Wind energy is one of the rapidly evolving renewable energy resources. It is particularly essential to built dependability and accessibility of wind turbines and additional to decrease the wind energy cost. Wind turbine blades are ought to be an important component among the other basic segments in the wind turbine framework since they transform the dynamic energy of wind into useable power. Wind turbine blades are manufactured from either carbon fiber reinforced polymer or glass fiber reinforced polymer. Damages and flaws are unavoidable either in the manufacturing process or the lifetime of a composite blade. Hence, structural health monitoring for wind turbine blade is essential to avoid failures and extend dependability in both fabrication quality control and in-service investigation. In this study, a three-bladed variable wind turbine was chosen and using histogram features, the condition of a wind turbine blade is inspected. The faults like hub-blade loose connection, blade crack, pitch angle twist, erosion and blade bend faults were studied and these faults are classified using various data mining algorithms. The main contribution of this study is to build and suggest a data-model for fault identification on wind turbine blade while in operation using machine learning classifiers like sequential minimal optimization (SMO) algorithm, simple logistic algorithm (SLA), multilayer perceptron (MLP), logistic algorithm (LA) and radial basis function (RBF). © School of Engineering, Taylor’s University.