الفهرس 
CHAPTER 1: INTRODUCTION .Only 14 pages are availabe for public view
 |  | from | 139 |  |  |
| | | from | 139 | | |
Abstract
In the current study, the wind turbine wake characteristics have been investigated numerically under thermally-stratified atmospheric flow conditions. The steady state three dimensional Reynolds-Averaged Navier–Stokes (RANS) equations are solved by employing ANSYS FLUENT 15. The turbulence modulation has been performed using the standard 𝑘-𝜀 turbulence model as well as two modified ones, namely as Crespo model, and El Kasmi model. Moreover, the full rotor method as well as the Actuator Disk Method (ADM) have been considered for the turbine rotor modeling, and the prediction of the wake behavior from both approaches is compared. The three types of the Atmospheric Boundary Layer (ABL) flows, which are neutral, stable and unstable regimes, are tested in the present work. In order to account for the buoyancy effect, two different methods are used and compared. In the first one (direct method), the energy equation is considered along with mass, momentum, and turbulence model equations. However, the stratification in the second method (indirect method) is virtually included by means of additional buoyancy production and dissipation terms, added to the turbulent kinetic energy and dissipation rate equations, instead of solving the energy equation. The effects of some operating conditions, including stability degree of the ABL, the inflow turbulence intensity level, and the inflow wind speed are studied and discussed. Furthermore, the amount of the wake deficits, as well as the available wind power in the wake region at different downstream positions from the turbine have been specified for the different of the mentioned operating conditions.
It has been found that, the predicted wake velocity field using the full rotor method, along with the standard 𝑘-𝜀 turbulence model, agrees well with the available experimental data from the literature. Also, the results of the ADM fairly agree with the available experimental measurements when using El Kasmi turbulence model, except a slight overestimation of the wake velocity distribution in the near wake region for some cases. It has been also reported that, there is no significant deviation between the direct and indirect approaches in simulating the thermally-stratified wake flow for the three
ii
implemented turbulence models. However, El Kasmi model is the best among the adopted models in the wake predictions when using the ADM. Hence, the indirect method, combined with El- Kasmi model, have been employed to study the impact of the operating conditions on the wake flow characteristics. It has been concluded that, the wake recovers faster when the atmosphere tends to the unstable conditions, while the stable cases delay the wake recovery. Additionally, the wake decay rate increases as the degree of the ABL stability decreases for both the stable and unstable ABL. Further, the wind power in the wake region increases with the downstream distance, and it also increases with the decreasing of the stability degree of the atmosphere. The upstream flow turbulence level affects also the wake flow structure for the two atmospheric regimes and its effect is more significant in the far wake region compared to the near wake region. As the turbulence intensity increases, the wake decays faster and the available wind power in the wake region increases. It has been also observed when testing the effect of the inflow wind speed, represented by hub height velocity, that, the wind power available in the wake region increases at the higher values of the flow velocity upstream the wind turbine. This strongly affects the implementation distance of the downstream turbines.