Among the renewable energies, wind energy presents the highest growth in installed capacity and penetration in modern power systems. This is why reliability of wind turbines becomes an important topic in research and industry. To this end, condition monitoring (or health monitoring) systems are needed for wind turbines. The core of any condition monitoring system (CMS) are fault diagnosis algorithms whose task is to provide early warnings upon the occurrence of incipient (small magnitude) faults. Thanks to the use of CMS we can avoid premature breakdowns and reduce significatively maintenance costs.<p><p>The present thesis deals with fault diagnosis in sensors of a doubly-fed induction generator (DFIG) for wind turbine (WT) applications. In particular we are interested in performing fault detection and isolation (FDI) of incipient faults affecting the measurements of the three-phase signals (currents and voltages) in a controlled DFIG. Although different authors have dealt with FDI for sensors in induction machines and in DFIGs, most of them rely on the machine model with<p>constant parameters. However, the parameter uncertainties due to changes in the operating conditions will produce degradation in the performance of such FDI systems.<p><p>In this work we propose a systematic methodology for the design of sensor FDI systems with the following characteristics: i) capable of detecting and isolating incipient additive (bias, drifts) and multiplicative (changes in the sensor<p>gain) faults, ii) robust against changes in the references/disturbances affecting the controlled DFIG as well as modelling/parametric uncertainties, iii) residual generation system based on a multi-observer strategy to enhance the isolation process, iv) decision system based on statistical-change detection algorithms to treat the entire residual and perform fault detection and isolation at once.<p><p>Three novel sensor FDI approaches are proposed. The first is a signal-based approach, that uses the model of the balanced three-phase signals (currents or voltages) for residual generation purposes. The second is a model-based approach<p>that accounts for variation in the parameters. Finally, a third approach that combines the benefits of both the signal- and the model-based approaches is proposed. The designed sensor FDI systems have been validated using measured voltages, as well as simulated data from a controlled DFIG and a speed-controlled induction<p>motor. <p><p>In addition, in this work we propose a discrete-time multiple input multiple output (MIMO) regulator for each power converter, namely for the rotor side converter (RSC) and for the grid side converter (GSC). In particular, for RSC<p>control, we propose a modified feedback linearization technique to obtain a linear time invariant (LTI) model dynamics for the compensated DFIG. The novelty of this approach is that the compensation does not depend on highly uncertain parameters such as the rotor resistance. For GSC control, a LTI model dynamics<p>is derived using the ideas behind feedback linearization. The obtained LTI model dynamics are used to design Linear Quadratic Gaussian (LQG) regulators. A single design is needed for all the possible operating conditions. / Doctorat en Sciences de l'ingénieur / info:eu-repo/semantics/nonPublished

Global wind power generation has grown rapidly in response to targets to reduce greenhouse gas emissions as part of efforts to mitigate climate change, and to increase energy security. While much of the focus in wind energy technology to date has been on wind farms, a relatively recent development is the expansion of the micro-wind sector (turbines generating < 100 kW), and there are now over 870,000 small wind turbines (SWTs) installed globally. However, official planning guidance for SWTs in the UK and elsewhere is lacking. This may be a barrier to SWT installations if there is confusion over the requirements to gain planning permission. One reason for the lack of planning guidance is that our understanding of the wildlife impacts of SWTs is limited and therefore it is difficult to make recommendations for their mitigation. There are a range of potential negative effects wind power can exert on wildlife, in particular on birds and bats, yet to date, there has been very little published research into the wildlife impacts of SWTs. Mortality rates of wildlife at SWTs appear to be relatively low, but disturbance of bats, highly protected species, near SWTs has been previously demonstrated. However, the extent (if any) of this disturbance at habitat features of known importance was unclear. Therefore this thesis used acoustic surveys of bat activity to quantify disturbance of use of linear features (e.g. hedgerows, treelines), habitat important to bats for commuting and foraging, caused by SWTs. Firstly, bat activity did decline after experimental installation of SWTs 5m away from linear features. This decline was species-specific with Pipistrellus pygmaeus showing declines in activity in close proximity to the SWT associated with SWT operation, while P. pipistrellus activity declined in response to installation both at the SWT site and 30m away. Secondly, bat use of linear features is lower when SWTs are located nearby. In particular, P. pygmaeus activity at linear features is lower the closer a SWT is to the feature, and at high wind speeds Myotis spp. use of linear features is similarly lower where SWTs are located nearby. This disturbance did not dissipate along the linear features away from the SWT for at least 60m. This is much further than previously documented disturbance of bats by SWTs, which appeared fairly localised, and may be due to the importance of linear features specifically for commuting between habitat fragments. If so, the cumulative impacts of such disturbance will be important in areas where suitable foraging and roosting habitats is limited and fragmented, and linear features suitable for commuting between habitat fragments are already rare. These results offer support for recommendations that SWTs should be subject to siting restrictions that create a buffer distance between them and important bat habitats such as linear features. Specifically, this thesis recommends that in landscapes with few alternative commuting routes or where particularly rare bat species are present SWT installations require buffer distances to ensure they are a minimum of 60m away from linear features. There has also been a lack of research into public attitudes towards SWTs, despite local attitudes towards wind farm developments having been linked to planning outcomes, implying attitudes can be a barrier to installations. This thesis presents the results of the first survey of public attitudes specifically towards SWTs. Generally attitudes towards SWTs were positive, with over half of respondents rating SWTs as acceptable across a range of landscape settings. However, as for wind power where public attitudes in general are positive but local wind farm developments may still face opposition, only 35% of respondents were in favour of having a SWT installed in sight of their home. A key finding of this survey was that acceptance of SWTs significantly differed between landscape settings, with those in hedgerows and gardens being less well accepted compared to those on road signs, buildings and fields. Respondent comments highlighted visual impacts, efficient use of technology, noise impacts, wildlife impacts and educational value as important factors in their decisions regarding SWT acceptability. Public concern about wildlife impacts appears to be responsive to context, being important to the lower acceptance of SWTs in hedgerows, which were perceived to be particularly risky for wildlife. Potential SWT owners are also shown to be concerned about wildlife impacts from SWTs. Using a choice experiment methodology, an economics technique that allows valuation of non-market goods, farmers (a group most likely to own SWTs in the UK) were found to be willing-to-pay, through loss of SWT earnings from electricity generation, to avoid disturbance of birds and bats or collision mortality of bats. These findings also support the recommendation of the use of buffer distances for SWTs. Buffer distances between SWTs and linear features will help to alleviate public and SWT owner concerns about wildlife impacts, and also increase public acceptance of SWTs by encouraging their installation away from some of the least accepted landscape settings such as hedgerows. Further, potential SWT owners were also found to have no significant preference for avoiding siting restrictions of SWT installations, suggesting they are open to the use of buffer distances, although the suggested distances were substantially smaller than those this thesis ultimately recommends. The findings presented in this thesis have implications for planning guidance, policy makers and developers, but also raise many questions that will require further study. A list of planning guidance recommendations and a list of recommendations for future SWT research are presented in the final section.

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.

Wind Power Project Development is a complicated, capital and resource-inclusive process, where a wide variety of factors have to be considered and several stakeholders have a significant say in the process. Decision making in such an environment is complex and has to be approached comprehensively. In order to sustain a structured and clear decision-making process, the sustainable energy industry has recognized Multi-Criteria Decision Analysis (MCDA) method as a suitable set of tools to aid in the decision-making process. One of the MCDA tools – PROMETHEE II, has been examined in this master thesis, to evaluate its eligibility as a decision making aid in wind power project development. To structurally and realistically evaluate the tool, it has been applied to a case study in Ventspils region, in Latvia. The author of this thesis has a preliminary agreement with the owners of the sites to develop the project, therefore, this thesis has a strong potential for a practical implementation in future. Four scenarios have been developed for an evaluation, contributing to four variations of the different amount of turbines erected, with two different hub heights, on two differently sized sites. The scenarios are assessed based on the interests of six key stakeholders. Their opinion on twelve criteria is examined. Input data for each criterion has been generated via WindPro and MS Excel software or by authors assessment based on the researched literature. PROMETHEE II is used to extrapolate a comprehensive and clear representation of the results. The evaluation of the MCDA method proved that MCDA tools, and PROMETHEE II, in particular, can provide excellent support in decision making in wind power development. A wide variety of input data, as well as the various and often contradicting interests by different stakeholders, can be taken into account, while, at the same time, a clear result that can assist in decision making, is generated.  Wind Power Project Development is a complicated, capital and resource-inclusiveprocess, where a wide variety of factors have to be considered and severalstakeholders have a significant say in the process. Decision making in such anenvironment is complex and has to be approached comprehensively. In order to sustaina structured and clear decision making process, sustainable energy industry hasrecognized Multi-Criteria Decision Analysis (MCDA) method as a suitable set of tools toaid in the decision making process. One of the MCDA tools – PROMETHEE II, hasbeen examined in this master thesis, to evaluate its eligibility as a decision making aidin wind power project development.To structurally and realistically evaluate the tool, it has been applied on a case study inVentspils region, in Latvia. The author of this thesis has a preliminary agreement withthe owners of the sites to develop the project, therefore, this thesis has a strongpotential for a practical implementation in future. Four scenarios have been developedfor an evaluation, contributing to four variations of different amount of turbines erected, with two different hub heights, on two differently sized sites. The scenarios are assessed based on the interests of six key stakeholders. Their opinion on twelve criteria examined. Input data for each criterion has been generated via WindPro and MS Excel software orby authors assessment based on the researched literature. PROMETHEE II is used to extrapolate a comprehensive and clear representation of the results.The evaluation of the MCDA method proved that MCDA tools, and PROMETHEE II in particular, can provide excellent support in decision making in wind power development. Wide variety of input data, as well as the various and often contradicting interests by different stakeholders can be taken into account, while, at the same time, a clear result that can assist in decision making, is generated.

Aus der Einleitung: "Der Anteil erneuerbaren Energien in der Energieversorgung soll auch in den kommenden Jahren weiter ausgebaut werden. Dadurch steigt der Bedarf an großen und effizienten Windkraftanlagen (WKA) mit immer höheren Anforderungen an Materialien und Strukturen. Die hohen Belastungen und limitierenden Faktoren in Bezug auf die Lebensdauer solcher Anlagen sind meist dynamisch und abhängig von Strukturschwingungen und Belastungsanregung. Hierzu werden vermehrt Simulationstechniken eingesetzt, die in der Praxis durch den Mangel an genauen Randbedingungen unpräzise sind bzw. als Modell durch Messungen validiert werden müssen. Während typischerweise zur Erfassung von Schwingwegen Beschleunigungsaufnehmer eingesetzt werden, gestaltet sich die Implementierung solcher Messtechnik in drehenden Strukturen meist komplex und aufwändig (Ozbek et al. 2010). Der Einsatz optischer Messtechnik ist im Vergleich einfacher und kann potentiell die Optimierung von Simulationsmodellen unterstützen (Ozbek et al. 2010, Schmidt-Paulsen et al. 2009, 2011)."

Coastal South African cities like Port Elizabeth are said to have a strong potential for wind energy. This study aims to model wind data in order to be able assess the power potential belonging to a given site. The main challenge in modelling wind direction data is that it is categorized as circular data and therefore requires special techniques for handling that are different from usual statistical samples. Statistical tools such as descriptive measures and distribution fitting, were re-invented for directional data by researchers in this field. The von Mises distribution is a predominant distribution in circular statistics and is commonly used to describe wind directions. In this study, the circular principles described by previous researchers were developed by using the statistical software, Mathematica. Graphical methods to present the wind data were developed to give an overview of the behaviour of the winds in any given area. Data collected at Coega, an area near Port Elizabeth, South Africa, was used to illustrate the models which were established in this study. Circular distributions were fit to the directional data in order to make appropriate conclusions about the prevailing wind directions in this area.

Coastal South African cities like Port Elizabeth are said to have a strong potential for wind energy. This study aims to model wind data in order to be able assess the power potential belonging to a given site. The main challenge in modelling wind direction data is that it is categorized as circular data and therefore requires special techniques for handling that are different from usual statistical samples. Statistical tools such as descriptive measures and distribution fitting, were re-invented for directional data by researchers in this field. The von Mises distribution is a predominant distribution in circular statistics and is commonly used to describe wind directions. In this study, the circular principles described by previous researchers were developed by using the statistical software, Mathematica. Graphical methods to present the wind data were developed to give an overview of the behaviour of the winds in any given area. Data collected at Coega, an area near Port Elizabeth, South Africa, was used to illustrate the models which were established in this study. Circular distributions were fit to the directional data in order to make appropriate conclusions about the prevailing wind directions in this area.

La problématique de la production d'énergie renouvelable, l'énergie du vent dans notre cas, est liée au réglage de l'amplitude et de la fréquence de la tension du générateur entraîné par l'éolienne. Ces réglages sont nécessaires pour injecter la puissance électrique produite sur le réseau électrique. Les avancées de l'électronique de puissance, en particulier les convertisseurs de puissances commandés (redresseur, onduleur, convertisseur boost, etc), ont aidé à résoudre ces problèmes de réglage. Le défi de la production éolienne est d'améliorer la stratégie de commande appliquée aux convertisseurs, afin d'augmenter la qualité de la puissance produite et de minimiser les pertes électriques et ainsi de réduire le coût de production. Dans cette thèse, nous avons développé la commande non linéaire fondée sur la platitude d'un système de production éolien et son couplage au réseau. Cette approche permet de faciliter la construction de la loi de commande. Nous avons exposé un réglage simple et intuitif des gains des contrôleurs, reposant sur très peu de paramètres. Les pertes électriques au stator du GSAP sont minimisées, la stratégie de MPPT est réalisée sur la partie de production. Le facteur de puissance de l'énergie injectée sur le réseau électrique est contrôlé à n'importe quelle valeur désirée comprise entre zéro et un. / The problem of the production of renewable energy, ``wind energy in our case'' is related to the adjustment of the amplitude and the frequency of the voltage of the generator driven by the wind turbine. These settings are required to supply the produced electrical power to an electrical network. The advancement of the power electronics used in the power converters (rectifiers, inverters, boost converters, etc.), helped to solve these two problems of adjustment. The challenge of wind generation is to improve the control strategy applied to the converters to increase the quality of the produced power and to minimize the electrical losses of the generator, and therefore to reduce the cost of production system. In this thesis, we developed the nonlinear flatness-based control of a wind generation system. This command is used to facilitate the construction of the control law. We have shown that this command allows the designer to choose the controller gains with few and intuitive tuning parameters. Electrical losses in the stator of the Permanent Magnets Synchronous Generator (PMSG) are minimized, the Maximum Power Point Tracking strategy (MPPT) is applied to the production system and the power factor of the apparent power injected to the grid is controlled to obtain any desired value between zero and one.

L'énergie éolienne jouant un rôle de plus en plus important dans le réseau électrique, elle ne peut plus être considérée comme une source d'énergie marginale. Par conséquent, l'impact sur le réseau de l'intermittence, inhérente à ce type d'énergie devient non négligeable. L'utilisation du stockage est une des principales solutions à ce problème d'intégration. Ce travail porte sur l'optimisation du système éolien/stockage en considérant la dynamique de l'éolien, la capacité de stockage et l'interaction avec le réseau. L'objectif consiste à répondre aux exigences du réseau en limitant les fluctuations, à fournir des services-systèmes tout en mettant en avant la rentabilité économique du système. La méthode de gestion proposée s'appuie sur deux niveaux de pilotage : l'anticipation et la gestion réactive. La première phase consiste à utiliser des informations de prévisions (météorologiques, contraintes du réseau, conditions du marché électrique...) afin de définir par avance le programme de fonctionnement optimal du système de stockage. La deuxième phase étudie le fonctionnement en temps réel, où le système doit faire face aux perturbations et respecter les règles du mécanisme d'ajustement. Le problème est complexe avec de nombreuses variables de contrôle discrètes et continues. La Programmation Mixte Linéaire (PML) est utilisée pour résoudre efficacement le problème. La stratégie de fonctionnement optimale proposée sera validée sur un simulateur hors temps réel et un simulateur en temps réel. / Wind energy playing an increasingly important role in the electrical network and it will no longer be considered as a marginal. Therefore, the impact on the electrical grid of its inherent intermittency becomes non-negligible. The use of storage means is one of key points in the integration problem. In this work, the optimization of the wind/storage system is addressed by considering the dynamics of the wind power, the storage capacity and the grid constraints. The main objective is to meet the grids requirements in limiting the fluctuations, providing possible ancillary services and highlighting the economic profitability of system. The proposed method relies on a two levels control approach: anticipation and reactive management. The first one consists in using forecast information (weather, grid constraints, electrical market conditions …) to define in advance the optimal operation schedule for the storage system. In the second one, on real time operation, the system has to deal with possible disturbances and take the right adjustment control with the actual capacity. The problem is complex with numerous discrete control variables and continuous ones. A mixed-integer linear programming (MILP) is used to efficiently solve the problem. The proposed optimal operation strategy will be validated with on an offline simulation (simulink/Matlab) and a real time simulator.

Cette Thèse présente une méthodologie pour la conception optimale multidisciplinaire des générateurs synchrones à aimants permanents appliqués à l'énergie éolienne. Telle méthodologie prend en compte les comportements électriques, géométriques, thermiques, électroniques, mécaniques et économiques de l'éolienne. La courbe d'occurrence du vent, modélisée par une fonction de densité statistique, est également prise en compte par la méthodologie de conception proposée. Considérer cette courbe dans la conception optimale permet de calculer la quantité d'énergie produite par l'éolienne, qui fonctionne sous vitesse et puissance variables par l'influence de la vitesse du vent. La proposition d'utiliser l'optimisation dans la conception est d'améliorer la compétitivité de l'énergie éolienne en concevant un générateur avec une bonne relation entre son coût et sa capacité de produire de l'énergie. Plusieurs études portent sur la conception optimale des machines électriques liées à l'énergie éolienne, mais peu ont présenté une méthode permettant d'estimer la production d'énergie éolienne, ce qui est obligatoire pour obtenir un générateur rentable. De plus, les aspects multidisciplinaires approfondis du générateur électrique ont été considérés par la méthode proposée plus en détail que d'autres études trouvées dans la littérature. Prendre en compte tous ces aspects entraîne un grand problème d'optimisation qui compte environ 2000 variables, dont environ 250 ont des contraintes imposées. Un algorithme d'optimisation déterministe a été choisi pour traiter ce problème. Un tel algorithme est capable de trouver la solution en quelques itérations et un court laps de temps grâce au calcul de la matrice jacobienne, contenant les gradients exacts du modèle (utilisé pour décrire le comportement de l'éolienne). Les résultats présentés explorent le potentiel méthodologique proposé. La première conception optimale réalisée minimise le coût du matériau actif du générateur, ce qui présente une réduction significative des coûts par rapport à un générateur non optimal. Ensuite, la contradiction entre le coût du générateur et son efficacité est analysée et l'importance d'inclure la production d'énergie éolienne dans la conception du générateur devient évidente. En utilisant l'énergie éolienne estimée générée par l'éolienne, le générateur avec un bénéfice net maximal (égal au produit de l'énergie éolienne moins le coût du générateur) a été obtenu. Ce générateur a un bon rapport entre son coût et sa capacité de produire de l'énergie. Ensuite, des analyses de sensibilité ont été effectuées pour vérifier comment le prix de l'électricité et la vitesse moyenne annuelle du vent influencent la conception. Enfin, deux éoliennes de différents diamètres et différents vitesses nominales du vent ont été utilisées, compte tenu de trois profils de vent distincts. Ces résultats fournissent les générateurs plus adaptés à chaque éolienne appliquée à chaque profil de vent, montrant que la méthodologie de conception proposée a la capacité de faciliter le choix de l'éolienne adéquate à chaque parc éolien. / This Ph.D. dissertation presents a multidisciplinary optimum design methodology to permanent magnet synchronous generators applied to wind power. Such a methodology considers the electrical, the geometrical, the thermal, the electronic, the mechanical and the economical behaviors of the wind turbine. The wind occurrence curve, modelled by a statistical density function, is also take into account by the proposed design methodology. Considering this curve in the optimal design allows to calculate the amount of energy produced by the wind turbine, which operates under variable speed and power by the influence of the actual wind speed. The proposal of using the optimization in the design is to enhances the competitiveness of wind power by designing a generator with a good relation between its cost and its capacity to generate energy. Several studies are addressed to electrical machines optimum design related to wind power, however, few have presented a method able to estimate the wind turbine energy generation, which is mandatory to obtain a cost-effective generator. Further, the electric generator in-depth multidisciplinary aspects have been considered by the proposed method in more detail than other studies found in literature. Take into consideration all these aspects results in a large optimization problem that has approximately 2000 variables, among which approximately 250 have constraints imposed on. A deterministic optimization algorithm has been chosen to deal with this problem. Such an algorithm is capable of finding the solution within few iterations and a short time thanks to the computation of the Jacobean matrix, containing the exact gradients of the model (used to describe the wind turbine behavior) outputs. The presented results explore the proposed methodology potential. The first optimal design performed minimizes the generator active material cost, presenting a significant cost reduction compared to a non-optimal generator. Next, the contradiction between the generator cost and its efficiency is analyzed and the importance of including the wind turbine energy generation into the generator design becomes evident. Using the estimated wind turbine generated energy, the generator with maximum net earnings (equal to the wind turbine energy proceeds less the generator cost) has been obtained. This generator has a good ratio between its cost and capability to generate energy. Then, sensibility analysis have been carried out to verify how the electricity price and the annual mean wind speed influence the design. Finally, two wind turbines with different diameters and rated wind speeds have been employed, considering three distinct wind profiles. These results provide the generators more suited to each wind turbine applied to each wind profile, showing that the proposed design methodology has the ability to helps choosing the wind turbine adequate to each wind farm.