Lithium-ion dominates short-duration storage. But the grid needs hours, days, and
even seasons of stored energy. No single battery chemistry can do it all.
The Duration Problem
Lithium-ion batteries are excellent at what they do: absorb and release energy over
minutes to a few hours. They respond in milliseconds, fit in a shipping container,
and their costs have dropped 90% since 2010. For short bursts of grid support -- frequency
regulation, peak shaving, solar time-shifting -- they are the clear winner.
But the grid does not only need short bursts. A winter Dunkelflaute (dark
doldrums) in Northern Europe can suppress solar and wind output for two weeks straight.
A summer heatwave can drive air-conditioning demand above normal for days. Seasonal
mismatches between when the sun shines and when people need heat can span months.
Lithium-ion cannot bridge these gaps. The batteries would be enormous, absurdly
expensive, and would degrade long before they paid for themselves.
The core issue is cost per duration. Lithium-ion stores energy in
the same cells that deliver power. To double the duration, you double the number of
cells -- and double the cost. Other technologies decouple power from energy: the
"engine" that delivers power is separate from the "tank" that holds energy. Add a
bigger tank, and duration goes up without adding more of the expensive parts.
No single technology covers the full spectrum. The future grid needs a portfolio --
fast-response batteries for seconds-to-hours, medium-duration storage for daily
cycling, and long-duration options that can bridge weeks of low renewable output.
The Storage Zoo
Researchers and engineers are working on dozens of storage technologies. Seven have
emerged as serious contenders, each with a different sweet spot on the duration
spectrum. Click any card below to learn how it works.
Pick Your Storage
Build a storage portfolio by mixing technologies. See how different combinations
cover the duration spectrum.
Grid frequency drops below 49.8 Hz. Storage must inject power within milliseconds to prevent cascading failure.
Duration needed: Seconds to minutesResponse time: Sub-second
Li-ion
excellentFast response, proven in frequency markets worldwide.
Flywheel
excellentSub-millisecond response. Built for this exact job.
Flow Battery
possibleFast enough, but oversized for seconds-scale needs.
Pumped Hydro
poorToo slow to ramp -- 30-90 seconds to respond.
CAES
poorMinutes to start. Not suitable for fast response.
Gravity
poorMechanical ramp time is too slow.
Thermal
poorCannot convert heat to electricity fast enough.
Hydrogen
poorFuel cells need minutes to ramp.
Technology
Response
Duration
Efficiency
$/kWh
Cycles
Li-ion
<1s
1-4h
90-95%
$150-300
3,000-5,000
Flow Battery
<1s
4-12h
70-80%
$200-500
20,000+
Pumped Hydro
30-90s
6-24h
75-85%
$50-150
50,000+
CAES
5-15min
8-24h
40-70%
$50-100
30,000+
Gravity
~30s
4-12h
80-85%
$150-250
35,000+
Thermal
10-30min
Hours-Days
50-95%
$20-60
30,000+
Hydrogen
1-10min
Days-Months
30-40%
$500-1500
20,000+
Flywheel
<0.1s
Sec-Min
85-95%
$1000-5000
100,000+
Compare Technologies
Select technologies to compare their key characteristics side by side.