I Started With the 'Sell Me the Cheapest One' Approach
When I first started managing facility purchases for a ~150-person manufacturing company in 2021, I assumed UPS sizing was simple. You check the nameplate wattage on whatever you're plugging in, match it to an APC unit, and you're done. Simple.
That assumption cost us about $8,000 in unplanned downtime and replacement equipment over the next year. Not great for my budget, and way worse for my relationship with our operations team.
Here's what I've learned since then: the question everyone asks is 'what's the wattage?' The question they should ask is 'what does that device actually do when it starts up?'
The Big Misconception: It's Just a Run Time Game
Most people—and I was definitely one of them—focus exclusively on runtime when sizing a UPS. 'I need 20 minutes to safely shut down the server.' Fair enough. But that's table stakes.
The real trap is inrush current.
You don't see it on the spec sheet. But that air compressor you're thinking about putting on a battery backup? The motors in a lot of HVAC and VFD gear? They can draw two to three times their running amperage for a split second when they kick on. If your UPS isn't rated to handle that surge, it'll drop into bypass or shut down entirely.
I learned this one the hard way. We had a small coolant pump on a critical CNC machine—nameplate said about 4.5 amps. A colleague suggested an APC Back-UPS 675 (which is rated for around 350-400 watts). Looked fine on paper (I told myself at the time). First time a tool change cycled and the inverter kicked on, that APC hiccupped, the machine faulted, and we lost about 40 minutes of production time.
(Note to my past self: motor loads aren't resistive loads. They're unpredictable.)
That single event is why I now build in a 1.5x to 2x safety factor on any load with moving parts or switching power supplies. If the nameplate says 500W, I'm sizing for 1000W of UPS capacity for that device. Period.
What This Means for Battery Chargers and Compressors
You see a lot of folks searching for a 'battery charger and air compressor' UPS solution. It makes sense—these are common items in maintenance shops or labs. But this is a classic blind spot scenario.
A battery charger for a forklift or pallet jack can pull a massive surge as it re-energizes those lead-acid plates. An air compressor's motor startup is similar.
Most buyers look at the average running amperage. They don't consider the locked-rotor amperage. That's the problem. If you're an admin buyer and you're reading this, I'd strongly suggest checking the LRA on your motor-driven equipment before even thinking about which APC unit to buy.
In our 2023 facility audit, we realized our main air compressor was pulling a 65A surge on startup. Running load was ~18A. The UPS we originally spec'd for the controller was barely keeping the lights on (I still kick myself for that oversight).
The best approach? Use APC's online sizing calculator, but don't trust it blindly for inductive loads. Call your vendor and ask for the locked-rotor amps. If they can't provide it, assume a 2.5x multiplier on the running load. Trust me on this one.
What About the Newer Tech? (Trane TR1 Series VFDs, for Example)
I'm seeing more facilities move to variable frequency drives (VFDs) for HVAC pumps and fans. The Trane TR1 series, for instance, is a solid unit. But a VFD is an active switching device. It creates harmonics on the power line.
Putting a standard 'office-grade' UPS on a circuit with a VFD used to be a recipe for nuisance alarms. 'Constant switching to battery' was the most common complaint I heard from colleagues in our 2024 vendor consolidation project.
The solution? It wasn't just buying a bigger UPS. It was buying a different kind of UPS—one with online double-conversion topology. An APC Smart-UPS (SUA or SRT series) handles dirty input power way better than a Back-UPS ever can.
I made that mistake on a pilot project in 2022. We put a standard line-interactive UPS on a pump controlled by a Trane VFD. It lasted about a week before the battery was worn down by constant switching. Replacing it with an online UPS solved the issue completely (and we finally stopped getting alerts at 2 AM).
If you have any variable frequency drives or motor controllers on your critical equipment, treat a standard UPS like a gamble. Get a Smart-UPS or an industrial-grade unit.
The Hybrid Threat: When 'Cheap' Fuel Pumps Ruin Your Plan
Another topic you might be researching: 'how to check a fuel pump.' If you're managing a site with backup generators, you know that the fuel pump is a single point of failure under the grid. But testing it usually involves running it under load.
I remember a situation in our 2020 load bank test (we do these every two years). We hooked up a portable UPS to the fuel transfer pump controller to prevent it from losing its programming during a power outage test. The pump itself was on a separate circuit. We checked the controller's draw—about 150 VA. No big deal.
What we missed: the pump's contactor had a slight arc when it de-energized during the test. That sent a voltage spike back down the line. Our cheap UPS (a small Back-UPS) didn't have a filter robust enough to snub that. It immediately popped a fuse. We lost the controller's programming anyway.
A frustrating day.
I now use a dedicated APC surge suppressor and a battery backup with proper AVR (Automatic Voltage Regulation) for any fuel pump controller—something like a Smart-UPS 1000. It's overkill for 150 VA. It's exactly right for the reality of a dirty industrial environment.
(The most frustrating part: you can't simulate this stuff in a spec sheet. You only learn it after you've bought one too many surge protectors that didn't quite handle the job.)
Countering the Obvious Objection: 'This All Sounds Too Expensive'
I can hear the finance guy in my head. This sounds like I'm advocating always buying the biggest, most expensive online UPS on the market.
No. Not what I'm saying.
What I'm saying is: size for the surge, not the idle state.
If you properly evaluate the starting load and harmonic rejection needs, you might find you actually need a high-end 1500VA unit instead of a cheap 675VA unit. The cost difference is a few hundred dollars. The cost of one unplanned production stoppage that I described?
- Lost labor time (~40 minutes of 3 operators) = ~$60 in direct cost
- Scrap parts from the machine fault = ~$200
- My time troubleshooting, ordering a better unit, and reinstalling = ~4 hours
- Total: way more than the cost diff on the UPS.
I'm not saying buy gold-plated everything. I'm saying understand what your device does at its peak, not its average.
Blindly buying for the nameplate wattage is like buying an insurance policy that only covers paper cuts. It looks smart until you cut an artery.
My Current Playbook for Admin Buyers
- List the motor loads separately. Anything with a fan, pump, compressor, or VFD is a special case.
- Get the Locked Rotor Amps (LRA) data. If the vendor can't give it to you, assume 2.5x running.
- Use an online UPS (APC Smart-UPS) for anything in a noisy electrical environment. It's a difference of topology, not just wattage.
- Build in buffer. Size to 150% of the worst-case surge. You won't regret the headroom. I promise you, you will regret buying undersized.
- Check the battery runtime curve. A larger UPS doesn't just handle surges; it usually supports a larger battery bank. That gives you real runtime for those critical 20-minute shutdowns.
So What's My Bottom Line?
I've been managing facility purchases for about 5 years now. I process maybe 60-80 orders annually, managing relationships with about 8 different vendors. The single purchase that has caused me the most 'oh no' moments? Underpowered power protection.
The UPS is the last line of defense. It's not a spec sheet game. It's a physics game. The inrush current, the harmonics, the dirty power—these are the things that kill your investment, not the runtime meter on the package.
The 'it's just a small pump' mentality will cost you. Trust me, I've paid that bill. Size your APC unit for the surge, handle your motor loads separately, and you'll sleep better.
I still kick myself for that first coolant pump mistake. But at least it taught me to ask better questions now.