Consideration of Co-Variance in Power Plant Test Uncertainty Calculations

One of the most important aspects of American Society of Mechanical Engineers (ASME) Performance Test Code (PTC) thermal performance testing is the proper determination of test uncertainty since the Uncertainty Analysis (UA) validates the quality of a test as well as demonstrates that the test meets code requirements. It can also carry a commercial relevance when test tolerances are linked to uncertainty figures. This paper introduces an approach to the calculation of the random component of uncertainty when covariance exists between certain primary measurements in thermal performance testing. It demonstrates how to identify parameters that are co-variant, provides a methodology for properly calculating the aggregated random uncertainty of co-variant measurements, and discusses the effect of co-variance on UA results.