Evaluation of meteorological measurements made on the nacelle of wind turbines in cold climate

Abstract

Operating wind turbines in cold climate leads to several challenges not encountered in regular climates such as frequent icing. These challenges are amplified by the difficulties associated with accurately quantifying icing without specialized instrumentation. Typically, the performance of wind turbines is monitored by comparing the turbine active power to the available power based on the wind speed measured on the nacelle. However, multiple other factors not related to cold climate can also lead to a reduction of turbine performance. Applying a temperature threshold under which ice accretion is deemed to be able to impact turbine performance is a popular tool to reduce the influence of these other factors. Currently, there is no generally accepted threshold temperature value. Using large data sets from 3 different wind farms equipped with different wind turbine models, this topic was investigated. Following this analysis, it has been found that the optimal strategy to define icing events is to consider that ice may remain on surfaces only if a minimal temperature of 1 °C in the previous 2-h window is observed. This strategy is supported by a temperature bias analysis, a turbine performance analysis and 3 case studies. An ice sensor can also be an efficient strategy to accurately identify icing events. Previous studies raised doubts on the accuracy of ice measurements made on a wind turbine nacelle due to the different conditions experienced by the blade tips. To verify this claim, the present paper provides comparisons of icing periods, defined by the wind turbine performance method, to the detection of an ice sensor installed on the nacelle. It was found that the start of 71 of the 74 identified icing events coincided with ice detection on the nacelle. The three remaining events were minor and could not be linked to the incapacity of detecting ice on the nacelle. In conclusion, sensitive nacelle-based sensors should be considered as suitable options for ice detection on wind turbine nacelle.