The simple answer, and why it is usually wrong
The first-order calculation is dividing available energy by load:
That is a useful starting point, but it overstates runtime for two reasons. First, most batteries cannot be discharged to zero without damage, so only the usable depth of discharge counts. Second, and more importantly, a battery's capacity is not a fixed quantity — it falls as the discharge current rises. This is the Peukert effect.
Understanding Peukert
Battery capacity is normally quoted at a specific discharge rate. For lead-acid that is the 20-hour rate: a 100 Ah battery delivers 5 A for 20 hours. Discharge it at 50 A and you will not get two hours — you will get noticeably less, because the internal resistance losses increase and the active material cannot keep up with demand.
Where H is the rated discharge period in hours, C is the rated capacity in amp-hours, I is the actual discharge current, and k is the Peukert exponent for the chemistry. A k of 1 would mean capacity is independent of current. Real values are 1.05 for LiFePO4, 1.15 for AGM, 1.25 for flooded lead-acid and around 1.45 for a small AGM in poor condition.
The practical lesson: a high current draw hurts lead-acid far more than lithium. If your load is bursty, lithium's flatter capacity curve is worth paying for.
Chemistry comparison
| Chemistry | Usable DoD | Peukert k | Cycle life | Notes |
|---|---|---|---|---|
| Flooded lead-acid | 50% | 1.25 | 300–500 | Cheapest per kWh, needs ventilation and watering |
| AGM / sealed | 60% | 1.15 | 400–700 | Maintenance free, better high-current performance |
| LiFePO4 | 90% | 1.05 | 2,000–5,000 | Highest upfront cost, lowest cost per cycle |
| Li-ion NMC | 90% | 1.03 | 500–1,000 | Lightest, but a fire risk without proper management |
On cost per delivered kilowatt-hour over the life of the bank, LiFePO4 usually wins despite the higher purchase price. On initial cost, flooded lead-acid wins easily.
Two worked examples
1. A 100 Ah AGM running a 300 W AC load
On a 12 V bank at 60% depth of discharge and 90% inverter efficiency, the usable energy is 100 × 12 × 0.6 = 720 Wh, and the load draws 300 ÷ 0.9 = 333 W from the battery. The simple answer is 2 hours 10 minutes. With the Peukert correction for AGM the realistic figure falls to roughly 1 hour 50 minutes — about 15% less. The discharge current is 28 A, a C-rate of 0.28, which is moderate for AGM.
2. The same load on LiFePO4
At 90% depth of discharge, usable energy is 1,080 Wh. Allowing for inverter losses the simple answer is 3 hours 15 minutes, and the Peukert correction barely moves it because k is 1.05. The same nominal capacity therefore delivers nearly 80% more runtime — which is the real reason lithium banks are increasingly specified even at several times the purchase price.
Factors the formula cannot capture
- Temperature. Lead-acid loses roughly 20% of its capacity at 0°C and much more below that. Lithium also suffers in cold, and LiFePO4 should not be charged below freezing at all without a heated battery.
- Age. A battery at end of life delivers a fraction of its rated capacity. The Peukert exponent rises as the plates sulphate, so an old battery behaves worse than a new one at the same current.
- Discharge cut-off. Inverters shut down at a low-voltage threshold, so the last few percent of the calculated capacity is often unreachable in practice.
- Standby loads. An inverter left switched on draws its own idle current, commonly 10 to 30 W, which can halve the runtime of a small system.
- Voltage sag. Under heavy load the terminal voltage falls temporarily, which can trip a low-voltage cut-off long before the battery is actually empty.
How this calculator is verified
The Peukert equation and the exponents used here follow the published behaviour of each chemistry at typical discharge rates. Where manufacturers publish their own Peukert exponent or a capacity-versus-current curve, use their figures — they are specific to the cell and are more reliable than a generic value.
- IEEE — IEEE 1188 and related standards for stationary battery qualification and testing.
- NIST — reference material on electrochemical energy storage measurement.
- NFPA 70, National Electrical Code — Article 480 for storage battery installation requirements.
Peukert exponents and worked examples last verified: 19 September 2026.