Backup runtime
How many hours your battery backup would really give you
Backup systems are sold on a runtime figure. Any runtime figure carries assumptions about the battery's age, the head the pump works against, and how much of the outage the pump spends resting. Change any one of those and the real number moves.
Runtime is not a property of a battery. It is a property of a battery, a pump, a head, and a storm — and the storm decides the duty cycle, which is what actually converts amp-hours into hours. This tool derives runtime from those, rather than repeating a marketing number back at you.
What your backup would really do
Illustrative assumption — not a market quote: every value filled in above is a worked-example starting point chosen by this site, not a measurement of your house. Replace each one with your own.
How long this backup keeps up
7.2 hours
Runs out before the outage ends.
Runtime, low to high 5.0 to 9.0 hours
Assumed, no direct source: the low-to-high band is 0.7x to 1.25x the runtime, a width chosen by this site.
Biggest swing factor Battery capacity at pump current. A battery can deliver less than its label at a pump's high current, by an amount that depends on the battery; this tool allows for it only through its assumed 50% usable fraction.
- This falls 2.8 hours short of the 10-hour outage entered. In this tool's arithmetic, doubling battery amp-hours doubles runtime; a different backup pump changes it in proportion to its flow per amp.
Not counted in this runtime
- Charger and controller draw during the outage, which this tool does not model.
- Cold-temperature capacity loss, which this tool does not model.
- Any claim that the backup will start. A backup that has never been tested under load is an assumption, not a system.
- Whatever the primary pump would have done. This figure is for the backup alone, on battery, with no utility power.
The duty cycle is the whole trick
In the example loaded above, a backup pump moving 14 gallons a minute against inflow of 6 gallons a minute does not run continuously. It runs about 43 percent of the time and rests the remainder, which is why the battery lasts about 2.3 times its raw pumping hours here. The raw pumping capacity is 3.1 hours of actual motor time; the duty cycle stretches that into the 7.2 hours quoted above.
That relationship also explains a failure mode. If inflow rises to meet the backup pump's capacity, the duty cycle goes to 100 percent, and runtime falls to the raw pumping hours. If inflow exceeds the backup's flow at your head, the water level rises regardless of battery size — at that point you do not have a runtime problem, you have a capacity problem, and adding batteries does not touch it.
This is the argument for checking the backup pump against the same head figure as the primary. A backup's flow at a lower head is not its flow at yours, and the shortfall shows up as duty cycle before it shows up as failure.
Why we do not trust the amp-hour label
Two separate deductions stand between a battery's printed capacity and the energy it will actually give a sump pump.
The first is depth of discharge. We assume 50 percent of the nameplate is usable (see the disclosure below) rather than treating the full label as available.
The second is rate. A battery discharged at a pump's high current can deliver less than its label. The example pump above draws 11 amps. We do not model the high-current capacity loss separately; the 50% usable assumption is meant to allow for it. The low-to-high band (0.7x to 1.25x of the runtime) is a width chosen by this site — assumed, no direct source.
In the example, the 4-year-old 100 amp-hour battery derates to 68 amp-hours at 8 percent per year, and then to 34 usable amp-hours after the discharge floor. The label said 100. That gap is the point of this section.
An untested backup is an assumption
A backup can fail without a flat battery. A system can be installed, never exercised, and quietly stop being a system — a charger that failed and never announced it, a float that stuck in place, a battery that went open-circuit and sat there looking installed.
Test it the way it will be used. Kill the power at the breaker, pour water into the basin, and watch the backup start, clear the pit, and shut off. Do that on a regular schedule and note the date somewhere you will find it. If the backup has an alarm, confirm the alarm sounds when the water reaches it, because an alarm nobody has ever heard is not evidence of anything.
A load test on the battery itself is worth more than any age-based rule, including ours. Ask a battery retailer or service shop for one.
Related tools
- Failure exposure — how deep a backup your own numbers actually justify.
- Pit inflow test — the inflow figure this page depends on.
- Head and discharge — the head your backup pump must also work against.