Wind Turbine Maintenance Costs Research Articles

Wind speed prediction

With the help of wind farms, wind energy is a vital renewable energy source that contributes significantly to the worlds energy balance. The lifespan and maintenance costs of wind turbines will be reduced with an accurate wind speed prediction. On the other hand, wind speed is highly volatile and unpredictable. Thus, it is essential to do research into creating complex models and algorithms for precise wind speed prediction. So far, some of the most promising models include Support Vector Machine (SVM), Artificial Neural Networks (ANN), and Autoregressive Moving Average (ARMA). Python, as an advanced and versatile programming language, is exceptionally suited for scripting the algorithms of these sophisticated models. This paper will use the data from Austin Texas and apply a Support Vector Machine (SVM) for wind speed prediction involves several stages, including data collection, data preprocessing, model selection, model training, parameter optimization, model validation, and prediction. Wind energy resource optimisation, maintenance cost reduction, and total wind farm efficiency can all be significantly improved by incorporating these models into predictive analytics and continuously improving them against changing data.

Fault diagnosis for wind turbines based on LSTM and feature optimization strategies

SummaryHigh penetration of renewable energy is the development trend of the future power system. As one of the clean energy sources, wind power generation has an increasing share in the energy market. However, due to the harsh working environment, the high fault rate and poor accessibility of the wind farms, resulting in the difficult maintenance process and high cost. This article proposes a fault diagnosis (FD) method based on long short‐term memory (LSTM) and feature optimization strategies for wind turbines (WTs), thus reducing the operation and maintenance costs of WTs. First, Pearson correlation coefficient analysis is performed on the collected data features to remove redundant features, and wavelet transform is adopted to remove the redundant data, so as to optimize the fault features and fault data. Then the selected features samples are used to train LSTM‐based FD model. Finally, the actual production data is adopted to verify the proposed method. The proposed method can effectively locate the faults, and provide data support for wind farms, thus improving the reliability, safety, and economic benefits of wind farms.

Pattern mining based data fusion for wind turbine condition monitoring

The profitability of wind turbine energy production is for an important part determined by the operation and maintenance costs of wind turbines. An important driver of these costs is currently the premature failure of components due to excessive wear. If it would be possible to accurately predict these failures, preventive maintenance can be made more effective, which should result in less downtime and expensive unexpected failures. This in turn should lower the operational and maintenance costs. The research presented here is a contribution to the research on condition monitoring and failure prediction for wind turbines. To this end, a methodology for failure prediction is designed that combines (fuses) multiple information sources (i.e. SCADA and status log data). The novelty of this research lies in the fact that pattern mining techniques are used to identify relevant rules for a rule-based failure classifier. The methodology is validated on generator bearing failure cases from a real operational wind farm. The results show that the methodology is able to predict generator bearing failures accurately well in advance. The rules on which the predictions are based are interpretable and correspond in general to expert knowledge on the matter.

Effects of Wind Conditions on Wind Turbine Temperature Monitoring and Solution Based on Wind Condition Clustering and IGA-ELM.

To reduce maintenance costs of wind turbines (WTs), WT health monitoring has attracted wide attention, and different methods have been proposed. However, most existing WT temperature monitoring methods ignore the fact that various wind conditions can directly affect internal temperature of WT, such as main bearing temperature. This paper analyzes the effects of wind conditions on WT temperature monitoring. To reduce these effects, this paper also proposes a novel WT temperature monitoring solution. Compared with existing solutions, the proposed solution has two advantages: (1) wind condition clustering (WCC) is applied and then a normal turbine behavior model is built for each wind condition; (2) extreme learning machine (ELM) is optimized by an improved genetic algorithm (IGA) to avoid local minimum due to the irregularity of wind condition change and the randomness of initial coefficients. Cases of real SCADA data validate the effectiveness and advantages of the proposed solution.

Integrated Structural Dependence and Stochastic Dependence for Opportunistic Maintenance of Wind Turbines by Considering Carbon Emissions

Wind turbines have a wide range of applications as the main equipment for wind-power generation because of the rapid development of technology. It is very important to select a reasonable maintenance strategy to reduce the operation and maintenance costs of wind turbines. Traditional maintenance does not consider the environmental benefits. Thus, for the maintenance problems of wind turbines, an opportunistic maintenance strategy that considers structural correlations, random correlations, and carbon emissions is proposed. First, a Weibull distribution is used to describe the deterioration trend of wind turbine subsystems. The failure rates and reliability of wind turbines are described by the random correlations among all subsystems. Meanwhile, two improvement factors are introduced into the failure rate and carbon emission model to describe imperfect maintenance, including the working-age fallback factor and the failure rate increasing factor. Then, the total expected maintenance cost can be described as the objective function for the proposed opportunistic maintenance model, including the maintenance preparation cost, maintenance adjustment cost, shutdown loss cost, and operation cost. The maintenance preparation cost is related to the economic correlation, and the maintenance adjustment cost is described by using the maintenance probabilities under different maintenance activities. The shutdown loss cost is obtained by considering the structural correlation, and the operation cost is related to the energy consumption of wind turbines. Finally, a case study is provided to analyze the performance of the proposed model. The obtained optimal opportunistic maintenance duration can be used to interpret the structural correlation coefficient, random correlation coefficient, and sensitivity of carbon emissions. Compared with preventive maintenance, the proposed model provides better performance for the maintenance problems of wind turbines and can obtain relatively good solutions in a short computation time.

Condition Monitoring in a Wind Turbine Planetary Gearbox Using Sensor Fusion and Convolutional Neural Network

To meet the requirements of reducing operations and maintenance costs of wind turbines, we present a machine learning framework for early damage detection in gearboxes based on the cyclostationary analysis of sensor data. The application focus is the condition monitoring of wind turbine gearboxes under varying load scenarios, in particular turbulent wind conditions. Faults in the rotating components in the gearbox can leave their signature in the vibrations that can be measured by accelerometers. We analyze data stemming from a simulated vibration response of a 5MW multibody wind turbine model in a healthy and damaged scenarios and under different wind conditions. With cyclostationary analysis applied on acquired sensor data, we generate cyclic spectral coherence maps that highlight signatures related to the fault damage and enable its identification. These maps are used to train convolutional neural networks that then identify faults, including those of small magnitude, in test data with a high accuracy. Benchmark test cases inspired by an NREL study are tested and faults successfully detected.

Condition monitoring of wind turbines using machine learning based anomaly detection and statistical techniques for the extraction of 'healthy data'

Premature failures caused by excessive wear are responsible for a large fraction of the maintenance costs of wind turbines. Therefore, it is crucial to be able to identify the propagation of these failures as early as possible. To this end, a novel condition monitoring method is proposed that uses statistical data analysis techniques and machine learning to construct a multivariate anomaly detection framework, based on high-frequency temperature SCADA data from wind turbines. This framework contains several steps. First, there is a preprocessing step in which relevant wind turbine states are extracted from the data. These states are the operating conditions and whether or not the turbine exhibits transient behavior. The second step entails anomaly detection on the temperature time series data. Fleet information is used to filter out exogenous (environmental) factors. Furthermore, multiple models are combined to get more stable and robust anomaly detections. By combining them the weaknesses of the individual models are alleviated resulting in a better overall performance. A limitation of machine learning-based anomaly detection on temperature data is the requirement that at least one year of “healthy” (meaning without anomalies) training data is available to account for seasonal effects. The lack of verified “healthy” data spread out evenly over the seasons generally means that the anomaly detection accuracy is severely compromised for the unrepresented seasons. This research uses smart retraining to reduce this limitation. Statistical techniques that leverage the information of the fleet are used to extract “healthy” data from at least one year, but preferably multiple years, of unverified data. This can then be used as training data for the machine learning-based models. To validate the pipeline, temperature and failure data of a real operational wind farm is used. Although the methodology is general in its scope, the validation case focusses specifically on generator bearing failures.

An opportunistic maintenance strategy for wind turbines

Maintenance strategies aim at reduction of the operation and maintenance cost of wind turbines. An opportunistic maintenance strategy offers cost advantages over the traditional preventive replacement maintenance strategy by considering cost dependency among maintenance activities. A new opportunistic maintenance strategy of wind turbines is proposed here. First, the proposed strategy includes two sets of reliability thresholds to handle subassemblies those are in operational and failed wind turbines. Using the reliability thresholds, ten maintenance modes and their service conditions are presented. Second, the Weibull distribution with an age reduction factor is introduced to describe reliability variation of subassemblies under imperfect maintenance. Third, the maintenance cost in each maintenance mode is expressed by a reliability variation-based function that is dependent on the stage of maintenance activities. Fourth, due to the randomness of failures, the expected total maintenance cost is determined with Monte Carlo simulation. The proposed approach has been validated with the data provided by a wind farm located in Northern China.

Deep domain adversarial residual neural network for sustainable wind turbine cyber‐physical system fault diagnosis

AbstractAs a popular renewable energy generation technology, wind turbine system has become a critical enabler for building the sustainable cyber‐physical system (CPS). The main shaft bearing is an important part of the wind turbine CPS and often runs under variable working conditions. Thus, the reliable bearing diagnosis method can timely discover the main shaft bearing fault, which reduces the maintenance cost of wind turbines. Inspired by the idea of domain adaptation, we combined domain adversarial neural network and residual network and proposed a novel deep domain adversarial residual neural network (DDA‐RNN) for diagnosing bearing fault and improving model performance on the unlabeled dataset. This proposed software and hardware co‐design method was evaluated by our bearing dataset, which was collected from two wind turbine CPSs from Sanmenxia in Henan Province. Besides, F1 score and accuracy are served as model metrics, which reflect the diagnosis performance. Compared with other methods, the experimental results show that DDA‐RNN can improve model performance. Meanwhile, DDA‐RNN extracts diagnosis knowledge from labeled dataset and improves the model performance on the unlabeled dataset under different working condition. Therefore, the proposed method can be potentially used to benefit many practical scenarios in the future.

A New Acoustic Emission Sensor Based Gear Fault Detection Approach

In order to reduce wind energy costs, prognostics and health management (PHM) of wind turbine is needed to reduce operations and maintenance cost of wind turbines. The major cost on wind turbine repairs is due to gearbox failure. Therefore, developing effective gearbox fault detection tools is important in the PHM of wind turbine. PHM system allows less costly maintenance because it can inform operators of needed repairs before a fault causes collateral damage happens to the gearbox. In this paper, a new acoustic emission (AE) sensor based gear fault detection approach is presented. This approach combines a heterodyne based frequency reduction technique with time synchronous average (TSA) and spectral kurtosis (SK) toprocess AE sensor signals and extract features as condition indictors for gear fault detection. Heterodyne techniques commonly used in communication are used to preprocess the AE signals before sampling. By heterodyning, the AE signal frequency is down shifted from MHz to below 50 kHz. This reduced AE signal sampling rate is comparable to that of vibration signals. The presented approach is validated using seeded gear tooth crack fault tests on a notational split torque gearbox. The approach presented in this paper is physics based and the validation results have showed that it could effectively detect the gear faults.

Optimal dynamic imperfect preventive maintenance of wind turbines based on general renewal processes

With the rapid growth in wind turbine technology worldwide, the high operational and maintenance costs of wind turbines have posed a major challenge to wind power operating companies. Considering the high replacement cost, imperfect maintenance measures are employed widely once downtime failures occur. However, as the metric to describe the effect of imperfect maintenance is a non-intuitive variable, the evaluation results obtained from existing research do not conform with actual wind turbine situations. To address this issue, the virtual age factor and failure intensity update factor are expressed by intuitive variables to illustrate the imperfect maintenance effect. Additionally, we propose a failure rate function update model considering the above factors. To minimise maintenance costs while ensuring the availability of the wind turbine, we investigate a periodic dynamic imperfect preventive maintenance decision model based on the proposed failure rate function update model. We also provide a brief illustration of the accuracy and feasibility of the proposed model through optimal solution and sensitivity analyses. The results obtained from the case analysis and strategies comparison, based on actual wind turbine maintenance data, demonstrate the economic advantages of our approach.

Diagnostic Framework for Wind Turbine Gearboxes Using Machine Learning

Operation and maintenance costs of wind turbines are highlydriven by gearbox failures, especially offshore were the logisticsof replacements are more demanding. It is therefore verycritical to foresee incipient gearbox faults before they becomecatastrophic failures. Wind turbine gearbox condition monitoringis usually performed using vibration signals comingfrom accelerometers installed on the gearbox surface. Thecurrent monitoring practice is a rule-based approach, wherealarms are activated based on thresholds. However, too muchmanual analysis may be required for some failure modes andthis can become quite challenging as the installed wind capacitygrows. Also, since false alarms have to be avoided,these thresholds are set quite high, resulting in late stage diagnosisof components. Given the fact there is a large amountof historic operating data with confirmed gearbox failure incidents,this paper proposes a framework that uses a machinelearning approach. Vibration signals are used from the gearboxsensors and processed in the frequency domain. Featuresare extracted from the processed signals based on the fault locationsand failure modes, using domain knowledge. Thesefeatures are used as inputs in a layer of pattern recognitionmodels that can determine a potential component fault locationand failure mode. The proposed framework is illustratedusing failure examples from operating offshore wind turbines.

Fault Diagnosis of Wind Turbine Gearboxes Based on DFIG Stator Current Envelope Analysis

The drivetrains of many utility-scale wind turbines have a gearbox connected with a doubly fed induction generator (DFIG). Since gearbox failure is a major contributor to the high maintenance cost of wind turbines faced by the wind industry, it is important to perform fault diagnosis for gearboxes. Among different gearbox fault diagnosis methods, those using current signals collected from generator terminals have shown their merits in terms of complexity, cost, and reliability. In this paper, a new method that uses a DFIG stator current signal for the fault diagnosis of wind turbine drivetrain gearbox in nonstationary conditions is proposed. In the proposed method, the dc offset and high frequency noise of the current signal are first eliminated. Then, the envelope of the current signal is obtained by using the Hilbert transform. The current envelope signal only contains nonstationary frequencies that are proportional to the DFIG shaft rotating frequency. Next, a synchronous resampling algorithm is designed to convert the current envelope signal with a constant time interval to a resampled current envelope signal with a constant phase angle interval. Finally, the power spectral density analysis is used to obtain the frequency spectrum of the resampled current envelope signal from which the constant characteristic frequencies can be easily identified for the purpose of fault diagnosis. Laboratory test data collected from a 1/3 hp DFIG wind turbine drivetrain test rig and field data obtained from three 1.6-MW DFIG wind turbines are used to show the superiority of the proposed method.

Improved criticality analysis method of equipment failures in wind turbines

High failure rate and maintenance cost of wind turbines are one of the key issues in wind energy. In view of this, reliability centred maintenance (RCM) has been introduced to distinctively treat equipment failures according to their criticality. Criticality analysis (CA) is a pivotal step of RCM, derived from the failure frequency and failure modes effects of equipment, which determines when and how to maintain the wind turbines. However, in traditional CA, failure data is roughly considered, e.g. failure frequency is categorised as different ranks instead of accurate numerical value. In addition, the method of calculating criticality by multiplying variables is sensitive to variable ranges and values, which is unable to distinctly distinguish the contributions of different variables. Aimed at these deficiencies, an improved CA method with expansibility is proposed to assess the criticality of equipment failures in wind turbines based on Euclidean distance of failure vectors. To assess disparities among criticality ranks from diverse CA methods, an inverse number based method is presented. Several existing CA methods and the proposed CA in this study are compared using the data of wind turbines in literature, and the results verify the effectiveness of the proposed method.

Detection and identification of windmill bearing faults using a one-class support vector machine (SVM)

The maintenance cost of wind turbines needs to be minimized in order to keep their competitiveness and, therefore, effective maintenance strategies are important. The remote location of wind farms has led to an opportunistic maintenance strategy where maintenance actions are postponed until they can be handled simultaneously, once the optimal opportunity has arrived. For this reason, early fault detection and identification are important, but should not lead to a situation where false alarms occur on a regular basis. The goal of the study presented in this paper was to detect and identify wind turbine bearing faults by using fault-specific features extracted from vibration signals. Automatic identification was achieved by training models by using these features as an input for a one-class support vector machine. Detection models with different sensitivity were trained in parallel by changing the model tuning parameters. Efforts were also made to find a procedure for selecting the model tuning parameters by first defining the criticality of the system and using it when estimating how accurate the detection model should be. Method was able to detect the fault earlier than using traditional methods without any false alarms. Optimal combination of features and model tuning parameters was not achieved, which could identify the fault location without using any additional techniques.

Assessing the Factors Impacting on the Reliability of Wind Turbines via Survival Analysis—A Case Study

The failure of wind turbines is a multi-faceted problem and its monetary impact is often unpredictable. In this study, we present a novel application of survival analysis on wind turbine reliability, including accounting for previous failures and the history of scheduled maintenance. We investigated the operational, climatic and geographical factors that affect wind turbine failure and modeled the risk rate of wind turbine failure based on data from 109 turbines in Germany operating for a period of 19 years. Our analysis showed that adequately scheduled maintenance can increase the survival of wind turbine systems and electric subsystems up to 2.8 and 3.8 times, respectively, compared to the systems without scheduled maintenance. Geared-drive wind turbines and their electrical systems were observed to have 1.2- and 1.4- times higher survival, respectively, compared to direct-drive turbines and their electrical systems. It was also found that the survival of frequently-failing wind turbine components, such as switches, was worse in geared-drive than in direct-drive wind turbines. We show that survival analysis is a useful tool to guide the reduction of the operating and maintenance costs of wind turbines.

A stall-regulated wind turbine design to reduce fatigue

Variable-speed stall-regulated (VS-SR) wind turbines can be designed to produce power as efficiently as variable-speed pitch-controlled (VS-PC) systems. However, amongst the main drawbacks of VS-SR systems high transient power and low predictability have been the primary factors in favour of adopting VS-PC system for multi-MW wind turbines. Cyclic and stochastic loads leading to fatigue failure is one of the prime considerations for large wind turbines. In contrast to the current trend of research, which is focused on load alleviation by integrating active flow controllers, this paper highlights the potential benefits of VS-SR wind turbines in reducing fatigue loads. Adopting the NREL 5 MW wind turbine as the baseline, blades are redesigned for stall-regulation. It is shown that a well-designed VS-SR wind turbine experiences significantly less fatigue loads compared to VS-PC systems. It also results in low power transients near and above rated wind speed. Taking into account added complexity, mass and maintenance costs of wind turbines utilising active flow controllers and in view of the recent progresses that have been made regarding the aeroelastic stability of stalled blades, VS-SR systems seem to have a role to play in the design of future wind turbines.

An approach combining data mining and control charts-based model for fault detection in wind turbines

Wind energy is growing to be one of main sources of renewable energy. As the operational and maintenance costs of wind turbines are adversely affected by the occurrence of faults, the early detection of potential faults can help reduce such costs. In this study, we propose a method for detecting potential faults sooner and identifying the probable variables contributing to the faults over a certain period as well as at a specific time. The proposed method uses data mining techniques to select the more important variables from the supervisory control and data acquisition (SCADA) systems of the turbine to improve the prediction accuracy and employs an exponentially weighted moving average (EWMA) model-based control chart to implement the residual approach, in order to remove the autocorrelation in the data. Both EWMA and multivariate EWMA (MEWMA) control charts are constructed so that their detection capabilities as well as the types of errors generated can be compared. We evaluated the proposed method by using both the SCADA data and the alarm log of a turbine. It was observed that the MEWMA chart is more suitable than the EWMA chart for the early detection and avoidance of errors.

For wind turbines in complex terrain, the devil is in the detail

The cost of energy produced by onshore wind turbines is among the lowest available; however, onshore wind turbines are often positioned in a complex terrain, where the wind resources and wind conditions are quite uncertain due to the surrounding topography and/or vegetation. In this study, we use a scale model in a three-dimensional wind-testing chamber to show how minor changes in the terrain can result in significant differences in the flow at turbine height. These differences affect not only the power performance but also the life-time and maintenance costs of wind turbines, and hence, the economy and feasibility of wind turbine projects. We find that the mean wind, wind shear and turbulence level are extremely sensitive to the exact details of the terrain: a small modification of the edge of our scale model, results in a reduction of the estimated annual energy production by at least 50% and an increase in the turbulence level by a factor of five in the worst-case scenario with the most unfavorable wind direction. Wind farm developers should be aware that near escarpments destructive flows can occur and their extent is uncertain thus warranting on-site field measurements.

Fault‐tolerant predictive power control of a DFIG for wind energy applications

This study presents a new fault-tolerant predictive power control strategy for doubly-fed induction generators (DFIGs) used in wind energy applications, with the rotor fed by a three-level neutral point clamped converter. The proposed control strategy is able to maintain the system in operation with a good performance after the occurrence of either open-circuit (OC) or short-circuit faults in the insulated-gate bipolar transistors (IGBTs) of the rotor-side converter (RSC), thus reducing the downtime and maintenance costs of wind turbines. The fault-tolerant strategy takes advantage of the discrete nature and flexibility of finite control set model predictive control strategies and restricts the possible switching states of the power converter according to the type of fault and its location. A diagnosis method for IGBT OC faults in the RSC, based on voltage errors, is also proposed for the considered system. This method uses the estimated rotor voltages from the DFIG model, thus avoiding the use of any extra sensors. Accurate and fast detection of OC faults in both interior and exterior IGBTs is achieved throughout the entire DFIG operational range. The effectiveness of the fault diagnosis method and fault-tolerant control strategy is validated with several experimental results.