Sustainable wind power generation by online sensor oil condition monitoring against premature gearbox bearing failures

Abstract

An online web-based oil sensor diagnosis system for continuous lubricant quality monitoring in wind turbine gearboxes is presented. The fundamental innovation is that bearing operating conditions, critical for premature failures, are detectable in an early stage before any raceway damage. The electrical oil properties are identified as a sensitive loading indicator. Common condition monitoring systems, such as vibration analysis, visual or endoscopic inspection, filters and particle counters, do not respond until initial spalling. With the novel sensor unit, the specific electrical conductivity and the relative permittivity of the oil are measured in real time. A thermocouple is integrated for accompanying online temperature measurement. Under the impact of vibration loading of the rolling contact, which represents the root cause of premature bearing failures, chemical aging of the lubricant by polycondensation reactions and degradation of its additives occur. The resulting conductivity increase due to, e.g., cracked oil molecules or forming acids thus provides a crucial indication of critical operating conditions of the monitored wind turbine gearbox before component damage. The gradual consumption of additives is reflected in a reduction of the electrical permittivity of the lubricant. As the conductivity of the oil rises with temperature, a self-learning adaptive compensation program is implemented. The measuring signals can be transmitted to a web-based condition monitoring system via LAN, WLAN or serial interfaces of the sensor.