Determining Prototypes and Signs for Automated Diagnostics of Steam Turbine Equipment

Abstract

Nowadays, turbine building plants try to equip their turbines with automatic control systems that include, in addition to their basic functions, automated fault-diagnosing procedures. The article presents the authors’ experience gained from developing diagnostic systems and evaluating state parameters both for individual components of power equipment and for a power unit and thermal power plant as a whole; in addition, the specific features pertinent to the development of such automated systems are described. Such systems can be used for the purposes of operative and postoperative diagnostics. The first group of applications includes, e.g., vibration diagnostics of turbines, and the second group may include assessing the state of the thermal expansion system components, control system, and turbine plant auxiliary equipment; determining the technical and economic indicators; etc. Development of a prototype is the basis of any diagnostic task. In comparing the equipment’s actual state with the state of the prototype, it is necessary to determine the values of the state parameters from which the diagnostic tasks are solved, i.e., the faults are detected. Diagnostic prototypes' development methods are identified: normative, statistical (expert), physical, and digital. These methods can be combined within a single diagnostic algorithm, due to which it becomes possible to achieve more reliable and efficient (less expensive) assessment of equipment malfunctions during its automatic diagnostics. The advantages and drawbacks of each method are described, and examples of implementing diagnostic tasks with different prototypes are given. Matters concerned with revealing a fault, evaluating the state of equipment components and residual life, and predicting the state parameters are solved. It is shown that one of the complex problems that has to be solved in developing a vibration diagnostic system is to substantiate the relationship between the sign and type of equipment malfunction. Using vibration diagnostics as an example, fault symptoms are grouped into boundary, factor, and correlation ones.