Modal Analysis of OWTs in Sand - Comparison between Measured and Predicted Values-Paper

Paper

Pdf Summary

The paper compares predicted and measured dynamic behavior of offshore wind turbines (OWTs) supported by monopiles in sand. Accurate natural frequencies and mode shapes are important for estimating fatigue loads, but conventional API soil models may underestimate the stiffness of sand for today’s large-diameter monopiles.<br /><br />The study analyzes 15 operating turbines at the Gode Wind 1&2 offshore wind farm in Germany, using measurements averaged over three months in 2016. Predictions from three soil models are compared: the standard API model, a modified API model with increased initial stiffness, and the PISA rule-based model calibrated for Dunkirk sand. Mode shapes are also examined for one instrumented turbine.<br /><br />The standard API model underpredicts measured natural frequencies by 1.6–3%. The modified API model improves the estimates but still underpredicts them by 1.1–2.2%. PISA provides the closest match, with deviations from −0.9% to +0.3%. For the lower section of the tower, the API models overpredict first-mode deflections by 8–14%, whereas PISA predictions are within about 1% of measurements.<br /><br />The results indicate that PISA more accurately represents soil–monopile interaction in sand and predicts both natural frequencies and mode shapes better than the API models. Because soil stiffness affects fatigue-load estimates, API’s underprediction of stiffness can lead to overly conservative designs. More accurate PISA predictions could support reduced fatigue loads and lower structural costs, particularly as turbine sizes increase.

Keywords

offshore wind turbines

monopile foundations

sand soil stiffness

natural frequencies

mode shapes

fatigue loads

API soil model

PISA model

Gode Wind 1&2

soil-monopile interaction