Rotary Gas Meters: Separate Repeatability, Indication Error and Laboratory Agreement

A rotary gas meter can repeat its reading closely and still exhibit a systematic indication error. Two laboratories can also report different numerical errors without their results being inconsistent within uncertainty. Treating repeatability, error and agreement as interchangeable makes it difficult to evaluate a calibration certificate or resolve a procurement disagreement.

Repeatability describes the spread obtained when a measurement is repeated under defined conditions. Indication error compares the meter's result with the reference quantity. Measurement uncertainty expresses the dispersion reasonably associated with the reported result, drawing on more than repeated readings alone. A narrow spread is therefore not a complete statement about accuracy.

Ask whether the reported error is an individual observation or an average of repeated measurements. Request the uncertainty, coverage information and the conditions to which it applies. A laboratory's general calibration capability is useful context, but the certificate should make clear how the actual reported result is supported.

Abdulrahman and colleagues used a DN50 rotary gas meter as a transfer instrument in a comparison involving nitrogen and carbon dioxide. Their normalized-error plots assess consistency between measurement results in relation to the comparison's uncertainty treatment. The selected nitrogen figure contains both satisfactory and more problematic points under that study's criteria. [1]

Normalized-error results for the rotary transfer meter with nitrogen. Warning and critical levels are those of this interlaboratory comparison, not generic meter acceptance limits. Source: Ara Abdulrahman, Gabriele Chinello, Revata Seneviratne, Kurt Rasmussen, Dennis van Putten and Pier Giorgio Spazzini (2026), Interlaboratory Comparison of SI-Traceable Flow Metering Calibration Facilities with Gaseous Carbon Dioxide, Figure 23. CC BY 4.0. Selected embedded source figure extracted and resized; no data or identifying technical labels changed. No editorial numbers added.

This graph is not a universal pass-fail curve for rotary meters. It evaluates a particular comparison, reference arrangement and set of flow conditions. Its warning and critical lines are the authors' comparison criteria. They must not be copied into a sales specification as if they were the maximum permissible indication error of every G16 meter.

When results differ, first align the gas, operating flow, pressure and temperature basis. Identify any transport, adjustment or maintenance between measurements. The transfer instrument itself may change, while the laboratories may use different reference methods and uncertainty models. A difference cannot automatically be attributed to one party's poor calibration.

A normalized comparison divides a difference by an uncertainty quantity constructed for the comparison. The correct construction depends on the chosen reference and any correlation or shared contributions. Avoid inserting two headline uncertainty values into a formula without checking those assumptions. A comparison against a weighted reference also needs to account for how that reference was formed.

For an African industrial-gas enquiry, state the required operating envelope and the intended legal or process use. Agree how certificate results will be assessed and what will happen if a point falls close to an acceptance boundary. That conversation should occur before delivery rather than after two different laboratories have produced reports.

Request raw repeats and the relevant configuration information alongside the final summary. The cover photograph shows an external laboratory installation from the published study, not the buyer's proposed station. Keeping the evidence and its interpretation separate allows a useful technical discussion without overstating what a good repeatability value or a single comparison statistic proves.

Send the following project inputs when requesting a configuration and evidence review:

Cover image: DN50 rotary transfer meter installed at NEL in the intermediate-scale comparison. Ara Abdulrahman, Gabriele Chinello, Revata Seneviratne, Kurt Rasmussen, Dennis van Putten and Pier Giorgio Spazzini (2026), Interlaboratory Comparison of SI-Traceable Flow Metering Calibration Facilities with Gaseous Carbon Dioxide, Figure 12. CC BY 4.0. Selected embedded source figure extracted and resized; no data or identifying technical labels changed. No editorial numbers added.