-
Who this checklist is for
-
Step 1: Measure the actual load before you spec a modular UPS
-
Step 2: Verify grounding and circuit continuity with a continuity tester
-
Step 3: Test the existing battery string with a multimeter and record a baseline
-
Step 4: Compare modular UPS systems on module replacement cost, not sticker price
-
Step 5: Kit out the maintenance station — including chargers for the tools
-
Step 6: Build the TCO on measured data, not vendor projections
-
Common mistakes and things to watch for
Who this checklist is for
This is for facility, IT, or procurement leads who are buying or replacing a UPS for a server room, edge cabinet, or small data center. I'm assuming you've already decided you need one. Now you're at the decision stage — comparing a Schneider Electric modular UPS against standalone APC by Schneider Electric units, or figuring out whether the battery string on your existing install is worth saving.
There are six steps. Budget half a day to a full day for the first pass, depending on how much of your test gear you already own. In my experience, skipping steps 2 or 3 is where the largest cost surprises come from — and I'm including the cost of the UPS itself in that statement.
One caveat up front: this checklist is tuned for mid-size sites, roughly 2 to 20 cabinets. If you're buying a single 600VA unit for two offices and a printer, most of this is overkill. If you're above 50 cabinets, you probably want a power distribution consultant rather than a blog post.
Step 1: Measure the actual load before you spec a modular UPS
Every UPS quote I've ever received started from a number the vendor gave me. That number was usually a recommendation, not a measurement. When I first started buying UPS equipment, I assumed vendor-recommended sizing was conservative in my favor. Three budget cycles later, I realized it's usually conservative in the vendor's favor — and that overspecifying a modular frame is expensive at the module level.
Here's what to do instead:
- Pull actual load readings from your PDU or rack metering over at least a 30-day window. Peak matters, but so does the sustained daytime average, because that's what drives efficiency and battery runtime.
- Add a growth buffer of 20–30% for a modular UPS frame. Modular architecture exists so you can buy capacity in steps; overpadding the initial frame negates the point.
- Match the spec to the manufacturer's published load charts, not to a sales rep's rule of thumb. Schneider Electric publishes load and runtime tables for the Galaxy and Smart-UPS lines, and those numbers are what you should be quoting back to yourself.
When I audited our 2023 spending across two rooms, the difference between a right-sized modular frame and an oversized one came to roughly $2,300 over a four-year life. That's a meaningful chunk of a facilities budget, and it was invisible on the original PO.
Step 2: Verify grounding and circuit continuity with a continuity tester
This is the step most buyers skip, and it's the one that bites hardest. An electric continuity tester — the low-voltage kind used by electricians to confirm a circuit is intact and properly bonded — belongs in your pre-deployment kit even if you're not the one pulling wire.
A continuity tester checks whether a circuit path exists where it should. On a UPS deployment, you're mainly interested in equipment grounding conductor continuity between the panel, the UPS frame, and the rack rails. If a ground connection is loose or missing, your UPS won't fail immediately — it'll just sit there until the first fault, and then you're dealing with damaged electronics and an insurance claim.
A few practical notes from having done this wrong once:
- Use a tester rated for the expected circuit voltage. Low-voltage continuity testers are for dead circuits only. Confirm the circuit is de-energized before testing — this is a non-negotiable.
- Test between the grounding bar and each rack rail. A reading of a few ohms is normal; an open reading means a missing bond.
- Document the readings. I keep a simple spreadsheet with panel location, test point, reading, and date. It takes ten minutes and has saved me two arguments with contractors.
According to IEEE 1100 (the IEEE Recommended Practice for Powering and Grounding Electronic Equipment), proper grounding and bonding is a foundational requirement for electronic load protection. That's the section I point to when a contractor tells me grounding tests are optional. They aren't.
Step 3: Test the existing battery string with a multimeter and record a baseline
If you're keeping any part of an existing UPS — or if you're inheriting a room — step 3 is how to test a battery with a multimeter without guessing. This sounds basic, and it is, but basic done consistently is what separates a planned battery replacement from an emergency one.
Here's the sequence I use:
- With the UPS in normal (not battery) mode, measure the full string voltage at the battery terminals. Compare it to the nominal specification in the manufacturer's manual.
- Measure individual battery block voltages. A healthy lead-acid block in a 12V string should read within about 0.2–0.3V of its neighbors. A block that reads noticeably low is a candidate for replacement.
- Note the readings and repeat quarterly. A block that drifts downward over two or three readings is failing well before it trips a self-test.
Two cautions. First, use a multimeter with a CAT III rating appropriate for the DC voltage present — battery strings in a UPS can carry significant energy even when isolated. Second, this is a health check, not a capacity test. Only a load bank or a manufacturer-run test tells you true available runtime.
Everything I'd read said battery strings either work or fail. In practice, I've found they degrade slowly and quietly, and the multimeter baseline is what caught a 9% capacity drop in year three — before it became a 5 a.m. page.
Step 4: Compare modular UPS systems on module replacement cost, not sticker price
This is where the Schneider Electric modular UPS conversation usually gets sideways. Initial quote comparisons are easy; they're also misleading. A modular unit's whole value proposition is that you can replace a single power module instead of the whole system — so the number that matters is what that module costs in year three, year five, and year seven.
What I actually ask for in the quote package:
- Per-module replacement price, quoted today, with any firmware or compatibility notes. APC by Schneider Electric modular units have generation-specific compatibility, and buying a module that doesn't match your installed generation is a common and expensive mistake.
- Warranty terms at the module level, not just the frame level. These are often different.
- Whether the unit supports hot-swap module replacement. If a module swap requires a load transfer, your maintenance window cost goes up substantially.
I recommend the modular approach for sites that expect load growth beyond 30% over five years, or that have a maintenance staff capable of swapping modules without a vendor visit. If your load is flat and your team prefers a single sealed unit, the modular calculus may not work in your favor — that's not a knock on the product, it's just the shape of the economics.
Step 5: Kit out the maintenance station — including chargers for the tools
This step feels small, but it's the one I underbudgeted on the first time. Once you've bought a modular UPS, you'll be doing rack work, panel work, and status checks regularly. That means power tools, and power tools means batteries that need charging between jobs.
A universal battery charger for power tools on the maintenance bench is, practically speaking, a continuity tool. If your drill is dead when the module swap window opens, the swap doesn't happen on schedule — and rescheduling a maintenance window usually costs more than the charger did. I keep two chargers on our bench so that one is always loaded while a tool is in use.
What I'd include in the station kit:
- A universal charger compatible with the tool battery platform you actually own — check compatibility before ordering, because "universal" is a marketing word, not a guarantee.
- A decent digital multimeter with CAT III/IV rating, kept in a case, calibrated on a schedule.
- An electric continuity tester for grounding checks, as covered in step 2.
- A simple log sheet for readings. Paper is fine. The point is consistency, not sophistication.
Total kit cost is usually a few hundred dollars. The avoided rescheduling cost is many multiples of that.
Step 6: Build the TCO on measured data, not vendor projections
The final step is arithmetic. Take the measured load from step 1, the grounding and battery findings from steps 2 and 3, the module replacement pricing from step 4, and the maintenance kit cost from step 5, and put them in one spreadsheet with a four-year or five-year horizon.
What usually falls out of that spreadsheet:
- Energy cost differences between efficiency tiers show up as real money at mid-size loads, and they're almost never in the initial quote comparison.
- Battery replacement in year three or four is a real line item. Budget it now or discover it later — those are the options.
- Module replacement costs, spread across expected failures, tend to be lower than frame replacement but higher than people assume.
One honesty note: this TCO model won't tell you which brand wins in the abstract. It'll tell you which configuration wins for your site. That's the useful answer, and it's the one that's harder to shortlist from a brochure. I can only speak to sites roughly our size — if you're running a hyperscale room, the model probably has different shape.
Common mistakes and things to watch for
Testing with an uncalibrated meter. A multimeter that's a few percent off will mislead you on battery block comparisons and grounding readings. Calibration is cheap; bad decisions built on bad readings are not.
Treating continuity tester results as binary. A reading in the single-digit ohms range on a grounding test is normal; an open reading is a fault. The tester doesn't tell you which is which without context — read the standards, or ask your electrician to interpret.
Skipping the battery baseline. This is the most common skip and the most expensive one. Even a rough baseline gives you a trend line next quarter.
Forgetting the tool layer. The UPS is infrastructure; the tools that keep it running are the maintenance layer. Underspending on the bench shows up as delayed windows later.
Assuming one checklist fits all sites. This list is calibrated for 2–20 cabinet rooms with a facilities team that does light electrical work. Larger sites need a PDU-level study; single-unit buyers need a much shorter list.
Prices and product lineup references in this article reflect information available as of Q2 2025 — verify current quotes and specification sheets before you commit a budget number to an approval request.
