Anomaly detection for wind turbine condition monitoring is an active area of research within the wind energy operations and maintenance (O & M) community. In this paper three models were compared for multi-megawatt operational wind turbine SCADA data. The models used for comparison were One-Class Support Vector Machine (OCSVM), Isolation Forest (IF), and Elliptical Envelope (EE). Each of these were compared for the same fault, and tested under various different data configurations. IF and EE have not previously been used for fault detection for wind turbines, and OCSVM has not been used for SCADA data. This paper presents a novel method of condition monitoring that only requires two months of data per turbine. These months were separated by a year, the first being healthy and the second unhealthy. The number of anomalies is compared, with a greater number in the unhealthy month being considered correct. It was found that for accuracy IF and OCSVM had similar performances in both training regimes presented. OCSVM performed better for generic training, and IF performed better for specific training. Overall, IF and OCSVM had an average accuracy of 82% for all configurations considered, compared to 77% for EE.

The goal of the present paper is to achieve the diagnosis of an in-service 1.5 MW wind turbine equipped with a doubly-fed induction generator through current signature and vibration analyses. Real data from operating machines have rarely been analysed in the scientific literature through current signature analysis supported by vibrations. The wind turbine under study was originally misdiagnosed by the operator, where a healthy component was replaced and the actual failure continued progressing. The chronological evolution of both the electrical current and vibration spectra is presented to conduct an in-depth tracking of the fault. The diagnosis is achieved through spectral analysis of the stator currents, where fault frequency components related to rotor mechanical unbalance are identified. This is confirmed by the vibration analysis, which provides insightful information on the health of the drive train. These results can be implemented in condition monitoring strategies, which is of great interest to optimise operation and maintenance costs of wind farms.

Health state assessment of wind turbine components has become a vital aspect of wind farm operations in order to reduce maintenance costs. The gearbox is one of the most costly components to replace and it is usually monitored through vibration condition monitoring. This study aims to present a review of the most popular existing gear vibration diagnostic methods. Features are extracted from the vibration signals based on each method and are used as input in pattern recognition algorithms. Classification of each signal is achieved based on its health state. This is demonstrated in a case study using historic vibration data acquired from operational wind turbines. The data collection starts from a healthy operating condition and leads towards a gear failure. The results of various diagnostic algorithms are compared based on their classification accuracy.

Operations and Maintenance (O&M) can make up a significant proportion of lifetime costs associated with any wind farm, with up to 30% reported for some offshore developments. It is increasingly important for wind farm owners and operators to optimise their assets in order to reduce the levelised cost of energy (LCoE). Reducing downtime through condition-based maintenance is a promising strategy of realising these goals. This is made possible through increased monitoring and gathering of operational data. SCADA data are useful in terms of wind turbine condition monitoring. This paper aims to perform a comprehensive comparison between two types of normal behaviour modelling: full signal reconstruction (FSRC) and autoregressive models with exogenous inputs (ARX). At the same time, the effects of the training time period on model performance are explored by considering models trained with both 12 and 6 months of data. Finally, the effects of time resolution are analysed for each algorithm by considering models trained and tested with both 10 and 60 min averaged data. Two different cases of wind turbine faults are examined. In both cases, the NARX model trained with 12 months of 10 min average Supervisory Control And Data Acquisition (SCADA) data had the best training performance.

This research aims to bring together thermal modelling and machine learning approaches to improve the understanding on the operation and fault detection of a wind turbine gearbox. Recent fault detection research has focused on machine learning, black box approaches. Although it can be successful, it provides no indication of the physical behaviour. In this paper, thermal network modelling was applied to two datasets using SCADA (Supervisory Control and Data Acquisition) temperature data, with the aim of detecting a fault one month before failure. A machine learning approach was used on the same data to compare the results to thermal modelling. The results found that thermal network modelling could successfully detect a fault in many of the turbines examined and was validated by the machine learning approach for one of the datasets. For that same dataset, it was found that combining the thermal model losses and the machine learning approach by using the modelled losses as a feature in the classifier resulted in the engineered feature becoming the most important feature in the classifier. It was also found that the results from thermal modelling had a significantly greater effect on successfully classifying the health of a turbine compared to temperature data. The other dataset gave less conclusive results, suggesting that the location of the fault and the temperature sensors could impact the fault-detection ability.

Wind turbine pitch system condition monitoring is an active area of research, and this paper investigates the use of the Isolation Forest Machine Learning model and Supervisory Control and Data Acquisition system data for this task. This paper examines two case studies, turbines with hydraulic or electric pitch systems, and uses an Isolation Forest to predict failure ahead of time. This novel technique compared several models per turbine, each trained on a different number of months of data. An anomaly proportion for three different time-series window lengths was compared, to observe trends and peaks before failure. The two cases were compared, and it was found that this technique could detect abnormal activity roughly 12 to 18 months before failure for both the hydraulic and electric pitch systems for all unhealthy turbines, and a trend upwards in anomalies could be found in the immediate run up to failure. These peaks in anomalous behaviour could indicate a future failure and this would allow for on-site maintenance to be scheduled. Therefore, this method could improve scheduling planned maintenance activity for pitch systems, regardless of the pitch system employed.