Le Chatelier's principle]] is the design rule: a more stable hydride stores more and releases less willingly. ### Aluminium [[Aluminium|Aluminium]] is used differently: not as a reversible hydride but as a consumable reductant that liberates hydrogen on contact with water, leaving an oxide behind. That gives hydrogen on demand with no pressure vessel at all, but the reaction is one-way — the store is regenerated in a smelter, not at a filling station — so it suits niches where carrying metal beats carrying gas. ### Magnesium [[Magnesium|Magnesium]] hydride is the type case of the hydride trade. Magnesium is light and cheap, and its hydride's gravimetric capacity is among the best of the simple binary hydrides — but the hydride is stable, so release needs temperatures far above what a vehicle's waste heat can supply. Work on [[Magnesium_alloy|magnesium alloys]] and [[Nanomaterials|nanostructuring]] attacks that from two directions: destabilise the hydride, or shorten the [[Diffusion|diffusion]] path so kinetics stop being the limit. ### Alanates-based systems Complex hydrides such as the alanates sit between the binary hydrides and the organic carriers: salts of an aluminium–hydrogen anion with a light cation such as [[Sodium|sodium]], holding more hydrogen per kilogram than a simple metal hydride. Release proceeds in stages, and it was the finding that traces of a transition-metal [[Catalysis|catalyst]] make those stages reversible at practical conditions that moved them from curiosity to candidate. ### Organic hydrogen carriers A liquid organic carrier is an aromatic molecule hydrogenated to a saturated one and dehydrogenated to release the gas again. The stored fluid is an ordinary ambient liquid, so it can use existing tanks, tankers and pumps — the largest practical advantage any storage route offers. The endothermic, catalytic release step decides the concept's efficiency, since the heat it demands is usually taken from the hydrogen itself. ### Ammonia and related compounds [[Ammonia|Ammonia]] is the most developed chemical carrier because it is already a global commodity made by the [[Haber_process|Haber process]] and moved in ordinary refrigerated tanks. It carries hydrogen densely, liquefies under mild pressure, and can be burned directly or cracked back to [[Nitrogen|nitrogen]] and hydrogen. Its drawbacks are toxicity and the energy cost of the cracking step. Related nitrogen compounds — hydrazine, amine boranes, and salts containing the [[Boron|borohydride]] anion — trade further up the same curve: more hydrogen per kilogram, harder regeneration. ## Physical storage Physical storage keeps hydrogen as molecules but binds them loosely to a surface, exploiting [[Van_der_Waals_force|van der Waals]] attraction instead of a chemical bond. The binding energy is small — a few kilojoules per mole rather than the tens or hundreds a hydride demands — so the hydrogen comes back readily, but for the same reason it stays only if the [[Temperature|temperature]] is low. Every material in this family is therefore a competition between [[Surface_science|surface area]] and cryogenic overhead. ### Zeolites Aluminosilicate frameworks with cages of molecular size were the first ordered porous solids tried, and they set the pattern: enormous internal surface, rigid well-characterised pores, and a framework mass that mostly does no storing. Their ambient-temperature capacity is small, and they now serve better as a reference material than as a candidate. ### Porous or layered carbon Activated carbons, [[Graphite|graphitic]] materials and [[Carbon_nanotube|carbon nanotubes]] offer the largest surface per unit mass of any cheap material, and carbon is light, which helps the gravimetric figure directly. The limiting physics is unforgiving: a surface holds roughly one layer of physisorbed molecules, so capacity tracks accessible area and falls with rising temperature, and at 300 K the coverage on any carbon is low. Reports of large ambient-temperature uptakes on carbon have a long history of not reproducing. ### Metal–organic frameworks Frameworks of metal nodes joined by organic linkers reach specific surface areas beyond anything else known, and pore size and chemistry are tuned by choosing the linker — which is why they dominate physisorption research. The same cryogenic limit applies: useful capacities are measured near liquid-nitrogen temperature, and raising the binding energy enough to work at ambient temperature without making desorption difficult is the open problem. ### Cryo-compressed Cryo-compressed storage combines the two established routes: cold hydrogen above its critical temperature, under pressure, in an insulated pressure vessel. It beats 700-bar ambient gas on density because the gas is cold, and beats a true liquid on boil-off because the vessel absorbs the pressure rise instead of venting. Its drawback is inheriting both parents' hardware. ### Clathrate hydrates Water [[Ice|ice]] can crystallise into cages that trap small guest molecules, and hydrogen is small enough to be one. The store is cheap and benign — water and hydrogen, nothing else — but the cages need high pressure and low temperature at once, and the water framework is heavy, so the gravimetric fraction is poor. ## Stationary hydrogen storage Stationary storage lifts every constraint that mass imposes and leaves only volume and cost, which is why the largest hydrogen stores are not tanks at all. ### Underground hydrogen storage Salt caverns, depleted gas fields and aquifers can hold hydrogen at the scale a seasonal store needs, and they are governed by the same reservoir engineering that governs [[Natural_gas|natural gas]] production and storage: porosity and permeability set the deliverability, and a caprock sets the containment.[^zeidouni] Hydrogen complicates the picture because its molecule is small and mobile, so leakage and reaction with formation minerals and microbes matter more than they do for methane. Solution-mined salt caverns are the preferred structure for that reason. ### Power to gas Power-to-gas is the reason stationary storage is being built. Surplus electricity drives [[Electrolysis_of_water|water electrolysis]] — in alkaline cells, 2H₂O + 2e⁻ → H₂ + 2OH⁻ at the cathode and 4OH⁻ → 4e⁻ + O₂ + 2H₂O at the anode — and the hydrogen is stored, injected into the gas network, or reacted on to methane.[^haverkort-electro] The efficiency is set by the cell voltage: 237 kJ/mol gives an equilibrium voltage of 1.23 V and 286 kJ/mol a [[Thermoneutral_voltage|thermoneutral voltage]] of 1.48 V, so a cell run at 1.8 V rejects 61.7 kJ of heat per mole of hydrogen made (derived), and the whole storage chain is, as the Portal Book puts it of storage in general, lossy by construction.[^haverkort-electro][^kerlin-lossy] ## Automotive onboard hydrogen storage Onboard storage is the hardest case, because the vehicle must carry the store and refill it in minutes. Everything that helps volumetric density — cold, pressure, a heavy lattice — hurts either mass, cost or refuelling time. ### Fuel cells and storage A [[Proton-exchange_membrane_fuel_cell|proton-exchange membrane fuel cell]] converts the stored hydrogen back to electricity, and its ceiling is thermodynamic: at most ΔG/ΔH ≈ 85 % of the fuel's [[Enthalpy|enthalpy]] can appear as work.[^haverkort-electro] Storage and cell are coupled through purity and pressure — the cell wants clean, humidified, modestly pressurised hydrogen — so a store that releases dirty or cold gas needs conditioning hardware that belongs on the storage side of the ledger.[^mitofsky] Taking the electrolyser above and the cell together, an optimistic enthalpy-basis round trip is (1.48/1.8) × 0.85 ≈ 0.70 before any storage loss at all (derived; ILLUSTRATIVE). ### Other advantages of nanomaterials in fuel cells [[Nanomaterials|Nanostructured]] materials help on the cell side as well as the storage side, because a porous electrode's usefulness is set by how deep the reaction penetrates it. Past a certain thickness only the front fraction of an electrode does any work, so a thicker electrode buys nothing, and the engineering goal is high surface area within a shallow layer rather than more material.[^haverkort-porous] ### Pressurized hydrogen gas The 700-bar composite tank is the production answer, for the reasons the Compressed hydrogen section gives: no new materials science, refuelling in minutes, a well-understood failure mode. It is also why a hydrogen car's tank shape is dictated by the vessel rather than the vehicle — a pressure vessel wants to be a cylinder with domed ends, which packages badly under a floor. ### Liquid hydrogen Liquid storage has been tried on road vehicles and abandoned over boil-off: a car standing for a week loses fuel it has already paid to liquefy. The method holds where the vehicle runs continuously and the quantity is large — heavy transport and launch vehicles rather than cars. ### Liquid organic hydrogen carriers (LOHC) For vehicles the organic-carrier route's attraction is that refuelling becomes pumping a room-temperature liquid, with the spent carrier pumped back out for regeneration — a closed loop that reuses the existing fuel-station form. The penalty is the onboard dehydrogenation reactor and its heat demand, which adds mass and start-up time exactly where a vehicle can least afford them. ## Research Research across all three families is aimed at one number: the binding energy per hydrogen. Physisorption binds too weakly, so it needs cryogenic temperatures; hydrides and carriers bind too strongly, so they need heat to release. The target sits between the two, at a binding energy strong enough to hold hydrogen at ambient temperature and weak enough to give it up with the waste heat a [[Fuel_cell|fuel cell]] already produces, and no material yet occupies that window with acceptable mass and kinetics. The second research front is the chain rather than the store. Hydrogen's case rests on its 33 kWh/kg, and every step — [[Hydrogen_production|production]], compression or liquefaction, storage, reconversion — takes a share of that.[^murphy-app][^kerlin-hydrogen] Against a battery's round trip the hydrogen chain loses decisively; against a battery's seasonal capability it does not compete at all, because no battery holds energy for six months. That is the comparison the [[Hydrogen_economy|hydrogen economy]] article draws, and it is why hydrogen research concentrates on the duration end of the storage problem rather than on the daily one.[^theis] ## See also - [[Hydrogen_economy]] - [[Hydrogen_vehicle]] - [[Hydrogen]] - [[Hydrogen_production]] - [[Energy_storage]] - [[Fuel_cell]] - [[Electrolysis_of_water]] - [[Ammonia]] ## References [^murphy-app]: Murphy, Thomas (2021). *Energy and Human Ambitions on a Finite Planet*. Appendices, pp. 378–431 (battery, gravitational and kinetic storage compared, and the ≈33 kWh/kg energy content of hydrogen; page to pin). Portal Book 097, https://open.umn.edu/opentextbooks/textbooks/energy-and-human-ambitions-on-a-finite-planet [^murphy-therms]: Murphy (2021), Chapter 5, p. 96 (1 therm = 10⁵ Btu = 1.055×10⁸ J; 100 ft³ of natural gas ≈ 1.036 therm; 1 gallon of propane ≈ 91,500 Btu) and p. 95 (1 Btu ≈ 1,055 J). The 7.08 kWh/L for propane and the 1.94 kWh/L for 700-bar hydrogen are computed from those figures, from 1 US gallon = 3.785 L, and from the ideal-gas density below. Portal Book 097. [^yan-gas]: Yan, Claire Yu (2022). *Introduction to Engineering Thermodynamics*. Chapter 3, "Ideal and Real Gasses", pp. 105–126: the ideal-gas model holds at high temperature and low pressure and using it unchecked is "a common mistake" (p. 109–110); the compressibility factor Z = p·v/(R·T), with v = Z·R·T/p and Z → 1 as p → 0 at any temperature (pp. 116, 122); the reduced properties Pr = p/p_crit and Tr = T/T_crit, and the generalized chart on which Z is smallest near Pr = Tr = 1 and tends to 1 as Pr → 0 or Tr → ∞ (p. 117). The equation displays in this book were lost in the text extraction behind the Portal Book index; these are the standard forms. Portal Book 115, https://open.umn.edu/opentextbooks/textbooks/introduction-to-engineering-thermodynamics [^yan-crit]: Yan (2022), Appendix E, "Critical Properties of Selected Fluids", p. 408, and Appendix F, "Triple Point of Selected Substances", pp. 409–412 (the table in which hydrogen's critical temperature and pressure are tabulated; page to pin for the hydrogen row — no critical constant for hydrogen is asserted here beyond its being a few tens of kelvin above absolute zero). Portal Book 115. [^kerlin-hydrogen]: Kerlin, Thomas (2013). *Future Energy: Opportunities & Challenges*. Chapter 4, pp. 192–216 (hydrogen as an energy carrier: production, storage and the chain losses between them; page to pin). Portal Book 048, https://open.umn.edu/opentextbooks/textbooks/future-energy-opportunities-challenges [^kerlin-lossy]: Kerlin (2013), *Future Energy*, Chapter 3, p. 188 ("storage by conversion is always lossy"). Portal Book 048. [^haverkort-electro]: Haverkort, Willem (2024). *Electrolysers, Fuel Cells and Batteries: Analytical Modelling*. Chapter 1, "Electrochemistry", pp. 25–28: ΔG = −n·F·V_eq with F ≈ 96,485 C/mol e⁻; the alkaline half-reactions 2H₂O + 2e⁻ → H₂ + 2OH⁻ and 4OH⁻ → 4e⁻ + O₂ + 2H₂O; 237 kJ/mol → 1.23 V and 286 kJ/mol → 1.48 V, with the cell absorbing heat below the thermoneutral voltage and rejecting it above; and p. 27 for the ΔG/ΔH ≈ 85 % fuel-cell ceiling and the ≈118 % ΔH-basis figure for electrolysis. The 61.7 kJ per mole at 1.8 V is n·F·(1.8 − 1.48), computed. Portal Book 053, https://open.umn.edu/opentextbooks/textbooks/electrolysers-fuel-cells-and-batteries-analytical-modelling [^haverkort-porous]: Haverkort (2024), Chapter 5, "Porous electrodes", pp. 78–101 (the effectiveness factor E = tanh(M)/M, falling to 1/M for large Thiele modulus: past M ≈ 3 only the front 1/M of the electrode reacts, so a thicker electrode buys nothing). Portal Book 053. [^zeidouni]: Zeidouni, Mehdi (2025). *Petroleum Reservoir Dynamics*. Chapter 3, pp. 2–167 (flow, porosity, permeability and containment in porous formations — the engineering that governs underground gas storage; page to pin). Portal Book 084, https://open.umn.edu/opentextbooks/textbooks/petroleum-reservoir-dynamics [^mitofsky]: Mitofsky, Andrea (2018). *Direct Energy*. Part I, "Survey of Energy Conversion Devices", pp. 33–254 (fuel cells among the direct converters, with their input-stream requirements; page to pin). Portal Book 055, https://open.umn.edu/opentextbooks/textbooks/direct-energy [^theis]: Theis, Tom; Tomkin, Jonathan, eds. (2015). *Sustainability: A Comprehensive Foundation*. Chapter 10, "Sustainable Energy Systems" (seasonal storage and the role of chemical carriers alongside variable renewable supply; page to pin). Portal Book 098, https://open.umn.edu/opentextbooks/textbooks/sustainability-a-comprehensive-foundation ## External links - [Future Energy: Opportunities & Challenges](https://open.umn.edu/opentextbooks/textbooks/future-energy-opportunities-challenges) (Kerlin, 2013) — Portal Book 048; Chapter 4 is the hydrogen chapter - [Electrolysers, Fuel Cells and Batteries: Analytical Modelling](https://open.umn.edu/opentextbooks/textbooks/electrolysers-fuel-cells-and-batteries-analytical-modelling) (Haverkort, 2024) — Portal Book 053 - [Introduction to Engineering Thermodynamics](https://open.umn.edu/opentextbooks/textbooks/introduction-to-engineering-thermodynamics) (Yan, 2022) — Portal Book 115; the real-gas chapter behind the 700-bar correction - The Wikipedia pair's external links list agency roadmaps and materials databases <!-- MATTERSIM:BEGIN g24 — Matter & Energy Cluster microsim (framework build, specs/sims/Hydrogen_storage.json); do not hand-edit inside --> **Microsim — three.js (Wikitube framework):** *Hydrogen storage* <div class="wt-sim" data-src="https://wikitube-3d-microsims.netlify.app/thury/Hydrogen_storage.html" data-title="Hydrogen storage"></div> *Reused from the Thury Hydrodynamics Compendium set (Compendium section 12), where it is already live; the Matter & Energy Cluster places the same build rather than rebuilding it.* <!-- MATTERSIM:END --> ## Wikipedia : Wikitube **Strict pair:** [Wikipedia](https://en.wikipedia.org/wiki/Hydrogen_storage) : [Wikitube](https://en.wikitube.io/wiki/Hydrogen_storage) · pinned revision [1373499572](https://en.wikipedia.org/w/index.php?oldid=1373499572) · 2026-09-11 ## Previous hub tags Hubs: `Life_Physics`. Portals: [[PORTAL_Energy]]. --- *Matter & Energy Cluster child articles, wave 1 · 2026-09-11 · drafted · Energy row E65 · sim live (reused thury/Hydrogen_storage).*