AXILINEOn-site pump balancing: impellers, rotors, and shafts of pump units
The pump shakes, the seal leaks more often than it should, the bearings run hot. Impeller imbalance is only one possibility, and on pumps it's confirmed less often than on fans: hydraulics often account for half the level. We come out with a two-channel vibration analyser, break the vibration down by frequency — separating the contribution from rotor rotation, blade passing, and cavitation — and balance the rotor on site if the numbers justify it.
Symptoms: what to call us about
Pump vibration tends to get noticed not through an instrument but through consumables: the seal starts leaking before its time, bearings get replaced twice a year instead of once every three years. No single sign like this means imbalance on its own. But each one means it's time to measure.
- Vibration at the bearing housing has risen above what's normal for this machine, even if it's formally still within limits.
- The mechanical seal leaks more often than once a year, and the reason for replacing it is the same every time.
- Pump or motor bearings fail faster than their rated life.
- The hum changed after a repair: after replacing the impeller, machining the wear rings, or removing and refitting the coupling.
- The pump got noisier after cleaning out deposits or after running on a contaminated medium.
- Vibration depends on the operating mode: it rises when the valve closes or when the level in the suction tank drops.
- The pipework has started shaking, and knocking is audible in the check valve.
- The pump was switched to a variable-speed drive, and vibration spikes sharply over part of the speed range.
The last two points most often turn out not to be imbalance. We'll still come out and measure: telling them apart from imbalance is only possible from the spectrum and the phase — the angle that shows at which point in the rotation the rotor is "hitting" — not from a description over the phone.
What pump equipment we balance
For balancing purposes a pump isn't one impeller, it's an assembly: impeller, shaft, coupling, sometimes a pulley and a cooling fan. Imbalance can sit in any of these elements, while you measure it at the bearing supports. So we work with the whole rotor and choose the correction plane based on what's actually reachable.
General-purpose centrifugal pumps
Centrifugal pump impellers, end-suction pumps, heating and cooling circulation pumps, water-supply pumps. Imbalance here is rarely a factory issue: it builds up from scale deposits on the vanes, corrosion, and wear on the leading edges. On end-suction pumps the imbalance contribution is higher, because the impeller is overhung beyond the supports and acts as a lever.
Process pumps in production
Process-line pumps, chemical-industry pumps, food-industry pumps. The specifics here are the restrictions on correction: welding a weight inside the wetted part is out, sanitary designs won't tolerate buildup or gaps, and lined, sealless pumps with a magnetic coupling give almost no access to the impeller.
Pump rotors, shafts, and multistage assemblies
Multistage pump rotors, pump rotors after a repair, pump-unit shafts together with couplings. The length-to-diameter ratio decides: a long rotor with several impellers almost always needs two correction planes, and some rotors like this behave as flexible.
Wastewater-treatment and contaminated-media pumps
Wastewater-treatment pumps, sludge and sewage pumps, machines with open and semi-open impellers. Imbalance builds up from fibre winding, fat deposits, abrasive wear, and edge chipping. Balancing gives a fast result, but it holds only until the next buildup.
Vacuum pumps and combined units
Vacuum pumps, liquid-ring and two-rotor types, plus rotors of pump-compressor units on a common bedplate. The rotors are elongated, speeds are higher, and two planes are almost mandatory. The main constraint is access: the rotor is enclosed in a sealed housing, and only the drive-side part, the pulley, or the coupling can actually be balanced.
If your pump isn't on this list, that's not a refusal. Send us the type, speed, power, and a photo of the coupling assembly. The photo shows whether there's a correction plane, and that's half the answer.
What we check before balancing, specifically on pumps
We fit the sensors on the pump's bearing supports: one on the coupling side, one on the impeller side. Magnet-mounted on a cleaned spot, direction radial-horizontal, plus an axial reading at the coupling. We aim the laser phase sensor at a reflective tape mark on the shaft or the coupling. Close-coupled pumps have no supports of their own, so we move the measurement points onto the motor housing next to the bearings.
Next comes a checklist that's longer on a pump than on a fan: axial force and seals get added on top of the mechanics.
- Impeller-to-shaft fit and the tightening torque of its nut: a worn fit gives an unstable 1x, and balancing it is pointless.
- Condition of the coupling, pins, and flexible bushes: worn elements produce their own vibration and spoil measurement repeatability.
- We check for shaft misalignment with a dial indicator. If it's present, the order is: fix the soft foot (a foot that doesn't sit flush against the frame) first, then align the pump and motor shafts, and only then balance.
- Pipe load on the nozzles: overtightened flanges and an unsupported pipe deform the casing and change the clearances.
- Axial play in the rotor and the thrust bearing: wear rings wearing down increases both axial force and radial load.
- Bearings, checked in the high-frequency region of the spectrum: balancing a machine with a failing bearing is dangerous.
- Resonance: on pumps with a variable-speed drive we check the whole operating speed range, not a single point.
Sources: ISO 13373-3:2015 · ISO 281:2007
Hydraulics disguised as imbalance
This is the main way a pump differs from any other rotor. A pump vibrates simply because it's moving liquid, and part of that vibration masks the running-speed component. That's why we ask you to fix the operating mode and record the numbers: flow rate, suction and discharge pressure, valve position, level in the suction tank, medium temperature.
The blade-passing frequency is calculated simply: the number of impeller vanes multiplied by the rotational frequency. If a peak sits there, no amount of weights will remove it. Operation far from the best-efficiency point, inlet recirculation, and a small gap between the vane and the volute tongue are fixed by the operating mode and by repairing the hydraulic section.
| What you observe | Likely cause | What we do |
|---|---|---|
| 1x dominates, phase is stable, the level doesn't depend on valve position | Rotor imbalance: erosion, buildup, a chipped vane, a shifted coupling | On-site balancing in one or two planes |
| A peak at the vane count times speed, rising as the valve closes | Operation away from the best-efficiency point, recirculation, a small gap at the volute tongue | Return to the best-efficiency flow rate, check the pump curve |
| A raised noise floor at high frequencies, a gravel-like sound at the suction side | Cavitation, insufficient suction head | The suction line, filter, tank level, operating mode |
| Strong 2x (vibration at twice the rotational frequency) and high axial vibration, phase difference across the coupling of about 180° | Shaft misalignment, a soft foot | Shims under the feet, shaft alignment, re-measurement |
| A comb of harmonics at 1x, 2x, 3x and above, the level jumps around from run to run | Loose fasteners, a worn impeller or coupling fit on the shaft | Retightening and fit repair before balancing |
| High-frequency peaks not a multiple of speed, growing over time | Bearing defects | Bearing replacement, balancing afterward |
| 1x is high, phase drifts from run to run, a sharp rise over a narrow speed range | Resonance in the frame or the pipework, a common case with variable-speed drives | Changing the operating speed or stiffening the supports |
If the process allows it, we take measurements at two operating modes. Vibration that changes together with flow rate is almost certainly hydraulic. Vibration that stays the same regardless of flow rate and pressure is almost certainly mechanical, and that's the kind balancing fixes.
How the work proceeds on site
- 01
Conversation before the visit
You describe the machine and what's changed. We tell you what to prepare: access to the bearing supports, a spot for the tachometer mark, the ability to start up at a stable mode. More detail is in our article on getting ready for a visit.
- 02
Measurement and the decision on whether to balance at all
Two accelerometers on the bearing supports, a laser phase sensor on the mark. We record overall vibration (the combined level across all frequencies), 1x with phase, speed, and spectrum at the operating mode. This is where it becomes clear what we're dealing with: imbalance, hydraulics, shaft misalignment, bearings, or resonance.
- 03
Mechanics and mode
We find loose fasteners, a soft foot, a worn impeller fit, or shaft misalignment: we fix it, or note it in the report and re-take the measurement. We won't balance on top of unresolved mechanical issues — the result wouldn't hold.
- 04
Trial weight
We fit a temporary weight of known mass at a known radius in whichever correction plane is accessible. A valid trial run changes 1x amplitude or phase by at least 20–30%. Less than that means the weight is too small, and we increase it.
- 05
Correction
The software outputs mass and angle for each plane. On pumps we more often work in fixed-position mode by the coupling bolts: instead of an angle we get a position number. Where add-on weights aren't allowed, we calculate metal removal by drilling.
- 06
Verification run and trim
We run at the same mode and compare. If the level has dropped but tolerance hasn't been met, we add a small correction to the weights already fitted. We save this machine's influence coefficients — its measured response to a fitted weight — so the next balancing runs without trial weights.
- 07
Report and recommendations
We hand over the before-and-after numbers, spectra, weight masses and positions, plus a list of what's left outside the scope of balancing: bearings, seals, operating mode, pipework.
One plane or two, and where to fit the weight on a pump at all
Rotor shape determines the number of planes. Rule of thumb: if the rotor's length in the correction zone is under roughly half the diameter, it behaves like a disc, and one plane is often enough. Anything elongated needs two, otherwise couple imbalance remains — a pair of unbalanced masses at opposite ends of the rotor — and the vibration at the second support won't go away.
Pump geometry is almost always elongated. A single-impeller end-suction pump can sometimes be closed out with one plane at moderate speed. An impeller between the supports, a multistage rotor, a vacuum-pump shaft, a pump-compressor unit rotor: two planes.
- The coupling and its bolts: the most common correction plane on pumps. Mass is built up with washers under the bolts, no welding needed.
- The impeller nut and the impeller face, if the casing comes apart quickly and the medium allows access.
- The pulley or the motor's cooling fan on belt-driven units.
- Balancing washers and spare holes, if the design provides for them.
- Metal removal by drilling where add-on weights are banned: food and chemical production, sanitary designs, product-contact zones.
An honest caveat. If the imbalance sits in the impeller but only the coupling is accessible, correcting there still reduces vibration as long as the rotor behaves as rigid. The further the correction plane is from the imbalance plane, and the longer the rotor, the worse this workaround performs. On multistage rotors it often doesn't work at all.
Sources: ISO 21940-11:2016 · ISO 21940-12:2016
When on-site balancing won't help
We prefer to say this before the visit, not after. Here are the cases where weights are useless or actively harmful.
- The running-speed component is small relative to overall vibration: the level is driven by bearings, looseness, shaft misalignment, or hydraulics, and balancing will remove only a small fraction of it.
- Cavitation and operation away from the best-efficiency point. Fixed by the suction line, suction head, tank level, and operating mode.
- Worn wear rings and elevated axial force: the rotor is loaded incorrectly, and the vibration will come back.
- The rotor's geometry is compromised: a bent shaft, chipped vanes, uneven wall thickness after erosion. Balancing will hide part of the level but won't restore head or service life.
- A worn impeller or coupling fit on the shaft: the 1x readings change from run to run, and an influence coefficient can't be obtained.
- Resonance in the frame, foundation, or pipework: 1x phase is unstable, and the result drifts with a small change in speed.
- The rotor is enclosed in a sealed housing, with no correction plane at all. This happens on vacuum pumps, sealless chemical pumps with a magnetic coupling, and lined casings. The right answer is dismantling and balancing the rotor on a stand.
- A stable mode can't be maintained: flow rate and pressure drift, the pump cycles on and off. Trial runs can't be compared against each other.
On-site balancing wins in one respect: the rotor is balanced in its own supports, together with the coupling, the shaft, and the frame's real stiffness. It loses where access inside the machine is needed. We'll tell you which option is cheaper for your case, including the option of "pull the rotor and take it to a stand".
What you get
The result of the work isn't "it got quieter" — it's numbers you can attach to a sign-off report and compare against six months from now.
- Overall vibration and the 1x running-speed component with phase, before and after, at each bearing support, in mm/s RMS (root mean square).
- Spectra before and after: showing exactly what went away and what remains.
- Speed and the mode at the time of measurement: flow rate, pressure, valve position.
- Masses and angular positions, or fixed-position numbers, of the fitted weights, with the radius stated.
- Residual imbalance and an assessment against the selected balance quality grade, if a G class is specified for the machine.
- An assessment of the machine's condition by zone from A to D (A — good, D — unacceptable), citing the standard part applied and the measurement conditions.
- A list of what balancing didn't resolve, and what's worth doing before next season.
- Saved influence coefficients: a repeat balancing of this pump will be shorter and cheaper.
Three different things shouldn't be confused. "Within tolerance" in the software means only that residual 1x is below the target value entered. An assessment of the machine's overall condition by zone, and confirmation of a balance quality grade, are separate criteria from different standards. In the report we state which criterion was applied and note the applicable part and edition of the standard: pump units have exceptions by power, speed, and support type.
Sources: ISO 20816-1:2016 · ISO 21940-11:2016 · Balanset-1A operation manual
How much it costs and how to book
vibration diagnostics with a report costs EUR 300 per unit, balancing adds from EUR 250. The minimum invoice for a visit is EUR 500, because the trip, sensor setup, and diagnostics take time regardless of the number of machines. The calculator on our website gives an exact figure for your case.
Several machines in one visit work out cheaper: the instrument and sensors are already on site, and the second and third rotor cost less than the first.
We're the engineers who design and manufacture the Balanset instruments, and we use them on-site ourselves. We're based in Vila Nova de Gaia near Porto, and we cover all of Portugal.
- Pump type: end-suction, impeller between supports, multistage, vacuum, wastewater-treatment pump.
- Speed and drive power, and whether there's a variable-speed drive or a belt drive.
- What changed and when: after a repair, after cleaning, gradually.
- Whether the bearing supports and coupling are accessible, plus a photo of the coupling assembly.
- Whether a stable mode can be held, and how long stoppages are allowed to be.
- Restrictions on correction: sanitary design, a welding ban, an explosion-hazard zone.
- Site address and a convenient time window.
If the photos and figures show that balancing won't solve your problem, we'll say so before the visit and suggest what to do instead.
Sources: Balanset-1A manufacturer specification
Frequently asked questions
Can a pump impeller be balanced without dismantling the unit?
Usually, yes. We balance the rotor in its own bearing supports in two or three runs at operating speed, and fit the correction mass on whichever plane is accessible: most often the coupling bolts. Dismantling is needed where there's no correction plane at all.
Why do you ask us to record flow rate, pressure, and valve position?
To separate imbalance from hydraulics. Vibration that changes together with flow rate comes from cavitation, recirculation, or operation away from the best-efficiency point, and a weight won't remove it. Without a fixed mode, trial runs also can't be compared against each other.
The pump is close-coupled and has no bearing supports of its own. What do we do?
We move the measurement points onto the motor housing, close to its bearings, and stick the reflective tape mark on the cooling fan or an accessible section of the shaft. The cooling fan or the coupling usually serves as the correction plane.
Will balancing help if the impeller is worn by erosion?
The vibration level will drop, but head and efficiency won't come back, and the uneven wall thickness will remain. We balance an impeller like that as a temporary measure and note in the report that its remaining life is limited and it's time to replace it.
Can a multistage pump rotor be balanced on site?
Sometimes. The rotor is long, imbalance is spread across the stages, and some rotors like this behave as flexible: a result at one speed doesn't guarantee the result at another. We decide based on the measurement, and quite often the right answer is dismantling and a balancing stand.
How long does a visit take, and for how long does the pump need to be stopped?
Diagnostics with measurement usually take a few hours. Two-plane balancing needs three runs plus stops to reposition the weight, 15–30 minutes each. We agree the exact window in advance so it fits your process.
Related content
On-Site Pump Impeller Balancing: Closed, Semi-Open and Open Impellers
Yes, we balance pump impellers on site, in their own bearing supports, without removing the impeller from the shaft. Three conditions apply: the running-speed component accounts for most of the vibration level, and its phase repeats from start to start; the impeller has been cleaned of deposits; and there is at least one correction plane — a place where mass can be added or removed — that is actually reachable. If erosion has eaten the blades down to uneven thickness, the disc has cracked, or the clearance on a semi-open impeller has gone beyond the manufacturer's tolerance, balancing will only postpone replacement. In that case we say so directly, before the visit.
Multistage Pump Rotor Balancing: What On-Site Work Can and Can't Do
Sometimes, but less often than with any other pump, and we say so before the visit. We balance a multistage rotor on site if it behaves as rigid. That means: the operating speed is well clear of the first critical speed — the speed at which the shaft goes into resonance and starts to bow; the dominant vibration is the 1x running-speed component, meaning vibration exactly at the rotation frequency; its phase repeats from run to run; and two correction planes — spots where a weight can be fitted — are accessible, at minimum the coupling half and the free end of the shaft. If the rotor behaves as flexible, or the imbalance is spread across the stages after wear and reassembly, weights at the shaft ends won't solve the problem. Then the visit turns into diagnostics: we measure, find the cause, give an assessment, and write up a work order for the workshop to balance the assembled rotor on a balancing machine.
Hydraulic Causes of Pump Vibration: Cavitation, Vane-Pass Frequency, and Operating Point
Hydraulics produces vibration that doesn't live at running speed, and weights have no effect on it. Cavitation produces a "gravel" crackle and a broadband rise across roughly 1 to 10 kHz, recirculation at low flow produces unstable subsynchronous components (vibration at frequencies below running speed), and vanes passing the diffuser or volute tongue produce a peak at the vane-pass frequency f_v = z · n / 60, where z is the number of vanes and n is the speed in rpm. Telling all this apart from imbalance is straightforward: hydraulics responds to valve position, flow rate, and suction head, while imbalance responds to nothing except speed. First restore the operating point and the NPSH margin (the cavitation pressure margin at the suction), then re-measure, and only then decide whether balancing is needed.
Describe your equipment and the problem
We'll answer your questions, clarify the details, and let you know what's needed for an estimate and a visit.