ISO 10816 / 20816 Vibration Severity Chart and Limits Explained

What ISO 10816 and ISO 20816 Actually Define

ISO 10816, formally titled Mechanical vibration. Evaluation of machine vibration by measurements on non-rotating parts, was first published by the International Organization for Standardization in the 1990s as a multi-part series covering different machinery classes. Its core contribution was a structured framework that moved vibration assessment away from ad-hoc engineering judgment toward a documented, repeatable zone system applicable across industries and jurisdictions.

ISO 20816, Mechanical vibration. Measurement and evaluation of machine vibration, supersedes ISO 10816 part by part from 2016 onwards. The key technical evolution is that ISO 20816 explicitly accommodates both bearing-housing measurements (non-rotating parts, inherited from ISO 10816) and shaft relative displacement measurements on rotating shafts. A technique used extensively for large turbomachinery equipped with proximity probes. This dual-measurement approach is more aligned with how modern condition monitoring systems actually operate in plant environments.

In practical terms, most mid-size industrial users (pump operators, compressor plant managers, HVAC engineers), will encounter ISO 20816-3 (industrial machines) most often. Heavy industry dealing with steam turbines above 50 MW references ISO 20816-2, while hydraulic power plant operators use ISO 20816-5. The standard explicitly does not cover reciprocating internal combustion engines, which fall under ISO 8528 and ISO 3046.

The Physics Behind Vibration Severity

Rotating machinery vibration arises from imbalance forces, misalignment, bearing defects, resonance, looseness, and aerodynamic or hydraulic excitation. The severity of these forces manifests as displacement (how far the machine surface moves, µm), velocity (how fast it moves, mm/s), and acceleration (the rate of velocity change, m/s² or g). Each quantity emphasises a different frequency range: displacement amplifies low-frequency content, velocity is broadly flat across the mid-range (10–1000 Hz) making it most useful for general severity assessment, and acceleration amplifies high-frequency content relevant to bearing inner/outer race defects.

This is why ISO 10816 and ISO 20816 select RMS vibration velocity as their primary severity metric. RMS (root mean square) integrates the energy content of the entire measured frequency band into a single number, making it both physically meaningful and operationally practical. A machine with broadband vibration velocity of 2.0 mm/s RMS is carrying roughly twice the vibratory energy flux per unit area compared to one at 1.0 mm/s RMS. The relationship to mechanical fatigue and bearing wear is approximately quadratic with velocity amplitude, making early detection of zone transitions critically important.

The relationship between velocity RMS and displacement amplitude at a given frequency is: v = 2πf × d, where v is peak velocity (mm/s), f is frequency (Hz), and d is peak displacement (mm). At 50 Hz (typical mains-frequency excitation from a two-pole motor), a displacement of 14.3 µm peak corresponds to approximately 4.5 mm/s peak, or about 3.2 mm/s RMS. Placing a rigid-mounted 100 kW motor squarely at the Zone B/C boundary under ISO 20816-3.

Zone Classification: Reading the Severity Tables

Both ISO 10816 and ISO 20816 organise machines into groups based on rated power and mounting rigidity, then assign zone boundaries to each group. The table below shows indicative zone boundary values from ISO 20816-3 for the two most commonly encountered groups in Singapore's manufacturing and utilities sectors. Always consult the published standard for the definitive values applicable to your specific machine configuration.

Flexible mounting (anti-vibration mounts, spring isolators) shifts zone boundaries upward (typically by a factor of 1.6×), because the isolators absorb a proportion of the vibration before it reaches the bearing housing measurement point. Misapplying rigid-mount limits to a flexibly mounted machine will generate systematic false Zone C and D readings, leading to unnecessary maintenance costs and operational disruption.

Measurement Methods: How to Take a Compliant Reading

A compliant ISO 10816 / 20816 measurement requires the right transducer, mounting method, measurement location, and operating condition. All four must be correct simultaneously. The standard specifies that measurements be taken at the main bearing housings in three orthogonal directions (vertical, horizontal, and axial), with the machine operating at normal load and thermal steady state (typically after 30 minutes of warm-up for motors and pumps).

Transducer Selection and Mounting

For the 10–1000 Hz frequency range required by ISO 10816-3 / 20816-3, a general-purpose accelerometer with flat response in this band is appropriate. Typical specifications: sensitivity 10–100 mV/g, resonance frequency above 15 kHz, weight below 50 g to avoid mass-loading effects on lightweight structures. Mounting method critically affects high-frequency response: stud mounting (threaded into the machine surface) gives the widest usable bandwidth; adhesive mounting reduces usable range to approximately 5 kHz; hand-held probe contact is adequate only below 1 kHz and introduces significant operator-to-operator variability. It should not be used for zone classification decisions.

The measurement instrument must integrate the accelerometer signal twice in frequency domain (or use a velocity transducer directly) to yield velocity in mm/s, then compute the RMS of the broadband signal. Portable vibration analysers from manufacturers such as Brüel & Kjær, SKF, Fluke, and Emerson all implement this processing, but their accuracy depends on the calibration state of the connected transducer. A certified calibration check verifies the complete measurement chain (transducer + cable + instrument), not just individual components.

Operating Condition Requirements

Both ISO 10816 and ISO 20816 require measurements at rated speed and load. For variable-speed drives (VSDs), the standard is typically interpreted as measurement at the maximum continuous rated operating point unless the application involves operation across a wide speed range, in which case measurements at multiple speed points are required. Transient start-up and coast-down data are useful for diagnostics but are not part of the zone severity assessment under the standard.

Calibration Implications for Vibration Instrumentation

The calibration certificate for a vibration transducer or meter documents its sensitivity deviation from the nominal value across a range of frequencies and amplitudes. For accelerometers, the primary calibration parameter is sensitivity in mV/g (or pC/g for charge-output types), determined by comparison against a reference accelerometer on a precision vibration exciter. A technique standardised in ISO 16063-11 (primary calibration by laser interferometry) and ISO 16063-21 (secondary calibration by back-to-back comparison).

Under ISO/IEC 17025, a calibration laboratory performing vibration transducer calibration must demonstrate traceability of its reference exciter's acceleration amplitude to the national measurement standard, in Singapore, the National Metrology Centre (NMC) at A*STAR. Unitest Instruments (SAC-SINGLAS Acc. No. LA-2023-0845-C) maintains this traceability chain, meaning a calibration certificate issued by us directly supports ISO 10816 / 20816 compliance documentation accepted by MOM inspectors and third-party auditors.

A calibrated vibration measurement system should carry an uncertainty statement on its certificate. Typically expressed as ±X% at k=2 (95% confidence). For a well-calibrated general-purpose accelerometer with stud mounting, expanded uncertainty values of ±2–5% across 10–1000 Hz are achievable. This uncertainty must be factored into zone boundary decisions: if your measured value is within the uncertainty band of a zone boundary, additional investigation or more accurate measurement is warranted before making a maintenance decision.

Singapore Regulatory Context

Singapore's industrial safety landscape is governed primarily by the Workplace Safety and Health (WSH) Act (Cap. 354A) and its subsidiary legislation, administered by the Ministry of Manpower (MOM). The WSH (General Provisions) Regulations require that machinery and plant be maintained in a condition that does not pose risk to workers. A requirement that in practice references internationally recognised standards including ISO 10816 and ISO 20816 for rotating equipment.

Factories operating higher-hazard rotating equipment (compressors, turbines, large pumps), under MOM's Major Hazard Installation (MHI) framework are required to maintain a Mechanical Integrity programme. Industry guidance aligned with MOM expects periodic vibration monitoring against severity criteria consistent with ISO 10816 / 20816 as part of this programme. Critically, vibration readings used as compliance evidence must come from calibrated instruments. MOM inspectors and third-party risk assessors (Approved Examiners under the WSH Act) have increasingly requested SAC-SINGLAS calibration certificates for measurement equipment used in condition monitoring programmes.

Singapore Standard SS 598:2014 (Code of practice for the maintenance of mechanical equipment) also references vibration monitoring best practices aligned with the ISO 10816 framework for rotating machinery in commercial and industrial buildings. Covering chillers, pumps, fans, and cooling towers. Building owners and facilities managers targeting BCA Green Mark Platinum or ISO 50001 energy management certification commonly use vibration monitoring as part of their predictive maintenance programme, where ISO 10816 / 20816 provides the assessment framework.

Common Mistakes and How to Avoid Them

The most frequent errors encountered when applying ISO 10816 / 20816 in practice fall into four categories. Understanding them prevents wasted maintenance budget and missed defect detection.

1. Applying the Wrong Machine Group

ISO 20816-3 organises machines by rated power and support rigidity into Groups 1 through 4. Applying Group 1 limits (15–75 kW) to a 250 kW motor incorrectly classifies the machine into Zone C when it may legitimately operate in Zone B under Group 2 limits. Always verify the rated power of the motor or driven equipment and the mounting type (rigid or flexible) before selecting the applicable severity zone table.

2. Using Hand-Held Probes for Zone Assessment

Hand-held vibration probes introduce contact force and angular variation that can cause ±20–30% measurement scatter at frequencies above 500 Hz. For trending and alarm purposes this may be acceptable, but for formal zone classification against ISO 10816 / 20816 boundaries (particularly for maintenance or insurance documentation), stud-mounted or adhesive-bonded transducers are required. Use a torque wrench to achieve the specified stud torque (typically 5–7 Nm for M5 studs) for repeatable coupling.

3. Measuring at the Wrong Operating Point

Motors measured during start-up, at partial load, or before reaching thermal equilibrium will give unrepresentative readings. At reduced load, imbalance forces are proportionally lower; at start-up, resonance traverse can give temporary Zone C or D readings that are not indicative of steady-state condition. Always document the operating speed (rpm), load (%), and elapsed run time when recording zone assessment data.

4. Neglecting Calibration Intervals

Many plant maintenance teams calibrate vibration instruments at purchase and then never again. Accelerometer sensitivity drifts due to repeated shock loading (particularly for sensors used near compressors or presses), temperature cycling, and connector corrosion. A 12-month calibration interval is the minimum recommended; high-shock environments should use 6-month intervals. The calibration certificate must document the measurement uncertainty and frequency range to ensure compliance with the ISO 10816 / 20816 measurement requirements. For detailed guidance on how often instruments should be sent for calibration, see our article on calibration intervals and how to set them correctly.

Vibration Severity vs Vibration Analysis: Two Different Disciplines

ISO 10816 and ISO 20816 severity zones answer a single, deliberately narrow question, is the overall vibration level acceptable, marginal, or dangerous, and they do this using a single broadband RMS velocity number per measurement location. This is not the same discipline as vibration analysis, which decomposes that same signal into its constituent frequency components (a spectrum) to diagnose the specific root cause of elevated vibration, imbalance typically shows up at 1× running speed, misalignment often produces strong 2× components, bearing defects generate characteristic high-frequency signatures tied to the bearing's geometry, and looseness frequently shows a broad harmonic series. A maintenance team using only the ISO 20816 zone number knows a machine has crossed from Zone B into Zone C, but they do not yet know why, and jumping straight to corrective maintenance without a spectral diagnosis often means guessing at the fix.

The two disciplines are complementary rather than competing, and mature predictive maintenance programmes in Singapore's industrial sector use severity zone monitoring as the trigger and vibration spectrum analysis as the diagnostic follow-up. A rising broadband RMS trend crossing toward a Zone B/C boundary prompts the maintenance team to pull a full frequency spectrum from the same accelerometer, which then reveals whether the root cause is a developing bearing defect, a coupling misalignment introduced during a recent overhaul, or simple imbalance from blade fouling on a fan. Both disciplines depend on the same calibrated transducer and the same traceability chain described above; the difference is entirely in how the recorded signal is processed and interpreted after the fact, not in the measurement hardware itself.

Building a Practical Vibration Monitoring Programme

Translating ISO 10816 / 20816 into an operating maintenance programme, rather than a document referenced only during commissioning, requires a few concrete decisions a facilities or reliability engineering team should make early. First, establish a baseline reading for every critical machine shortly after commissioning or a major overhaul, while the equipment is in known-good condition, since severity zone tables describe absolute limits but a machine's own historical baseline is frequently a more sensitive early-warning indicator than the zone boundary itself, a machine that has crept from 1.2 mm/s to 2.0 mm/s RMS while remaining comfortably inside Zone A is still worth investigating even though it has not technically crossed any zone limit. Second, define measurement points and directions explicitly and mark them physically on the machine, since ISO 10816 / 20816 zone limits assume consistent measurement locations and directions (vertical, horizontal, axial at each bearing housing), and a monitoring programme that measures at slightly different points each visit introduces variability that masks genuine trends. Third, set a monitoring frequency proportional to the machine's criticality and its historical stability, monthly or continuous online monitoring for machines whose failure would stop a production line or create a safety hazard, quarterly or semi-annual for less critical rotating equipment with a stable trend history. Fourth, and most often skipped, tie the monitoring programme's calibration schedule directly into the same equipment register used for the machines themselves, so a transducer's calibration due date is never overlooked simply because it lives in a separate instrument list from the plant maintenance schedule it supports.