The objective of this thesis is to assess the optimal power production and flight trajectories of crosswind, ground-generation or pumping-mode airborne wind energy systems (AWES), subject to realistic onshore and offshore, mesoscale-modeled wind data as well as LiDAR wind resource assessment. The investigation ranges from small scale AWES with an aircraft wing area of 10 m^2 to utility scale systems of 150 m^2. In depth knowledge of the wind resource is the basis for the development and deployment of any wind energy generator. Design and investment choices are made based on this information, which determine instantaneous power, annual energy production and cost of electricity. In the case of AWES, many preliminary and current analyses of AWES rely on oversimplified analytical or coarsely resolved wind models, which can not represent the complex wind regime within the lower-troposphere. Furthermore, commonly used, simplified steady state models do not accurately predict AWES power production, which is intrinsically linked to the aircraft's flight dynamics, as the AWES never reaches a steady state over the course of a power cycle. Therefore, leading to false assumption and unrealistic predictions. In this work, we try to expand our knowledge of the wind resource at altitudes beyond the commonly investigated lowest hundreds of meters. The so derived horizontal wind velocity profiles are then implemented in to an optimal control framework to compute power-optimal, dynamically feasible flight trajectories that satisfy operation constraints and structural system limitations. The so derived trajectories describe an ideal, or at least a local optimum, and not necessarily realistic solution. It is unlikely that such power generation can be reached in practice, given that disturbances, model assumptions, misalignment with the wind direction, control limitations and estimation errors, will reduce actual performance. We first analyze wind light detection and ranging (LiDAR) measurements at a potential onshore AWES deployment site in northern Germany. To complement these measurements we generate and analyze onshore and offshore, mesoscale weather research and forecasting (WRF) simulations. Using observation nudging, we assimilate onshore LiDAR measurements into the WRF model, to improve wind resource assessment. We implement representative onshore and offshore wind velocity profiles into the awebox optimization framework, a Python toolbox for modelling and optimal control of AWES, to derive power-optimal trajectories and estimate AWES power curves. Based on a simplified scaling law, we explore the design space and set mass targets for small to utility-scale, ground-generation, crosswind AWESs. / Graduate

The wind power industry is one of the fastest-growing renewable energy industries in the world. Since more energy can be extracted from wind when the density is higher, a lot of the investments made in the wind power industry are made in cold climates. But with cold climates come harsh weather conditions such as icing. The icing on wind power rotor blades causes the aerodynamic properties of the blade to shift and with further ice accretion, the wind power plant can come to a standstill causing a loss of power, until the ice is melted. How big these losses are, depend greatly on site-specific variables such as elevation, temperature, and precipitation. The literature claims these ice-related losses can correspond to 10-35% of the annual expected energy output. Some studies have been made to standardize an ice loss determining method to be used by the industry, yet a standardization of calculating these losses do not exist. It was therefore interesting for this thesis to investigate the different methods that are being used. By using historical Supervisory Control and Data Acquisition (SCADA) data for two different sites located in Sweden, a robust ice determining code was created to identify ice losses. Nearly 32 million data points are being analyzed, and the data itself is provided by Siemens Gamesa which is one of the biggest companies within the wind power industry. A sensitivity analysis was made, and it was shown that a reference dataset reaching from May to September for four years could be used to clearly identify ice losses. To find the ice losses, three different scenarios were tested. The three scenarios use different temperature intervals to find ice losses. For scenario 1 all data points below 0 degrees are investigated. And for scenario 2 and 3 this interval is stretching from 3 degrees and below versus 5 degrees and below. It was found that Scenario 3, was the optimal way to identify the ice losses. Scenario 3 filtered the raw data so that only data points with a temperature below five degrees was used. For the two sites investigated, the annual ice losses were found to lower the annual energy output by 5-10%. Further, the correlation between temperature, precipitation, and ice losses was investigated. It was found that low temperature and high precipitation is strongly correlated to ice losses.

This thesis investigates if the profitability of a wind farm can be increased by reducing itscontracted grid capacity. Two years of SCADA data is cleaned from non- and partialperformance which is used to estimate a wake reduced annual power time series. Stochasticmodels of production losses are applied to translate the wake reduced annual power timeseries. Ice losses are modelled with a 3-state Markov chain. The statistical properties arecalculated by identifying ice events in the SCADA. With the IEA task19 IceLoss algorithm areice events identified in the SCADA signal. An ice loss factor of 86 % is estimated for Juktanduring 2019. The results indicate that profitability can be increased by reducing the (contracted)grid capacity. Furthermore, the optimized grid capacity is shown to have low sensitivity to powerprice and ice losses. This finding is valuable since the power price market and weather areinherently difficult to predict. It follows that the prediction uncertainties of these inputs are lesssignificant when calculating the optimized grid capacity.

Les modèles énergétiques des bâtiments à l’échelle du quartier sont généralement simplifiés pour faire face au manque de données et pour réduire le coût de calcul. Cependant, l’impact de ces simplifications sur la validité des modèles n’est pas systématiquement analysée, en particulier lorsqu’on s’intéresse à la courbe de charge. Pour combler ce manque, une méthodologie permettant de quantifier la validité des simplifications, notamment vis-à-vis de la courbe de charge, est proposée. Cette méthodologie est appliquée aux simplifications couramment utilisée pour les modèles thermiques d’enveloppe de bâtiments grâce à une plateforme numérique développée dans le cadre de cette thèse. Cette plateforme permet de générer et simuler automatiquement des modèles énergétiques de bâtiments, avec différents niveaux de détails, à partir de données issues de systèmes d’information géographique. La parallélisation des simulations énergétiques des bâtiments est utilisée à l’échelle du quartier, afin de tirer avantage de la structure du modèle global et de réduire les temps de calculs. La définition d’indicateurs spécifiques selon l’objectif de simulation apparait clairement comme l’étape essentielle lorsque l’on s’intéresse à la courbe de charge. Les résultats indiquent que la puissance est plus sensible aux simplifications que la consommation annuelle d’énergie. Les différents effets induits sont quantifiés et analysés physiquement. La capacité de l’échelle du quartier à atténuer les impacts des simplifications et d’intégrer les données statistiques est démontrée. La quantification des impacts des simplifications permet de guider l’adaptation des modèles vis-à-vis des objectifs de simulation et vis-à-vis des contraintes techniques. Cette contribution a pour objectif d’améliorer la performance des simulations énergétiques à l’échelle de la ville, et de favoriser leur développement, afin de répondre aux enjeux futurs. / District-scale building energy models are generally simplified to cope with a lack of data and to reduce computational cost. However, the impacts of these simplifications on model accuracy are not systematically studied, particularly when considering power demand. The present manuscript introduces a methodology to determine the suitability of any simplifications, notably those at the district scale, and considering the power demand. This methodology was applied to usual simplifications of the building envelope model thanks to a specific platform developed in the frame of this thesis. This platform enables automatically generating and simulating building energy models with different modelling levels of detail from geographical information systems. The parallelisation of the building energy simulations was notably implemented at the district scale in order to benefit from the model structure and to efficiently reduce the computational duration. The definition of indicators related to specific simulation objectives appears to be a necessary step when focusing on power demand. The results show a higher sensitivity to simplifications of the power demand than the annual energy consumption. These effects are quantified and physically analysed. The district-scale ability to attenuate the impacts of simplifications and to integrate statistical sources of data were demonstrated. The resulting quantification of the impacts of the simplifications made it possible to guide the adaptations of models to the simulation objectives and to the technical constraints. Such contribution aims to increase the efficiency and to favour the development of city-scale energy simulations, which are particularly needed to cope with future challenges.

This work proposes a methodology to evaluate the performance of operating wind farms via the use of Supervisory Control and Data Acquisition System (SCADA) and modeled data. The potential annual energy is calculated per individual turbine considering underperforming/loss events to have their power output in accordance with a representative derived operational power curve. Losses/underperformance events are calculated and categorized into several groups aiming at identifying and quantify their causes. The methodology requires both anemometry data from SCADA system as well as modeled data. The discrepancy of the data representing the valid points of the power curve is taken into consideration as well when assessing the performance, i.e. wind speed vs power output of events that are not loss/underperformance. Production loss and relative standard deviation of power output of what is defined as “valid sample” in this work (per each turbine) are the main results obtained in this work. Finally, a number of optimization measures are suggested in order to enhance the performance, which can lead to a boost in the financial output of a wind farm. Aiming at judging the reliability of the proposed methodology, a case study is conducted and evaluated. The investigated case study shows that the methodology is capable of determining potential energy and associated losses/underperformance events. Several questions were raised during the assessment and are discussed in this report, recommendation for optimization measures are presented at the end of the study. Also, a discussion on the limitations and uncertainties associated to the presented methodology and the case study.

Many parameters used for Wind Resource Assessment (WRA) have uncertainty and variability, yet are input into the process as single values. The extent of the uncertainty or variance may not be known, and may or may not be significant enough to affect output. This Thesis focused on the energy calculation element of WRA, to assess the affect that errors (uncertainty) in three key user inputs had on the energy results. A parameter was chosen from each of the main groups influencing the energy calculation: wind speed (atmosphere), surface roughness (site conditions), and power curve (turbine technology). Reasonable variation due to uncertainty for wind speed and power curve were taken from other studies and their application simplified. Roughness change was assessed over the 5 classes (Class 0 (water) to 4 (dense forest/city)). WindPRO software was used to calculate the Annual Energy Production (AEP) and applied to three different wind turbine generators at the same coordinate. A sensitivity analysis was done on the AEP results using a hybrid One-At-a-Time Local Sensitivity Analysis by determining percentage changes from baselines and an overall rate of change for those key input parameters. The results showed that roughness class change effect was not linear. Changing from Class 0 to 1, AEP was on average -8±1%. Class 1 to 2 change was on average ‑12±1%. Class 2 to 3 change was on average -20±2%. Class 3 to 4 change was on average -29±2%. The wind speed change effect was found to be roughly linear. If mean wind speed has an error of ±10%, the AEP could be expected to be out by approximately +18/‑17% with a standard deviation of +4/-3%. The power curve change effect was also roughly linear. A PC±9% error leads to an approximate +6/-7% AEP error with a standard deviation of ±1%. Roughness class change was the most sensitive parameter to AEP with a 14.5 average rate of change, followed by wind speed at 1.8, then power curve with a 0.8 rate. Results compared reasonably well with other relevant studies.

To face climate change and transform the electricity supply to an environmentally friendly generation, wind plays an important role. Due to a yearly increase in installed wind power turbines, in the European Union, the need for maintenance increases as well. For reducing the maintenance times and, with that, the standstill time and resulting economical losses, the time for troubleshooting must be reduced. This work aims to show that the troubleshooting process of wind turbines can be reduced to a minimum with the automation. This can be reached by creating a scatter plot of the active power over the wind speed curve and investigating the data points where the turbine is not performing as it should. The data is extracted from a wind farm located in Finland for the wind year 2021. The methodological approach taken in this study is to build a normalised threshold power curve and compare it to monthly binned power curves of two selected turbines. The deviation between the threshold and the monthly power curve is investigated, and the months with a high deviation are chosen for further analysis, which includes the separation of the outlier data into four different categories. The outlier in bins with a higher deviation than 5 % are selected. The four categories are further inspected, and the reasons for the curtailments are extracted and analysed. In summary, these results show that the analysis of curtailment reasons based on a scatter plot of the active power of a wind turbine is possible. Moreover, the troubleshooting process can be reduced in time. Due to practical constraints, this work cannot provide an analysis with a threshold power curve built with data from more than one year. This makes the results less objective since fluctuations, which can occur during only one year, cannot be minimised.

The need for optimizing wind farms’ production and maximizing the profitability of projects necessitates power performance analysis. Nowadays, the use of remote sensing devices for this purpose becomes more and more popular due to many advantages this technology has over traditional met masts. The main objective of this study is to assess the performance of a wind turbine in wake and wake-free conditions through measurements performed with a nacelle mounted lidar. The analysis is based on the data obtained during a Power Curve Measurement campaign performed on an onshore wind farm in Sweden. The power production and power curves are compared for a range of wind direction sectors in order to assess turbine performance in different wake conditions. Surprisingly low power output is observed in wake-free sectors [180°, 240°) and [240°, 300°), whereas production in wake wind directions [300°, 360°) is relatively high. The main reason for this is the wind speed distribution, however the terrain complexity and roughness should also be considered as possible factors. Generally, the wind speed distribution seems to have more influence on the results than the wake conditions. Moreover, the correlation between the met mast and lidar datasets is investigated in the study. The results indicate a good agreement between the wind speed measurement from the two devices, however, a poor correlation is found for turbulence intensity and wind shear exponent. Additionally, the influence of turbulence intensity and wind shear on the power production was analyzed. Generally, the results were in line with the reviewed literature: at low and moderate wind speeds the power production was higher for higher TI values, while the opposite was observed for higher wind speeds, where the higher TI resulted in lower production. As for the wind shear, a pattern is observed for moderate wind speeds, where the higher wind shear resulted in lower power production.

Vertical-axis wind turbines (VAWTs) have with time been outrivaled by the today more common and economically feasible horizontal-axis wind turbines (HAWTs). However, VAWTs have several advantages which still make them interesting, for example, the VAWTs can have the drive train at ground level and it has been argued that they have lower noise emission. Other proposed advantages are suitability for both up-scaling and floating offshore platforms. The work within this thesis is made in collaboration between Halmstad University and Uppsala University. A 200-kW semi-guy-wired VAWT H-rotor, owned by Uppsala University but situated in Falkenberg close to Halmstad, has been the main subject of the research although most results can be generalized to suit a typical H-rotor. This thesis has three main topics regarding VAWTs: (1) how the wind energy extraction is influenced by turbulence, (2) aerodynamical noise generation and (3) eigenfrequencies of the semi-guy-wired tower. The influence from turbulence on the wind energy extraction is studied by evaluating logged operational data and examining how the power curve and the tip-speed ratio for maximum Cp is impacted by turbulence. The work has showed that the T1-turbine has a good ability to extract wind energy at turbulent conditions, indicating an advantage in energy extraction at turbulent sites for VAWTs compared to HAWTs.The noise characteristics are studied experimentally, and models of the two most likely aerodynamic noise mechanisms are applied. Here, inflow-turbulence noise is deemed as the prevailing noise source rather than turbulent-boundary-layer trailing-edge noise (TBL-TE) which is the most important noise mechanism for HAWTs. The overall noise emission has also been measured and proven low compared to similar sized HAWTs. The eigenfrequencies of a semi-guy-wired tower are also studied. Analytical expressions describing the first-mode eigenfrequency of both tower and guy wire has been derived and verified by experiments and simulations.

The wind power prediction plays a vital role in a wind power project both during the planning and operational phase of a project. A time series based wind power prediction model is introduced and the simulations are run for different case studies. The prediction model works based on the input from 1) nearby representative wind measuring station 2) Global average wind speed value from Meteorological Institute Uppsala University mesoscale model (MIUU) 3) Power curve of the wind turbine. The measured wind data is normalized to minimize the variation in the wind speed and multiplied with the MIUU to get a distributed wind speed. The distributed wind speed is then used to interpolate the wind power with the help of the power curve of the wind turbine. The interpolated wind power is then compared with the Actual Production Data (APD) to validate the prediction model. The simulation results show that the model works fairly predicting the Annual Energy Production (AEP) on monthly averages for all sites but the model could not follow the APD trend on all cases. The sensitivity analysis shows that the variation in production does not depend on ’the variation in roughness class’ nor ’the difference in distance between the measuring station and the wind farm’. The thesis has been concluded from the results that the model works fairly predicting the AEP for all cases within the variation bounds. The accuracy of the model has been validated only for monthly averages since the APD was available only on monthly averages. But the accuracy could be increased based on future work, to assess the Power law exponent (a) parameter for different terrain and validate the model for different time scales provided if the APD is available on different time scales.