When a fan, pump impeller, or motor rotor develops unbalance, the classic remedy is to send it to a balancing machine, but that means removing it from service, dismantling the equipment, and losing days of production. Field balancing avoids all of that by balancing the rotor in place, on its own bearings, while it stays installed. This guide explains how trial weights and the influence-coefficient method let a technician cancel the once-per-revolution unbalance vibration without removing the rotor, when single-plane versus two-plane balancing is needed, and how residual unbalance is verified against tolerance.
Field balancing in one line: Field balancing, also called in-situ balancing, is the practice of balancing a rotor while it remains installed in its own bearings rather than on a dedicated balancing machine. Using the measured once-per-revolution vibration and phase, a technician applies trial weights and the influence-coefficient method to calculate the correction weight that cancels the unbalance. Single-plane balancing suits narrow rotors while two-plane is needed for longer ones, and the residual unbalance is verified against a tolerance to confirm the job is complete.
Unbalance exists when a rotor's mass is not evenly distributed about its axis of rotation, so a heavy spot sits somewhere on the wheel. As the rotor spins, that heavy spot generates a centrifugal force that rotates with it, pushing the bearings outward once for every revolution. This is why the vibration signature of unbalance is dominated by the component at exactly the running speed, the once-per-revolution or 1x frequency: the force cycles precisely once each turn, and the vibration follows. A strong, steady 1x vibration, with its phase locked to shaft rotation, is the classic fingerprint that separates unbalance from other faults.
That fingerprint is what makes field balancing possible, because both the size and the location of the heavy spot are encoded in the 1x vibration. The amplitude of the 1x vibration reflects how much unbalance is present, and its phase, the angle at which the vibration peaks relative to a reference mark on the shaft measured with a tachometer or phase sensor, points toward where the heavy spot is. Reading amplitude and phase together tells the technician not just that the rotor is unbalanced but roughly how badly and in which direction, which is the raw material for calculating a correction.
It is worth confirming that 1x really is unbalance before balancing, because a few other faults, notably certain misalignment conditions and a bent shaft, can also raise the 1x component. A quick check that the problem is dominated by 1x, is consistent, and behaves like unbalance, rather than being accompanied by the strong 2x of misalignment or the axial signature of a bent shaft, avoids the wasted effort of trying to balance out a problem that weights cannot fix. Assuming every 1x vibration is unbalance is a common trap.
The core problem in field balancing is that the technician can measure the 1x vibration but does not initially know how the rotor will respond to a correction weight of a given size at a given position. The influence-coefficient method solves this empirically with a trial weight. First the original vibration, its amplitude and phase, is recorded. Then a known trial weight is attached at a known angular position on the rotor and the machine is run again, and the new amplitude and phase are recorded. The change between the two runs shows exactly how the rotor responded to that known weight at that known place.
That measured response is the influence coefficient: it relates a weight at a position to the vibration it produces, capturing the rotor's particular sensitivity and phase lag. Once it is known, the technician can work backward from the original unbalance vibration to calculate the size and angular position of the correction weight needed to cancel it. Modern balancing instruments and even smartphone-based analysers do this vector arithmetic automatically, taking the before and trial-run readings and computing where and how much correction weight to add. The trial weight may be left in place as part of the correction or removed, depending on how the calculation works out.
In practice a first correction rarely lands perfectly, so field balancing is iterative. After the calculated correction weight is fitted, the machine is run again and the residual 1x vibration is measured; if it is not yet within tolerance, a further trim correction is calculated from the new readings and applied. Each pass uses the response information already gathered, so the vibration converges toward the target over a small number of runs. This measure, add weight, re-measure loop is the practical rhythm of a field balance, and a good result is usually reached in a few iterations.
Whether one plane or two is needed depends on the rotor's geometry. A narrow rotor, such as a single thin fan wheel or a grinding wheel, can usually be corrected in a single plane, because its unbalance acts effectively in one plane and one correction weight can cancel it. This is the simpler case, requiring measurement at one bearing and one trial and correction weight. Longer rotors, such as multi-stage pumps, long fans, and motor rotors, can carry unbalance distributed along their length that also creates a rocking, couple unbalance, and correcting this needs weights in two separate planes, with the two planes interacting so the calculation handles them together.
Two-plane balancing follows the same influence-coefficient logic but with more measurements and more bookkeeping. Vibration is measured at two bearings, trial weights are applied in each plane, and the response of both measurement points to weights in both planes is captured, giving a set of influence coefficients that the instrument uses to solve for correction weights in both planes at once. It is more involved than single-plane work, which is why the technician chooses the simplest approach the rotor allows, using two planes only when the geometry or the presence of couple unbalance requires it.
The job is finished not when the vibration merely looks better but when the residual unbalance is verified against a tolerance. Balance quality standards, such as the ISO framework for rotor balance grades, express acceptable residual unbalance in terms of the rotor's mass and operating speed, with faster and more critical machines demanding tighter tolerances. The technician confirms the final 1x vibration, or the calculated residual unbalance, falls within the grade specified for that machine before declaring it balanced. This verification step is what turns a field balance from an improvement into a documented, defensible result. Because unbalance shows up so clearly as 1x vibration, a cloud SCADA and monitoring platform such as Merobix that trends vibration across a fleet of rotating equipment can flag a machine whose 1x is climbing, prompting a field balance before the bearings suffer, and can hold the before-and-after vibration so a reliability team confirms the correction held and catches unbalance if it returns.
Single-plane balancing corrects unbalance with weights in one plane and suits narrow rotors like a single fan wheel whose unbalance acts effectively in one plane. Two-plane balancing uses weights in two separate planes and is needed for longer rotors that can carry unbalance distributed along their length, creating a rocking couple that one plane cannot fix. Technicians use the simpler single-plane approach when the rotor geometry allows it and switch to two planes only when required.
A trial weight is a known mass attached at a known position on the rotor to learn how the machine responds to a correction. By comparing the vibration before and after fitting it, the technician measures the influence coefficient that relates a weight at a place to the vibration it produces. That response is then used to calculate the exact size and position of the correction weight needed to cancel the unbalance, which cannot be known in advance without it.
The residual unbalance is checked against a tolerance from a balance-quality standard, such as the ISO balance-grade framework, which sets acceptable residual unbalance based on the rotor's mass and operating speed. Faster and more critical machines require tighter grades. The rotor is considered balanced when the final 1x vibration or calculated residual unbalance falls within the grade specified for that machine, which is verified and documented rather than judged by feel.
Merobix reads your field devices into a cloud SCADA - the real thing behind these terms, live in days from any browser.