Uncertainty and calibration analysis

All measurements contain some deviation from the true value which is being measured. In the common vernacular this deviation between the true value and the measured value is called an inaccuracy, an error, or a mistake. Since all measurements contain errors, it is necessary to accept that there is a limit to how accurate a measurement can be. The undertainty interval combined with the confidence level, is one measure of the accuracy for a measurement or value. Without a statement of uncertainty (or a similar parameter) it is not possible to evaluate if the accuracy of the measurement, or data, is appropriate. The preparation of technical reports, calibration evaluations, and design calculations should consider the accuracy of measurements and data being used. There are many methods to accomplish this. This report provides a consistent method for the handling of measurement tolerances, calibration evaluations and uncertainty calculations. The SRS Quality Assurance (QA) Program requires that the uncertainty of technical data and instrument calibrations be acknowledged and estimated. The QA Program makes some specific technical requirements related to the subject but does not provide a philosophy or method on how uncertainty should be estimated. This report was prepared to provide a technical basis to support the calculation of uncertainties and the calibration of measurement and test equipment for any activity within the Experimental Thermal-Hydraulics (ETH) Group. The methods proposed in this report provide a graded approach for estimating the uncertainty of measurements, data, and calibrations. The method is based on the national consensus standard, ANSI/ASME PTC 19.1.