Stanley tool article workspace
Stanley Notes

Repair or Replace? A Procurement Comparison for Facility Pumps

2026-11-06 · Darius Campbell

The comparison that comes up more often than you'd think

When our facilities manager handed me a requisition for a new sump pump in early 2024, my first question wasn't which brand to buy. It was whether we needed a new pump at all.

I'm the office administrator for a roughly 180-person company, and I handle all facilities procurement—about $140,000 a year across maybe 12 vendors. That includes pumps, parts, and the assorted infrastructure that keeps a building running. I don't install any of it, but I do decide what gets ordered, from whom, and whether the repair quote sitting in my inbox is actually the right move.

The repair-or-replace question comes up more than you'd expect. So after six years of this, I've built a four-dimension comparison I run every time. Each dimension has a clear winner depending on the situation, and at least one of them surprised me when I started tracking it.

Dimension 1: Actual total cost (repair vs. replace)

This one seems obvious until you do it wrong once.

Everything I'd read about repair-versus-replace said the 50% rule is the standard—if a repair costs less than half a replacement, fix it. In practice, that rule has been basically useless for pumps, because it doesn't price in the second failure.

Example. Our Little Giant 1/3 HP sump pump had been in service for about four years when it started short-cycling in February 2024. The repair quote was $190 for motor and seal against a replacement unit around $280. By the 50% rule, I repair. But the quote didn't cover the float switch that was already showing corrosion, or the strainer basket with a hairline crack. Real cost for a repair that would actually last: $340. Which is more than the replacement.

That's the pattern I keep running into. Repair quotes are priced for the failure mode, not for the whole unit's condition. Replacement prices are for a known-good unit with a fresh warranty clock.

Winner: replace, for any pump that's already hit 3-4 years of service and has more than one wear component involved.

Dimension 2: Downtime profile

Repair and replacement have completely different downtime curves, and this matters more than the invoice for anything critical.

A repair is usually one service visit—sometimes two if the parts aren't on the van. A replacement is one visit plus the logistics of getting the right unit in, which sounds worse. But in practice, replacement has been way more predictable in my experience.

Repairs have a tendency to turn into "we need a second visit." The first visit diagnoses. The second visit does the work. Sometimes there's a third because the part number the tech wrote down was for the previous model year. Replacement is a known quantity: order the correct unit, schedule the swap, done.

Across the roughly 15 pump work orders I've processed since 2019, repairs averaged 4.2 days from initial call to equipment back online. Replacements averaged 3.5 days. The gap isn't huge, but the variance is—one repair stretched to nine days because of a back-ordered seal kit. I don't think I've had a replacement stretch past five.

Winner: replace, mostly because the downside is bounded. A repair can always be a second repair.

Dimension 3: Documentation and warranty burden

This is the dimension nobody puts in a spreadsheet, and it's the one that's bitten me more than cost ever has.

Here's what most people don't realize: the documentation burden is actually lower for a replacement than for a repair, even though it feels like it should be the other way around.

A new unit is covered by a fixed warranty term from a specific purchase date. You file the invoice, you note the serial, done. A Little Giant submersible pump with float switch, for example, ships with installation documentation that makes the warranty process straightforward—there's a single phone number and a single claim form.

A repair, by contrast, is covered by a patchwork: parts warranty from the supplier, service guarantee from the technician, and often a separate installer warranty on labor. I had a $340 repair job get partially rejected in 2023 because the service ticket said "pump fixed" instead of listing the part number. That was a $90 mistake on a line item I'd approved in good faith.

(Should mention: the same logic applies to just about any facility equipment. We keep a 20 gallon air compressor manual on file for exactly this reason—not to operate the compressor, but so we can look up its service intervals and understand whether a repair quote is actually a "too late" signal. If the manual says the separator needs replacing at 500 hours and the compressor is at 900, a repair quote on that separator is a warning, not a fix.)

Same discipline applies to measurements. If a repair requires fitting, you need to know the dimensions before you approve anything. This is the same rule as when you need to measure a hydraulic cylinder for replacement—measure first, order second. If I don't have the measurement in the requisition, the requisition gets returned to the tech. Not because I'm difficult, but because return shipping on a $200 part that doesn't fit is a budget line I've already paid once.

Winner: replace, despite the counterintuitive logic. One warranty to track beats three.

Dimension 4: Risk and redundancy

This is the one that should override everything else, and it's the one I got wrong early on.

If the pump in question is the only one protecting anything—basement, server room, inventory storage—the decision isn't really about cost or downtime. It's about whether you can live with a unit that's already proven it fails. A pump that has failed once is statistically more likely to fail again, because it's the same motor, the same seal, the same wear profile.

For anything critical, I default toward replace. For non-critical applications—decorative fountains, secondary circulation—repair is usually right because the downside of a second failure is basically zero.

I don't have hard data on how often repaired pumps fail a second time within the first year. What I can say anecdotally, based on the 15 or so work orders I've run, is that roughly a third of them did. That's not a big enough sample to publish, but it's enough for me to have a default.

Winner: depends entirely on the application. Criticality decides this one, not economics.

So which do you actually choose?

Here's the decision rule I use, written down so I don't have to relitigate it on every requisition:

  • Repair when: the application is non-critical, the repair is a single wear component, and the unit is under three years old.
  • Replace when: the application is critical, the unit has already failed once, or the repair quote touches more than one component.
  • Upgrade when: the repair is failing because the spec was wrong for the application. That's how our sump application ended up moving from a 1/3 HP unit to a 1 HP water pump after the volume requirement went up—the repair kept coming back because the pump was undersized, not because it was defective.

If you've ever had a repair quote come back twice for the same unit in the same year, you already know which of these categories you're in.

One last thing. The total spend on that pump replacement was about $280—maybe $265, I'd have to check the PO—but the decision wasn't really about the $280. It was about not wanting to be the person who approved a repair in February and a replacement in June. That's the version of the comparison that actually matters.

Share on WhatsApp
Darius Campbell

Darius Campbell

Darius Campbell is an independent power tool and cordless systems analyst covering drills, impact drivers, impact wrenches, grinders, circular saws, rotary hammers, and jigsaws. He uses IEC 62841-1 safety requirements while comparing rated input, battery voltage, torque, no-load speed, duty cycle, vibration, dust control, runtime, and accessory compatibility. His evaluation articles help contractors and buyers match tool platforms to workload, mobility, service conditions, and total battery-system cost.

Leave a Reply