Solar-Power Energy Supply for Wind Turbine Blade Sensors in Northern Climate

Sammanfattning

Wind power is a growing industry and plays a central role in the green transition. In colder climates, ice formation on wind turbine blades is a common and serious issue. Ice reduces efficiency, increases wear on the blades, and poses a safety risk when it detaches. Several de-icing techniques have been developed to address this, but they are often energy-intensive. It is therefore crucial that de-icing is only activated when necessary, making efficient ice detection systems essential. A literature review revealed that existing research in this area primarily focuses on data collection and signal processing, while energy supply and consumption are often treated superficially. Furthermore, most studies are conducted in regions closer to the equator, missing the unique conditions and challenges present in polar regions. Consequently, this project focused on the energy aspect of ice detection systems – both energy production and consumption. Initially, the PVGIS ERA 5 dataset was used to estimate potential solar energy production. However, due to inaccuracies during midwinter, a custom model was instead developed based on experimental data. Measurements were taken in sunny, cloudy, and shaded conditions when the sun was between -5 and 15 degrees above the horizon. This data was used to build a model estimating daily energy production between two dates for a given location and weather condition. For a 35.3 cm2 solar panel with 25% efficiency, the model showed that energy production in Luleå from October 20, 2024, to February 20, 2025, ranged from an average of 2.4 Wh/day to as little as 24 mWh/day in sunny conditions. Corresponding values under cloudy conditions ranged from 47 mWh/day to 1 mWh/day. An ice detection system was designed, but due to time constraints, only the energy harvesting subsystem was both built and tested. The system’s energy consumption was analysed using component datasheets, followed by a battery level simulation for the same time period in Luleå (October 20 to February 20). The simulation compared daily energy production under cloudy conditions with estimated consumption and losses. The results showed that for an 10 Wh battery, the charge level dropped to a minimum of approximately 91.4% by late January and rose to about 93.3% by February 20. During this period, the system transmitted over 6600 bits of data every five minutes. The conclusion is that while available solar energy during the polar winter is low, it is not negligible and is sufficient to power a low-energy ice detection system. Several uncertainties remain – particularly regarding wireless data transmission and battery efficiency – but the key question is not whether the system works, but how frequently it can be active and how much data it can transmit each day without reaching critically low battery levels. Additionally, it was concluded that self-discharge likely is the largest energy loss, followed by wireless transmission. This meant that a larger battery would not increase the systems safety margins noticeably, and instead the production should be increased by adding more PV cells in order to improve the margins.