Skip to main content

How long will your UPS actually last?

Not what the box claimed when you bought it — what the one under your desk will do today, with what's actually plugged into it. You don't need to know anything about batteries.

1 · Which UPS do you have?

Type a brand or model — “APC 1500”, “CyberPower CP”. It's on the label, usually on the back.
The input people forget — and it roughly halves runtime by year five.

3 · What's plugged into it?

W
W
W
W

Total load: 166 W — that's 18% of this unit's 900 W rating.

40 min

is roughly what you'd get at 166 W on a battery no idea. ⚠️ This is below the lightest load the manufacturer publishes, so it's an estimate beyond their figures.

What unplugging something buys you

Runtime bought by removing load is free. Runtime bought with batteries is not — and because of how batteries behave, halving the load more than triplesthe time, rather than doubling it. That's the median across 399 certified small-office units.

Take this off the UPSSavesYou'd getGain
Desktop PC100 W2h 54m+2h 13m
Network switch30 W55 min+15 min
Router / firewall20 W50 min+9 min
Monitor (24–27")16 W47 min+7 min

Monitors are usually the easy win: a screen going dark during a cut costs you nothing, and it is often the second-biggest draw on the strip. Printers should not be on a UPS at all — a laser printer's heater pulls more on its own than everything else combined.

Is that long enough?

That's the question the number can't answer for you. 40 min is plenty to save your work and shut down properly. It is not enough to keep trading through an outage.

Rough is fine — lost billable time, staff sitting idle, work pushed to tomorrow.

Why we ask for the model instead of the battery

Every other runtime calculator asks for battery amp-hours and voltage. It's a fair thing to ask an electrical engineer and a useless thing to ask a business owner, which is why most people give up and guess.

There's a better route. Manufacturers already publish, for each certified model, how long it runs at full load and at half load. Those figures sit in a public US Environmental Protection Agency dataset covering 851 models from 32 brands. So we ask which unit you own and read its own published numbers, then fit the curve between them. The answer comes from your manufacturer, not from our formula.

The thing that surprises people

Runtime doesn't scale the way you'd expect. Halve the load and you might assume you get double the time. Across the certified small-office units in that dataset, the median is more than three times the runtime, not double — batteries give up proportionally more capacity the harder you drain them.

That is why the table above matters more than the headline number. Taking two monitors off a UPS is free and can add more minutes than buying a larger unit would. Runtime you buy by removing load is the cheapest runtime there is.

What this can't tell you

Battery age is an estimate. A sealed lead-acid battery loses capacity as it ages, and by year five a unit often holds around half what it did new — but the only way to know yours is to test it. Most UPSs have a self-test, and many report the real figure on their display or management software. If the number above matters to you, test rather than trust it.

Below half load we're also extrapolating past what the manufacturer publishes, and real curves flatten out there. We flag it on screen when that applies rather than quietly printing a confident number.

Where the numbers come from

Model runtimes and watt ratings: ENERGY STAR certified product data, published by the US EPA and in the public domain, last refreshed 2026-09-18. Device wattages are a mix of measured ENERGY STAR figures and clearly-labelled estimates — there is no public measured dataset for network gear or phones, and we'd rather say so than pretend otherwise.

If the answer worries you

A UPS that gives you eight minutes is doing its job — that's enough to save work and shut down cleanly. If you need to keep trading through an outage, that's a different conversation, and usually a cheaper one than people expect once you separate what genuinely has to stay up from what merely happens to be plugged in.

That's the sort of thing we sort out for clients. If you'd like a second pair of eyes on it, get in touch — there's no charge for a conversation. Related: what Windows 10 now costs your business and the boring thing that saves your business.

Book a consult

Let's talk about what's not working

The first conversation is free and pressure-free. You talk, we listen, and by the end you'll have at least one concrete thing you can act on — whether you work with us or not.

30 minutes, via phone or video

We come prepared if you share your brief first

Flexible scheduling, including evenings and weekends

What to expect

0–5 min

Context

We learn about your business and what's not working

5–20 min

Diagnosis

We ask specific questions and share our initial read

20–30 min

Next steps

You leave with at least one concrete recommendation

“James created something that I have no doubt is the reason my practice has been at capacity for years.”

— Donovan Bigelow, LMHC