Offshore wind energy is expanding rapidly to meet global decarbonisation targets, and realising this transition, together with its socio-economic benefits, depends on large capital investments that are only committed when projects can demonstrate a reliable energy yield. Accurate energy yield assessments are therefore central: they secure competitive financing and limit financial risk. Yet standard power performance models typically assume idealised, neutral atmospheric conditions, while real offshore environments feature dynamic atmospheric boundary layers in which thermodynamic stability alters the wind profile and turbulence intensity. This mismatch introduces uncertainty into yield predictions under real-world stratification.

This research investigates how atmospheric stability influences the aerodynamic power performance of a modern offshore wind turbine, and evaluates the extent to which these thermodynamic effects can be decoupled from interconnected kinematic characteristics.

To isolate these aerodynamic impacts, the study processed continuous operational and meteorological data from a commercial wind farm located in the North Sea. The campaign used a forward-looking nacelle-mounted Light Detection and Ranging (Lidar) system to characterise the incoming wind field, alongside thermodynamic sensors to measure temperature, pressure, and humidity. The analysis was conducted within a wake-free sector, applying instrument availability and operational filters to ensure clean, undisturbed free-stream data. The dataset was then categorised into stable, neutral, and unstable regimes using kinematic proxies such as wind shear and Turbulence Intensity (TI), along with thermodynamic metrics, primarily the Bulk Richardson Number. Standardised empirical power curves were then constructed to quantify performance deviations between regimes.

The majority of the analysis relied on the Bulk Richardson Number classification. Because it directly incorporates thermodynamic information, it provided the most physically consistent basis for classification among the methods tested. Wind shear served as a viable low-cost alternative proxy, while Turbulence Intensity was less reliable due to its high volatility. In terms of aerodynamic performance, the data showed clear deviations in the partial-load operational region: unstable, convective conditions were consistently associated with overperformance, yielding an energy surplus of approximately +1.80% relative to the manufacturer's baseline, while stable, highly stratified conditions corresponded to a near-symmetric deficit of about -1.81%. In the rated power region, the turbine's active pitch control absorbed these stability-driven differences, bringing power anomalies close to zero. Seasonal variations in atmospheric stability, along with other aerodynamic phenomena, were also examined to provide a more complete evaluation of the turbine's performance across the year.

The research concludes that atmospheric stability significantly affects offshore wind turbine efficiency, and that standard neutral baseline power curves tend to overestimate energy production in stable offshore environments. The most plausible explanation is that unstable conditions promote strong convective mixing, producing a more uniform wind profile that improves kinetic energy extraction across the rotor swept area. To reduce financial risk, project developers should integrate stability-weighted probability distributions into pre-construction Energy Yield Assessments rather than relying on generic wind distributions. This methodology might also be suitable for resource assessment planning by deploying floating Lidar systems that capture local thermodynamic data. Future research should use vertically profiling Lidars to implement Rotor Equivalent Wind Speed (REWS) corrections and apply machine learning algorithms for continuous, multivariate stability classification. ... Read more