Watts, VA and power factor: read the watts
UPS and inverter boxes lead with VA because it is the bigger number. What your equipment uses is watts, and watts = VA × power factor. The power factor varies a lot by model: APC's Back-UPS BX1100C-IN is rated 1100 VA and 660 W (a power factor of 0.6), while Luminous's Zelio+ 1700 is 1.5 kVA and 1260 W (0.84). Two boxes that both say “1.1 kVA” can differ by more than 200 W in what they actually deliver.
So the first check is not backup time at all. Add up the watts you will plug in and compare it with the watt rating on the label. If the load is higher, the UPS will trip the moment mains fails, however big the battery is. The calculator shows this as a warning above the results.
Why a 150 Ah battery does not give you 150 Ah
Lead-acid capacity is quoted at a slow discharge, usually over 20 hours (C20) or 10 hours (C10). Pull current faster and you get fewer amp-hours out before the voltage hits the cut-off. Exide's own table for its NXT65-12 sealed battery shows it: 65 Ah over 20 hours, 58.5 Ah over 10, 49 Ah over 3, 36 Ah over one hour and 29.5 Ah over 30 minutes. At a one-hour drain you get a little over half of the label. This is Peukert's effect, and we model it with an exponent of 1.25, which tracks Exide's published points to within 13% — a little pessimistic between 1.5 and 5 hours, a little optimistic at 30 minutes.
The second deduction is depth of discharge. A battery that is drained flat every day dies early. Exide rates its InvaTubular range at 1,200 cycles to 80% depth of discharge; its NXT sealed UPS battery's cycle-life chart roughly doubles the cycles at 50% compared with 80%. Victron rates its LiFePO4 batteries at 2,500 cycles at 80% and 5,000 at 50%. The calculator plans around 80% for tubular and lithium and 50% for SMF/VRLA, and also shows the time to a full cut-off so you know what is left in an emergency.
Heat matters in Uttar Pradesh. Exide's float-life chart for the NXT series falls from roughly four to five years at 20 °C to under two years at 40 °C. A battery rack in a closed, unventilated cupboard in a Lucknow summer will not reach its rated life, and an aged battery holds less than this calculator assumes.
Worked example: 10 desktops and a router through a 4-hour cut
Say you measured each office desktop with its monitor at about 60 W on a plug-in meter while people were working, and the router and switch together draw 15 W. That is 10 × 60 + 15 = 615 W. A 1.5 kVA inverter rated 1260 W carries it at 49% of its rating, which is comfortable.
With 85% inverter efficiency the batteries supply 615 ÷ 0.85 ≈ 724 W. On a 24 V bank that is about 30 A. For four hours at 80% depth of discharge, the batteries must be able to run 5 hours to empty at 30 A. Plain watt-hour arithmetic says 151 Ah per battery. With the high-current derating, the calculator asks for about 199 Ah at C20 — so two 200 Ah tubular batteries, the size Exide sells as the InvaTubular IT750, rather than the 150 Ah pair the simple sum suggests.
Two practical points from the same sum. First, four hours is the edge for that bank; if your area's cuts run longer in summer, size for the longest cut you actually see, not the average. Second, a long cut followed by a short return of mains will not recharge 200 Ah batteries — a lead-acid bank needs hours of charging, so back-to-back cuts get shorter backup each time.
What this calculator ignores
It assumes a new, fully charged battery at around 25–27 °C. It does not model battery ageing, cold weather, the inverter's own idle consumption, voltage sag on long cables, or start-up surges from motors, compressors and laser printers. It treats the load as constant; real offices draw less at lunch and more at month-end. The Peukert exponent is fitted to one Exide sealed battery's table; tubular flooded batteries behave similarly but we have not fitted a tubular datasheet, and below about one hour the estimate can be around 10% optimistic.
Treat the answer as a sizing estimate, then measure: run the actual load on battery for an hour and check the battery voltage against the manufacturer's discharge curve.

