Implementing conservation of mass to wind flow field solvers to bridge the analytical wake model physics gap

Currently, full computational fluid dynamic (CFD) simulations are required to simulate fluid flow through wind farms. These CFD simulations are computationally expensive and therefore limit wind farm optimization analysis. Conversely, analytical wake models that model turbine wake deficits have been developed and implemented into industrial-grade software over the past several years, such as FLORIS from NREL. These models allow for extremely fast flow and power estimation for different wind turbine configurations, making them ideal for optimization and agile decision-making processes. Unfortunately, the resulting fields of these models do not necessarily conserve mass, nor momentum. This study investigates the impact of imposing mass conservation as a first-order improvement to analytical wake models on flow fields and power estimates. To do so, a rapid mass-conserving workflow is used, where results are compared to those obtained both through a full scale Large-Eddy Simulation (LES) approach, and through the superposition of analytical wake flow models. Results show that the newly tested mass-conserving approaches generate flow fields and power estimates more similar to those obtained with LES while simultaneously maintaining simulation times on the order of seconds. The comparatively low computational cost of the workflow highlights the potential of incorporating additional physics into analytical wake models while maintaining their computational efficiency.