Home » Hydrogen Storage: Tanks, Salt Caverns and Buffer Sizing
Hydrogen Storage: Tanks, Salt Caverns and Buffer Sizing
10 min read · Sep 16, 2025 · By Alex
Keep the long, large volumes in geological storage, where salt caverns typically cost less than 0.6 USD per kg of H2 and have run at around 98 percent efficiency since the chemical sector first used them, on 2019 figures, and keep the short, small buffers in tanks at around 99 percent efficiency on the same figures. Size the choice on four things only: volume, duration, discharge rate and how close your site sits to a salt formation. For a project producing green hydrogen, those four lines are the whole of the storage decision, which is why hydrogen storage solutions have to be sized before the cost model freezes rather than after it. I size buffers for developers in feasibility, and the answer is shorter than most people expect.
I will work through the four criteria that pick the asset, what has actually been built, what the region has committed, what a battery buffer does instead, how I size the buffer, and what I put in front of the lender. Each step closes the same way, with the question of what changes for you.
In short
- Medium: Volume, duration, discharge rate and geographic availability pick the asset, and nothing else does (2019).
- Built: Just over 7 percent of announced underground storage reached a final investment decision or was under construction in 2026.
- Region: Committed production capacity in APEC passed 4.2 million tonnes per year in 2024, and demand is still the binding constraint.
- Battery: A battery covers the short buffer and the daily swing, and it cannot hold a season or a strategic reserve.
- Lender: I separate the contracted part of my storage from the part that is still an option (2025).
In this article
What storage is actually doing for the project
Buffer storage is the inventory that covers the gap between what I produce and what I deliver on a given day. Geological storage means putting the molecule into the rock itself, and a salt cavern is a void washed out of a salt deposit and sealed by that rock. Discharge rate is how fast I pull the gas out, state of charge is the inventory still in the store, load factor is the share of installed capacity I use across a year, and seasonal storage is inventory carried from a surplus season into a deficit one. Energy density is the energy per unit of volume, and delivered cost is what a kilogram costs at the gate once every upstream step is paid for.
I buy storage for four needs: a short buffer against a delivery that slips, a daily swing against a plant that ramps, a seasonal swing against the weather, and a strategic reserve against an outage I do not control. Each need has its own best asset, and paying for one of them with the asset built for another is how a margin disappears.
What this means for you: write those four needs as four separate lines in the schedule, because one blended storage line hides which need you are actually buying.
The four criteria that pick the asset
The right medium depends on the volume to store, the duration, the required speed of discharge and the geographic availability of each option, with geological storage best for large volumes and long durations and tanks better for short, small storage, as The Future of Hydrogen set out in 2019. I run those four tests in the same order every time, and they feed my technology maturity and scale-up page.
- Salt caverns: used by the chemical sector in the United Kingdom since the 1970s and in the United States since the 1980s, they typically cost less than 0.6 USD per kg of H2, run at around 98 percent efficiency, carry a low contamination risk and give high discharge rates thanks to their pressure (2019).
- Tanks: compressed or liquefied hydrogen gives high discharge rates at around 99 percent efficiency for small, short applications, while compressed hydrogen at 700 bar holds only 15 percent of the energy density of gasoline, which is close to seven times the space for the same energy (2019).
What this means for you: if your volume is large and your duration is long, the cavern carries the cost advantage, and if your buffer is short and small, the tank carries the discharge advantage, so run the four tests before you price either.
How much storage is actually being built
Announced underground hydrogen storage could provide 11 TWh of capacity by 2035, equal to 335 kt of hydrogen, but just over 7 percent has reached a final investment decision or is under construction, equal to 0.6 percent of the estimated throughput from committed low-emissions projects, according to Global Hydrogen Review 2026. That 11 TWh is an announcement, and I never book it as capacity I can call on.
Few commercial-scale underground facilities are under construction, and large salt caverns are being built in the United States, Germany and China (2026). The same gap between brochure and bankable shows up in transport, where announced hydrogen pipeline projects exceed 40 000 km by 2035 while only 9 percent of that length is operational, under construction or committed (2026). I read the two gaps together, because a project that depends on both is exposed twice.
What this means for you: when you size a buffer, treat only the committed share as real inventory and treat the announced share as a future option that may never arrive.
What this region has committed, and what that does to buffer sizing
The APEC region reached more than 4.2 million tonnes per year of committed production capacity in 2024, a 68 percent growth driven by renewable hydrogen in China and natural gas with carbon capture in North America, with state-owned enterprises playing a major role in delivery, as the APERC Hydrogen Report 2024 records. China holds the largest share of renewable hydrogen projects in APEC at 1.8 million tonnes per year committed as of 2024.
Demand uncertainty remains the primary barrier to market maturity and creates risks for long-term investment (2024), so a buffer sized against announced demand is a buffer sized against nothing. For a site in this part of the world the power and feedstock choice sits beside the storage choice, which is why I keep the technology choices for Southeast Asia in the same file.
What this means for you: size the buffer against the demand you can document today rather than the demand announced for 2030, or your inventory will be paid for by volume that may never come.
What a battery buffer does instead
Batteries and electrolysers both convert electricity into chemical energy and back; they share electrochemistry, materials and manufacturing processes, and battery cost progress is more advanced while electrolyser manufacturing sits at an earlier stage with more room for near-term cost cuts, as Batteries and hydrogen technology sets out in 2020. The power sector uses batteries more and more to integrate variable renewables such as wind and solar into electricity systems (2020). Electrolysers can also clean up existing hydrogen supplies in industrial clusters such as ports, where much of today’s production sits (2020).
Of my four storage needs, a battery covers the short buffer and the daily swing, because it takes in surplus power and gives it back inside the same operating day. It cannot cover the seasonal swing or the strategic reserve, since battery inventory is measured in hours rather than in weeks of molecules. That trade lands on the delivered price, so I read it against green hydrogen cost in Southeast Asia in 2026.
What this means for you: buy the battery for the swing you can time, buy the cavern or the tank for the swing you cannot, and do not let one line item claim to do both.
How I size the buffer
Weeks of storage protect users against mismatches between supply and demand, while much larger and longer storage is needed to bridge seasonal swings or to give system resilience (2019). The most appropriate medium follows volume, duration, discharge speed and geography, and the energy a 3 percent hydrogen blend takes out of a gas pipe, around 2 percent, is a reminder that every conversion has an efficiency you pay for (2019). Tanks and caverns carry different efficiencies, so I size the buffer on the energy I can actually withdraw, not on the volume I can actually inject (2019).
The withdrawal test is the one I carry into every model built during feasibility and financial model work, because a vessel that fills quickly and empties slowly changes the cash flow rather than the hardware list.
What this means for you: write your buffer in energy you can withdraw on the worst day, then convert back to volume, because volume is what your supplier quotes and withdrawal is what your buyer pays for.
What I put in front of the lender
Uncertain demand continues to pose a critical challenge to the viability of the sector, because producers need reliable offtakers to justify large-scale investments, which is the finding I carry from Global Hydrogen Review 2025. Cumulative firm offtake agreements stand at less than 2 Mtpa, about 5 percent of the production announced projects could reach by 2030 (2025). Challenges remain in high production costs, demand and regulatory uncertainty, and infrastructure gaps (2025).
So I show the lender which part of the storage I rely on is contracted and which part is an option, and I apply the same split in my contracting and risk management notes so that the offtaker and the storage schedule describe the same inventory.
What this means for you: put two columns in front of the lender, one for storage backed by a firm offtake and one for storage backed by nothing, and make sure the second column is visibly smaller than the announcement.
The five lines I write in the storage schedule
- The four criteria that pick the asset: volume, duration, discharge rate and geographic availability (2019).
- The efficiency I can actually withdraw at, tank or cavern (2019).
- The committed share of the announced capacity I depend on (2026).
- The demand that covers the inventory (2025).
- The cost per kilogram the buffer adds to the delivered price (2019).
The order matters as much as the content, because the asset choice should be visible before the cost line.
What this means for you: keep these five lines in this order, size on the energy you can withdraw and on the demand that is contracted, and treat every uncommitted kilogram as an option rather than as inventory.
Sources
- IEA, The Future of Hydrogen https://iea.blob.core.windows.net/assets/9e3a3493-b9a6-4b7d-b499-7ca48e357561/The_Future_of_Hydrogen.pdf, 2019.
- IEA, Global Hydrogen Review 2026 https://iea.blob.core.windows.net/assets/af4b6218-0311-4963-91ab-903cb18ca0ca/GlobalHydrogenReview2026.pdf, 2026.
- IEA, Global Hydrogen Review 2025, Five key questions about hydrogen https://www.iea.org/reports/global-hydrogen-review-2025/five-key-questions-about-hydrogen, 2025.
- IEA, Batteries and hydrogen technology: keys for a clean energy future https://www.iea.org/articles/batteries-and-hydrogen-technology-keys-for-a-clean-energy-future, 2020.
- Asia Pacific Energy Research Centre, APERC Hydrogen Report 2024 https://aperc.or.jp/file/2025/3/18/APERC_Hydrogen_Report_2024_final.pdf, 2025.
Frequently asked questions
What is the cheapest way to store hydrogen?
In geological storage, salt caverns typically cost less than 0.6 USD per kg of H2, have around 98 percent efficiency and a low contamination risk, and have served the chemical sector since the 1970s in one country and the 1980s in another (2019). Tanks are quicker to site and need no formation, but they suit small buffers held for short periods, so the cheapest option is the one whose volume and duration match the asset rather than the lowest capital line (2019).
Should I build a salt cavern or a tank?
It depends on four things only: volume, duration, required discharge speed and geographic availability. Geological storage suits large volumes held over long periods, while tanks run at around 99 percent efficiency for short, small buffers, and compressed hydrogen at 700 bar holds only 15 percent of the energy density of gasoline, which is close to seven times the space for the same energy (2019).
How much hydrogen storage is actually available today?
Announced underground capacity could reach 11 TWh by 2035, equal to 335 kt of hydrogen, but just over 7 percent has reached a final investment decision or is under construction, equal to 0.6 percent of the estimated throughput from committed low-emissions projects (2026). What is genuinely available today is therefore a small part of the announcement, and the announced figure is not inventory you can schedule.
Does a battery replace a hydrogen buffer?
Only partly. Batteries are used more and more to integrate variable renewables such as wind and solar into electricity systems, so they cover the short buffer and the daily swing (2020), while seasonal and large-volume storage still needs geological capacity, and a battery does not give you a molecule at the gate (2019).
Not sure your project would survive a credit conversation?
A short call is usually enough to see where your project stands.
If you want a second pair of eyes on the storage schedule before it enters the cost model, I provide technology maturity and scale-up support for teams at exactly this stage, and I work from your volume, duration, discharge rate and geography rather than from a catalogue.
I do not provide investment advice or arrange investments. I act as a strategic advisor and facilitator.
Alex, Founder of WeConnect Asia
I help low-carbon fuel and carbon removal project developers in Southeast Asia secure the commercial and financial partners needed to reach FID.
Related articles

Pitching a Climate Tech Company to Strategic Investors: What the Corporate Has to Believe Before It Writes a Cheque
How to pitch a climate tech company to strategic investors: the metric cluster, the readiness evidence and the ask that corporates fund.

Biomethane: What It Is, How It Is Made and Who Buys It
Biomethane is biogas cleaned to pipeline quality. I explain how it is made, how it differs from biogas and who buys it in Southeast Asia.

CO2 Sourcing for e-SAF Projects: What the Rules Allow and What the Contract Must Say
Biogenic CO2, direct air capture or industrial flue gas: which sources an e-SAF plant may use, what they cost and which contract terms make the fuel certifiable.

Operating Performance Guarantees for SAF Plants: What Lenders Ask For
Minimum guarantees, performance damages, retesting and who stands behind them: the operating guarantees a SAF plant needs to keep debt and offtake intact.

EPC and Construction Risk in SAF Plants: What a Lender Actually Tests
FEED scope, convertible EPC, performance guarantees and damages caps: the construction risk questions a SAF project must answer before financial close.

Matching Feedstock and Offtake Risk Back-to-Back in a SAF Project: How Lenders Read the Chain
Lenders do not buy a feedstock plan and an offtake plan separately. They buy the link between the two, and the link is where most SAF files break.