The model behind the numbers
A naive runtime estimate divides battery energy by load: two 12 V 9 Ah batteries hold 216 Wh, so at 432 W you might expect 30 minutes. Real 1500 VA units are rated closer to 10 minutes at that load. The gap has three causes, and the calculator models each one.
1. Rate-dependent capacity (Peukert effect)
A VRLA battery's amp-hour rating is measured over 20 hours. Discharged in 10 minutes, the same battery delivers well under half of that capacity, because the chemical reaction cannot keep up and voltage collapses to the cutoff sooner. We use Peukert's relationship referenced to the 20-hour rating:
For small VRLA blocks, an exponent k of about 1.25 reproduces published UPS runtime charts reasonably well. LiFePO4 cells are modeled with k = 1.05, reflecting their much flatter behavior at high rates, and 95% usable capacity.
2. Inverter losses and overhead
Converting 24 V DC into 120 V AC costs energy. The efficiency field covers conversion loss; a fixed 8 W covers the control board, display and fans.
3. Battery health
Capacity fades with age and heat. A battery at 70% health gives roughly 70% of the rate-adjusted capacity. The battery life estimator shows how quickly that happens in your conditions.
Worked example: one UPS, three loads
| Load | Watts | Naive estimate | Modeled runtime |
|---|---|---|---|
| Router, ONT and NAS | 100 | about 110 min | about 57 min |
| Office PC and two monitors | 250 | about 46 min | about 19 min |
| Gaming PC at full load | 450 | about 26 min | about 9.5 min |
At 100 W the model reaches the energy ceiling of the battery minus losses, which is why light loads gain so much: going from 450 W to 100 W (4.5 times less load) gives about six times more runtime.
Plan for the battery you will have in year three
Sizing against a brand-new battery is the most common reason a UPS "suddenly" fails during an outage. Set battery health to 70% and check whether the runtime still covers your shutdown. If it does not, size up now or schedule earlier replacements.
When the estimate will be wrong
- Cold rooms: below about 59 °F (15 °C) lead-acid capacity drops noticeably; expect shorter runtimes in garages and unheated spaces.
- Very light loads: UPS self-consumption dominates. Some units also shut down after a fixed time or below a minimum load.
- Low-battery shutdown settings: many UPS units and shutdown agents trigger when a few minutes remain, not at zero.
- Parallel strings and external packs: enter the equivalent: for two identical strings in parallel, double the amp-hours. See extended runtime packs.
Frequently asked questions
How long will a 1500 VA UPS last?
A typical 1500 VA consumer unit with two 12 V 9 Ah batteries lasts roughly 3 to 5 minutes at full load, about 10 minutes at half load, and 40 minutes to an hour at 100 W. Runtime depends far more on your load in watts than on the VA rating.
Why does halving the load more than double runtime?
Lead-acid batteries deliver less of their rated capacity when drained quickly, an effect described by Peukert's law. At lower current more of the stored energy becomes available, so runtime grows faster than the load shrinks.
How accurate is this runtime estimate?
For a healthy battery near room temperature it is typically within about 20% of manufacturer runtime charts for common consumer units. Real results vary with the exact battery, inverter design, temperature and age. Use the battery health field to model an older battery.
Does the UPS itself use power on battery?
Yes. The inverter, control board, display and fans consume power, typically several watts to a few tens of watts depending on size. The calculator includes a fixed 8 W overhead and the inverter efficiency you set.
What inverter efficiency should I use?
On battery, most small UPS inverters convert roughly 85 to 92% of battery power to AC output. The default 88% is a reasonable middle value. Larger business units are often more efficient at moderate loads.