Wind Farm Dynamics over a Diurnal Cycle: Analysis of a Comprehensive Large Eddy Simulation, Web-Services Accessible Dataset

TL;DR

This study uses a large-eddy simulation with a concurrent precursor inflow and a 1D soil conduction boundary to model an 8-turbine wind farm over a full diurnal cycle, avoiding prescribed surface temperature or heat flux. The resulting dataset, integrated into the JHTDB-wind platform, enables analysis of how wakes modify thermal fields and turbine power during evening and morning transitions. Key findings include night-time surface warming downstream of the farm under stable stratification and a morning-transition anomaly where downstream turbines can temporarily outpace front-row turbines due to enhanced wake mixing and blockage effects. The public dataset, including 3D fields, soil temperature, turbine forces, and rotor-blade data, facilitates broad future investigations into ABL–wind-farm interactions with high fidelity boundary conditions.

Abstract

The atmospheric boundary layer undergoes significant changes throughout a diurnal cycle, affecting wind turbine performance and wakes in wind farms. Wind farm Large Eddy Simulations (LES) under such conditions provide rich datasets to study the underlying dynamics and identify important trends. Here, we describe a comprehensive open dataset generated using LES of an 8-turbine wind farm consisting of four rows of two turbines. To avoid specifying either prescribed surface temperature or heat flux, a local 1D soil heat conduction model is used with time-periodic solar surface heating, coupled to LES. After several days of low-resolution LES, an approximately time periodic behavior is achieved, after which high-resolution LES is continued during a 24-hour period. Analysis of the LES data reveals that wind turbine wakes have a significant impact on the temperature field and spatial surface heat flux patterns and exhibiting increased surface temperature behind the wind farm at night under the specific conditions of the simulation (dry unvegetated soil, clear sky). We observe that for a few morning hours the first row of wind turbines generates less power compared to the last row. Detailed analyses of the data using innovative web-services facilitated data access tools reveal that during the morning transition, the presence of a low-level jet and the wind farm blockage effect combine to cause cooling and a reduction in wind speed at hub height upstream of the wind farm. In addition, larger turbulence levels exist downstream in the wind farm, explaining the larger power production of downstream turbines.

Figures (31)

• Figure 1: Schematic representation of the computational simulation domain (not to scale), showing: (a) top view ($x$–$y$ plane) and (b) front view (x-z plane). The precursor computational domain consists of the regions denoted as "$P$", the precursor mapping region "$P_M$", and the precursor spanwise shifting region "$P_S$". The wind farm computational domain includes the windfarm region "$W$" and the fringe region "$W_F$" near the outlet. Both precursor and windfarm computational domains include a Rayleigh damping region at the top (region $R$"). The turbine diameter $D = 126$ m and hub height $z_h = 90$ m are also indicated.
• Figure 2: Ingredients of the soil conduction model used in LES. (a) Sketch of the thermal energy partition of radiative heat-flux at the ground surface partitioned into thermal conduction in the soil and thermal convection in the air. (b) Time series of imposed radiative heat-flux with zero mean. (c) Illustration of the 1D grids used to solve the 1D heat conduction equation at every LES grid point separately. Horizontal conduction is neglected as justified by the thermal penetration depth during a 24 hour cycle being much thinner than the horizontal grid spacing in the LES.
• Figure 3: Time series of three quantities from the coarse-resolution simulation in the precursor domain during the initial five diurnal cycles, exhibiting convergence towards 24-hour periodic behavior. (a) Horizontally averaged wind direction relative to the streamwise direction in the precursor domain. The orange line shows the results from the coarse resolution, the green line shows the results after switching to fine resolution and ADM to ALM wind turbine modeling, while the solid blue line shows the result for the last 24 hours using fine resolution, starting at 15:00 of day 5 when data collection takes place. (b) Horizontally averaged heat flux into the air in the precursor domain from the precursor and coarse resolution run. (c) Horizontally averaged ground surface temperature in the precursor domain during the coarse resolution run.
• Figure 4: Schematic representation of the database domain. It represents the physical domain available in the database, merging the precursor domain ($P+P_M$) up to the end of the mapping region at $3/4 L_x$, with the windfarm domain ($W$) but excluding the fringe region. Turbines are numbered from 1 to 8 as shown. The domain dimensions are $1.625 \, L_x$ (streamwise) = $3L_{x}/4$ (precursor) + $7L_{x}/8$ (wind farm), $L_y$ (spanwise), and $L_z$ (wall-normal).
• Figure 5: Snapshots of x-direction ($u$) velocity snapshots at four times during the diurnal cycle, combining fields in both the precursor and wind farm domains. (a) Mid afternoon (15:00 hr). (b) Late afternoon (18:00 hr). (c) Evening (22:00 hr). (d) Early morning (06:00 hr). The boundary between the precursor and wind farm domains is located at $x=5{,}953.5$ m, with seamless transition of velocity fields between the two domains. The bottom plane shown is at height $z = 70\,\text{m}$. The arrow at the top indicates the imposed geostrophic wind $G$, oriented in such a way that the daily average flow direction at hub height is in the $x$ direction.