Development of a self-contained device for rapid detection of volatile impurities in the oil system of a turbine

The research is devoted to development of a self-contained device for rapid detection of volatile impurities in the oil system of a turbine and testing it using the operating equipment. The device consists of a remote sensor, whose sensitive element is a 3–5-mm long wire probe 20 microns in diameter, and a measurement unit that comprises a microcontroller with a set of peripherals. The design of the device enables automation of the measurement procedure with a minimum number of preset settings and real-time output of information to the operator console. The software of the device provides two-stage pulse heating of the wire probe and a resistance temperature detector. The two-stage mode proves to be the most sensitive to appearance in the system of moisture, including its trace amounts. The characteristic time of the heating is of the order of 10 ms. The measurement procedure is based on a method that consists in automatic search for spontaneous boiling-up temperature of the oil accompanied by a characteristic response signal. The results were interpreted by formal correlation of the measured values with an array of calibration data obtained in similar experiments with well-defined oil samples. An experimental method for application of the device has been developed that takes into account technological factors, such as comparatively high values of the flow rate and the temperature of the oil in locations of the oil drain from bearings, the variability of these values, and the variety of noise types that accompany the operation of the thermal power equipment that complicate the online measurements. Tests of the device were carried out in locations of oil drain from supporting bearings. The test results have demonstrated the possibility of applying the device directly in the oil system of a turbine and provided a practical basis for development of a system of multipoint control of the technological scheme in real time.