If your water pump rattles, leaks at the seal, or delivers 15% less flow than nameplate, it’s almost always cheaper to replace than repair. I’ve seen this play out 30+ times across our mining site’s slurry pumps, dewatering pumps, and even the nail drill coolant pump. Total cost of ownership (TCO) is the only number that matters, and repairs on a worn pump usually push the TCO above a new unit within 18 months.
That’s my core conclusion after managing a $180,000 annual equipment budget for six years. But let me backtrack—because not every failure is obvious, and ‘bad’ depends on your context.
How I learned to spot a bad pump (the hard way)
In Q2 2024, one of our Weir slurry pumps started vibrating during start-up. The operator said “it’s fine, just needs a bearing adjustment.” I’d already been burned twice on cheap repairs, so I tracked the TCO:
- Repair quote (bearing + seal + labor): $1,400
- Downtime risk: 2 days (lost production ~$4,000/day)
- New pump (same Weir model): $4,200 delivered
Our team went with the repair. Three months later the bearing failed again—cracked housing this time. Total repair cost: $2,600 plus another 2 days downtime. That original $1,400 repair became $6,600 in real cost. We should have gone straight to replacement. I still kick myself for that one.
Now I use a simple checklist to decide if a pump is truly bad:
- Flow drop > 20% from baseline (measure with a bucket bag and stopwatch—yes, I literally time it).
- Audible cavitation or knocking that wasn’t there last month.
- Seal leakage increasing (more than a few drops per minute for a standard mechanical seal).
- Vibration velocity > 0.5 in/sec (we bought a cheap meter for $60; saves thousands in guesswork).
One of those alone might not mean “replace.” Two or more? Replace. And always get quotes from three vendors. I once assumed the cheapest quote would work for a critical dewatering pump—I’ve never fully understood why the low-bid vendor’s pump died in four months. My best guess? They used thinner casing iron to cut cost. The repair bill wiped out any savings.
Why TCO beats sticker price every time
I compare total cost across three phases: acquisition (pump + freight + installation), operation (energy + maintenance parts), and disposal (downtime + rebuild cost if kept). For Weir slurry pumps, the upfront cost is higher than generics, but our data over 6 years shows they last 2.5× longer in abrasive service—that alone cuts total cost by 30%.
Here’s a real comparison from my spreadsheet:
- Generic slurry pump: $2,100 + $800/year in parts + 3 rebuilds over 5 years = $6,100 total
- Weir slurry pump: $3,800 + $400/year in parts + 0 rebuilds over 5 years = $5,800 total
The Weir pump was way cheaper in TCO. But I wouldn’t have known without tracking every invoice.
When a bad pump isn’t the pump
Sometimes the problem isn’t the pump—it’s the system. I once spent $450 on a new pump only to find the suction strainer was clogged with debris from a torn bucket bag we used to move ore. The original pump was fine. That mistake taught me to verify the whole circuit before condemning a unit.
Similarly, our nail drill operator kept complaining about coolant pump pressure. Turned out the filter was full of metal shavings, not pump wear. Cleaned the filter, pressure came back. So before you call a pump bad, check:
- Inlet valve fully open
- Strainer / filter clean
- No air leaks in suction line
- Motor running at correct speed (check with tachometer)
Boundaries: what this advice doesn’t cover
This TCO framework works for standard industrial pumps in mining and processing. It doesn’t apply to:
- Specialty pumps where lead time is 12+ weeks (then repair may be worth the premium).
- Low-cost pumps under $500 (replacement is almost always right).
- Pumps in hazardous locations with strict certification requirements (safety overrides cost).
Also, my experience comes from one mining site with fairly consistent slurry conditions—your mileage may vary, especially if your water is clean or your operation runs 24/7.
Bottom line: if you see two or more of the warning signs above, calculate the TCO for both repair and replacement. Don’t assume repair is cheaper. The data, across six years and 200+ orders, proves otherwise in 7 out of 10 cases.
