Energy

Why a Flywheel Is Great for Seconds and Useless for Hours

By Lucian — builder & engineer, LK Forge

A spinning flywheel can charge and discharge almost instantly, but the same bearing and windage drag that lets it respond so fast also bleeds the stored energy away the moment it has to wait. We swept the Flywheel Energy Storage lab's model across hold time and standby rate for a 500 kg, 0.5 m rotor to measure exactly how fast that self-discharge eats round-trip efficiency, and what actually sets how long the energy lasts.

 ·  5 min read  ·  measured from the standby self-discharge benchmark

84.64%
round-trip efficiency with no hold at all (charge, then discharge immediately)
7.68%
round-trip efficiency after an 8-hour hold at a 30%/h standby rate
13.86 h vs 0.87 h
stored-energy half-life at a 5%/h vs 80%/h standby rate
10.281 kWh
usable energy capacity of the 500 kg, 0.5 m rotor at 12,000 rpm

Why waiting has a physical cost

A flywheel stores energy as motion, not as a chemical or gravitational potential that can sit still indefinitely. The moment the rotor is up to speed, bearing friction and windage drag start pulling kinetic energy off it, and both forms of drag scale with how fast the rotor is spinning — so the loss is exponential in hold time, not linear. There's no idle mode: every hour the rotor spends waiting between charge and discharge is an hour it spends losing energy at a rate set entirely by the standby rate.

round-trip % = (machine efficiency)² × exp(−standby rate × hold hours)

That's why every row in the benchmark below moves the same direction: nothing about the rotor changed between hold-time rows except how long it waited, and nothing about the hold time changed between standby-rate rows except how leaky the bearings were assumed to be.

Round-trip efficiency and retained energy fall together

Round-trip efficiency (amber) and retained energy (blue), plotted against hold time for the 500 kg rotor at a 30%/h standby rate. Both curves decay exponentially, with round-trip efficiency always slightly below retained energy because it also carries the fixed machine-efficiency penalty.

0% 25% 50% 75% 100% 02468 hold time (hours) percent
retained energy round-trip efficiency

Lower drag stretches the half-life

Stored-energy half-life against standby rate. Lower standby rates (better bearings, harder vacuum) buy a much longer half-life — the relationship is inverse, not linear, so the biggest gains come from cutting drag when it's already low.

0 4 8 12 16 020406080 standby rate (%/h) hours

The exact numbers

Every row below is a direct sample from the lab's model: a 500 kg, 0.5 m rotor spinning to 12,000 rpm, 0.92 machine efficiency, 30%/h standby rate, for the hold-time sweep.

Hold time (h) Round-trip Retained energy
0 84.64% 100%
0.1 82.14% 97.04%
0.25 78.52% 92.77%
0.5 72.85% 86.07%
1 62.7% 74.08%
2 46.45% 54.88%
4 25.49% 30.12%
8 7.68% 9.07%

And the separate standby-rate sweep, for the same rotor:

Standby rate (%/h) Half-life
5 13.86 h
10 6.93 h
20 3.47 h
40 1.73 h
80 0.87 h

Rotor: 500 kg mass, 0.5 m radius, 12,000 rpm top speed, 0.92 machine efficiency, solid-cylinder inertia. Generated 2026-09-11.

Try it yourself

Open the lab, set your own rotor mass, radius, top speed, hold time, and standby rate, and watch round-trip efficiency and retained energy move in real time.

Open the Flywheel Energy Storage lab →
Share this X Facebook Reddit

Related reading

Common questions

What is round-trip efficiency for a flywheel, and why does it fall so fast with hold time?

Round-trip efficiency is the fraction of the electrical energy used to spin the rotor up that you get back out when it spins the generator back down. In this model it comes from two factors multiplied together: machine efficiency (motor and generator losses, applied once each direction) and standby retention (how much kinetic energy survives bearing and windage drag during the hold between charge and discharge). Machine efficiency is fixed once the hardware is built, so it doesn't change with hold time; retention does, and it decays exponentially the whole time the rotor just sits there. For this 500 kg rotor at a 30%/h standby rate, charging and discharging immediately round-trips 84.64% of the energy, but holding for 8 hours drops that to 7.68% — 76.96 percentage points lost purely to sitting idle, with the machine efficiency, mass, and speed all unchanged.

Why do flywheels self-discharge so much faster than a battery sitting on a shelf?

Because the loss mechanism is mechanical drag, not a slow chemical side-reaction. A spinning rotor loses kinetic energy continuously to bearing friction and windage (air or residual gas resistance), and both scale with speed, so the energy bleeds off exponentially for as long as the rotor keeps spinning — there's no way to "pause" a flywheel the way you can leave a battery unpowered. The benchmark shows how fast this adds up: at a 30%/h standby rate, retained energy falls from 100% at zero hold to 97.04% after just 0.1 hours, 86.07% after 0.5 hours, and 9.07% after 8 hours — a loss of 90.93 percentage points over a single day. That is exactly why flywheels are built for short, frequent cycles (grid frequency regulation, ride-through, seconds-to-minutes buffering) rather than hours-long storage: the standby rate itself is the whole story, and a lower-drag design (better bearings, a harder vacuum) directly buys a longer half-life. Cutting the standby rate from 80%/h to 5%/h stretches the energy half-life from 0.87 hours to 13.86 hours, a 15.9× improvement, with nothing else about the rotor changed.

What actually sets a flywheel's energy capacity — is it the mass, the radius, or the speed?

All three, but not equally. Stored kinetic energy is proportional to mass and to the square of radius (through the rotor's moment of inertia), but to the square of angular speed — so top speed dominates. Doubling mass or radius roughly doubles or quadruples capacity respectively, but doubling top speed roughly quadruples it too, which is why real flywheels are engineered to spin as fast as the rotor material can tolerate rather than built simply bigger or heavier. This 500 kg, 0.5 m rotor spinning up to 12,000 rpm stores 10.281 kWh of usable energy (energy above the minimum useful speed). None of that capacity number depends on the standby rate or hold time in the charts above — capacity is a property of the rotor and its top speed; retention and round-trip efficiency describe what happens to that stored energy while it waits.