Most pump PM programs aren't bad because they're missing obscure checks. They're bad because nobody standardized the obvious ones. The same ten inspections that would catch 80% of pump failures before they become failures — missed alignment, early seal wear, suction conditions going wrong — are the ones that get skipped, abbreviated, or never written down in the first place.
That's not a technician problem. That's a program design problem.
This post covers the foundational checks every pump PM program should include regardless of pump type, service, or criticality level — the baseline that should be non-negotiable before you start layering in equipment-specific tasks.
The broader context for how your pump program fits together lives in centrifugal pumps as the starting point.
Check 1: Baseline the Operating Conditions First
Every pump inspection should start the same way: document what the pump is actually doing right now.
Flow rate. Discharge pressure. Suction pressure. Differential pressure across the system. Amp draw. Temperature at the bearing housings and seal area.
Not because you expect something to be wrong. Because you need a number to compare to. A pump running 12 psi higher discharge pressure than it did six months ago is telling you something. A pump pulling 4 amps more than its baseline is telling you something louder. Without documented baselines, every inspection is just a snapshot with no frame of reference.
Record actual values. Not "normal." Not "OK." Numbers.
Check 2: Suction Conditions
Suction problems kill pumps slowly and get blamed on everything else.
Check suction pressure against design specifications. Verify adequate NPSH margin — if you don't have the design NPSH readily available, flag it as a documentation gap and find it. Low suction pressure creates cavitation. Cavitation creates impeller erosion, vibration, and accelerated bearing wear. By the time those symptoms are obvious, the damage has been accumulating for months.
Check for air entrainment. Bubbling, crackling, or irregular discharge pressure fluctuations at normal flow rates are worth investigating. So is any change to suction line configuration since the last inspection — temporary fittings, isolation valves partially closed, strainers that haven't been serviced.
The suction side is where most pump failures begin. Inspect it like it matters.
Check 3: Seal Condition and Leakage
Mechanical seals announce their decline before they fail outright. The announcement is easy to miss if nobody's looking for it.
Check the seal gland area for weeping, dripping, or encrusted residue from previous leaks. A dry face seal should show no liquid at the gland under normal operation — even intermittent dripping is a finding worth documenting and trending.
For packed gland seals, document the leakage rate. A properly adjusted packed seal weeps slightly — that's by design. No weeping means it's overtightened and burning the packing. Heavy dripping means it needs adjustment. Record what you observe rather than adjusting blindly.
Note any changes in leakage pattern since the last inspection. Sudden increase in leakage from a previously stable seal is a stronger signal than gradual progression.
Seal failures are their own failure mode category — see the early warning signs that most PM programs overlook.
Check 4: Bearing Temperature
Bearing housings have a normal operating temperature range for each pump. If you don't have it documented, establish it during commissioning or early operation.
Check bearing housing temperature at every PM interval using a contact thermometer or infrared thermometer — not your hand. Document both drive-end and non-drive-end temperatures separately. A 15°F rise from baseline warrants investigation. A 30°F rise is a finding that changes the urgency of your next corrective action.
Temperature alone doesn't tell you the failure mode. But it tells you something is changing, and that matters.
For grease-lubricated bearings, overtemperature after a recent relubrication event can indicate over-greasing — the grease is churning, not lubricating. For oil-lubricated bearings, a temperature spike often points to oil level, contamination, or lubrication system problems first.
Check 5: Vibration
Vibration is the most information-dense signal a pump produces. Most PM programs either ignore it entirely or treat it as a pass/fail with no context.
Standardize a vibration measurement point for every pump in your program. Measure at bearing housings — axial, radial horizontal, and radial vertical. Record amplitude in velocity (inches per second or mm/s, not just acceleration). Use the same instrument and the same points every time.
You don't need a vibration analyst to establish a baseline and trend against it. You need consistent measurements and a place to record them.
A pump that's vibrating at 0.15 in/s at commissioning and climbs to 0.35 in/s over eighteen months is telling you something about alignment, impeller condition, or bearing wear. A pump that jumps from 0.15 to 0.40 in/s between inspections is telling you something more urgent.
Trend the numbers. Don't just record them.
Check 6: Alignment Verification
Pump-to-driver alignment is not a one-time installation activity. It drifts. Thermal growth, pipe strain, soft foot, routine maintenance that required disconnecting the coupling — any of these can shift alignment from acceptable to failure-accelerating without triggering any obvious symptoms until bearing and seal wear is already underway.
Alignment checks should be part of every PM interval for coupled equipment. Laser alignment tools make this faster than dial indicators but require the same rigor: document the found condition before any corrections, document the final alignment values after.
Acceptable alignment tolerances depend on speed and coupling type. Know the specification for each pump in your program. "Looks parallel" is not an alignment check.
For a complete look at what misalignment actually costs and how often it gets skipped, see the real reasons alignment gets deprioritized.
Check 7: Coupling Condition
The coupling is the component most PM programs treat as invisible until it fails loudly.
Inspect flexible coupling elements at every PM interval. On jaw couplings, look for compression, cracking, or extrusion of the spider element. On disc couplings, look for cracking or fatigue at the disc pack. On gear couplings, check for lubrication condition and fretting at the gear teeth.
Coupling wear accelerates when alignment is off. This makes coupling condition a useful indirect indicator of alignment problems, especially on equipment that gets inspected less frequently than it should.
Document the condition. Replace coupling elements proactively when wear is visible — not after the element fails, not at the next major overhaul, when you see it.
Check 8: Lubrication
Lubrication failures are preventable every time. That doesn't stop them from being one of the leading causes of pump bearing failures.
For grease-lubricated bearings: verify grease type matches the specification for that equipment. Verify interval compliance. Look for signs of contamination — discoloration, grit, or moisture in purged grease. Document the volume applied so you have a reference for the next interval.
For oil-lubricated bearings: check oil level against the sight glass reference mark with the pump running. Check oil color for contamination — dark, cloudy, or milky oil is a finding. Sample oil from long-interval systems for particle count and water content rather than relying on visual checks alone.
One of the most common errors: using the right interval with the wrong lubricant, or vice versa. Both create failures that look like bearing problems but trace back to the lubrication program.
Check 9: Foundation, Baseplate, and Fastener Integrity
A pump mounted to a degraded baseplate is a pump that will never hold alignment regardless of how often you check it.
Inspect grout condition around the baseplate perimeter at every PM interval. Look for cracking, voids, or separation at the grout-to-baseplate interface. Grout failures are common in environments with thermal cycling, vibration, and chemical exposure — which describes most pump installations.
Check anchor bolt torque. Soft foot is often a baseplate problem or a fastener problem as much as it is a shimming problem. A loose anchor bolt on one corner will defeat any alignment work done at the coupling.
Inspect for any evidence of excessive movement — fretting marks around fasteners, cracked paint at the baseplate edge, vibration patterns that are inconsistent with historical data.
Check 10: System Conditions and Downstream Equipment
A pump doesn't fail in isolation. It fails in response to what the system demands of it.
At every PM, verify that isolation valves on both suction and discharge are in their correct operating positions. Check that discharge pressure reflects expected system conditions — significant deviation often means something in the system has changed, not something in the pump. Verify that any pressure gauges or flow instrumentation associated with the pump are reading and functioning.
If the pump is equipped with a pressure relief valve, verify it hasn't been bypassed, locked open, or set outside its calibration range.
System changes — new piping runs, added branch loads, valve configurations that shift flow balance — can push a pump into off-design operation. Off-design operation accelerates wear at the impeller, seal, and bearings. The inspection should account for whether the system the pump is running in today is the same system it was designed for.
What Standardizing These Checks Actually Looks Like
Standardizing means the same checks, the same measurement points, the same documentation format, every time — regardless of who does the inspection.
That means written procedures with defined measurement locations. It means actual numerical values recorded, not condition descriptors. It means trend charts that get reviewed, not just filled in.
The checks above aren't novel. Most experienced maintenance techs would recognize all ten. The problem is that without standardization, half of them get done inconsistently, a quarter get abbreviated under time pressure, and the documentation that would allow trending doesn't get created.
A PM that doesn't generate trending data is a checklist, not a program.
Start with the checklists built from these principles:
- Centrifugal pump PM tasks
- Positive displacement pump PM tasks
- Diaphragm pump PM tasks
- Piston/plunger pump PM tasks
- Rotary pump PM tasks
- Rotary vane vacuum pump PM tasks
- Liquid ring vacuum pump PM tasks
- Diaphragm vacuum pump PM tasks
- Submersible pump PM tasks
- Sump and sewage pump PM tasks
- Magnetic drive pump PM tasks
- Peristaltic and hose pump PM tasks
- Thermal fluid circulation pump PM tasks
- Oil and fuel transfer pump PM tasks
- Grease and lubrication pump PM tasks
- Fire suppression and sprinkler pump PM tasks
The checks aren't the hard part. The standardization is. Get that right and everything downstream gets easier.