The impact of wind conditions on wind turbines

Sammanfattning: The world is facing global warming and the challenge to reduce greenhouse gas emissions. Wind power is a renewable source of energy with no greenhouse gas emissions when operating. Therefore, it could contribute in this challenge. Vattenfall is a Swedish energy company that invests in the wind power business and have many wind turbines in operation. On behalf of Vattenfall this study has been performed with the aim to find how wind conditions affect wind turbines and how it is correlated to amount of alarms, time loss, energy availability and time availability in a wind turbine. Improving this knowledge will be an advantage when making investments in wind turbines and their maintenance. A statistical analysis was conducted in order to examine correlations of wind speed and turbulence intensity with the number of alarms, time loss, energy availability and time availability of a turbine. A case study of Lillgrund, an offshore park in Sweden, was performed since the park layout is tightly spaced and hence interesting in sense of turbulence intensity. Lillgrund suffered to a lot of blade vibration alarms and therefore these alarms were investigated deeper in terms of the wind conditions one hour, one day and one week before the alarm occurred. Four additional parks with other turbine types and manufacturers than Lillgrund’s were also included in the correlation analysis. The amount of alarms per year of each turbine was examined in order to compare this with the failure curve of a wind turbine. The purpose was to see if the wear-out period had started earlier for turbines with certain wind conditions but unfortunately the turbines were not old enough to draw a conclusion. The analysis resulted in positive correlations for high turbulence and high amount of alarms and blade vibration alarms for the Siemens SWT-2.3-93 turbines investigated while there were no clear correlations for time loss, energy availability and time availability. Also, the other turbine types had no strong correlations for the investigated parameters. From the results it can be concluded that there are no strong correlations for the wind conditions investigated and energy availability and time availability. It can also be said that the amount of alarms in Siemens SWT-2.3-93 turbines increases with higher turbulence. Hence, alarms do not influence the energy availability noticeably for this type of turbine.