Battery Runtime Calculator: Sizing for Autonomous Loads
Battery Runtime Calculator: Sizing for Autonomous Loads
“How long will my battery run this load?” is the most common question in autonomous and backup design, and the answer is rarely the simple division everyone tries first. Runtime depends on usable capacity, the discharge rate, temperature and the cutoffs your BMS enforces. This guide gives you a working runtime method and a sizing approach for loads that must run unattended—sensors, telecom, remote pumps—where a miscalculation means a silent failure in the field. A lithium battery manufacturer will quote nameplate amp-hours, but the runtime you actually get is the number that matters, and it is almost always smaller than the catalog implies once real conditions are applied.
The basic runtime formula
Start from nameplate capacity in amp-hours, divide by the load current to get raw hours, then multiply by usable depth-of-discharge and by a Peukert-style efficiency factor. For a 100 Ah LFP pack running a 5 A load: 100 / 5 = 20 h raw; at 90% DoD that is 18 h; at a moderate 0.05C rate the efficiency factor is near 1.0, so real runtime is about 18 h. The trap is assuming 100% of nameplate is usable—on lithium you safely get 80–90%, and on lead-acid only 50%, which halves runtime for the same label. Always compute in usable amp-hours, not catalog amp-hours, because the catalog number is a best-case laboratory figure that no field deployment ever fully realizes.
Why C-rate eats runtime
Capacity is stated at a test C-rate, often 0.2C. Pull harder and delivered capacity drops. A 100 Ah cell rated at 0.2C might deliver only ~92 Ah at 1C and less at 2C because internal resistance and heating consume energy. For an autonomous load drawing a steady 10 A from a 100 Ah pack (1C), expect roughly 9 h of usable runtime, not 10. Request the discharge curve at your actual C-rate; the curve is the honest spec. A battery application solutions design that ignores C-rate will over-promise runtime and under-deliver in the field, which is precisely where unattended loads cannot tolerate surprises and a dead sensor becomes a blind spot nobody notices until the data gap is already expensive.
Worked runtime table
| Load current | Pack | C-rate | Usable DoD | Realistic runtime |
|---|---|---|---|---|
| 5 A | 100 Ah LFP | 0.05C | 90% | ~18 h |
| 10 A | 100 Ah LFP | 0.1C | 90% | ~9 h |
| 10 A | 100 Ah lead-acid | 0.1C | 50% | ~5 h |
| 20 A | 100 Ah LFP | 0.2C | 90% | ~4.5 h |
The table makes the DoD penalty obvious: the lead-acid pack at 50% usable delivers barely half the runtime of LFP at 90% for the same label. C-rate and chemistry together decide the number you can bank on, and the field engineer who remembers both will never be caught out by a “100 Ah” pack that dies at hour five.
Sizing for autonomous telecom loads
A remote telecom site drawing 200 W at 48 V (about 4.2 A) with a 24 h autonomy target needs: 4.2 A × 24 h = 100.8 Ah raw; at 90% DoD that is 112 Ah nameplate, rounded to a 120 Ah 48 V rack. Add 15% for converter losses and cold and you land near 140 Ah. The load is steady and low C-rate, so LFP is ideal; a home energy storage system rack repurposed for the site works well because the duty is identical to stationary solar buffering. Size to the worst winter temperature and confirm the BMS low-temperature charge limit if the site sits below freezing, because a telecom outage during a cold snap is exactly when the bank is most stressed and least able to recharge.
Sizing for remote sensors and IoT
Low-power sensors running milliampere loads can stretch tiny packs for months. A 10 Wh/day sensor on a 100 Wh LFP pack with 90% DoD has 90 Wh usable, giving ~9 days—too short—so you either enlarge the pack or add a small solar trickle. Autonomy math is the same as for big banks, just smaller numbers, and the failure mode is identical: a dead sensor is a blind spot. Because these loads are unattended, build in 30% runtime margin rather than 10%, since you cannot swap the pack on a whim. A sodium cell could help in cold sensor housings where LFP charge-limiting would otherwise stall the recharge, and the larger sodium footprint is irrelevant when the whole device is the size of a lunchbox.
Sizing for backup pumps
A backup sump or well pump may sit idle for months then draw a heavy burst. Size for the burst C-rate, not the average, because the peak is what the BMS and connectors must survive. A 400 W 24 V pump draws ~17 A; on a 100 Ah 24 V pack that is ~0.17C continuous, comfortable, but the startup surge may hit 50 A for seconds and needs busbars and a BMS shunt rated for it. A semi-solid-state battery is unnecessary here—weight is irrelevant for a fixed pump—so plain LFP is the rational choice, sized for the surge and the required run-hours, with connectors torqued to spec because a loose lug at high C-rate is a heater, not a conductor.
Temperature derating in runtime
Cold reduces both capacity and charge acceptance. A pack that delivers 18 h at 25 °C might deliver 14 h at 0 °C and refuse to charge without low-temperature management. For autonomous Arctic or alpine sites, either heat the enclosure or specify sodium, which keeps more capacity and charges more reliably in the cold. A sodium-ion battery in an unheated cabinet can be the difference between a site that self-sustains and one that needs a winter service call. Derating is not pessimism; it is what the field demands, and a runtime calculator that ignores temperature will be wrong by exactly the margin that triggers the failure you were trying to prevent.
BMS cutoffs and the “dead” threshold
Your runtime ends at the BMS low-voltage cutoff, not at true zero. Set that cutoff conservatively—say 2.8 V per LFP cell—and remember the inverter’s own low-voltage trip may arrive first. The effective runtime is bounded by the higher of the two thresholds, so coordinate them. A distributed BMS with state-of-health reporting lets you see capacity fade and re-size before a load goes dark, which is the whole point of autonomous design: predict the failure, don’t experience it. The cutoff is also a warranty boundary—deeply discharging past it can void the cell guarantee, so the BMS that trips early is protecting both the runtime and the contract.
Building a simple runtime calculator sheet
Turn the method into a one-page calculator: enter nameplate Ah, usable DoD, load current, and a C-rate efficiency factor from the discharge curve. It outputs usable amp-hours, raw hours, and realistic hours. Distribute it to your field engineers so every autonomous spec is computed the same way instead of by guess. The sheet also exposes mismatches: a “100 Ah” pack that delivers 80 Ah at your 1C duty is an 80 Ah pack in practice, and pricing it as 100 Ah is where margins quietly leak. A ten-minute calculator prevents a mis-specification that surfaces only after installation in a place nobody wants to visit, and it gives you a defensible number to show the client when the load runs precisely as predicted.
Field validation and re-rating
After deploying, log the first month of actual runtime against the prediction and re-rate the pack. Real loads rarely match the estimate—a sensor that draws 12 Wh/day instead of 10 shortens life by 20%—and the gap tells you whether to enlarge the bank or trust the model. A disciplined operator treats the first seasonal cycle as calibration, then sets the replacement schedule from measured data rather than the datasheet. This closes the loop between the calculator and the field, and it is the difference between a runtime number you believe and one you merely hoped was correct when the site was commissioned months ago.
Frequently asked questions
Why is my runtime shorter than the math? You likely used nameplate Ah; switch to usable Ah at your C-rate and temperature and the number drops to reality.
Does C-rate really change runtime? Yes—pulling above the test C-rate sheds 5–15% of capacity through resistance and heat.
How much margin for autonomous loads? 30% on unattended sites versus 10% on serviced ones, because you cannot swap the pack quickly.
Is sodium good for cold autonomous sites? Yes—it keeps more capacity and charges better below freezing than LFP without heater mats.
Written by Karl at China Battery Technology. Give us your load current and autonomy target and we will size it—request a quote for a runtime-verified pack.
Related Reading
- battery application solutions hub
- How to Choose Lithium for Your RV: A Step-by-Step Method
- How to Size a Forklift Battery: Throughput-Based Method
