Effects of wind climate anisotropy and wake modeling assumptions on wind farm layout optimization

Wind farm layout optimization (WFLO) is commonly performed under simplified wind-climate assumptions that neglect directional variability and wind-speed modeling choices, potentially leading to suboptimal wind-farm configurations. This study presents a genetic-algorithm-based WFLO methodology under anisotropic wind-climate conditions using the Jensen wake model and a cost-normalized objective function. The proposed approach integrates directional wind-rose modeling, empirical and Weibull wind-speed representations, sensitivity analysis of wake decay and thrust coefficients, inflow-direction/grid-alignment effects, and multi-run repeatability assessment. The results demonstrate that realistic directional wind characterization is the dominant factor governing best-found layout performance and wind-farm performance. Compared with the uniform directional assumption, the site-representative wind rose increases annual energy production (AEP) by approximately 60% while reducing the Cost/AEP objective by about 19%. The Weibull model closely reproduces the empirical optimization results, with only about 2.3% difference in AEP and less than 1% difference in the Cost/AEP objective. Sensitivity analysis identifies a favorable wake-recovery regime near k ≈ 0.04, where the best cost-normalized performance is obtained among the tested wake-decay cases. Oblique inflow directions, particularly around 30°–60°, improve AEP by up to 59% relative to axis-aligned inflow cases because they reduce persistent wake superposition. Robustness denotes repeatability under fixed inputs. Layouts are best-found solutions without guaranteed global optimality; directional differences compare separately optimized climate scenarios.