UPS or Generator? A Procurement Manager's Guide to Backup Power by Scenario

The question isn't "UPS or generator." It's what kind of outage you're actually planning for.

I'm a procurement manager at a mid-sized manufacturing firm (around 400 people). Over six years I've managed roughly $180,000 annually in facility power backup spending. I've negotiated UPS contracts, generator installs, and battery maintenance agreements across 11 vendors, and logged every invoice in our procurement system.

My first mistake: in 2021 I let our team decide based on what was on the UPS quote sheet. We under-bought on the generator side, and when an 11-hour regional outage hit, we lost $42,000 in overtime and scrapped material. That kind of pain doesn't show up on a spreadsheet.

What I've pushed internally ever since is a framework that asks a different question: which scenario are you actually in? Most procurement teams I've worked with end up in one of four — and the right answer looks very different depending on which one fits you.

Scenario A: You can't tolerate a single second of downtime

This is the UPS scenario — and specifically the kind where transfer time matters more than runtime.

Take the APC Smart-UPS C 1500VA 2U 120V, a model that comes up constantly in apc smart-ups c 1500va 2u 120v ups reviews. Per Schneider Electric's published spec sheet (apc.com, 2024), runtime is roughly 3 to 8 minutes under a "network-critical" load — and that phrasing matters. That's not "my whole building" and it's not "my motor starter."

Here's where most people get it wrong: they see 1500VA and assume it handles anything up to 1500VA. But a 1500VA heater drains that battery in under a minute, while a 1500VA server load holds up far longer because of power-factor differences. I've had two separate departments ask for the same unit for wildly different loads, and we had to size them down both times.

On apc ups battery test: my team now runs a self-test quarterly, with a log. Not annually — quarterly. The first time we ran quarterly tests, three units flagged as replacement-ready that our annual inspection had never caught. Battery replacements cost $2,100. A three-hour outage on the units that would have failed would have cost us, by our own estimate, $11,000 in search fees and lost billable time. That's not counting the client relationship we nearly lost.

Scenario B: You need hours, not minutes

Once your need is measured in hours, you're not shopping for a UPS anymore — you're shopping for a generator. And usually not a small one.

A 1000 kVA diesel generator is in the range that can carry a mid-sized plant through a full-grid outage. In 2022 Q3 we installed one, and it ran us about $340,000 landed. I should say: that's a 2022 Q3 price. Steel and copper costs have moved since.

The hidden costs I didn't fully price in: fuel quality testing (~$1,400/year), monthly 10-minute load test, and a service visit every 500 run-hours. The annual maintenance envelope on that unit runs $8,000–10,000.

Now — confession. I almost went with a cheaper generator. One non-discount supplier came in 19% lower than the unit we eventually chose. Then I asked what wasn't in the quote, and found no annual service package and a lower efficiency tier. Factoring seven-year TCO, the "cheap" option was about $23,000 more expensive. That's not a quote comparison — that's a guessing game.

To be fair, their base unit was solid. The gap was in documentation and ongoing support terms. But that's exactly the point: with generators, the line item that's missing tells you more than the one that's there.

Scenario C: You need power somewhere that isn't a building

This is where portable generator vs inverter generator comes into play. And most procurement teams I've seen focus on the wrong spec.

For sensitive loads — test equipment, laptops, small diagnostic devices — THD (total harmonic distortion) matters more than peak kW. General rule of thumb: inverter units typically produce THD under 3%, which is fine for electronics. Traditional portable generators can easily hit 15%+ THD. Fine for power tools. Not fine for anything with a logic board.

I ran the numbers myself: a portable unit plus one or two fried laptop chargers came out roughly even with the premium on a decent inverter generator. The inverter unit we put at one site cost about $400 more than a comparable portable. But it's kept two laptops and one diagnostic device alive over two years. I'll pay that $400 any day.

Scenario D: You already own a UPS — and it's struggling

This is the counterintuitive one.

The instinct in our industry is to replace the whole unit when it underperforms. Most of the time, replacing the battery is the smarter move.

This is where apc ups battery test earns its keep. If the self-test log is clean and output voltage is right, a short runtime isn't a card to buy a new unit — it's a card to examine the battery pack. And more specifically, check the battery charger ends — the terminal contact at the end of the charging lead. A rusted or loose charger end can quietly cut an 8-minute reserve down to 90 seconds and still pass a basic self-test.

We had exactly that happen in 2023. The self-test passed. The actual load test didn't — the battery pack had dropped to 47% of rated capacity, and we'd been running on a ragged edge for months.

The math here is worth sitting with: a replacement battery cartridge runs $350–700. A new UPS plus install is $1,800–5,000. Unless you genuinely need more capacity, replace the battery. And add the load test. Schneider's own service guidance recommends periodic load testing for VRLA batteries — not just self-test. It's the reason we caught it.

How to figure out which scenario you're in

I ask three questions before I sign off on any backup-power purchase:

1. How clean does your power need to be? If the load is precision equipment that can't survive a switching event, Scenario A. If it's lighting and tools, B or C.
2. How long does your load need to run? Minutes: UPS. Hours: generator. "Depends on the outage": UPS + right-sized generator, but factor in fuel logistics.
3. How old is the existing UPS, and has it been load-tested? If it's over 4–5 years old and hasn't been, run the test first. If the battery test is clean, you saved money. If not, replace the battery before you replace the unit.

If two of those three answers are "not sure," grab a service partner who will put the numbers on paper. The price of transparency is almost always lower than the cost of guessing.