Analytical system for predicting cracks in hydraulic turbines

Today, many hydropower plants face serious problems associated with the sudden appearance of dangerous cracks in key components of hydraulic turbines. This is especially true for large hydraulic units whose design lifetime has already expired. In recent years, hydraulic units have been progressively used under off-design modes of operation, which causes increased loads and stresses in their components, accelerates the appearance and growth of defects, stimulates premature failures, and ultimately can lead to serious accidents with large losses. The currently used diagnostic systems for hydraulic turbines often do not allow tracking the appearance of fatigue cracks in runner blades, guide vanes, shafts, etc. The strong individuality of hydraulic units also excludes the use of any statistical methods for determining the time of cracks initiation. The predictive analytics system is proposed as an additional tool for predicting the appearance and growth of dangerous operational cracks. The system implements an analytical algorithm based on the evaluation of fatigue strength of hydraulic turbine components under frequently changing operating conditions and allows summing up damage from various external loads and in various time ranges. The initial information for this predictive system is the information about the actual operating time at each operating mode and the expected regime parameters for the upcoming period. The output data is information about the exhausted and residual lifetime. All individual features are taken into account by an individual digital model of the matrix type as a combination of equipment responses to external influences. The proposed predictive analysis system not only reduces the frequency of accidents and unplanned shutdowns but is also a way of smart maintenance for hydraulic turbines by developing the most efficient, reasonable, maximum timesaving equipment use strategy.