💧 Pump Cavitation Troubleshooting: Diagnosing and Fixing the Real Cause

If you found this page while a pump is actively rattling in your plant right now, you probably don't need another explanation of vapor pressure and Bernoulli's principle — you need to know what to check first, and why the same-sounding noise can have two completely different root causes that need two completely different fixes. The underlying physics of why cavitation happens — local pressure dropping below the liquid's vapor pressure, bubbles forming and then collapsing — is covered in detail in the Concept Explainer, Cavitation: How Fast-Moving Water Can Literally Boil at Room Temperature. This guide picks up from there and stays entirely on the practical side: symptom recognition, diagnosis, and the fix.

The Field Symptoms, Quickly

Before splitting cavitation into its two types, it helps to have the shared symptom list in one place, because these are what first get a technician's attention regardless of which type is occurring:

• A distinctive rattling, crackling, or "gravel and marbles" noise that seems to originate inside the pump casing near the impeller.
• Noticeable vibration at the pump, sometimes severe enough to be felt through the baseplate or piping.
• Fluctuating or "hunting" discharge pressure and flow — gauges that won't settle on a steady reading.
• Reduced pump performance overall: less head and flow delivered than the pump curve says it should produce at that speed and power draw.
• Over time, physical evidence on teardown: pitted, eroded impeller vanes, casing wear near the eye or volute tongue, and premature seal or bearing failure from the vibration.

Where field diagnosis commonly goes wrong is stopping right there. Every one of those symptoms is common to both major categories of pump cavitation, and treating them as one problem with one fix is the single most common mistake made in the field. The two categories have different root causes, occur under different operating conditions, and require different corrective actions. Applying the wrong fix doesn't just fail to help — it can waste a maintenance cycle, or in the recirculation case, make things measurably worse.

Two Different Problems That Sound the Same

Suction Cavitation

Suction cavitation is what most people picture when they hear the word: the pump simply is not getting enough usable pressure at its inlet to keep the liquid from vaporizing as it accelerates into the impeller eye. In formal terms, the Net Positive Suction Head Available (NPSHA) at the suction flange has fallen below the Net Positive Suction Head Required (NPSHR) that this specific pump needs at its current flow rate, as published on its performance curve.

Typical causes on the suction side include:

• A clogged or partially blocked suction strainer, screen, or foot valve — adds friction loss that isn't visible without inspecting it directly.
• Suction lift that is too high, or a suction source elevation that has dropped (a tank or wet well running lower than the level the system was designed around).
• Restricted, undersized, or overly long suction piping, or too many elbows and fittings crammed close to the pump inlet, disturbing the flow profile entering the impeller.
• A low tank or sump level approaching the point where a vortex can form at the suction inlet and pull entrained air into the line — which compounds the problem alongside true vapor cavitation.
• Excessive fluid temperature, which raises the liquid's vapor pressure and directly erodes the NPSH margin even if nothing else has changed.

The telltale operating signature of suction cavitation is that the pump's actual output — head and flow — comes in noticeably below what its curve predicts, alongside the noise and vibration. The pump isn't just noisy; it's underperforming, because it genuinely cannot draw in enough liquid without vaporizing part of it first.

Discharge (Recirculation) Cavitation

The second category is easy to overlook because it happens for almost the opposite reason: the pump has plenty of NPSH margin on the suction side, but it's being operated far off its best efficiency point (BEP) — usually at very low flow against a discharge head that is much higher than the system was designed for at that flow rate. Common triggers are a nearly closed discharge valve, a system curve that has shifted (a downstream valve throttled back, a filter or strainer fouling and adding resistance, a process demand dropping), or a pump that was oversized for the duty it now sees.

At very low flow relative to BEP, the flow pattern inside the impeller passages breaks down. Instead of moving cleanly from eye to discharge, liquid recirculates internally at the impeller inlet and/or discharge tips, and that internal recirculation creates its own localized low-pressure zones — separate from the suction-line NPSH story — where vapor bubbles form and collapse. This is often called suction recirculation or discharge recirculation depending on where in the impeller it originates, but from the operator's chair both are grouped as discharge-side or off-BEP cavitation.

The operating signature here is different: the pump is often still building close to its rated shutoff head — it isn't starved — but it's doing so at a flow rate well below where the curve says it should be running efficiently. The noise and vibration are present, sometimes even worse than typical suction cavitation, but they occur specifically as flow is throttled down, and often ease off as flow is increased back toward BEP.

The Diagnostic Sequence

Because the two types point toward opposite fixes, guessing from noise alone is a bad bet. Work through the check in this order:

• 1. Compare NPSH available to NPSH required at the actual operating flow. Calculate or measure NPSHA at the suction flange (atmospheric or source pressure, plus or minus elevation, minus suction friction losses, minus vapor pressure at the actual fluid temperature) and compare it against the pump's published NPSHR curve at the flow rate the pump is actually running, not its rated flow. A thin or negative margin here points straight at suction cavitation.
• 2. Inspect the suction side physically. Check strainer and foot valve condition, confirm actual tank or wet well level against what the design assumed, look for suction-line restrictions, and verify no air-entraining vortex is forming at the inlet. Don't rely on a "looks fine" visual pass — a strainer half-clogged with debris can look adequate at a glance but still be adding several feet of unaccounted head loss.
• 3. Plot the actual operating point against the pump curve. Read the real discharge pressure and flow rate off the gauges/meters and mark that point against the manufacturer's head-flow curve. If the point sits far to the left of BEP — low flow, head close to shutoff — and NPSH margin from step 1 looked adequate, that's the signature of recirculation cavitation, not a suction starvation problem.
• 4. Cross-check against recent changes. Cavitation that appears suddenly usually followed a change: a valve that got throttled back, a tank level that's been running lower than usual, a fouled downstream filter raising system resistance, a warmer process fluid, or a strainer that hasn't been cleaned on schedule. Identifying what changed often confirms the diagnosis faster than instrumentation alone.

Fixing Suction Cavitation

Once suction starvation is confirmed, the fix is to increase NPSH available, reduce NPSH required, or both:

• Clean or replace a clogged suction strainer, screen, or foot valve on a defined maintenance interval, not just when a problem is already audible.
• Reduce suction lift, or where feasible, raise the source liquid level or lower the pump elevation relative to the source.
• Shorten, enlarge, or simplify suction piping — larger diameter and fewer close-coupled fittings both reduce suction friction losses.
• Address low-level vortexing with proper minimum submergence, anti-vortex baffles or fittings, or a bell-mouth suction inlet.
• If fluid temperature has crept up, address the source of the heat gain; a few degrees can meaningfully cut into vapor pressure margin.
• If the system genuinely can't provide more NPSHA, look at reducing NPSHR instead: a lower-NPSHR pump model, an inducer, or simply operating at reduced flow if the process allows it.

Fixing Discharge / Recirculation Cavitation

Once the diagnosis points to off-BEP operation rather than suction starvation, the fix is about restoring flow toward the pump's efficient operating range, not chasing NPSH:

• Open the discharge valve or otherwise reduce artificial throttling if the process can tolerate the added flow — this is the one case where opening the discharge valve is the correct move, which is exactly why confirming the diagnosis first matters.
• Investigate and correct whatever shifted the system curve — a fouled filter, a partially closed downstream valve, scaling in the discharge piping — that pushed the duty point away from BEP.
• If the pump is simply oversized for its current duty (a common result of process changes over a facility's life), consider trimming the impeller, installing a smaller pump, or adding a minimum-flow bypass line back to a suction tank to keep the pump above its minimum stable flow.
• Check for a variable-speed drive as an option — slowing the pump down to match actual demand keeps it operating closer to BEP across a range of conditions instead of forcing it to throttle against a fixed-speed curve.

Why Getting the Diagnosis Wrong Costs You Twice

The reason this distinction matters in practice, not just in theory, is that the two fixes can actively work against each other. Opening a discharge valve to relieve recirculation cavitation is the right move for that problem — but doing the same thing on a pump that's actually suffering suction starvation increases the flow demand on an already-inadequate suction supply, worsening the NPSH margin and making the cavitation louder, not quieter. Conversely, spending a maintenance cycle chasing suction-side fixes — cleaning strainers, checking tank levels — on a pump that's actually cavitating from being throttled far off its BEP will produce no improvement at all, because the suction side was never the problem. Running the diagnostic sequence above, in order, before touching anything is what prevents that wasted effort.

When It Isn't Cavitation At All

Worth a final flag: not every rattling, vibrating pump is cavitating. Mechanical issues — a bent shaft, worn bearings, impeller imbalance, or a coupling misalignment — can produce noise and vibration that superficially resembles cavitation, and entrained air pulled in through a leaking suction gasket or a vortexing wet well can sound and behave similarly without any true vapor-pressure event occurring. If the NPSH margin checks out, the operating point sits close to BEP, and the suction side inspects clean, it's worth widening the investigation to mechanical condition and air ingress before assuming cavitation is recurring for a reason not yet found.