Kazuo Yamamoto

Many countries are actively promoting the large-scale deployment of wind power generation, both onshore and offshore. However, damage to wind turbines caused by winter lightning has become a growing concern in Japan. Japan has made efforts since an early stage to establish legal frameworks for reducing lightning damage; nevertheless, lightning damage to wind turbines remains a problem that has not been completely eradicated. After a wind turbine has been struck by lightning, it is restarted only after its structural integrity has been verified; however, the current method relies on visual inspection by workers, making accurate and rapid inspections difficult. One approach to solving this problem is to use anomaly detection techniques based on SCADA data. Research is currently underway to implement this approach. However, anomaly detection methods based on SCADA data have been criticized for their limited ability to accommodate multiple operating modes, including de-rated operation. In this study, we propose the “scaled power curve” as a robust feature that is less affected by operating modes, with its effectiveness verified through anomaly detection. This method showed improved anomaly detection accuracy compared to using the original power curve as a feature; moreover, in the present case, the method remained effective under de-rated operation. By using this feature, it is expected that a lightning damage detection model can be developed, contributing to improved availability of wind turbines.

There have been numerous reported accidents of lightning strikes damaging wind turbine blades, which poses a serious problem. In certain accidents, the blades that were struck by lightning continued to rotate, resulting in breakage due to centrifugal force. Considering this background, wind turbines situated in Japan have been mandated to be equipped with emergency stop devices. Consequently, upon detection of a lightning strike by the device installed on the wind turbine, the turbine is promptly stopped. In order to restart the wind turbine, it is necessary to verify its soundness by conducting a visual inspection. However, conducting prompt inspections can be difficult due to various factors, including inclement weather. Therefore, this process prolongs the downtime of wind turbines and reduces their availability. In this study, an approach was proposed: a SCADA data analysis method using an autoencoder to assess the soundness of wind turbines without visual inspection. The present method selected wind speed and rotational speed as effective features, employing a sliding window for pre-processing, based on previous studies. Besides, the assessment of a trained autoencoder was conducted through the utilization of the confusion matrix and the receiver operating characteristic curve. It was suggested that the availability of wind turbines could be improved by employing this proposed method to remotely and automatically verify the soundness after lightning detection.

We have developed and deployed a 3D Interferometer-type Lightning Mapping Array (InLMA) for observing winter lightning in Japan. InLMA consists of three broadband interferometers installed at three stations with a distance from 3 to 5 km. At each interferometer station, three discone antennas were installed, forming a right triangle with a separation of 75 m along their two orthogonal baselines. The output of each InLMA antenna is passed through a 400 MHz low-pass filter and then recorded at 1 GS/s with 16-bit accuracy. A new method has been proposed for finding 3D solutions of a lightning mapping system that consists of multiple interferometers. Using the InLMA, we have succeeded in mapping a positive cloud-to-ground (CG) lightning flash in winter, particularly its preliminary breakdown (PB) process. A study on individual PB pulse processes allows us to infer that each PB pulse process contains many small-scale discharges scattering in a height range of about 150 m. These small-scale discharges in a series of PB pulses appear to be continuous in space, though discontinuous in time. We have also examined the positive return stroke in the CG flash and found a 3D average return stroke speed of 7.5 × 10⁷ m/s.

This article discusses the wideband modeling of an isolation transformer aiming at studying the transference of differential-mode voltage surges from the primary side to the secondary side. A modeling approach based on the independent fitting of the branch admittances associated with the equivalent pi-circuit of the transformer is proposed and validated. The resulting model is described either as a Norton-type equivalent circuit for time-domain simulations or as an equivalent circuit. The proposed modeling approach is used to calculate differential voltages transferred to the secondary side of the transformer for the application of impulse voltages at the primary side for different load conditions. The model performance is shown to be comparable to that of a pair of models based on the fitting of all elements of the admittance matrix of the transformer with a single set of poles, with the advantage of requiring a reduced number of circuit elements. It is shown that the tested transformer has a resonant behavior that amplifies differential-mode voltage surges with frequency content around 205 kHz. Also, the time-domain response of the transformer presents a damped oscillatory behavior in the no-load condition and a damped behavior if resistive loads are added at the secondary side.

There have been many reports of damage to wind turbine blades caused by lightning strikes in Japan. In some of these cases, the blades struck by lightning continue to rotate, causing more serious secondary damage. To prevent such accidents, it is a requirement that a lightning detection system is installed on the wind turbine in areas where winter lightning occurs in Japan. This immediately stops the wind turbine if the system detects a lightning strike. Normally, these wind turbines are restarted after confirming soundness of the blade through visual inspection. However, it is often difficult to confirm the soundness of the blade visually for reasons such as bad weather. This process prolongs the time taken to restart, and it is one of the causes that reduces the availability of the wind turbines. In this research, we constructed a damage detection model for wind turbine blades using machine learning based on SCADA system data and, thereby, considered whether the technology automatically confirms the soundness of wind turbine blades.

Many failures of wind turbine blades and electrical equipment due to powerful lightning strikes have occurred. The lightning-induced wind turbine failures that have occurred in recent years can be classified into two main categories. The first is a mechanical component failure. The second is a failure or malfunction of electric or electronic equipment inside or around the wind turbine due to an increase in the potential of the wind turbine grounding system or the electromagnetic induction caused by the lightning strike. In the latter case, the equipment often fails owing to the increase in the potential of the grounding system, the method of suppressing the potential rise by laying a grounding wire vertically or horizontally to the wind turbine foundation is adopted. We propose an empirical expression for the effective length of counterpoises connected to a large grounding electrode such as a wind turbine foundation, and verify the expression using finite-difference time-domain (FDTD) method.