Pumped Hydro and Hydrogen Storage Primer — Choosing Between Storage That Spans Hours and Storage That Spans Seasons

The more solar and wind power grows, the more the grid faces mismatches across several distinct timescales — not just "the next few seconds" but "this evening," "this week's bad weather," and "demand that spans the seasons." Lithium-ion batteries can design power (kW) and capacity (kWh) fairly independently, which makes them strong for adjustments over minutes to a few hours (for battery fundamentals, see the Power Battery Primer). But trying to store energy over days to seasons with batteries alone means the required capacity balloons quickly, and the cost stops making sense. Pumped hydro storage has long filled that longer timescale, and in recent years hydrogen has drawn attention as the technology that reaches further still. This article leaves the basic physics of hydraulic turbines themselves to the Hydroelectric Power Primer, and instead lines up pumped hydro and hydrogen storage side by side as two answers to a single problem: "store energy, and turn it back into electricity later."

The Bottom Line in 30 Seconds

• Pumped hydro storage, which shuttles water back and forth between an upper and a lower reservoir, is an established large-capacity, long-duration storage technology. Round-trip efficiency runs roughly 70–87%, with a central figure around 80%.
• Variable-speed pump-turbines can continuously adjust the power consumed while pumping, and by quickly throttling back the pumping load when frequency drops, they can support the grid in the same direction as increasing generator output (see the Hydroelectric Power Primer for details).
• Hydrogen doesn't store "electricity as electricity" — it electrolyzes water into hydrogen, stores it, and converts it back to electricity when needed via a fuel cell or gas turbine. Round-trip efficiency currently sits in roughly the 30–50% range, lower than pumped hydro.
• In exchange, hydrogen lets you design the storage medium (tanks or underground storage) and the energy-conversion equipment (electrolyzers, fuel cells) separately, so in principle it doesn't depend on terrain and suits long-term storage spanning weeks to seasons.
• Thinking of pumped hydro as "efficient storage over hours to days" and hydrogen as "less efficient but location-agnostic long-term storage" captures a relationship that's closer to complementary than competitive.

1. Sorting the "Long-Duration Storage" Problem by Timescale

Renewable variability can't be described on a single timescale. A few minutes of variation from passing clouds, a several-hour ramp as evening demand rises and solar output falls together at sunset, a multi-day shortfall in wind from a stalled low-pressure system, seasonal differences in sunlight and rainfall patterns — the character of the storage needed changes with the timescale in play.

Batteries can design output (kW) and capacity (kWh) relatively freely through cell count, but cost grows nearly in proportion as capacity is stretched. Pumped hydro and hydrogen store energy in the form of "water's potential energy" or "hydrogen molecules' chemical energy," so the cost curve for expanding storage capacity has a different structure from batteries. That's why both can play a leading role in long-duration storage.

2. Pumped Hydro Storage — Revisiting an Established Large-Capacity Technology

The basic principles of pumped hydro and the physics of the turbine were covered in the Hydroelectric Power Primer. Here the focus narrows to its performance as a storage technology.

The amount of energy that can be stored can be estimated using water volume V, effective head H, and overall efficiency \eta:

The capacity equivalent to a battery's kWh is set by the effective storage volume of the upper and lower reservoirs and the head, while output (kW) is set by the rating of the pump-turbine and the motor-generator. Being able to design these two independently is something pumped hydro shares with batteries, but for pumped hydro the constraints of terrain and water rights are a precondition for capacity design.

Round-trip efficiency — the share of power that comes back after one full cycle of pumping (charging) and generating (discharging) — is the product of pump efficiency, turbine efficiency, motor-generator efficiency, and piping losses, and measured values fall roughly in the 70–87% range, with a central figure around 80%. This is a mature figure backed by decades of operating experience, at a level comparable to or slightly below lithium-ion battery round-trip efficiency.

Fixed-speed pump-turbines draw roughly constant power while pumping, which makes them poor at tracking surplus renewable power that comes in awkward, in-between quantities. Variable-speed pumped storage uses power electronics to continuously vary rotational speed, allowing fine adjustment of pumping load — and when frequency drops, it can quickly throttle back the pumping volume, supporting the grid in the same direction as increasing generation output. In Japan, the Kannagawa Power Station (TEPCO) in Gunma Prefecture is known as one of the world's largest pumped-storage plants, designed with an effective head of 653 m and, on paper, six motor-generator units rated at 470,000 kW each — for a maximum output of 2.82 million kW with all units running. At present, units 1 and 2 have begun operation, giving a current maximum output of 940,000 kW. For the technical details of pumped hydro, including variable-speed operation and capacity scale-up, see the relevant section of the Hydroelectric Power Primer.

Figure 1 — Pumped hydro shuttles water between upper and lower reservoirs, switching between pumping and generating with a reversible pump-turbine. Round-trip efficiency runs roughly 70–87%, centering around 80%.

3. Long-Duration Storage via Hydrogen — Storing Electricity as Molecules

The basic idea behind hydrogen storage is to use surplus power to electrolyze water into hydrogen, store it in compressed, liquefied, or underground form, and convert it back to electricity when needed via fuel cells or hydrogen-fired/co-fired gas turbines. This whole process is called Power-to-Gas-to-Power (P2G2P). Hydrogen made using power derived from renewables such as solar or wind is called "green hydrogen," and its distinguishing feature is low CO_2 emissions during production.

Electrolyzer types include alkaline water electrolysis, polymer electrolyte membrane (PEM) water electrolysis, and solid oxide electrolysis cells (SOEC), which operate at higher temperatures. Alkaline water electrolysis leads on scale-up and track record, while PEM is said to offer superior responsiveness, making it a good fit for variable renewable power. The hydrogen produced is sometimes converted into another chemical form — high-pressure gas, liquid hydrogen, or ammonia — for storage and transport.

On the reconversion side, in addition to generating power with polymer electrolyte fuel cells (PEFC) or solid oxide fuel cells (SOFC), power generation using hydrogen-fueled gas turbines is also being considered for large-scale applications. The round-trip efficiency combining electrolysis and reconversion stays around 48% for a simple configuration (roughly 80% electrolysis efficiency × roughly 60% fuel-cell efficiency), and reported figures for full systems vary in the literature from roughly 28% to 52%. Some research points to figures in the 60s of a percent for more advanced electrolyzers combined with solid-oxide systems, and even above 80% for future high-efficiency systems, but as things stand, the mainstream configuration has lower round-trip efficiency than pumped hydro.

A leading demonstration site in Japan is the Fukushima Hydrogen Energy Research Field (FH2R), which opened in March 2020 in Namie, Fukushima Prefecture. It pairs a 20 MW solar power facility with what's said to be one of the world's largest alkaline water electrolysis units in the 10 MW class (6 MW rated output, 10 MW maximum power consumption), capable of producing 1,200 Nm^3 of hydrogen per hour at rated operation. That's said to be enough to fill roughly 560 fuel cell vehicles (FCVs), and the facility is positioned as a demonstration site for validating the full chain of technology that converts surplus renewable power into hydrogen, then stores and uses it.

Figure 2 — Hydrogen stores electricity by converting it into a molecular form. Because there are losses at both electrolysis and reconversion, round-trip efficiency is lower than pumped hydro, but separating the storage medium from the generation equipment suits long-term, large-scale storage.

4. Comparing the Two Technologies Side by Side

Looking at efficiency alone, pumped hydro comes out ahead, but that's a conditional advantage — "if suitable terrain exists." Hydrogen trades away efficiency in exchange for the ability to deploy on flat land or in areas with existing infrastructure, and for versatility that lets it share value with uses beyond electricity (industrial feedstock, fuel cell vehicles, future fuel use). Rather than one replacing the other, it's more practical to see them as splitting roles depending on terrain, demand, and existing infrastructure.

5. How This Looks From the Grid's Perspective

From a grid-operations standpoint, neither pumped hydro nor hydrogen storage is a "power plant" — both are "devices that shift time." Second-to-minute-scale response, like frequency regulation, is best handled by batteries or variable-speed pumped hydro with their superior response speed. Hour-to-day-scale energy shifting is best handled by pumped hydro, which has the edge in round-trip efficiency. And for long-duration variability that battery or pumped-hydro storage capacity alone can't fill — several days of sustained bad weather, or a seasonal supply-demand gap — molecular storage like hydrogen becomes a candidate.

In market designs such as the balancing-power market, different product categories of regulating power are set up for each targeted response time, and matching a storage technology's ideal timescale to the market's product design is the key to getting the most out of each technology. The details of this market system are covered in the VPP (Virtual Power Plant) Primer and the Grid Inertia and Frequency Regulation Primer.

Summary: Looking at the Trade-Off Between Efficiency and Siting

Pumped hydro, armed with established technology and high round-trip efficiency, has carried energy shifting on the timescale of hours to a few days. Hydrogen trades away efficiency, but the flexibility of decoupling the storage medium from the generation equipment gives it room to reach long-term, large-scale storage that pumped hydro can't, and to expand into uses beyond electricity.

When evaluating a large-scale storage project, it's worth checking the following:

1. What timescale is targeted — hours, days, or seasons?
2. How does round-trip efficiency compare with pumped hydro?
3. How much does the design depend on siting conditions (terrain, land, existing infrastructure)?
4. Can it share value with uses beyond electricity (industrial use, use as fuel)?

Keeping these four points in mind lets you read pumped hydro and hydrogen storage not as an "old versus new" rivalry, but as complementary technologies suited to different timescales.

References

• U.S. Energy Information Administration — Utility-scale batteries and pumped storage return about 80% of the electricity they store
• NREL — Pumped Storage Hydropower, Annual Technology Baseline
• TEPCO Renewable Power — Representative hydroelectric plants (Kannagawa Power Station)
• Toshiba Energy Systems & Solutions — Fukushima Hydrogen Energy Research Field (FH2R)
• NEDO — Materials on the Fukushima Hydrogen Energy Research Field
• Agency for Natural Resources and Energy — Hydrogen Policy
• IEA — Hydrogen
• International Hydropower Association — Pumped Storage