Lubricant Performance in Wind Turbines: A Data Study in Real-Use Conditions

An extensive statistical study was conducted on a 10-year dataset (2014–2023) containing the lubrication condition monitoring results from wind turbine gearboxes in the Iberian Peninsula. The dataset includes two mineral and two synthetic lubricants; all four were sampled and analyzed regularly in accordance with the ISO 14830-1 standard. This dataset comprises over 25,000 records across 24 distinct parameters, as defined by the standard maintenance procedure of lubricants in wind turbine gearboxes. To reduce dimensionality, the analysis begins with principal component analysis, followed by discussion of Spearman correlations and finishing with comparation of trends. Performance differences, correlations and typical trends will be evaluated and compared for mineral and synthetic lubricants. This data-driven study presents a significant contribution to supporting maintenance decision-making processes in wind farms. Despite the differences, P, Pb and K have been identified as having a major influence on the variance in the dataset for all the lubricants, although no significant correlations with these elements have been found. Mineral lubricants showed very few correlations between elements and lubricant parameters, with Fe as the major element to be considered. Meanwhile, synthetic lubricants showed several correlations between elements (Fe, Mn, Cu, Ba) and lubricant parameters denoting complex tribochemical reactions.