Dans le cadre de la maintenance conditionnelle, l'analyse vibratoire temporelle repose sur 50 ans d'expérience et jouit, de fait, d'une représentativité importante dans le secteur éolien. Cette analyse, appliquée sous des conditions de fonctionnement non stationnaires, fait néanmoins preuve d'un manque de précision dans la détection de défauts mécaniques. Afin de synchroniser l'échantillonnage avec les défauts des éléments tournants de la ligne d'arbre, il est alors nécessaire de s'appuyer sur une discrétisation angulaire du vecteur d'information. L'objet de cette étude est le développement d'un outil de surveillance disposant de cette qualité et s'appuyant sur les variations que présente la vitesse instantanée de la machine tournante pour évaluer l'état des éléments tournants. La vitesse angulaire instantanée, associée à l'analyse spectrale, est assujettie à des perturbations dont les origines sont identifiées et les influences quantifiées. Ainsi, il est envisageable de dimensionner un système de mesure en fonction des caractéristiques de la chaîne cinématique sous surveillance. Ce document propose également plusieurs outils de traitement originaux qui permettent, sous des conditions de fonctionnement non stationnaires, d'améliorer suffisamment l'observation pour que l'utilisation d'un seul capteur soit envisageable pour assurer la surveillance de la ligne d'arbre dans sa totalité. Ces outils mettent en exergue la dualité temporelle angulaire de la surveillance de machines tournantes et tirent parti des conditions de fonctionnement non stationnaires par la connaissance de la vitesse instantanée. Les procédés d'acquisition combinés aux outils de traitement sont mis à l'épreuve sur plusieurs banc d'essais. Sensible au degrés de désalignement, au type de défaut d'engrènement et à la gravité d'un écaillage, la vitesse angulaire instantanée est dorénavant une alternative confirmée à la surveillance vibratoire, même angulaire. Une vaste campagne de mesures a été lancée en parallèle et en continu sur une éolienne dans l'objectif ambitieux de mettre en place un procédé de pronostic. Les diverses conditions de fonctionnement sont prises en compte par une approche statistique permettant d'utiliser avantageusement leurs influences sur les mesures qui en sont issues. Ces travaux constituent une première étape dans la réalisation d'un modèle de surveillance multi-niveau adapté aux machines à fonctionnement variable dont les éoliennes sont un parfait exemple de complexité. / The overall background for the thesis is the need to develop new methods for monitoring machines subject to nonstationary operating conditions of speed and load, as typified by wind turbines. This has become particularly important because of the increasing prevalence of wind turbines for the sustainable supply of electrical power. Because of the cost of installing and maintaining wind turbines, often situated in remote areas and increasingly used off-shore, it has become imperative to use the most up-to-date monitoring methods to avoid unforeseen failure. For the same reason there has been a tendency to increase the size of individual units, but this gives problems for the transmissions used to convert the low speed input at a fraction of a Hz to the speed of conventional generators in the range 16-30 Hz or so. The power delivered by a turbine is proportional to the square of its diameter, but the input torque is proportional to the cube, and a large part of the cost of the gearbox is related to the torque. Anecdotally, this appears to have resulted in a situation where many transmissions are failing in a shorter time than their design life, and this means not only that their design must be improved, but also that they must be monitored more closely than many other machines, to detect and diagnose incipient faults at the earliest possible stage. The thesis seeks to help to redress the situation by proposing a number of innovative methods to improve the detection and diagnosis of faults in machines with variable speed and load, and especially wind turbine transmissions, not only to make the monitoring more efficient, but also more economical.

Grandes bague de roulement utilisés dans éolienne sont l'un des éléments de transmission de charge importantes de ces machines tournantes. Ces roulements fonctionnent grâce à des cycles de charge et de la fréquence et de l'expérience des défis complexes tribologiques sévères. Le coût de remplacement de ces paliers est très élevé et conduit aussi à quantité importante de temps d'arrêt. Il est donc important de comprendre certains des principaux problèmes de conception et tribologiques de ces roulements. Quatre points type de roulement de l'anneau de contact de rotation a été considéré comme une base de référence pour cette étude pour démontrer les questions de contact de troncature et d'échec de la cage pour les roulements de hauteur. Un palier de contact à deux points de remplacement est proposé d'éliminer le contact troncature et de réduire la force de la cage accumulation. Les méthodes de conception et d'analyse démontré dans cette étude peuvent être facilement étendus à lacet paliers ainsi que d'autres grands roulements utilisés dans l'industrie. / Large slewing ring bearings used in wind turbine are one of the important load transmitting elements of these rotating machines. These bearings operate through complex load and frequency cycles and experience severe tribological challenges. The cost of replacement of these bearings is very high and also leads to significant amount of down-time. It is therefore important to understand some of the major design and tribological issues in these bearings. Four-point contact slewing ring bearing type has been considered as a baseline for this study to demonstrate contact truncation and cage failure issues for pitch bearings. An alternate two-point contact bearing is proposed to eliminate contact truncation and reduce the cage force build-up. The design and analysis methods demonstrated in this study can be easily extended to yaw bearings as well as other large bearings used in the industry.

L'étude développée dans cette thèse concerne le contrôle des performances et des lâchers tourbillonnaires au cours du cycle de rotation d'une hydrolienne à axe vertical de type Darrieus. L'élaboration d'une famille de lois de commande d'incidence de pales exploitant le principe de conservation de la circulation autour de profils en mouvement permet ici le contrôle du fonctionnement de l'hydrolienne ainsi que la maîtrise de son sillage tourbillonnaire afin de préserver l'environnement.L'écoulement 2D est simulé à l'aide du solveur incompressible de Star CCM+ afin de mettre en évidence l'effet de ce type de contrôle sur le rendement de la turbine pour différents points de fonctionnement. Ce modèle CFD a été utilisé pour améliorer l'analyse analytique en ce qui concerne l'extraction de l'énergie, la compréhension de l'écoulement autour de l'hydrolienne et le contrôle des tourbillons générés. La nouveauté de cette étude est l'élaboration de lois de commande de pales d'hydrolienne, basées sur des valeurs constantes et transitoires de la circulation, afin d'augmenter la puissance de la turbine tout en garantissant un contrôle efficace de la vorticité et ainsi prévenir de l'interaction entre les tourbillons et les pales. Une bonne comparaison est réalisée entre les résultats analytiques et numériques concernant les forces hydrodynamiques.En outre, une campagne d'essais a été menée afin d'acquérir des mesures quantitatives sur une hydrolienne de type Darrieus à pales fixes en terme de puissance, mais aussi des résultats qualitatifs pertinents comme la visualisation de l'écoulement autour des pales à différentes positions et pour différents points de fonctionnement. La mise en place complète d'un système PTV pour les mesures qualitatives et les étapes de traitement sont discutées et les divers paramètres obtenus à partir des études CFD sont validées en utilisant ces résultats PIV.L'étude expérimentale dans la présente recherche appo11e des informations détaillées sur les gradients de pression et de vitesse, les contours de vorticité et le critère Q qui ont servi à valider les visualisations obtenues numériquement. / With key applications in marine renewable energy. the vertical axis water turbine can use current or tidal energy in an eco-friendly manner. However, it is difficult to reconcile optimal performance of hydrokinetic turbines and compliance wilh the aquatic environment as the main drawback of the turbines is the formation of non-linear flow structures caused by the unsteady movement of the blades. Eddies in the flow are advected and can interact with other blades, which leads to a reduction in power output. To limit this phenomenon, the turbines operate at high speeds, which are likely to reduce the shaft power. High speeds of rotational so forbid the passage of aquatic animais, and are the cause of a suction effect on the sediments.The objective of this thesis work is twofold. First, it aims to develop a blade pitch control to get the flow adjusted around the blade profile at any given flow configuration by incorporatin.g the profile's motion with respect to incident flow. Such a system intends to achieve the objective of operating at reduced speeds without vortical releases, which should allow achieving a high torque without causing damage to the environment.This thesis work is mainly carried out in three phases. ln the first phase, the irrotational flow over an arbitrary profile is formulated using conforma] mapping. Prospective potential flow application on the basis of Couchet theory (1976) is involved in the development of a control law that decides the blade pitching in a constant circulation framework. In the second phase, a numerical validation of the developed analytical work is presented using CFD to examine how the theoretical fomulation can be effectively applied to Darricus turbines. In the final phase, two prototypes are developed, one is classical Darrieus turbine with fixed blades, and other is the turbine with pitching blades for experimental measurements of performance as well as flow fields(by PIV) in order to validate the computational results.

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / The constant human need for electricity, as well as the increasing demand for energy causes the use of finite sources, as oil and coal. The search for alternative energy sources as wind harvesting is the motivation for this master dissertation. Wind energy is a consolidated technology, mostly based on three blades wind turbine. This master thesis constitutes an analysis of a non-conventional wind machine, named Magnus wind turbine. The turbine s operation is based on the Magnus effect and its difference with respect to the traditional type is that it has rotating cylinders instead of fixed blades, which benefits an increase of generated power, since this turbine can produce power from low wind speeds. A method to track the maximum power point based on the control of the cylinders rotation is proposed, whereas in literature only fixed values were found for this variable. This Magnus turbine is used to drive a permanent magnet synchronous generator as a way to prove the theoretical predictions. The electrical generator also uses a specific method to track the maximum power point for the whole set. Simulated and practical results are demonstrating the performance of the Magnus wind turbine driving a permanent magnet synchronous generator. / A constante necessidade de energia elétrica, assim como o aumento de demanda, implica no uso de fontes de energia que são finitas, como o petróleo e o carvão. A busca por fontes alternativas de energia como a eólica é a motivação para este trabalho, também por a energia eólica ser uma tecnologia já consolidada, utilizando principalmente turbinas de três pás para a geração de energia. Esta dissertação trata da análise de uma turbina eólica não convencional, conhecida como turbina eólica Magnus para acionamento de geradores elétricos. A operação desta turbina é baseada no efeito Magnus e seu grande diferencial é que ao invés de pás fixas possui cilindros giratórios que ajudam no incremento de potência, já que a turbina pode produzir potência mesmo com baixas velocidades de vento. Um método de rastreamento de máxima potência é proposto baseando no controle da velocidade de rotação dos cilindros, já que na literatura são considerados apenas valores fixos para esta variável. A turbina Magnus é acoplada a um gerador síncrono a imã permanente, como forma de demonstração da precisão do método proposto na determinação do seu ponto de máxima potência. Paralelo a isto, o gerador também utiliza um método de rastreamento de máxima potência para que o conjunto inteiro trabalhe no ponto de máxima potência. Os resultados de simulação e práticos são apresentados para demonstrar o desempenho da turbina Magnus acoplada a um gerador síncrono a imã permanente.

Abstract<p><p>This thesis deals with the operation of and winding short-circuit fault detection in a Doubly-Fed Induction Generator (DFIG) based Wind Turbine Generator System (WTGS). Both the faulted and faultless condition of operation has been studied, where the focus is on the electrical part of the system. The modelled electrical system is first simulated and the developed control system is then validated on a test bench. The test-bench component dimensioning is also discussed.<p><p>The faultless condition deals with the start-up and power production mode of operation. Control design based on the Proportional Integral (PI) control technique has been compared for power and torque control strategies against the Linear Quadratic Gaussian (LQG) control technique, at different operating points through the variable-speed region of WTGS operation following the maximum power curve of the system. It was found that the torque control strategy offered less degradation in performance for both the control techniques at operating points different for the one for which the control system was tuned. The start-up procedure of the DFIG based WTGS has been clarified and simplified. The phase difference between the stator and the grid voltage, which occurs due to the arbitrary rotor position when the rotor current control is activated, is minimized by using a sample-and-hold technique which eliminates the requirement of designing an additional controller. This method has been validated both in simulation and experiments.<p><p>The faulted condition of operation deals with the turn-turn short-circuit fault in the phase winding of the generator. The model of the generator, implemented using the winding-function approach, allows the fault to be created online both in a stator and a rotor phase. It has been demonstrated that the magnitude of the current harmonics, used extensively in literature for the Machine Current Signature Analysis (MCSA) technique for winding short-circuit fault detection, is very different when the location of the fault is changed to another coil within the phase winding. This makes the decision on the threshold selection for alarm generation difficult. Furthermore, the control system attenuates the current harmonics by an order of magnitude. This attenuation property is also demonstrated through experiments. The attention is then shifted to the negative-sequence current component, resulting from the winding unbalance, as a possible fault residual. Its suitability is tested in the presence of noise for scenarios with different fault locations, fault severity in terms of the number of shorted-turns and grid voltage unbalance. It is found that due to the presence of a control system the magnitude of the negative-sequence current, resulting from the fault, remains almost the same for all fault locations and fault severity. Thus, it was deemed more suitable as a fault residual. In order to obtain a fast detection method, the Cumulative Sum (CUSUM) algorithm was used. The test function is compared against a threshold, determined on the basis of expected residual magnitude and the time selected for detection, to generate an alarm. The validation is carried out with noise characteristics different from the ones used during the design and it is shown that the voltage unbalance alone is not able to trigger a false alarm. In all the scenarios considered, the detection was achieved within 40 ms despite the presence of measurement filters. / Doctorat en Sciences de l'ingénieur / info:eu-repo/semantics/nonPublished

Ce travail de thèse porte sur les propriétés de stabilité des tourbillons hélicoïdaux, structures que l'on retrouve notamment dans le sillage des rotors d'hélicoptères et d'éoliennes.Dans une première partie, le développement spatio-temporel de l'instabilité d'appariement est caractérisé à l'aide d'un code numérique pseudo-spectral pour une allée infinie d'anneaux tourbillonnaires. On montre que ce modèle axisymétrique d'écoulement est en effet une bonne approximation du système hélicoïdal dans la limite des grands rayons et petits pas d’hélice. Dans ces conditions, et en utilisant un adimensionnement judicieux, on obtient également que le résultat théorique pour le taux de croissance spatio-temporel obtenu pour une double allée de tourbillons ponctuels s’avère être une bonne prédiction pour le cas hélicoïdal.Dans une seconde partie, on décrit comment un ou plusieurs tourbillons hélicoïdaux ont pu être générés de façon très peu perturbée à l’aide de modèles réduits de rotors dans le canal hydrodynamique du laboratoire. Grâce à l’introduction de perturbations d’amplitudes et de fréquences soigneusement contrôlées, le taux de croissance de l’instabilité d’appariement a pu être mesuré et comparé aux résultats théoriques. L’évolution non linéaire de ces perturbations ainsi que d’autres modes instables, à plus petites longueurs d’onde, ont également pu être observés expérimentalement pour la première fois.Enfin, ces résultats ont été appliqués au cas des rotors d’hélicoptères pour la prédiction du régime de Vortex Ring State (VRS) et à la transition vers la turbulence du sillage des éoliennes. / This thesis is devoted to the stability properties of helical vortices, which are of interest for applications such as helicopter and wind turbine wakes.In a first part, the spatio-temporal development of the pairing instability is characterised for an infinite array of vortex rings, using a pseudo-spectral numerical code. We show that this axisymmetric flow model is indeed a good approximation of the helical system in the limit of large helix radius and small pitch. Under these assumptions, and by using appropriate dimensionless variables, we also show that the theoretical result concerning the spatio-temporal growth rate for a double row of point vortices represents a good prediction for the helical case.In a second part, we describe how one or several helical vortices were generated in a carefully controlled way using small-scale rotor models in the water channel of the laboratory. Introducing perturbations with well-defined amplitudes and frequencies, the growth rate of the pairing instability could be measured experimentally and compared to theoretical predictions. The non-linear evolution of these perturbations, as well as other unstable modes of smaller wavelengths, were also observed experimentally for the first time.Finally, these results were applied to helicopter wakes for the prediction of the Vortex Ring State (VRS) regime and to the transition to turbulence in wind turbine wakes.

In this paper, a new algorithm for utilizing energy storage is proposed and applied on floating wind turbine economics. The proposed algorithm’s decision making on storing energy or selling electricity onto the grid is based on the electricity price, which makes it unique and different from similar algorithms. From the literature review, it was concluded Ocean Renewable Energy Storage to be most suitable with the Spar-Type and Semi-Submersible floating wind turbine to which the paper is based upon. The objective of this paper is to find the suitable ratio of energy storage versus wind farm, find the product of increase in wholesale, and evaluate whether the proposed method makes the hybrid economically sound. The algorithm was applied on spot-price data from Denmark due to its large share of wind energy with wind data from off the coast of Morro Bay in California, USA. Additionally, a sensitivity analysis is applied to evaluate to energy storage cost impact as well as evaluate the algorithm by lowering the required energy storage size.   Using the algorithm, the wind farm must account for nine days’ worth of energy production with a product of energy storage versus wind farm ratio of 1.42. The wholesale price increased with 11.9-21.5% for the four years studied, however, all financial results favored not utilizing energy storage. By the results derived from the sensitivity analysis, it was concluded that with future cost reductions, the algorithm will still favor no energy storage. However, by fine tuning the algorithm to reduce the need for storage, positive financial result might be achievable. The key to achieve a profitable result seems to rely on minimizing the need for energy storage, to which the proposed algorithm fail to achieve. Conclusively, spot-price decision-based energy storing is not economically sound.

Significant advances in wind turbine technology have increased the need for maintenance through condition monitoring. Indeed condition monitoring techniques exist and are deployed on wind turbines across Europe and America but are limited in scope. The sensors and monitoring devices used can be very expensive to deploy, further increasing costs within the wind industry. The work outlined in this thesis primarily investigates potential low-cost alternatives in the laboratory environment using vibration-based and modal testing techniques that could be used to monitor the condition of wind turbine blades. The main contributions of this thesis are: (1) the review of vibration-based condition monitoring for changing natural frequency identification; (2) the application of low-cost piezoelectric sounders with proof mass for sensing and measuring vibrations which provide information on structural health; (3) the application of low-cost miniature Micro-Electro-Mechanical Systems (MEMS) accelerometers for detecting and measuring defects in micro wind turbine blades in laboratory experiments; (4) development of an in-service calibration technique for arbitrarily positioned MEMS accelerometers on a medium-sized wind turbine blade. This allowed for easier aligning of coordinate systems and setting the accelerometer calibration values using samples taken over a period of time; (5) laboratory validation of low-cost modal analysis techniques on a medium-sized wind turbine blade; (6) mimicked ice-loading and laboratory measurement of vibration characteristics using MEMS accelerometers on a real wind turbine blade and (7) conceptualisation and systems design of a novel embedded monitoring system that can be installed at manufacture, is self-powered, has signal processing capability and can operate remotely. By applying the conclusions of this work, which demonstrates that low-cost consumer electronics specifically MEMS accelerometers can measure the vibration characteristics of wind turbine blades, the implementation and deployment of these devices can contribute towards reducing the rising costs of condition monitoring within the wind industry.

Wind turbine modelling using doubly-fed induction generators is a well-known subject. However, studies have tended to focus on optimising the components of the system rather than considering the interaction between the components. This research examines the interaction of the control methods for a doubly-fed induction generator (DFIG) in a wind turbine application integrating them with the crowbar protection control and DC-link brake control to make the best use of the converter. The controls of the rotor-side and the grid-side converters of the DFIG model are both well established and have been shown to work. Typically the crowbar protection is designed in order to protect the rotor-side converter and the power electronic components of the DFIG system from high currents occurring in the rotor due to the faults. The DC-link brake-overvoltage protection is also designed to prevent the overcharging of the DC-link capacitor placed between the rotor-side converter and the grid-side converter. In order to show that these protection schemes work and with thought can co-ordinate with each other, tests consisting of a number of balanced three-, two- and one-phase voltage sags are applied to the network voltage. The main contributions of this thesis are establishing operational tuning and design limits for the controllers and system subassemblies. This is to minimise the electrical subsystem interaction while maintaining adequate performance, and have an improved DC-link control. This work also includes a full electrical system study of the wind turbine and an essential literature review on significant references in the field of the DFIG wind turbine system modelling, control and protection. Specifically this research project makes a number of novel contributions to the literature: enhanced DC voltage control including operating point sensitivity analysis and dynamic stiffness assessment, sensitivity and robustness analyses of the power loop control and control loop segmentation by appropriately tuning the controller loops.

The increasing penetration of wind power into the power system dominated by variable-speed wind turbines among the installed wind turbines will require further development of control methods. Power electronic converters are widely used to improve power quality in conjunction with the integration of variable speed wind turbines into the grid. In this thesis, a detailed model of the Predictive Current Control (PCC) method will be descripts for the purpose of control of the grid-connected converter. The injected active and reactive power to the grid will be controlled to track their reference value. The PCC model predicts the future grid current by using a discrete-time model of the system for all possible voltage vectors generated by the inverter. The voltage vector that minimizes the current error at the next sampling time will be selected and the corresponding switching state will be the optimal one. The PCC is implemented in Matlab/Simulink and simulation results are presented.