Wind power is the fastest-growing energy source and the offshore wind industry is expected to grow by over 80 GW through 2024. There is a need to reduce all associated costs to be competitive in a market that might be fully subsidy-free soon. Adapting a predictive or condition-based maintenance strategy can help in reducing O&M costs, as it accounts for more than 30% of the lifetime costs in offshore turbines.This master thesis focuses on developing a novel gearbox condition monitoring technique using raw Supervisory Control And Data Acquisition (SCADA) data from an offshore wind farm. The developed approach attempts to leverage the use of physics of failure technique and artificial neural network models for continuous monitoring of gearbox condition. Utilizing SCADA data for monitoring and fault detection can help in eliminating retrofitting additional expensive sensors such as an online oil monitoring system, vibration monitoring systems and so on for the same purpose.Firstly, a physics of failure methodology is implemented to monitor turbines on a farm level. Using Failure Mode, Effects Analysis (FMEA), the operating regimes under which gears and bearings can potentially get damaged are ascertained. A correlation between the damage operating conditions and the root cause of failure is analyzed using boxplots. In particular, a correlation between operating conditions and, the unusual and unexplained White Etching Cracks or axial cracking failure that occur frequently in the high-speed shaft bearings is studied. Additionally, this methodology is also used to complement the Artificial Neural Network model developed for anomaly detection.Finally, an Artificial Neural Network (ANN) based on a Recurrent Neural Network (RNN) is developed for fault detection. The model is trained on data when the turbine was operating normally, to predict gear bearing and gearbox lubrication oil temperatures. A stacked Gated Recurrent Unit (GRU), a type of RNN model with two hidden layers resulted in predictions with the lowest mean squared error (0.14640퐶) and with good generalization. The difference/residual between the measured and ANN predicted temperatures are used to flag anomalies. The ANN-based condition monitoring method is validated through case studies using real data from wind turbines. The results from the case studies indicate that the RNNbased fault detection method can detect a failure in the wind turbine gearbox components as early as 3-6 months before it fails.

In the Dutch coastal waters more and more wind turbines on monopile foundations are built. An important design parameter is the first natural frequency of the wind turbine. The first natural frequency is dependent on environmental conditions and different soil conditions such as scour. This research gives a better insight into the sensitivities of these influences on the first natural frequencies.
The first objective was to determine the sensitivity of different influences on the first natural frequencies in a full scale test environment. The first natural frequencies of four wind turbines (two protected against scour and two not protected against scour) in the Eneco Luchterduinen wind farm are constantly monitored and the natural frequencies are identified. It was found that only the water level has an identifiable influence on the identified first natural frequency and that the design natural frequencies and identified natural frequencies differ significantly.
After the analysis from the full scale test different influences on the first natural frequency in a numerical computer model are investigated. The computer model predictions confirm that the water level has an identifiable effect and only small effects of backfilling of the scour hole on the first natural frequency are found.
The third objective was to compare the sensitivity of different influences on the first natural frequency obtained from the computer model and from the full scale test. The effects of the water level are comparable. And the effect of backfill of the scour hole was found in the model but not in the full scale test. Therefore no conclusion is made on the effect of backfill nor on the presence of backfill. It was found that in order to mitigate the differences in first natural frequency of the test turbines and the model the soil should be modelled twenty times stiffer.
The main influence on the first natural frequency of the offshore wind turbines in the Eneco Luchterduinen wind farm is the soil stiffness. The only other identifiable influence was the water level. Also in the design of the wind turbine the soil stiffness is underestimated significantly, implying overconservative designs.