The DOE A2e Mesoscale to Microscale Coupling Project

The Department of Energy Atmosphere to Electrons (A2e) program seeks to better model the flow physics that impact the energy produced by wind farms. To that end, beginning in FY2015, they initiated a Mesoscale to Microscale Coupling (MMC) project to look at the best methods to model the interface between the mesoscale systems and the microscale effects of the local wind farm. The purpose of the MMC Project is to develop, verify, and validate physical models and modeling techniques that bridge the most important atmospheric scales that determine wind plant performance and reliability. The current generation of tools is insufficient for this purpose and this project has brought together a team of subject matter experts to address modeling gaps. This project has involved six DOE laboratories (Argonne Lawrence Livermore, Los Alamos, National Renewable Energy Lab, Pacific Northwest, and Sandia National Laboratories), plus the National Center for Atmospheric Research.

For the first year of the project (FY15), the team was charged with: • Establishing metrics as well as best practices for coupled simulation, • Validating various mesoscale and microscale simulations for different atmospheric boundary layer states, to include convective, neutral, and stable conditions, • Setting up a validated model for inflow conditions at the Sandia Scaled Wind Farm Technology (SWiFT) test facility in preparation for wind plant validation activities, • Providing DOE with recommendations on the merits and constraints of the models used in the code comparison, highlighting potential candidates with superior capabilities.

The team worked together over the course of six months to establish the metrics and select canonical cases of stable, neutral, and convective conditions at the site. It was found that neutral conditions are rare at the SWiFT site, with most of those conditions occurring in the transitions from convective to stable at dusk or the transition from stable to convective in the morning. Stable conditions are predominant and produce much of the available wind power, although they frequently exhibit low level jets that typically produce a shear across the rotor blade, likely decreasing the power output. The summer daytime conditions are dominated by convective conditions that exhibit a deeper well-mixed boundary layer.

For the purposes of assessing the current-state-of-the-science of mesoscale and microscale modeling for canonical stable, neutral, and convective cases, the MMC team exercised a series of models and provided a formal assessment of their performance in comparison with the SWiFT tower and wind profiler data. The mesoscale modelers studied the sensitivity to boundary layer schemes, boundary and initial conditions, and use and type of data assimilation for nearby data. The microscale modelers focused the initial sensitivity studies on the convective and neutral cases, studying sensitivity to grid spacing, the geostrophic wind speed, roughness height, aspect ratio, order of the advection scheme, and the subgridscale turbulence model employed. A major finding is that when tuned for the cases, the models performed similarly and the differences between models was smaller than the uncertainty in the SWiFT tower data that was being used for comparison. It was also found that it is quite difficult to set up a high quality stable boundary layer case with the appropriate characteristics.