A proof-of-principle for frequency-domain vibration analysis

The economic and regulatory pressures on operating utilities to improve efficiency, availability, and reliability and to extend the effective operating life of existing electric power generating plants have created a real need for instrumentation, systems, and programs that can monitor, analyze, and provide the means to correct incipient machine and process problems. Predictive maintenance programs, properly utilized, can be one key to achieving these goals. Recent developments in microprocessor technology have provided the ability to routinely monitor the actual mechanical condition of all rotating and reciprocating machinery and process variables (i.e., pressure, temperature, flow, etc.) of other process equipment within an operating electric power generating plant. This direct correlation between frequency-domain vibration and actual mechanical condition of machinery and trending process variables of nonrotating equipment can provide the key to improving availability, reliability, and thermal efficiency and can provide the baseline information necessary for developing a realistic plan for extending the useful life of power plants. The premise of utilizing microprocessor-based predictive maintenance to improve power plant operation has been proven by a number of utilities