Field Balancing: How to Balance Rotating Equipment On-Site
Key Takeaways
- Field balancing corrects rotating equipment in-place using vibration phase and amplitude data, avoiding the cost and disruption of full rotor removal.
- The influence coefficient method is the dominant analytical approach: it uses vector math from trial mass runs to calculate the exact corrective weight, eliminating iterative guesswork.
- ISO 1940-1 balance quality grades (G0.4 to G40) define the allowable residual unbalance for a given rotor mass and service speed; most industrial rotating equipment targets G2.5 to G6.3.
- Continuous vibration monitoring after balancing confirms that corrections hold over time and flags re-developing imbalance before it causes bearing damage.
What Is Field Balancing?
Field balancing addresses a fundamental problem in rotating machinery: unbalanced rotors generate centrifugal forces at running speed that load bearings, increase vibration, and accelerate wear. Unlike shop balancing, which requires removing the rotor and placing it on a dedicated balance machine, field balancing is performed with the rotor still mounted in the machine and operating under normal bearing conditions.
The technique is applicable to a wide range of rotating equipment including pump impellers, fan wheels, motor rotors, compressor wheels, and turbine blades. It is a core skill in predictive maintenance programs and is often the first corrective action when vibration analysis identifies elevated 1x (once-per-revolution) vibration as the dominant fault signature.
Static vs. Dynamic Imbalance
Understanding imbalance type determines how many correction planes are needed and shapes the entire balancing procedure.
Static Imbalance
Static imbalance exists when the rotor's center of mass is displaced from its geometric center, but the principal inertia axis remains parallel to the rotational axis. If you placed the rotor on frictionless bearings with no rotation, it would roll until the heavy side faced downward. A single correction weight added in the same plane as the heavy spot eliminates the unbalance force. Thin, disc-like components such as grinding wheels and fan impellers with low axial length-to-diameter ratios typically exhibit primarily static imbalance.
Dynamic Imbalance
Dynamic imbalance occurs when the principal inertia axis is tilted relative to the rotational axis. This tilt creates a couple: two equal, opposite forces separated axially that produce a rocking motion during rotation. Dynamic imbalance is invisible at rest and can only be detected and corrected during rotation. Most wide rotors exhibit a combination of static and couple imbalance, which is why two-plane correction is the standard for pumps, long motor rotors, and multi-stage impellers.
Single-Plane vs. Two-Plane Balancing
The selection of single-plane or two-plane balancing depends on the rotor geometry and the dominant imbalance type.
Single-Plane Balancing
Single-plane balancing, sometimes called static balancing, places all corrective weight in one transverse plane. It is appropriate when the axial length-to-diameter ratio is 0.5 or less, the rotor runs below the first critical speed, and the mass distribution is reasonably uniform across the width. Typical candidates include fan impellers, pump impeller discs, and V-belt pulleys. The procedure requires one reference run and one trial run, making it the faster field procedure.
Two-Plane Balancing
Two-plane balancing applies corrections in two axially separated planes to cancel both the net unbalance force and the couple moment. It is required for rotors with an axial length-to-diameter ratio above 0.5, rotors operating at or near critical speed, and any component where bearing proximity allows the couple component to generate significant forces at the bearing locations. Two-plane procedures require a reference run and two trial runs (one per plane), plus the cross-effect corrections described in the influence coefficient method below.
The Influence Coefficient Method
The influence coefficient method is the mathematical foundation of modern industrial vibration analysis based field balancing. It treats the vibration response as a linear system and computes how the vibration vector at each measurement point changes per unit of mass added at each correction plane.
Procedure: Single-Plane Case
- Reference run. Record the 1x vibration vector (amplitude A and phase angle P) at steady-state operating speed with no trial weight. This is the original unbalance vector O.
- Trial mass run. Stop the machine, attach a known trial mass T at a known angular position (typically 0 degrees, referenced to the trigger mark). Restart and record the new vibration vector O+T.
- Compute the influence coefficient. The influence coefficient alpha is: alpha = (O+T - O) / T. In vector terms, subtract the original vector from the trial vector to get the effect of the trial mass alone, then divide by the trial mass value. This gives the vibration change per unit of added mass at that plane.
- Compute the corrective mass. The corrective weight C = -O / alpha. The magnitude of C is the weight to add, and the angle of C (adjusted for the sign reversal) gives the angular position. Place the weight at that angle and remove the trial mass.
- Verification run. Run the machine and confirm the residual 1x vibration is within the acceptance limit. Trim corrections iterate the same math if residual remains above target.
Two-Plane Extension
In two-plane balancing, four influence coefficients are needed because a correction at Plane 1 affects vibration at both Bearing 1 and Bearing 2 (and symmetrically for Plane 2). The coefficients are:
- alpha11: vibration change at Bearing 1 per unit mass at Plane 1
- alpha21: vibration change at Bearing 2 per unit mass at Plane 1
- alpha12: vibration change at Bearing 1 per unit mass at Plane 2
- alpha22: vibration change at Bearing 2 per unit mass at Plane 2
These four coefficients are determined from two trial runs (one weight at each plane), and the corrective weights are solved from the resulting 2x2 linear system. Modern vibration analyzer instruments perform this matrix calculation internally and display the corrective weight magnitude and angle directly.
ISO 1940 Balance Quality Grades
ISO 1940-1 provides the internationally accepted specification for permissible residual unbalance. The standard assigns quality grades labeled G0.4 through G40, where the G value equals the product of the permissible eccentricity (mm) and the maximum service angular velocity (rad/s).
| Grade | G Value | Typical Rotor Types | Application Examples |
|---|---|---|---|
| G0.4 | 0.4 mm/s | Precision spindles, gyroscopes | High-speed grinding spindles, precision instrument rotors |
| G1.0 | 1.0 mm/s | Gas and steam turbines | Turbine rotors, turbocharger wheels |
| G2.5 | 2.5 mm/s | Large electric motors, compressors | Motors above 15 kW, centrifugal compressors, process blowers |
| G6.3 | 6.3 mm/s | Centrifugal pumps, industrial fans | Process pumps, HVAC fans, general industrial rotors |
| G16 | 16 mm/s | Agricultural machinery, crushers | Grain fans, hammer mills, slow-speed crushers |
| G40 | 40 mm/s | Crankshaft assemblies, diesel engines | Multi-cylinder engines, slow crankshafts |
Worked Example: Pump Impeller at G6.3
Consider a process pump impeller with the following parameters:
- Rotor mass: 50 kg
- Service speed: 1,500 rpm (157 rad/s)
- ISO 1940 grade target: G6.3
Step 1: Calculate the permissible eccentricity. From the grade definition: e_per = 1,000 x G / omega = 1,000 x 6.3 / 157 = 40.1 micrometers.
Step 2: Calculate the permissible residual unbalance per plane. For a single-plane impeller: U_per = e_per x m = 40.1 micrometers x 50 kg = 2,005 g-mm.
Step 3: Express as a practical weight at a correction radius. If the correction radius is 150 mm, the allowable correction weight is: W = U_per / r = 2,005 / 150 = 13.4 grams.
After balancing, if the residual vibration calculation indicates the remaining unbalance mass is below 13.4 grams at the 150 mm radius, the job is complete to G6.3.
Vibration Acceptance Limits
Beyond ISO 1940 unbalance mass tolerances, plant engineers typically specify post-balancing vibration limits in velocity (mm/s RMS) or displacement (microns peak-to-peak) at the bearing housing. ISO 10816-3 (machinery in the 15 kW to 15 MW range) classifies vibration severity into four zones:
- Zone A (0 to 2.3 mm/s): New machinery in good condition.
- Zone B (2.3 to 4.5 mm/s): Acceptable for long-term operation.
- Zone C (4.5 to 7.1 mm/s): Short-term operation tolerated; plan corrective action.
- Zone D (above 7.1 mm/s): Vibration severe enough to damage the machine; immediate action required.
A successful field balance brings the 1x component below Zone B limits and total broadband vibration into Zone A or B. Vibration monitoring after balancing tracks whether the machine remains within acceptable zones or trends toward re-developing imbalance due to material build-up, erosion, or thermal effects.
Field Balancing vs. Shop Balancing
Both methods correct rotor imbalance, but each is suited to different operational scenarios. The choice affects both the achievable precision and the total downtime required.
| Factor | Field Balancing | Shop Balancing |
|---|---|---|
| Rotor removal | Not required | Required |
| Downtime | Lower; multiple run-up cycles needed | Higher; includes disassembly and reassembly |
| Achievable grade | G2.5 to G6.3 typically achievable | G0.4 to G2.5 achievable with precision machine |
| Bearing influence | Actual service bearings used; more representative | Tooling bearings used; may need field trim |
| In-service build-up | Can correct recurring build-up without removal | Cannot account for in-service conditions |
| Equipment required | Portable vibration analyzer with phase input | Dedicated dynamic balance machine |
| Best use case | Verification, correction between overhauls, recurring imbalance | New rotor commissioning, major overhaul, tight tolerance requirements |
Tools Required for Field Balancing
Successful field balancing depends on having the right instrumentation and hardware.
Vibration Analyzer with Phase Measurement
A vibration sensor and analyzer capable of displaying the 1x vibration component as a polar vector (magnitude and phase angle) is the core instrument. The analyzer must accept a tachometer trigger input to synchronize phase measurement to shaft rotation. Many modern route-based data collectors and online monitoring platforms include this functionality.
Once-Per-Revolution Trigger
Phase measurement requires a consistent reference point on the shaft. The standard setup uses a strip of reflective tape attached to the shaft and an optical or laser sensor that generates one pulse per revolution. The phase angle is measured as the delay between the trigger pulse and the vibration waveform peak, referenced to the trigger mark location.
Stroboscope (Optional)
A stroboscope can visually confirm the angular position of the trigger mark during rotation and cross-check the calculated heavy-spot location before drilling, grinding, or adding weights. It is a useful secondary verification tool, particularly when the analyzer's phase display is unfamiliar or when the correction position needs to be physically marked on the rotor.
Trial Masses and Correction Weights
Trial masses must be accurately known in grams and securely attached during the trial run. Common materials include calibration weights, bolts, hose clamps, or weld beads. Final corrections are made by drilling, grinding, or adding permanent weights via tapped holes, clip-on balance weights, or weld material, depending on the rotor type and site safety standards.
Common Mistakes in Field Balancing
Field balancing errors fall into a few predictable categories, most of which are preventable with proper procedure.
Incorrect Phase Reference Setup
If the trigger mark and sensor are not positioned consistently between runs, the phase angles recorded in the reference and trial runs are referenced to different angular positions, making the influence coefficient calculation invalid. The trigger mark must remain in the same position on the shaft throughout all runs, and the sensor must not move.
Trial Mass Too Small or Too Large
The trial mass must produce a measurable change in the vibration vector. A trial mass that changes vibration by less than 15 to 20 percent of the original amplitude produces results dominated by measurement noise. A trial mass that changes the vibration by more than 100 percent may enter nonlinear response territory. As a starting rule: select a trial mass that produces a centrifugal force equal to 10 to 15 percent of the estimated rotor weight at running speed.
Measuring at a Resonant Speed
If the running speed is near a structural or rotor critical speed, the phase and amplitude data become highly sensitive to small speed changes and do not accurately represent the imbalance condition. When possible, avoid balancing at speeds within 20 percent of a known resonance. If this is unavoidable, multiple measurement speeds and additional analysis are required.
Mistaking Other Faults for Imbalance
Elevated 1x vibration can result from shaft misalignment, bent shaft, looseness, eccentric coupling, or electrical forcing at line frequency. Attempting to balance a machine with a bent shaft or misalignment adds corrective weight that partially cancels the fault signature without fixing the underlying problem. Confirm that the dominant failure mode is imbalance through spectrum analysis and phase measurements at both bearings before beginning the balancing procedure.
Unsafe Weight Attachment
Trial masses and correction weights that come loose during rotation become projectiles. All weights must be mechanically secured and checked for security before restart. Follow the lockout/tagout procedure between every run to ensure the rotor is fully stopped before anyone approaches it.
Residual Unbalance Verification
After completing the balancing procedure, residual unbalance verification confirms that the corrected rotor meets the ISO 1940 grade specification. The two approaches used in the field are:
Vibration-Based Verification
Calculate the equivalent unbalance mass from the residual 1x vibration amplitude using the machine's known sensitivity (influence coefficient). If the calculated equivalent unbalance mass is below U_per for the target grade, the balance is accepted. This is the most practical method in the field because it requires no additional instrumentation.
Direct Measurement with Known Weights
Add a small test mass at the calculated heavy-spot angle and confirm that it increases vibration. Remove it and add the same mass at the calculated light-spot angle and confirm that it decreases vibration. This confirms the correction is centered on the true heavy spot and that the residual is at or below the measurement sensitivity of the instrument.
Condition monitoring after balancing is equally important. Even a correctly balanced rotor can develop imbalance over time due to material build-up on fan blades, erosion of pump impeller vanes, thermal bow during startup, or blade cracking. Continuous vibration monitoring provides the early warning that allows re-balancing before bearing damage accumulates.
Frequently Asked Questions
What is the difference between static and dynamic imbalance?
Static imbalance occurs when the rotor's center of mass is displaced from its rotational axis, but the principal inertia axis remains parallel to it. It is detectable at rest. Dynamic imbalance occurs when the principal inertia axis is tilted relative to the rotational axis, producing a couple that only appears during rotation. Most real-world rotors exhibit both types simultaneously, which is why two-plane correction is standard for wide rotors such as pump impellers and long motor rotors.
What is single-plane versus two-plane field balancing?
Single-plane balancing adds or removes weight in one correction plane and suits thin, disc-like rotors with an axial length-to-diameter ratio below 0.5, such as fan impellers and pulleys. Two-plane balancing applies corrections in two axially separated planes to eliminate both the static and couple imbalance components. It is required for wide rotors such as pump impellers, long motor rotors, and multi-stage compressor wheels. Two-plane procedures require one additional trial run compared to single-plane.
How many trial runs does the influence coefficient method require?
The influence coefficient method requires one reference run, one trial run for single-plane balancing, and two trial runs (one per plane) for two-plane balancing. After applying the calculated correction, one verification run confirms the result. If residual exceeds the acceptance limit, a trim correction run may follow. Total machine starts typically range from three to five for a straightforward field balance.
What do ISO 1940 balance quality grades mean?
ISO 1940-1 quality grades run from G0.4 (most precise) to G40 (least precise). The G number equals the permissible eccentricity in mm multiplied by the service angular velocity in rad/s. Lower grades apply to precision spindles and turbines; G2.5 to G6.3 covers most industrial rotating equipment including motors, pumps, and fans. The grade determines the maximum allowable residual unbalance mass for a given rotor weight and operating speed.
When should field balancing be used instead of shop balancing?
Field balancing is preferred when removing the rotor would cause excessive downtime or disassembly cost, when the rotor develops recurring imbalance due to build-up or erosion, or when shipping and reinstallation risk disturbing the balance. Shop balancing offers higher precision (G0.4 to G1.0 achievable) and is better suited for new rotor commissioning, major overhauls, or applications with tight grade requirements. Many reliability programs combine both: shop balancing at overhaul and field balancing for in-service correction.
What is residual unbalance and how is it calculated?
Residual unbalance is the remaining imbalance after correction. ISO 1940-1 defines the permissible residual unbalance as U_per = e_per x m, where e_per = 1,000 x G / omega (omega in rad/s) and m is rotor mass in kg. For a 50 kg pump impeller at 1,500 rpm (157 rad/s) to grade G6.3: e_per = 1,000 x 6.3 / 157 = 40.1 micrometers, and U_per = 40.1 x 50 = 2,005 g-mm. At a 150 mm correction radius, this equals 13.4 grams of allowable residual weight.
What tools are required for field balancing?
Field balancing requires a vibration analyzer with phase measurement capability, a once-per-revolution tachometer trigger (reflective tape and optical sensor), trial masses for the test runs, and permanent correction weights (bolts, weld, or balance clips). A stroboscope is optional but useful for visually confirming the angular position of the trigger mark. The vibration analyzer must display the 1x vibration component as a polar vector showing both amplitude and phase angle.
The Bottom Line
Field balancing is a precision corrective technique that extends bearing life, reduces vibration-induced fatigue, and lowers energy consumption in rotating machinery without the downtime and cost of full rotor removal. The influence coefficient method provides a rigorous mathematical framework that eliminates trial-and-error, and ISO 1940-1 quality grades give engineers a clear, testable acceptance criterion for every balancing job.
Achieving a good balance is only part of the equation. The other part is maintaining it. Build-up, erosion, and operating condition changes cause imbalance to redevelop over time. Without continuous vibration monitoring, imbalance goes undetected until bearing temperatures rise or vibration becomes audible. At that point, the damage is already accumulating. Pairing field balancing capability with real-time condition monitoring creates the closed loop that keeps rotating equipment running within specification between planned maintenance events.
Monitor Rotating Equipment Health in Real Time
Tractian sensors detect imbalance, misalignment, bearing faults, and other failure modes continuously, giving your team the data to plan field balancing before damage occurs.
See How Tractian Monitors Rotating EquipmentRelated terms
IIoT (Industrial Internet of Things): Definition
IIoT connects industrial sensors, devices, and systems to collect real-time data that improves equipment reliability, enables predictive maintenance, and drives Industry 4.0 transformation.
Kaizen: Definition
Kaizen is the Japanese philosophy of continuous, incremental improvement involving every person in an organization. Learn the PDCA cycle, Kaizen events, the eight wastes in maintenance, and how Kaizen drives OEE improvement.
Kitting: Definition
Kitting is the process of pre-assembling all parts, materials, and tools for a maintenance work order before the job begins. Learn how kitting increases wrench time, reduces downtime, and integrates with CMMS planning.
Lead Time: Definition
Lead time is the total elapsed time from initiating a process to receiving its output. Learn how procurement lead time affects spare parts inventory, reorder points, safety stock calculations, and JIT strategy.
Leak Detection and Repair (LDAR): Definition
LDAR is a regulatory compliance program requiring facilities to monitor equipment components for fugitive VOC and HAP emissions and repair detected leaks. Learn EPA requirements, covered components, and monitoring methods.