In our previous post on Energy Storage Systems, we discussed about grid services, how grid stability spans multiple timescales, and how different storage technologies match different grid services. This post takes that a step further. It looks at each storage technology in detail — how it works, where it fits on the grid, and where its limits lie.
Following are the Four Main Energy Storage Technologies
Supercapacitors and Flywheels: Milliseconds to Seconds
Supercapacitors and flywheels are the sprinters of the energy-storage world. These technologies respond the fastest. They store little total energy, but they discharge and charge at very high power almost instantly.
Supercapacitors store energy electrostatically, in the electric field between two plates. Flywheels store it kinetically, in a spinning mass, usually turning at 10,000–50,000 RPM in a vacuum enclosure to cut friction losses. Neither relies on a chemical reaction, so response time stays extremely short — typically under 10 milliseconds.
Both technologies have low energy density but very high power density. A typical grid-scale flywheel array delivers 1–20 MW of power but stores only 15–30 minutes of energy at full power, often far less. Supercapacitors sit at a similar ratio. This makes them a poor fit for energy shifting, but ideal for repeated, high-power pulses.
Cycle life is a real advantage here. Flywheels and supercapacitors tolerate millions of charge-discharge cycles with minimal degradation, since there’s no chemical wear. Li-ion cells wear down chemically with every cycle. This is why flywheels often serve applications with constant, rapid cycling — data center backup, and short-interval frequency regulation markets.
Grid role:
Synthetic inertia and power smoothing. They mimic the stabilizing effect that spinning turbines once gave for free, and they absorb fast transients before those transients spread. In some markets, flywheel farms compete directly in fast frequency response auctions, alongside Li-ion BESS.
Limitation:
These systems sustain output for seconds to a couple of minutes. They cannot shift energy across hours, and they add cost when a project only needs a small power boost, since the equipment is comparatively expensive per kW.
Li-ion BESS: Milliseconds to 4+ Hours
Li-ion batteries dominate current grid-scale deployments. They combine fast response with real energy capacity. Lithium-ion BESS has become the most versatile storage technology for many grid applications.
BESS respond fast enough for FFR and frequency reserves. They also hold enough energy for voltage support and ramp-rate control over longer windows. Most utility-scale systems today run 1 to 4 hours of duration, with 4-hour systems as the common default. Sizing usually uses the C-rate convention: a 1C system discharges its full rated energy in one hour, while a 0.25C system — a 4-hour design — discharges more gently and suits daily energy shifting better.
The Battery Technology That Dominates-LFP vs. NMC
Two chemistries dominate grid-scale projects today. Lithium iron phosphate (LFP) leads new installations, because it offers better thermal stability and longer cycle life than nickel manganese cobalt (NMC). NMC still appears in some earlier projects and in applications where energy density per unit volume matters more. LFP systems typically reach 6,000–10,000 cycles at 80% depth of discharge before falling to 80% of original capacity. This makes LFP the safer default for utility-scale tenders, especially where fire risk and site footprint both matter.
Each BESS also needs a battery management system (BMS) and a power conversion system (PCS). The BMS monitors cell voltage, temperature, and state of charge, and it balances cells to prevent thermal runaway. The PCS converts DC battery output to grid-synchronous AC, and it’s the component that actually executes fast frequency response commands. Response speed in the field depends as much on PCS control logic as on the battery cells themselves.
Grid role:
the versatile middle of the spectrum. Li-ion covers FFR, reserve replacement, voltage support, ramp control, and shorter-duration energy shifting. A single asset can often stack several of these services — for example, providing frequency response during normal operation, then discharging into the evening peak.
Limitation:
economics beyond 4–6 hours. Extending duration means adding more cells, since energy scales with cell count while power electronics stay largely fixed. Past 4–6 hours, cost outpaces benefit compared to other technologies. Degradation also accelerates with deeper, more frequent cycling, and thermal management gets harder as system size and ambient temperature both rise — a relevant factor for installations in high-ambient climates.
Flow Batteries: Minutes to Hours
Flow batteries separate power and energy. This is their key design difference from Li-ion.
Power capacity depends on the size of the electrochemical stack. Energy capacity depends on the size of the electrolyte tanks. Engineers can scale each independently. This decoupling avoids the cost penalty Li-ion faces at long duration — adding hours of storage means adding tank volume, not more expensive stack hardware.
Vanadium redox is the most mature flow battery chemistry today. Unlike Li-ion, the electrolyte doesn’t degrade chemically over the system’s life, so flow batteries can sustain very high cycle counts — often 15,000–20,000 cycles — with minimal capacity fade.
Grid role:
renewable shifting and congestion relief. A flow battery can move several hours of solar or wind output to a different part of the day. It can also manage flow on constrained transmission corridors, by charging when a line is underused and discharging as it nears its thermal limit.
Limitation:
response speed and footprint. Flow batteries respond slower than Li-ion — typically seconds rather than milliseconds. They also need a larger site footprint for the same energy capacity, since electrolyte tanks take up considerably more space than an equivalent battery rack.
Pumped Hydro and Hydrogen: Hours to Seasonal
These two technologies anchor the long-duration end of the spectrum.
Pumped hydro is the incumbent that has been the dominant grid-scale storage technology for decades. It uses elevation change and water mass to store large amounts of energy — pumping water uphill to a reservoir when power is cheap or surplus, then releasing it through turbines when power is needed. It requires specific geography — an elevation difference and two reservoir sites — and long construction lead times, often five to ten years for a large project.
Hydrogen is the newer option. Electrolysis produces the hydrogen. Storage holds it, typically as compressed gas or in underground caverns for large volumes. Fuel cells or turbines reconvert it to power. Hydrogen needs less specific geography than pumped hydro, but its round-trip efficiency remains lower today — often 35–50%, compared to 70–85% for pumped hydro.
Grid role: energy adequacy. These technologies provide the long-duration reserve that covers extended renewable droughts or seasonal demand swings. They don’t manage day-to-day frequency.
Limitation: agility and cost. Neither technology responds at millisecond speed. Both carry high capital cost and long development timelines, which makes them a planning-horizon decision rather than a project-level one.
What’s Next
Part 3 covers the practical side: how this technology mix affects substation design and tendering scope.