UNCERTAINTY ANALYSIS OF THE PERFORMANCE OF CENTRIFUGAL COMPRESSORS - NUMERICAL MODEL AND MEASURED UNCERTAINTIES

摘要

Compressed air and gas are the lifeline of power plants. Deficient or unstable supply of air and gas can result in huge costs and revenue losses of plants. Thus, the accurate determination of performance of the compressors plays important role in prediction of the plant performance. In order to provide reliable and low cost operation for end users, an uncertainty analysis of volume flow, pressure ratio, and power consumption have been investigated and implemented to accurately determine the effect of the compressors to the plant performance. Mathematical models for the uncertainty treatments are proposed based on the ASME test code, and both the systematic errors originating from the measurement instruments and random errors rooted from the raw data are taken into consideration. Moreover, the approach of the uncertainty propagation is also presented through data reduction equations in this paper to evaluate the final performance. Both the rigorous numerical model and sophisticated data acquisition system instrumented in the test facility are employed to conduct the uncertainty analysis for a multi-stage centrifugal compressor. Comprehensive error sources such as ambient conditions, inter-stage pressures and temperatures, and rotational speeds are identified and studied for the final tolerance of the pressure ratio and power consumption of the whole compressor. The test uncertainty results of the compressor can help to im- prove the power plant field design and demonstrate quality assurance and quality control. Moreover, the tolerance analysis introduced in this paper can be extended to each component of the power plant system to optimize the performance of the whole power plant.