Wind turbines are relatively new and complex systems, consisting of a wide range of technologies from mechanical gears and bearings, aerodynamics, electrical generators to composite (fiber reinforced) structures. As the wind turbine fleet ages, damage accumulates that can lead to catastrophic failures and long, expensive operational downtimes. Understanding how and why these damage mechanisms and/or failures occur can help prevent, mitigate and assess the risk of future events.
Failure analysis on any component requires application of forensic techniques. To perform the analysis, a series of questions need to be answered to determine what failed, how it failed, and why it failed. The series of questions are outlined below.
STEP 1: What Component Failed?
Perform Site Inspection: a detailed incident site inspection can determine which of the following occurred: hub or blade detachment, blade fracture, tower damage or collapse, generator fire, etc. Figure 1 shows an example of a catastrophic failure where a blade detached from the hub and hit the tower. Upon site inspection it was discovered that the blade impact dented the tower.
STEP 2: Where Did the Failure Originate?
Fractography: visual and macroscopic examination of the fracture features can pin-point specific areas of interest for detailed destructive and microscopic analysis. Recently, we were retained to provide analysis of a wind turbine failure where the main shaft fractured causing the hub and blades to fall to the ground. Figure 2 shows an example of macroscopic inspection of the shaft’s fracture surface that pin-pointed the fracture origin area. This area was then cut-out and inspected in a Scanning Electron Microscope (SEM).
STEP 3: What Was the Driving Failure Mode?
Microscopic analysis: different materials exhibit specific microscopic features at fracture surfaces that provide information on the mode of failure such as tension, compression, torsion due to an overload or fatigue. This process can also reveal the presence of cracking, inclusions, voids, porosity, oxidation and other damage and material degradation features. For example; wind turbine blades are made of laminated composites and cracking can occur between layers, referred as delamination, and at ply termination points (ply drops at resin pockets) and are sometimes only found microscopically as shown in Figure 3.
These 3 steps are based on physical evidence and are the basis for identification of the failure mechanism. To get to the root cause and mitigation of future failures, additional questions need to be answered:
STEP 4: What Was the Component’s Operational Loading & Environment?
The loading conditions such as wind speeds, humidity, lightening, yaw and pitch angles and rotor speed (rpm) can be obtained from operational data from inspection reports, Supervisory Control and Data Acquisition (SCADA), Condition Monitoring System (CMS), weather, etc. This information can be the inputs to component stress and fatigue analysis. The results of such analysis can be compared to the OEM’s design allowable and to industry standards/guidelines such as those from DNV-GL and IEC.
STEP 5: What Was the State of the Component Before Failure?
The design of the component was based on specific geometry and material properties but regularly these are different or have degraded due to exposure to different environments. A careful review of manufacturing procedures, inspection reports, SCADA data, and Balance of Plant (BOP) performance can determine component structural degradation, pre-existing damage or manufacturing defects. Mechanical properties can be tested following testing standards to get actual material properties and assess if the material is compromised. Material degradation or composition can be evaluated via metallography and chemical composition analysis for metallic components while FTIR (Fourier Transform Infrared Spectroscopy), DSC (Differential Scanning Calorimetry) or GC-MS (Gas Chromatography Mass Spectroscopy) analyses can be performed for composite materials. Figure 4 shows a cross-section of a blade root to skin transition area where “waves” resulted from the manufacturing process resulting in large resin pockets at ply drops, which can weaken the overall structure/material.
STEP 6: What Is the Effect of Pre-Existing Damage or Manufacturing Defects?
If damage or defect is found, it is important to understand if and how it will affect the component’s performance. Stress analysis and fracture mechanics that simulate the presence of damage/defect can estimate the effect of these on the component’s stress distribution and strength. An example of such stress analysis is shown in Figure 5 for a damaged wind turbine tower. Following stress analysis, a revised component lifetime can be estimated using various fatigue analysis methods providing the information needed to plan inspection, repair or replacement intervals.
STEP 7: Are These Damage/Defects Persistent Among the Fleet?
Last, but not least, is to evaluate whether this is an isolated event or is it a systematic issue. Depending on the component, damage/defect type, area of interest and accessibility, different non-destructive inspection techniques may be used such as dye penetrant or phased array ultrasonic examination. The key at this stage is the interpretation of these results and a technique’s effectiveness in identifying the damage/defect in other components. Another benefit of such inspections is to gather information of the damage/defect size and specific locations per component/turbine, which allows for lifetime estimation and informed operations and maintenance scheduling.