Ice detection using thermal infrared radiometry on wind turbine blades

Abstract

Abstract Wind farms are located in areas with a high probability of ice occurrence. Icing involves problems such as energy losses, mechanical failures and downtimes. The priority is to detect icing in order to avoid these problems. Icing detection is a complex procedure for example, low temperatures are not a guaranty of ice formation, and other variables may affect. Different techniques have been recently proposed to detect ice in wind turbine blades. They are mainly based on damping of ultrasonic waves on the blade surface, or measuring the resonant frequency of a probe. But these methods have some drawbacks that may cause the system to fail, e.g. the behaviour of ultrasonic waves in composite materials is difficult to predict due to different fibre orientations, and the ice detection by changes in the resonance frequency could lead to false alarms due to variations in working conditions. This paper takes advantage of the remote sensing techniques to propose a novel approach for icing detection without physical contact. The approach is based on the drastic emissivity change that it is produced over a surface characterized with a low emissivity value when ice appears. An experiment was conducted using a broad-band thermal radiometer and a section of a wind turbine blade. Radiometric temperature measurements were collected over the blade with and without an aluminium foil patch. The piece of blade was cooled down and different scenarios were considered, including frozen with and without ice. This study was completed with a sensitivity analysis of the approach to dust accumulation, accounting for real operation conditions. Results show the feasibility of this technique to detect ice formation and discern between frozen and icing conditions.