Article
Energy & Renewables · Storage · ⏱ ~12 min read · Last updated: 9 July 2026

Energy Storage: Pumped Hydro vs Battery Economics

As wind and solar supply a growing share of electricity, the question of how to store energy when the sun sets or the wind drops becomes central to grid planning. Two technologies dominate the conversation: pumped-storage hydropower, which has quietly stored most of the world's grid energy for over a century, and lithium-ion batteries, whose costs have collapsed so fast they now dominate every new storage project announcement. Neither wins on every metric — the right choice depends on duration, geography, and how fast the energy needs to move.

1. How Each Technology Stores Energy

Both technologies store electrical energy by converting it into another form and later converting it back, but the physical mechanism could not be more different in scale and character.

Pumped hydro: gravitational potential energy E = m · g · h Two reservoirs at different elevation; excess grid power pumps water uphill, later released through turbines downhill to regenerate electricity on demand Example: 1 GWh plant, head height h = 400 m Required water mass ≈ E / (g·h) ≈ 9.2 × 10⁸ kg ≈ 920,000 m³ (roughly 370 Olympic swimming pools of water, moved once) Lithium-ion battery: electrochemical potential energy Energy stored via lithium ion intercalation between graphite anode and metal-oxide cathode layers Energy density: ~150-260 Wh/kg (pack level) Same 1 GWh requires roughly 4,000-6,700 tonnes of battery pack

Pumped hydro moves an enormous mass of water a modest height; batteries pack a comparatively tiny mass of reactive chemistry with very high energy density. This single difference in physical mechanism explains almost every trade-off that follows.

2. Round-Trip Efficiency

TechnologyRound-trip efficiencyMain loss mechanisms
Pumped hydro70-85%Pump/turbine hydraulic losses, pipe friction, transformer losses
Lithium-ion battery85-95%Internal resistance (I²R heating), inverter conversion losses
Compressed air (CAES)40-70%Heat lost during compression unless recovered (adiabatic designs)
Hydrogen (power-to-gas-to-power)30-45%Electrolysis, compression/storage, and fuel-cell/turbine losses compound

Every percentage point of round-trip efficiency lost is effectively a fuel cost — the grid operator must generate and pay for extra electricity to compensate for what storage wastes. Batteries' efficiency edge over pumped hydro is one reason they win decisively for short-cycle, frequent daily charging applications, where the efficiency gap compounds every single day.

3. Capital Cost per kWh and per kW

Storage systems have two separate cost components that matter for different applications: the cost per unit of power (kW, how fast it can charge/discharge) and the cost per unit of energy (kWh, how much it can store in total).

Pumped hydro power cost

$1,000-2,500 per kW of turbine/generator capacity — dominated by civil works, tunnels and powerhouse construction.

Pumped hydro energy cost

As low as $5-100 per kWh of storage capacity once reservoirs exist — extremely cheap to add hours of duration, since it's "just" a bigger reservoir.

Battery power cost

$150-300 per kW of inverter/power electronics capacity — cheap and fast to scale for power-focused applications.

Battery energy cost

$100-250 per kWh of pack capacity (2020s pricing, still falling) — cheap for short duration but the total bill scales linearly with every added hour.

This cost structure explains the duration crossover: pumped hydro's huge upfront civil cost is justified only when amortised over very large total energy capacity and a very long asset life, while batteries' low fixed cost and fast deployment win for short-duration, quick-turnaround projects where a decade-long permitting process for a dam would be commercially unacceptable.

4. Levelized Cost of Storage (LCOS)

LCOS combines capital cost, efficiency, lifespan, and cycling frequency into a single number: the average cost of every megawatt-hour the system will ever discharge over its whole life.

Simplified LCOS formula: LCOS = (CAPEX + Σ OPEX_t / (1+r)^t) / Σ (Energy_discharged,t / (1+r)^t) CAPEX = upfront capital cost OPEX = annual operating + replacement cost r = discount rate Energy_discharged,t = MWh delivered in year t, after accounting for round-trip efficiency losses Illustrative 2024-2025 published ranges (highly site- and scale-dependent, treat as orders of magnitude only): 4-hour lithium-ion battery: $150-300 / MWh cycled New-build pumped hydro: $150-250 / MWh cycled (long-duration) Existing (already-built) pumped hydro: often under $50 / MWh

The headline takeaway from most published LCOS studies is that new lithium-ion batteries are now cost-competitive with, and often cheaper than, newly built pumped hydro for short-duration applications — but pumped hydro plants already built decades ago, with their capital cost long since depreciated, remain some of the cheapest storage in existence per MWh delivered today.

5. Duration: The Metric That Decides the Winner

Duration — how many hours a system can discharge at full rated power — is arguably the single most important variable in choosing between the two technologies, because the cost of adding duration scales completely differently for each.

Rule of thumb used by many grid planners: lithium-ion wins decisively below about 4 hours of duration; pumped hydro (where a suitable site exists) tends to win above about 8-10 hours; the 4-8 hour range is genuinely competitive and increasingly contested by both technologies plus emerging long-duration alternatives.

6. Siting, Lifespan and Environmental Factors

FactorPumped hydroLithium-ion battery
Site requirementTwo reservoirs, 200-800 m elevation difference, water rightsAny flat land or building near grid connection
Typical build time5-10+ years (permitting-dominated)6 months to 2 years
Operating lifespan50-100 years (civil works, refurbishable)10-20 years (chemical degradation)
End-of-life impactMinimal beyond original construction footprintRequires recycling infrastructure for cathode metals
Land/water footprintLarge (reservoirs, watershed effects)Small (compact enclosures)

Pumped hydro's century-scale lifespan means its lifetime cost keeps falling the longer it operates past its original payback period, while batteries must be at least partially replaced within 10-20 years, adding a recurring capital cost that pumped hydro's civil infrastructure largely avoids.

7. What Comes Next: Long-Duration Storage

Neither incumbent technology is a perfect fit for the emerging need for multi-day or seasonal storage — the kind needed to shift summer solar surplus into winter, or to ride through a multi-day wind lull. This gap has spurred a wave of new approaches.

Iron-air batteries

Extremely cheap, abundant materials (iron, air, water); low round-trip efficiency (~50-60%) but very low cost per kWh, targeting multi-day duration.

Flow batteries (vanadium redox)

Store energy in liquid electrolyte tanks — power and energy capacity scale independently, and lifespan can exceed 20 years with minimal degradation.

Compressed air energy storage (CAES)

Uses underground caverns or purpose-built vessels; adiabatic designs recover compression heat to raise efficiency toward 70%.

Gravity-based systems

Raise and lower solid concrete or rock blocks instead of water, aiming to replicate pumped hydro's low-cost duration scaling without needing a river valley.

None of these has yet reached the commercial maturity or scale of lithium-ion or conventional pumped hydro, but all are explicitly targeting the multi-day duration gap between the two established technologies — the next decade of grid storage economics will likely be shaped by which of these alternatives, if any, can undercut both incumbents at scale.

Frequently Asked Questions

What is round-trip efficiency and why does it matter for storage economics?

Round-trip efficiency is the fraction of energy put into storage that comes back out usable, and it directly sets the fuel cost of storing energy: pumped hydro typically achieves 70-85% round-trip efficiency while lithium-ion batteries achieve 85-95%, meaning batteries waste less energy per storage cycle even before considering capital costs.

Why is pumped hydro still the largest form of grid energy storage worldwide?

Pumped hydro accounts for roughly 90% of installed grid storage capacity globally because it was built out over many decades at a very low cost per kWh of stored energy once the dam and reservoir infrastructure exists, and a handful of very large plants (each storing gigawatt-hours) can outweigh thousands of smaller battery installations in total capacity terms.

Why has battery storage grown so much faster than new pumped hydro in the last decade?

Lithium-ion battery pack costs fell by roughly 90% between 2010 and the mid-2020s due to electric-vehicle-driven manufacturing scale, while pumped hydro projects require suitable geography (two reservoirs at different elevations), face 5-10 year permitting and construction timelines, and often meet environmental and community opposition — batteries can be deployed in months almost anywhere.

What does "duration" mean in grid storage, and why does it change which technology wins?
Duration is how many hours a storage system can discharge at its rated power before running empty. Most grid-scale lithium-ion batteries are economically sized for 2-4 hour duration (matching daily solar-to-evening-peak shifting), while pumped hydro plants are typically sized for 6-20+ hours, making pumped hydro structurally better suited to multi-day or seasonal storage needs that batteries cannot economically match at current costs.
What is the levelized cost of storage (LCOS) and how do the two technologies compare?
LCOS spreads a storage system's total lifetime cost (capital, operations, replacement, financing) over all the energy it will ever discharge, in dollars per MWh cycled. For short-duration (2-4 hour) applications lithium-ion batteries now typically beat pumped hydro on LCOS, while for long-duration (8+ hour) and very large-scale applications pumped hydro's much longer lifespan (50-100 years vs 10-20 for batteries) can still give it the edge where suitable sites exist.
How does energy density differ between pumped hydro and batteries?
Lithium-ion batteries store roughly 150-260 Wh per kilogram of pack, while pumped hydro stores only about 0.5-1.3 Wh per kilogram of water moved (though the water itself is nearly free and reusable) — batteries win dramatically on mass and volume density, which is precisely why batteries work for vehicles and portable devices while pumped hydro only makes sense as large, fixed civil infrastructure.
What are the main geographic requirements for a pumped hydro plant?
A conventional pumped hydro plant needs two reservoirs with a substantial elevation difference (typically 200-800 metres) located close enough together to keep tunnel or pipe costs manageable, along with reliable water rights and access, which is why suitable sites are geographically limited and increasingly already developed in many countries.
Do batteries or pumped hydro degrade faster over their operating life?
Lithium-ion batteries degrade chemically with every charge cycle and with calendar time, typically losing 70-80% of their original capacity after 10-20 years or a few thousand deep cycles, whereas pumped hydro's mechanical and civil components (turbines, dams, tunnels) can operate for 50-100 years with periodic refurbishment, giving it a fundamentally longer asset life.
Is there a role for both technologies on the same grid?
Yes — modern grid operators increasingly deploy batteries for fast, short-duration response (frequency regulation, daily solar shifting) and rely on pumped hydro or emerging long-duration technologies for multi-day and seasonal balancing, treating the two as complementary tools rather than direct competitors for the same job.
What newer storage technologies are competing with both pumped hydro and lithium-ion?
Iron-air batteries, flow batteries (vanadium redox), compressed-air energy storage, and gravity-based solid-mass systems are all being developed specifically to fill the multi-day, long-duration storage gap between short-cycle lithium-ion and site-constrained pumped hydro, though none has yet reached the same commercial scale as either incumbent.