Vibration-based Ice Detection of Rotor Blades in Wind Turbines—The Industrial Realization of an SHM-System

Abstract

Since the efficiency of wind turbines (WT) is primarily reflected in their ability to generate electricity at any time, the downtimes of WTs due to “conventional” inspections for damage or ice detection are costly and undesirable for WT investors. For this reason, the Wolfel Group has developed a wide range of products with vibration-based SHM systems for damage and ice detection on rotor blades, foundation and tower monitoring (onshore and offshore), load monitoring, vibration reduction systems, etc., to give wind turbine operators the opportunity to reduce the number of WT inspections and to increase the inspection intervals. Since one of the latest and most interesting topics regarding the availability of wind turbines in winter times is fully automatic ice detection, this paper will only present the background of Wolfel’s vibration-based ice detection system IDD.Blade®. First, the great importance of ice detection on rotor blades will be discussed, taking into account the risk of ice throw and the requirements of wind power plant certifiers. In the next step, available ice detection systems known from industrial application will be mentioned, followed by a short presentation of the functionality, the electronics and software components of the IDD.Blade® system. The centerpiece of this paper is the presentation of the principles (algorithms) behind the system. In this context it will be explained how - by means of a fully automatic approach - it is possible to ensure a clear distinction between the structural response due to icing and the “normal” vibration of the blades. The basis for this approach is a combination of operational modal analysis (adapted to deal with nonlinear system behavior and non-stationary time series) and statistical pattern recognition techniques. The latter are mainly used to compensate effects of changing environmental and operating conditions (EOC) on the structural dynamic features extracted from the measured signals. At least some examples of long-time monitoring regarding ice detection at different wind turbines will be discussed. These examples illustrate the validation of IDD.Blade® results also by means of camera pictures. doi: 10.12783/SHM2015/352