![]() Physical models are far more reliable than data-based ones when cause-effect relationships must be determined. Overall, physical models are a valid option for better understanding the inner workings of turbine components and generating new knowledge about them. The works presented in 7, 8 attempt a different approach by first determining the thermal network that describes gearbox conditions. Such work requires a dynamic study of the gear conditions and Finite Element Method simulations. For instance, in 6, a physical simulation of the loads of a turbine gearbox is proposed, showing that it can determine the effect that varying loads have on the component’s lifespan. Nonetheless, such models have been presented in various studies in the literature. Sometimes, even WT owners do not have all the required information, as manufacturers do not always share the details of the turbine’s inner systems. ![]() Building physical models of wind turbines requires deep knowledge and expertise in operation principles and a deep knowledge of the WT components. Monitored components are modelled into systems of physical equations that describe their behaviour from a thermodynamic, electrical, or mechanical perspective. The physical model approach is useful for determining and capturing how the various components of turbines work. The early deterioration of WT’s systems and subsystems can be detected using their physical models or building models from their generated data. Adopting effective methodologies and tools that assist in this process can significantly benefit wind farm owners by increasing energy production, availability, and cost savings. Identifying the root causes of failures leading to turbine downtime is essential in reducing inactivity and promptly addressing critical failures. WF operators have adopted a wide range of measures to extend the operative time of their assets, as mentioned in 5. This is particularly relevant for those turbines that have been installed in the 1990s and early 2000s that are approaching the end of their lifetime. The life expectancy of WT is commonly estimated at around 20 years, and on average one week of downtime per year is required due to maintenance 4. Preventive maintenance schedules for wind turbines are insufficient to detect and predict device conditions and anticipate failures. The unexpected shutdown of turbines incurs substantial costs, especially considering the logistical challenges of remote locations and the time required for component replacement and on-site repairs. However, relying solely on preventive maintenance is insufficient to detect and predict device conditions and anticipate potential failures. To ensure effective management of wind farms (WF), wind turbines (WT) are scheduled for preventive maintenance every 2500 to 5000 hours. ![]() By reducing these costs, wind farms can become more competitive with fossil fuels and expedite the transition to sustainable energy 3. However, the operation and maintenance (O&M) costs of wind farms, which range from 10% to 35% of overall generation costs, pose a challenge 2. Wind energy is essential for meeting the EU Commission’s ambitious climate and energy targets for 2020, which include generating at least 20% of electricity from sustainable sources 1. By using this normality modelling approach, it is possible to detect rotor malfunction when the estimate differs from the actual measured value. In this example, different gearbox variables are selected to estimate a target variable to detect whether or not the estimate differs from the actual value provided for the sensor. To demonstrate the usefulness of this database, an illustrative example is given. Finally, a set of functions to download specific subsets of the whole database is freely available in Matlab, R, and Python. The database also contains the alarm events, indicating the system and subsystem and a small description. The reported values for each sensor are minimum, maximum, mean, and standard deviation. The database contains 312 analogous variables recorded at 5-minute intervals, from 78 different sensors. With the aim of helping researchers to develop intelligent operation and maintenance strategies, in this manuscript, an extensive 3-years Supervisory Control and Data Acquisition database of five Fuhrländer FL2500 2.5 MW wind turbines is presented.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |