SpecCalc Hub Guide
UPS runtime: why real runtime differs from simple math
UPS batteries, runtime curves, power factor, and aging explained.
Runtime curves
Manufacturers often publish runtime charts because efficiency changes with load.
Battery aging
Old or hot batteries can deliver much less energy.
Why it matters
A UPS is rated in VA and often also in watts. The connected load, power factor and battery health determine whether the UPS can support the load and for how long.
Formula or method
The simplified model compares load power with UPS VA and power factor, then estimates runtime from usable battery energy and efficiency. It flags load conditions that exceed VA x PF planning capacity. This is useful for fast screening, but real UPS behavior can still depend on battery cutoff logic, internal temperature, aging, transfer design and manufacturer runtime curves.
Worked example
A 1000 VA UPS with 0.6 PF has about 600 W of planning watt capacity. A 400 W load is within that simplified capacity, but runtime still depends on battery energy, age and UPS efficiency. A seemingly small change from 400 W to 550 W can reduce runtime sharply because UPS runtime curves are usually nonlinear near the upper part of the load range. For instance, a 1000 VA UPS might run a light 100 W networking load for 40 minutes, but if you increase the load to 550 W (near its 600 W limit), the runtime might crash to just 3-4 minutes instead of proportionally scaling down. This non-linear drop is why manufacturer runtime curves always win over simple math.
How to use in practice
Use this page when someone expects a UPS label to guarantee a precise backup time. It helps you explain why the same UPS can behave differently under server, router, workstation or mixed office loads, and why the manufacturer runtime chart remains important.
Comparison table
| Topic | Value | Note |
|---|---|---|
| VA | Apparent power rating | Not the same as watts. |
| W | Real load power | Often the practical limit. |
| Runtime curve | Manufacturer data | Best source when available. |
Checklist before using the result
- Check the units, equipment nameplate and real operating scenario first instead of trusting the nearest rounded number.
- Write down which factors the method models directly and which still need separate checking: losses, installation conditions, tariff structure or manufacturer behavior.
- Be explicit about the use case for the estimate: quick planning, option comparison, budgeting, team explanation or preliminary technical review.
- Before practical use, compare the result with equipment documentation, local rules and the real limits of the site.
Common mistakes
- Using VA as if it were watts.
- Ignoring battery age and temperature.
- Expecting linear runtime at all load levels.
Limitations
- The model does not replace manufacturer runtime charts.
- Battery condition can dominate the result.
- The result is not a guarantee of backup duration.
- It does not model battery replacement history, inverter waveform sensitivity or load inrush after an outage event.
What to check next
After reading the guide, open the related calculator and test at least two realistic scenarios: a normal case and an edge case. Then compare the result with equipment documentation, local rules, site conditions and the factors that this simplified method intentionally does not model.
When to use the calculator
Use the UPS Runtime Calculator to check whether a load is within simplified UPS capacity and to estimate runtime before consulting manufacturer data.
FAQ
Why does real UPS runtime differ from the estimate?
Efficiency, battery age, discharge rate, temperature and UPS cutoff behavior can all change runtime.
Can I use VA as watts?
No. Convert using power factor or use the UPS watt rating when available.
Why can the runtime drop faster than load seems to increase?
UPS runtime is rarely linear. As the load approaches the practical limit, efficiency, battery stress and cutoff behavior can make the available minutes fall much faster than a simple proportion suggests.
Related guides
Last reviewed: 2026-05-29