Anemometer Calibration Wind Tunnels: ISO 17025 & MEASNET

Most wind tunnels are test facilities: the model is the subject, and the tunnel is the environment. A calibration tunnel inverts that relationship. Here the tunnel is the instrument, and the object in the flow is the thing being judged against it.

That inversion changes everything about how the facility is specified. Nobody asks a calibration tunnel for impressive speed or a large section. They ask it for a reference wind speed that can be defended, with an uncertainty budget, to an accreditation body.

1. What the standards actually require

For the wind energy sector — the largest single driver of anemometer calibration demand — the operative documents are IEC 61400-12-1 and the MEASNET Anemometer Calibration Procedure . The current Version 3 (December 2020) applies Annex F of IEC 61400-12-1, Edition 2 (2017); earlier procedure versions tracked the 2005 edition, so it is worth confirming which revision a given lab or client is actually referencing before quoting compliance.

Two flow-quality requirements sit at the centre of it:

Axial turbulence intensity below 2% at the anemometer position.
Flow uniformity to 0.2% across the area covered by the anemometer.

Alongside those, the laboratory itself is expected to operate under ISO/IEC 17025 accreditation for the relevant measurands, with national accreditation (for example DAkkS in Germany) and MEASNET membership as the route to recognised calibrations. Comparability is verified by round robin exercises between institutes, where an average compliance value of 1% or better is required.

Note the shape of these numbers: they are tight, but they are not the sub-0.1% turbulence figures quoted for low-turbulence aerodynamic research. A calibration tunnel is optimised for stability and traceability, not for the lowest achievable turbulence.

2. Why those numbers are harder than they look

The instrument blocks its own reference. A cup anemometer in a small section is not a negligible obstruction. Section size must be chosen so that blockage stays within the procedure's limits for the instruments you intend to calibrate — which is why "we'll just use a small tunnel" fails on the first accreditation review.

The reference speed must be traceable. Speed is derived, not observed: from differential pressure, air density and therefore from calibrated pressure, temperature and humidity instrumentation, each with its own traceability chain and contribution to the uncertainty budget.

Stability matters more than range. Calibration points typically sit in a modest speed band, and each point must hold steady long enough for averaging. Fine, repeatable speed control at the low end is a harder engineering problem than a high top speed.

Uniformity is measured, not asserted. The 0.2% figure applies across the area the anemometer occupies, which means the section has to be mapped and the mapping maintained.

The building is part of the uncertainty. Laboratory temperature stability, barometric measurement and even personnel movement affect results at this level.

3. Design consequences

A calibration tunnel is a small machine built to instrument-grade tolerances:

High contraction ratio with honeycomb and multiple screens, sized to deliver the required turbulence intensity and uniformity at the measurement plane.
Section sized for blockage compliance with the instrument family, not for the biggest sensor imaginable.
A defined, repeatable mounting for the device under test, with fixed geometry relative to the reference probe.
Reference instrumentation designed in — pressure, temperature, humidity — with calibration intervals and traceability planned from the start.
Drive and control tuned for stability at low speed, including how quickly the tunnel settles after a change of set point.
Environmental control of the room , not only of the flow.
Automation and data handling that produce a certificate with an uncertainty statement, because throughput determines whether the lab is a business or a hobby.

4. The business case

Search volumes for this topic are small; the buying intent is close to absolute. Wind farm developers, independent measurement institutes and turbine OEMs need calibrated anemometry, recalibration is periodic rather than one-off, and accreditation creates a defensible position: an accredited lab in a region without one captures work that currently travels abroad.

The same facility usually serves a wider sensor population than cup anemometers alone — ultrasonic anemometers, pitot tubes, flow sensors, HVAC and process instrumentation — provided the section, mounting concept and uncertainty budget were designed with that in mind.

5. What to specify when procuring one

The accreditation you intend to hold , and the procedures you must satisfy. This determines the entire specification and cannot be added later.
The instrument families to be calibrated, with their dimensions — this sets the section size through blockage limits.
The speed range and the required stability , including settling time per point.
The target uncertainty budget , decomposed by contribution.
Throughput , which drives automation, mounting design and how much of the process is manual.
The round robin plan , because comparability with peer institutes is what makes your certificates worth issuing.

6. Common mistakes

Buying a general-purpose tunnel and hoping to accredit it. Flow quality at a point is achievable; documented uniformity across the instrument area, with traceability, is a design decision.
Treating uncertainty as a post-processing task. It is an architecture requirement.
Under-sizing the section. Blockage limits are unforgiving, and enlarging a finished section is not a modification, it is a new tunnel.
Ignoring the laboratory environment. Excellent flow in an unstable room produces unstable results.
Planning for cup anemometers only. The sensor mix changes; the tunnel will not.

7. Specifying with confidence

A calibration tunnel is a small, precise, boring machine that earns money for decades — provided its specification starts from the accreditation and works backwards.

We design and build calibration and low-turbulence tunnels around exactly that logic: required uncertainty first, then flow quality, then geometry, then drive. Related reading:

improving flow quality

and our

comparison of wind tunnel types

Tell us which accreditation and which instruments and we will return a section concept, a flow-quality target and an uncertainty structure to work from.

Sources and notes

MEASNET Anemometer Calibration Procedure, Version 3 (December 2020), which states it applies Annex F of IEC 61400-12-1 Edition 2 (2017), with earlier procedure versions (V1 1997, V2 2009) tied to the 2005 edition: MEASNET Anemometer Calibration Procedure.
Axial turbulence intensity below 2% at the anemometer position and flow uniformity of 0.2% across the area covered by the anemometer reflect the figures used in wind-energy calibration practice; the current MEASNET procedure itself refers the reader to the IEC 61400-12-1 Edition 2:2017 standard text for the exact clauses, which is a paid document we have not reproduced here. Confirm the current numeric limits against that standard text before writing them into a specification.
ISO/IEC 17025 accreditation and MEASNET membership as the basis for recognised calibrations: Ammonit wind tunnel, DIN EN ISO/IEC 17025 and MEASNET, IDR-UPM anemometer calibration laboratory.
Design guidance (section sizing for blockage, reference instrumentation, stability at low speed, automation) is TunnelTech engineering practice.

Calibration Wind Tunnels: Meeting ISO/IEC 17025, IEC 61400-12-1 and MEASNET Requirements