Uncertainty of Measurement (UoM)

In today's digital world the assumption is often made that an everyday measurement has no uncertainty. For every measurement made however, there is always uncertainty. Luckily for us, in most countries around the world there is a harmonised measurement & traceability system that ensures UoM will be relatively low. This ensures that when purchasing products by mass or volume, we get a fair deal. Internationally, traceability of measurement that most countries subscribe to is administered by BIPM International Bureau of Weights and Measures in France.

Reference

JCGM 100:2008 Evaluation of measurement data (GUM with minor updates) - Guide to the expression of uncertainty in measurement
Internationally accepted rules for evaluating and expressing UoM.

Definition

"an estimate characterising the range of values within which the true value of a measurand lies"
OR
"parameter, associated with the result of a measurement, that characterises the dispersion of the values that could reasonably be attributed to the measurand"

Measurand

Official technical term for a quantity that is being determined by Measurement, or more simply put; the true measured value.

UoM involves defining what confidence is associated with the Measurand based on a series of measurements and a statistical analysis of the repeatability. Hence, repeatability of measurements is critical to minimising uncertainty. Also, when designing and selecting flow measurement equipment engineers need to fully evaluate all of the component uncertainties that contribute to overall UoM. Note that this includes calibration equipment and any tolerances used/specified in verification procedures.

The following graph from ISO 5168 - Figure 1 shows a normal or Gaussian distribution of typical raw measurement results statistically analysed, X1 is standard deviation, X2 is coverage factor, Y is percent of readings. Coverage factor (confidence level) defines the factor by which standard uncertainty is multiplied by to express a level of confidence in any measurement.

Pasted Graphic 3

For measurement applications a coverage factor of 2 (95% confidence level) is usually used.

Uncertainty Analysis:

When performing an uncertainty analysis GUM recommends taking either a Type A or Type B approach. Type A usually applies to estimating the uncertainty in a large amount of experimental data from primarily one source. Type B usually applies to estimating uncertainty from disparate other sources such as manufacturers data but can include some data from actual measurements such as calibration of a transmitter. For flow measurement systems we recommend an uncertainty analysis is performed over the flow range expected to validate that the meter has been sized and selected correctly for the intended duty. It will also validate the design of all secondary equipment used for generating a reported quantity.

Our capability in this area:

An analysis of any particular measurement system can provide a report that details sources of uncertainty, their contributions and the combined UoM.

Design advice on reducing uncertainty of measurement whether it's in the instruments themselves, the way they are installed or the way they are maintained and calibrated.

Preparation of Accuracy Verification Testing procedures and Measurement Management documentation in accordance with AS/NZS ISO 10012.

When analysing this measurement system:

MeasSystem1

The following graph is the result of the uncertainty analysis performed on the Venturi meter and associated flow measurement equipment:

UncertGraph
Both mass flow rate (red trace) and uncertainty (blue trace) are shown plotted against measured differential pressure (x-axis). In this case dual range differential pressure (flow) transmitters are used hence the turndown is approximately 10:1. Uncertainty rises dramatically as the minimum flow point (Qmin) is approached.

Real World Sources of Mis-Measurement

Flow measurement always presents operating challenges:

Sulphur 010Sulphur 029

Contamination like this in meter tubes can cause significant financial loss. In this case a sulphur buildup in a meter tube with a turbine meter and upstream flow conditioner caused a significant mis-measurement. Trace quantities of contaminants in a gas stream can build up over time hence the importance of regular maintenance and inspection.