A carregar...
Publicação
Análise Espectral da Velocidade do Vento com Vista à Determinação de Interacções entre Turbinas num Parque Eólico
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 7.94 MB | Adobe PDF |
Autores
Orientador(es)
Resumo(s)
As razões para o desenvolvimento da energia eólica em alguns locais estar a florescer, enquanto
que noutros não atinge a totalidade do seu potencial são complexas. A variabilidade
da velocidade do vento e da sua direcção são das características mais marcantes deste recurso.
A potência disponível no vento varia com o cubo da sua velocidade, de modo que a
compreensão das características desta variável são fundamentais para todos os aspectos da
exploração desta forma de energia eólica.
Com o objectivo de contribuir para a compreensão da variabilidade da velocidade e direção
do vento, foram observados dados provenientes de dois parques eólicos, localizados em
zonas distintas de Portugal continental. Estes conjuntos de dados, registados entre 2011 e
2013 por um sistema SCADA contêm informação relativa a diversas variáveis das turbinas,
sendo a velocidade do vento o principal foco desta dissertação.
Assim, este trabalho consiste na análise e tratamento de dados eólicos reais recorrendo
à análise espectral e ferramentas de estatística. Para além disso, foi avaliada a segunda
hipótese de similaridade de Kolmogorov, assim como a validade do modelo de turbulência
de von Kármán modificado.
Os principais resultados extraídos dos diversos estudos efectuados são: a distinção de
dois grupos de turbinas num dos parques quer pela análise estatística preliminar efectuada,
quer pela análise espectral; a verificação da existência de interação entre as turbinas de
um dos grupos identificados, o que pode levar a que algumas destas não produzam tanta
energia como esperado; a observação da segunda hipótese de similaridade de Kolmogorov
nos espectros de ambos os parques com erros inferiores a 1%, e o modelo de turbulência de
von Kármán modificado ter um erro geral de ⇡ 15% em relação aos espectros de ambos os
parques.
The reasons for the development of wind power in some places being flourishing while in others not reaching its full potential are complex. The variability of the wind speed and its direction are the most outstanding characteristics of this resource. The power available in the wind varies with the cube of its velocity, so the understanding of the characteristics of this variable is fundamental for all aspects of the exploration of this energy. In order to understand this variability, data from two wind farms located in di↵erent areas of mainland Portugal were observed. These datasets, recorded between 2011 and 2013 by a SCADA system, contain information on several variables of these turbines, with wind velocity being the main focus of this dissertation. Thus, the scope of this work is the analysis and treatment of real wind data using spectral analysis and statistical tools. In addition, the second Kolmogorov similarity hypothesis was evaluated, as well as the validity of the modified von Kármán turbulence model for these data. The main results obtained from the various studies carried out are: the possibility of distinguishing two groups of turbines in one of the farms by the preliminary statistical analysis carried out and by the spectral analysis; The verification of the existence of interconnection between the turbines of one of the identified groups, which may lead to some of them not producing as much energy as expected; The observation of the second Kolmogorov similarity hypothesis in the spectra of both farms with errors below 1 %, and the modified von Kármán turbulence model having an overall error of ⇡ 15% relative to the spectra of both farms.
The reasons for the development of wind power in some places being flourishing while in others not reaching its full potential are complex. The variability of the wind speed and its direction are the most outstanding characteristics of this resource. The power available in the wind varies with the cube of its velocity, so the understanding of the characteristics of this variable is fundamental for all aspects of the exploration of this energy. In order to understand this variability, data from two wind farms located in di↵erent areas of mainland Portugal were observed. These datasets, recorded between 2011 and 2013 by a SCADA system, contain information on several variables of these turbines, with wind velocity being the main focus of this dissertation. Thus, the scope of this work is the analysis and treatment of real wind data using spectral analysis and statistical tools. In addition, the second Kolmogorov similarity hypothesis was evaluated, as well as the validity of the modified von Kármán turbulence model for these data. The main results obtained from the various studies carried out are: the possibility of distinguishing two groups of turbines in one of the farms by the preliminary statistical analysis carried out and by the spectral analysis; The verification of the existence of interconnection between the turbines of one of the identified groups, which may lead to some of them not producing as much energy as expected; The observation of the second Kolmogorov similarity hypothesis in the spectra of both farms with errors below 1 %, and the modified von Kármán turbulence model having an overall error of ⇡ 15% relative to the spectra of both farms.
Descrição
Palavras-chave
Energia eólica Turbulência Velocidade do vento Análise espectral Segunda hipótese de similaridade de Kolmogorov Modelo de turbulência de von Kármán modificado