Power Plant Vibration Monitoring and Optimization Using Functional Design of Experiments

In a typical pressurized water reactor nuclear power plant, the condensate pumps contribute to the majority of the feedwater flow to the steam generators, which is converted into steam that spins the turbine generator to produce electricity. Adequate monitoring of the condensate pump feedwater flow is essential not only for the pump’s reliability but also for efficient plant operations overall. This study applies the functional principal component analysis in characterizing the vibration signals generated under six different levels of flow rates of the pump, an environmental factor. The functional design of experiments is then applied to obtain an optimal flow rate according to the target vibration curve. The obtained flow rate is found comparable to the theoretical pump curve’s best efficiency point and is recommended to be used for optimal pump performance and reliability.