Given the enormous capital value of power transformers and their integral role in the electricity network, increasing attention has been given to diagnostic and monitoring tools as a safety precaution measure to evaluate the internal condition of transformers. This study overcomes the fault diagnosis problem of power transformers using an ultra high frequency drain valve sensor. A convolutional neural network (CNN) is proposed to classify six types of discharge defects in power transformers. The proposed model utilizes the phase-amplitude response from a phase-resolved partial discharge (PRPD) signal to reduce the input size. The performance of the proposed method is verified through PRPD experiments using artificial cells. The experimental results indicate that the classification performance of the proposed method is significantly better than those of conventional algorithms, such as linear and nonlinear support vector machines and feedforward neural networks, at 18.78%, 10.95%, and 8.76%, respectively. In addition, a comparison with the different representations of the data leads to the observation that the proposed CNN using a PA response provides a higher accuracy than that using sequence data at 1.46%.

In this paper, the amount of SF₆ decomposition gases due to the partial discharge (PD) was studied in the SF₆ gas-insulated transformer. The long-term PD degradation experiment was performed while controlling the discharge magnitude using the surface discharge, and the gas generation amount was measured by using gas chromatography for SO₂F₂, SOF₂, SO₂, CO, and CF₄. In addition, to investigate the relationship between the partial discharge energy and the decomposed gas generation amount, partial discharge energy was calculated by a data processing program and converted to the unit of joule per mole. With the finite element method (FEM), the electric field distribution and SF₆ gas decomposition mechanism were explained for the partial discharge energy effect on the gas generation. This study helps understand the relationship between the partial discharge energy and the decomposed gas generation ratio with the experimental results and can be used for the diagnosis of PD and maintenance process for the gas-insulated transformers.

Remanufacturing has emerged as a way to solve production problems, as raw material costs increase and environmental pollution caused by discarded equipment occurs. The process can extend product lifetime and prevent waste of resources. In particular, it has economical efficiency for large equipment such as GIS (Gas Insulated Switchgear). The crucial points in remanufacturing are determining replaceable parts and economic valuation. To address these issues, we propose a framework for remanufacturing GIS with remaining lifetime prediction. We construct a regression model for remaining useful life (RUL) in the proposed framework using GIS sensor data. The cost of the replacement parts is estimated with the selected sensors. To validate the effectiveness of the proposed framework, we conducted accelerated life testing on a GIS for data acquisition and applied our framework. The experimental results demonstrate that the tree-based RUL regression model outperforms the others in prediction accuracy. In the simulation of part replacement, the important sensor-based decision-making improves RUL significantly.

The increase in renewable energy and the development of Silicon Carbide (SiC) and Gallium Nitride (GaN) power switches have led to higher voltage magnitudes, slew rates, and switching frequencies, imposing greater electric stress on insulating materials. However, it is hard to recognize the partial discharge signal from background noise due to electromagnetic interference when diagnosing electrical equipment with high switching frequency. Therefore, we numerically analyzed negative corona discharge under a superposition of direct current (DC) and pulse-width modulation (PWM) voltage, with a switching frequency of 80 kHz in a needle-plane geometry varying rising times of 100, 200, and 500 ns. The analysis model was coupled with Poisson’s equation and the drift-diffusion model. We considered three types of charge carriers—electrons, positive ions, and negative ions—and the generation and loss of each charge carrier. The discharge current was calculated from Poynting’s theorem. The discharge current pulses were regularly sustained, like the Trichel pulses in the previous research. However, the periods between pulses and the magnitude of the pulses changed depending on the applied voltage and rising time. With the increase in rising time, both the current pulse peak value and the pulse period decreased, affecting the charge distribution, especially for the ionization region and negative ions.