Posters

Abstract

Real-weather, LES-based simulations are becoming increasingly utilised for high fidelity, late-stage Annual Energy Production assessments (AEPs). Advances in GPU hardware and highly efficient solvers have significantly reduced the computational cost of LES, enabling wind farm simulations to capture detailed flow dynamics. However, budget constraints typically restrict LES runs to 1 year of data, usually executed late in the project lifecycle near the Final Investment Decision (FID). This limitation prevents early-stage project phases from fully benefiting from LES insights. To address this challenge, RWE and Whiffle are developing a cost-effective LES-based approach designed for early-stage AEPs. This approach leverages a selectively-averaged model (SAM) approach, using a limited number of strategically chosen LES days that statistically represent the long-term climate. Two LES simulations of N sampling days are ran, i) wind resource only and ii) wind turbines present with full aerodynamic losses. The difference of these two simulations provides an turbine interaction loss and AEP efficiency. SAM runs provide high-resolution, real weather-based data, including wind characteristics, energy production, aerodynamic losses, and additional meteorological variables (e.g., turbulence intensity), on a 100m horizontal grid for a statistically representative period. The presentation will demonstrate the methodology behind this innovative approach, explore its sensitivity to the number of (N) sampling days, and provide a comparison with industry-standard AEP methodologies, including both LES- and non-LES-based Wind Resource Grid methods. Additionally, a case study of a wind farm project will highlight the product’s application and benefits in a real-world context. This work showcases how the SAM approach enables tailored, cost-efficient solutions, enhancing decision-making from early project development to FID.