Subsoil stress reconstruction for fatigue monitoring of offshore wind turbine using accelerometers on the tower

Subsoil constraints are seen as a black box. Correct modeling of soil dynamics is still part of research and widely used methods do not necessarily apply to upscaled turbines (e.g. py curves [1]). However, direct measurements are not common since expensive and subsoil sensors being difficult to mount and maintain. Still information about subsoil stresses are crucial with fatigue critical welds often found below mudline. This paper presents a combined strategy. Starting from subsoil strain measurements on three operating offshore wind turbines located in the Belgian North Sea much is learned about characteristics of subsoil stresses. Sensor data revealed a high impact of wave load for welds low on the monopile as well as notable contribution of the second order mode on fatigue damage below the mudline. In a second step subsoil stresses are estimated rather than measured. A virtual sensing approach is used to extrapolate measurements from tower accelerometers and one level of strain gauges mounted to the TP to areas close to the seabed. This way stress histories of all fatigue hot spots can be reconstructed, enabling to accurately assess consumed fatigue life. A validation of this concept[2] is already performed for virtual sensors above the water level using accelerometers and strain gauges on the tower. It can be shown that despite uncertainties of dynamic properties a good estimation quality can be achieved for sensors over the entire length of the monopile.