Turbine Load Control Instability Fault and Its Diagnosis Method with Big Data Fusion Model

Abstract

Turbine load control instability fault has a great impact on the thermal power unit's primary and secondary frequency modulation performance. A new type of load instability fault was found by checking the actual operation condition of a 660 MW supercritical steam turbine, that is, the actual load rejection amount of the regulating valve is as high as 200 MW under the condition of small action. Through comprehensive analysis, the physical mechanism of the actual fault is found: One of the high-pressure steam control valve servo cards had a problem, which caused the valve to close completely due to abnormal voltage. Feedback monitoring of the valve showed that it was in normal condition, but the valve was found to be completely closed on site. Based on the above fault mechanism, this paper establishes a fault diagnosis model, and realizes the effective identification of such faults based on the fusion of actual running big data. Finally, an effective solution for this fault is proposed, which improves the fast and accurate load-changing capacity of high-power steam turbines, and has certain reference significance for similar units.