Combined Cycle Performance Test Uncertainty Validation by Comparing Monte Carlo Analysis With Monovariate Perturbation Results

This paper focuses on the calculation of the test uncertainty of an ASME PTC 46 [1], overall plant performance test of a combined cycle by two separate methods. It compares the combined cycle corrected plant output and heat rate systematic uncertainty results that are generated using monovariate perturbation analysis with the Monte Carlo method. The Monte Carlo method has not been used widely in power plant performance testing applications. It offers insights into the results of the Monte Carlo analysis method, which is less intuitive than the conventional method. This study shows that utilizing two distinctly different methods of calculation of test uncertainty serves to corroborate assumptions, or to isolate flaws in one or both methods. In developing the method for calculation of test uncertainty, the authors conclude that it is prudent to validate the calculation method of choice of test uncertainty, and to consider the correlations in measurement uncertainties. Also discussed in detail are the impact of correlated uncertainty assumptions, and recommendations on their application. Correlated uncertainty has not been extensively discussed in the literature concerning specific applications in performance testing, although it should be a critical consideration in any uncertainty analysis. Details of determination of instrumentation uncertainty, measurement uncertainty of a parameter, and calculation of sensitivity factors are included in this paper.