Posters

Abstract

The linear flow model of WAsP is the industry standard for wind resource assessment, however, its theoretical scope is limited to simple terrains, where rugdness index (RIX) is nearly zero. When this model is applied to complex terrain, flow separation occurs and results in significant prediction errors.  Previous research (Mortensen et al., 2006) identified a log-linear relationship between these errors and the orographic performance indicator ΔRIX , proposing a simple yet efficient empirical correction. The relation obtained was derived from a single case study of five sites and lacked broad validation. Therefore, this study will address that gap by establishing a large-scale validation data pool to refine and extend the ΔRIX correction method, aiming to reduce the uncertainty in complex terrains.  The study utilized an extensive industrial dataset of wind measurement stations situated in complex terrains, provided by USENS Energy Solutions. Which will follow three validation steps. Firstly, cross-predictions are performed for stations in the same site, and the logarithmic wind speed error "ln(Up/Um)" and Weibull parameters (A and k) are regressed against  to validate and extend the correction proposed by (Mortensen et al., 2006). Then, a sensitivity analysis is performed on the  calculation itself by fine tuning the critical slope θc between 0.3 to 0.5 to find the value that maximises the correlation across the data pool. Finally, to address Measnet and FGW TR6 requirements regarding representativeness, the prediction residuals were analysed as a function of categorized horizontal (e.g., 0–3.5 km vs. >3.5 km) and vertical separation distances, which will decouple.  The initial hypothesis is supported by preliminary results of the expanded data pool, which states that the ΔRIX indicator accurately predicts the sign and magnitude of WAsP errors in complex terrain. The log-linear correction will be tested if it holds up well in a variety of topographies, possibly lowering prediction uncertainty by amounts comparable to around 70% improvement observed in the initial Portuguese case study. Additionally, preliminary sensitivity testing indicates that the ideal critical slope θc for characterising flow separation might be greater than the conventional WAsP default of 0.3, consistent with the 0.4–0.45 range noted in previous studies. A distinct "validity radius" where this correction works is anticipated to be defined by the distance analysis. Additionally, for a subset of highly complex sites, WAsP results are benchmarked against CFD simulations to isolate errors strictly attributable to the linearisation limit versus those inherent to the measurement campaign. By measuring prediction errors across a large data set, this study extends the WAsP operational envelope by converting the ΔRIX correction from a site-specific observation into a validated, generalised process. Ultimately, this research seeks to identify if additional geometric components beyond standard (RIX) can be integrated to further linearise the prediction error.