Presentations

Abstract

Accurate and efficient representation of wind-farm wakes remains a central challenge in offshore wind energy, particularly as development progresses toward large, closely spaced wind farm clusters. Interacting wakes across multiple farms introduce complex atmospheric dynamics that are difficult to capture with conventional engineering wake models and are only partially resolved in mesoscale simulations. The multi-scale coupled model (MSC), which combines meso-scale modelling with wake and local induction models, aims to capture these wakes in a fast, reduced-order method. However, its performance remains limitedly validated, especially against comprehensive field measurements of offshore clusters. This study evaluates the capability of the MSC model to reproduce wake behavior across the Borssele offshore wind farm cluster, leveraging the extensive observational dataset collected within the BeNeWakes project, operated by the Netherlands Enterprise Agency (RVO). The measurement campaign integrates fixed, floating, and long-range scanning lidars strategically deployed around and downstream of the cluster, providing high-resolution observations of wind-speed deficits, turbulence intensity, wake meandering, recovery rates, and atmospheric stability over a wide range of spatial scales. To represent realistic offshore atmospheric conditions, output from mesoscale models is used to provide initial and boundary conditions for the MSC model. This approach allows simulations to inherit representative stability regimes, inversion height, and other atmospheric conditions. The MSC model is then applied to a selection of representative periods spanning different seasons, wind directions, atmospheric stability classes, and degrees of cluster-to-cluster wake interaction. Model performance is assessed using a structured validation framework that compares MSC output directly against the multi-platform lidar measurements. Key evaluation metrics include hub-height wind-speed deficits, turbulence intensity levels, wake width and lateral spreading, and wake recovery behavior from in-farm locations to far-wake regions extending several tens of kilometers downstream. By systematically analyzing model skill across contrasting atmospheric and wake regimes, the study identifies conditions under which MSC provides robust agreement with observations, as well as regimes where discrepancies persist and further model development is required. This work constitutes one of the first comprehensive, measurement-based evaluations of MSC performance for wake modelling in a commercial-scale offshore wind farm cluster. The BeNeWakes dataset provides a uniquely rich reference for model benchmarking, while the combination of fixed, floating, and scanning lidar observations enables an unprecedented assessment of wake dynamics across multiple spatial scales. The results are expected to support model developers, wind-farm designers, and operators in selecting and improving modeling tools for offshore applications, particularly in the context of future multi-gigawatt cluster developments, and to demonstrate the value of coordinated large-scale measurement campaigns for advancing validation of atmospheric models.