Most wind turbines are installed in wind farms, as the costs associated with a single turbine for maintenance and energy production are much higher than installing multiple in the long run. There is a growing need to create more compact wind farms ever in the face of a reduction in costs linked to the impact on the territory and the development of ever more efficient machines. As a result, turbines are generally close enough to interact with each other and cannot be arranged randomly. The flow at a turbine has a reduced average speed and a greater intensity of turbulence; this flow region behind the wind turbine is called the wake and is the current that directly affects all the machines placed behind the first one. Much past research has focused on methodologies for re-energizing the wake. The task is to ensure that machines immersed in this region of the current can receive optimal flow and increase performance. However, the presence of squat bodies such as spacecraft and tower, which are essential for developing the current in the wake, has rarely been considered, especially in CFD simulations. The objective of this study is precisely to provide a preliminary and quantitative assessment of the impact of nacelle and tower on the wake of a machine and the performance of a turbine placed behind another. Dynamic Induction Control (DIC) techniques are adopted to re-energize the wake, here understood as Pitch Collective Motion (PCM). The two PCM methods chosen are a full-blade sine offset called S-PCM and a full-blade Gaussian offset called G-PCM. The LES simulations, based on OpenFOAM, are coupled with the aero-elastic structural solver OpenFAST, and a model for the atmospheric boundary layer is also included. It has been seen that maximum wake recovery with PCM methods, compared to a case with a standard control technique, occurs in the close wake, especially three diameters away. A mini wind farm consisting of two turbines placed in series, three diameters apart, is then analyzed and compared with similar scenarios where the nacelle and tower structures are missing. Only the machine upstream will be equipped with controls on the pitch of the blades, while the turbine located downstream of the first will be equipped with only the basic controller. It can be seen that, considering the spacecraft and the tower in the model, the average speed in the wake has a slight percentage loss of 1% compared to the cases in which these two structures are missing. As a result, the downstream turbine will produce less power. Here, global energy production is the primary performance indicator and, although the speed loss is minimal, the percentage loss of global power produced reaches 4%. The second part of the thesis focuses on analyzing a new blade pitch control strategy called helix. The latter is compared with the PCM strategies presented above. Without going through the LES simulations, it was considered appropriate to directly exploit the aero-elastic software OpenFAST for the simulations on the single isolated turbine. The helix method provides good power with minimal fluctuations in rotor thrust. However, to its detriment, it is extreme from a structural point of view, especially on the loads felt at the top of the tower.
La maggior parte delle turbine eoliche sono installate in parchi eolici, poiché i costi associati ad una singola turbina, per la manutenzione e la produzione di energia, sono molto più elevati rispetto all'installazione di molteplici sul lungo periodo. Cresce l'esigenza di realizzare parchi eolici sempre più compatti a fronte di una riduzione dei costi legati all'impatto sul territorio e allo sviluppo di macchine sempre più efficienti. Di conseguenza, le turbine sono generalmente abbastanza vicine da interagire tra loro e non possono essere disposte casualmente. Il flusso in corrispondenza di una turbina ha una velocità media ridotta e una maggiore intensità di turbolenza; questa regione di flusso dietro la turbina eolica è chiamata scia ed è la corrente che investe direttamente tutte le macchine poste dietro la prima. Molte ricerche passate si sono concentrate su metodologie per rienergizzare la scia. Il compito è far sì che le macchine immerse in questa regione della corrente possano ricevere un flusso ottimale e aumentare le prestazioni. Tuttavia, la presenza di corpi tozzi come navicella e torre, che sono importanti per lo sviluppo della corrente nella scia, è stata raramente presa in considerazione, specialmente nelle simulazioni CFD. L'obiettivo di questo studio è proprio quello di fornire una valutazione preliminare e quantitativa dell'impatto di navicella e torre sulla scia di una macchina e sulle prestazioni di una turbina posta dietro un'altra. Vengono adottate tecniche di Dynamic Induction Control (DIC) per la rienergizzazione della scia, qui intese come movimento collettivo del passo delle pale (PCM, Pitch Collective Motion). I due metodi PCM scelti sono: un offset sinusoidale full-blade chiamato S-PCM e un offset gaussiano full-blade chiamato G-PCM. Le simulazioni LES, basate su OpenFOAM, sono accoppiate con il solutore strutturale aeroelastico OpenFAST, e viene incluso anche un modello per lo strato limite atmosferico. È stato visto che il recupero massimo della scia con i metodi PCM, rispetto a un caso con una tecnica di controllo standard, si verifica nella scia vicina, specialmente a tre diametri di distanza. Viene quindi analizzato un mini parco eolico costituito da due turbine poste in serie, distanti tre diametri e confrontato con scenari simili dove mancano le strutture della navicella e della torre. Solo la macchina a monte sarà dotata dei controlli sul passo delle pale, mentre la turbina posta a valle della prima sarà dotata del solo controller di base. Si vede che, inserendo la navicella e la torre nel modello, la velocità media in scia ha una leggera perdita percentuale dell'1% rispetto ai casi in cui queste due strutture mancano. Di conseguenza, la turbina a valle produrrà meno potenza. Qui, la produzione globale di energia è il principale indice di performance e, nonostante la perdita di velocità è minima, la perdita percentuale di potenza globale prodotta raggiunge il 4%. La seconda parte della tesi si concentra sull'analisi di una nuova strategia di controllo sul passo della pala chiamata helix. Quest'ultima è confrontata con le strategie PCM presentate sopra. Senza passare per le simulazioni LES, si è ritenuto opportuno sfruttare direttamente il software aero-elastico OpenFAST per le simulazioni sulla singola turbina isolata. Il metodo helix fornisce una buona potenza con fluttuazioni minime nella spinta del rotore. A suo discapito risulta estrema dal punto di vista strutturale, soprattutto sui carichi avvertiti alla sommità della torre.
A preliminary study of the impact of the dynamic induction control on the wake of a wind turbine
Arcari, Valerio
2020/2021
Abstract
Most wind turbines are installed in wind farms, as the costs associated with a single turbine for maintenance and energy production are much higher than installing multiple in the long run. There is a growing need to create more compact wind farms ever in the face of a reduction in costs linked to the impact on the territory and the development of ever more efficient machines. As a result, turbines are generally close enough to interact with each other and cannot be arranged randomly. The flow at a turbine has a reduced average speed and a greater intensity of turbulence; this flow region behind the wind turbine is called the wake and is the current that directly affects all the machines placed behind the first one. Much past research has focused on methodologies for re-energizing the wake. The task is to ensure that machines immersed in this region of the current can receive optimal flow and increase performance. However, the presence of squat bodies such as spacecraft and tower, which are essential for developing the current in the wake, has rarely been considered, especially in CFD simulations. The objective of this study is precisely to provide a preliminary and quantitative assessment of the impact of nacelle and tower on the wake of a machine and the performance of a turbine placed behind another. Dynamic Induction Control (DIC) techniques are adopted to re-energize the wake, here understood as Pitch Collective Motion (PCM). The two PCM methods chosen are a full-blade sine offset called S-PCM and a full-blade Gaussian offset called G-PCM. The LES simulations, based on OpenFOAM, are coupled with the aero-elastic structural solver OpenFAST, and a model for the atmospheric boundary layer is also included. It has been seen that maximum wake recovery with PCM methods, compared to a case with a standard control technique, occurs in the close wake, especially three diameters away. A mini wind farm consisting of two turbines placed in series, three diameters apart, is then analyzed and compared with similar scenarios where the nacelle and tower structures are missing. Only the machine upstream will be equipped with controls on the pitch of the blades, while the turbine located downstream of the first will be equipped with only the basic controller. It can be seen that, considering the spacecraft and the tower in the model, the average speed in the wake has a slight percentage loss of 1% compared to the cases in which these two structures are missing. As a result, the downstream turbine will produce less power. Here, global energy production is the primary performance indicator and, although the speed loss is minimal, the percentage loss of global power produced reaches 4%. The second part of the thesis focuses on analyzing a new blade pitch control strategy called helix. The latter is compared with the PCM strategies presented above. Without going through the LES simulations, it was considered appropriate to directly exploit the aero-elastic software OpenFAST for the simulations on the single isolated turbine. The helix method provides good power with minimal fluctuations in rotor thrust. However, to its detriment, it is extreme from a structural point of view, especially on the loads felt at the top of the tower.File | Dimensione | Formato | |
---|---|---|---|
main_tesi.pdf
accessibile in internet solo dagli utenti autorizzati
Dimensione
6.87 MB
Formato
Adobe PDF
|
6.87 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/177220