Improve Turbine Reliability Research Articles

Automatizacija praćenja stanja reduktora vetroturbina - uporedna studija tehnike mašinskog učenja zasnovane na analizi vibracija

Wind turbines play a role in the adoption of renewable energy production, but they are susceptible to shutdowns that require thorough monitoring. Gearbox failures are an issue leading to maintenance and operational downtime. This study investigates the application of machine learning methods to enhance the diagnosis of gearbox problems using vibration analysis. Through the application of fault scenarios that impact bearings and gears, the researchers successfully extracted time domain features from vibration data of a 750 kW turbine testbed in order to detect indications of damage. Support Vector Machine (SVM), Naive Bayes, and K Nearest Neighbour (KNN) machine learning models were used to classify gearbox faults. Among these models, Naive Bayes achieved an accuracy rate of 95.7%, which exceeded the established benchmarks. The probabilistic approach was able to successfully associate symptom characteristics with fault patterns. Intelligent monitoring systems could improve maintenance efficiency. This data-driven approach highlights the potential of machine learning in supporting wind power development by eliminating gearbox inefficiencies and improving turbine reliability, and further research is being conducted to ensure that this approach works in concert with diversity and in the real world. This shows how machine learning is contributing to advances in renewable energy by helping to analyze predictive problems and prevent costly gearbox failures.

Anomaly detection and fault analysis of wind turbine components based on deep learning network

Abstract Continuous monitoring of wind turbine health using early fault detection methods can improve turbine reliability and reduce maintenance costs before they reach a catastrophic stage. To achieve anomaly detection and fault analysis of wind turbine components, this paper proposes a deep learning method based on a deep auto-encoder (DAE) network using operational supervisory control and data acquisition (SCADA) data of wind turbines. First, a component DAE network model using multiple restricted Boltzmann machines (RBM) was constructed. Previously collected normal SCADA data from wind turbines were used to train this multilayer network model layer-wise to extract the relationships between SCADA variables. Then, a reconstruction error (Re) was calculated by using the DAE network input and its output reconstruction value, which was defined as the condition detection index to reflect the component health condition. Due to the acute changes and disturbances of wind speed in actual operation, the calculated detection index always has an extreme distribution that can cause false alarms. Therefore, an adaptive threshold determined by the extreme value theory was proposed and used as the rule of anomaly judgement. The method can not only implement early warning of fault components but also deduce the physical location of a faulted component by DAE residuals. Finally, the effectiveness of the proposed method was verified by some reported failure cases of wind turbine components.

Investigation of using multi-shockwave system instead of single normal shock for improving radial inflow turbine reliability

A variable nozzle turbine (VNT) has become an important technology for diesel engine applications. During the development and design process, one of the most important tasks is to minimize the risk of high cycle fatigue (HCF) failure. When a VNT operates with high inlet pressure and small nozzle opening, a shock wave may appear near nozzle vane trailing edge. The shock wave introduces strong excitation force on the downstream turbine blades and has a potential to damage the turbine wheel. To improve turbine reliability, a method of generating multi-shockwave system by using grooved surface nozzle was investigated. It was found that, in the multi-shockwave system, the intensity of the shock wave is weakened in comparison to original single normal shock wave, and its impact on the turbine wheel is therefore lowered. However, with the use of the grooved surface, penalty on the turbine efficiency is numerically found, but very small at the research point.

Effects of turbulence on tidal turbines: Implications to performance, blade loads, and condition monitoring

Laboratory scale testing of tidal turbines has generated valuable datasets to support optimised turbine design and numerical model validation. However, tidal sites are highly turbulent with a broad range of length scales and turbulence intensities that are site specific. In this work we describe an experimental campaign using static grids to generate turbulence and investigate its impact on a model tidal turbine in a circulating water flume. Length scales, energy spectra and turbulence dissipation rates are first considered for centre point measurements before full flow characterisation of the ambient conditions across the turbine rotor area. Six different cases were chosen to observe the performance of a 1/20th scale 0.8m diameter turbine subjected to these flows. The rotor thrust and torque, and flapwise and edgewise blade root bending moments were measured. It was found that the thrust and power coefficients were sensitive to the estimate of ambient velocity. In the most extreme case the Betz limit could be ‘exceeded’ depending on which estimate of ambient velocity was used. Overall variations in the peak power coefficient of over 10% were observed, demonstrating the significance turbulence has on turbine performance. It was also found that there is a strong correlation between fluctuations in blade root bending moments and the rotor loads. As a result we proposed that fatigue loads acting on the blades may be estimated from the fluctuations in power output of the turbine. Therefore maintenance operations maybe optimised from real-time fatigue monitoring of blade loads without the need to install additional instrumentation on the turbine blades. Under this proposed regime the cost of energy will be reduced due to reductions in turbine costs and following optimisation of the maintenance requirements and operational costs. This could also improve turbine reliability which would have significant implications for large multi turbine arrays.

Electrical Discharge and Its Impact on Drivetrains of Wind Turbines

Defects in drivetrains of wind turbines are typically very costly due to the difficulty in accessing the defective components for repair and maintenance. Electrical discharge has been identified as one of the key root causes to defects in wind drivetrains. Improving the reliability of key wind turbine components has been attracting a great research interest since the launch of the first wind turbine. In this paper, a study to understand the electrical discharge and its impact on wind drivetrains is presented. The root cause of electrical discharge in wind turbines is discussed. A field test is carried out on 1.5-MW doubly fed induction generator wind turbines in practical environments. Suggestions are provided on the design of the wind turbine grounding and protection system to improve turbine reliability and reduce defect rates due to electrical discharge.

Wind Turbines Condition Monitoring and Fault Identification using Acoustic Emission Testing Method

Acoustic Emission (AE) technique can successfully applied for condition monitoring of low speed rotating components such as rolling bearing and gearbox of Wind Turbines. This technique is able to detect very small energy released rates from incipient defect in a very early stage. Wide range of signal processing methods can be apply for diagnosing faults and fatigues in AE spectrums and the changes in wave forms are very significant to recognize the failures. Condition monitoring and Fault identification (CMFI) of wind turbine health using automated failure detection algorithms can improve turbine reliability. AE testing is based condition monitoring system uses data already collected at the wind turbine controller. It is an effective way to monitor wind turbines for early warning of failures and performance issues. CMFI methods are classified into modelbased and signal-based methods. They can be implemented with or without the use of artificial intelligence. The object of this thesis is to design a model-based CMFI scheme for WTs, which can be used under normal operation conditions.