Capacitive high‐voltage bushing is one of the typical oil‐immersed paper insulation equipment, which is an indispensable external connection component for power transformers. In order to evaluate the effectiveness of frequency domain spectroscopy (FDS) for diagnosing moisture content of oil‐immersed paper insulation bushing, this article studied the frequency‐domain dielectric characteristics of oil‐immersed paper insulation samples and field bushing. At the same time, the basic principle of interfacial polarisation is used to analyse the samples and field bushing with non‐uniform moisture content. The results show that the moisture content affects the whole frequency band of the FDS curve. When the distribution of moisture content is non‐uniform, the paper with high moisture content has a great influence on the whole FDS curve of oil‐immersed paper insulation. In addition, the FDS curves will show a significant loss peak. The frequency of the peak point gradually moves towards the high frequency as the non‐uniformity of the moisture content distribution increases. The experimental results show that the variation law of oil‐immersed paper insulation can be used to quantitatively evaluate the moisture content and moisture type of oil–paper insulation power equipment.

Epoxy‐impregnated paper transformer bushings are prone to water absorption due to the presence of water‐attracting moieties in the cellulose chemical structure and epoxy resin, leading to moisture‐induced insulation defects. To understand the mechanism of moisture's impact on epoxy‐impregnated paper, this study developed a testing platform that integrated moisture diffusion, adsorption, and dielectric properties of insulating materials. Insulation specimens made of epoxy resin and epoxy‐impregnated paper were prepared and their water diffusion characteristics were investigated using the Fick diffusion model and the Langmuir model, respectively. The Guggenheim–Anderson–de Boer equation was used to examine the vacuum vapour adsorption characteristics of the materials, providing a theoretical foundation for examining the form of water present within them. The dielectric response of materials with varying moisture levels was then tested to study the effect of water on the dielectric spectrum curve. Using the Extended Derivative Method, the characteristic frequency reflecting the water content of the epoxy‐impregnated paper was extracted. The results demonstrate that 0.01‐Hz characteristic frequency provides superior accuracy for moisture assessment in epoxy resins compared to 50 Hz, and extracting and fitting the relaxation peak characteristic frequency in epoxy‐impregnated paper moisture evaluation yields higher accuracy than using tan δ at the power frequency.

Cellulose insulating paper is widely used in the power industry for its good electrical insulating properties. Moisture sharply increases its conductivity, which directly leads to the weakening of insulation performance and greatly increases the risk of subsequent electric field distortion and insulation breakdown. This paper focuses on the microscopic mechanism of moisture changing the characteristics of charge transport in cellulose insulation and attempts to reveal the related conductivity mechanism. To achieve this purpose, microscopic and macroscopic perspectives are integrated and several simulation and experimental methods are utilized comprehensively. The molecular dynamics simulation results showed that most water molecules in damped cellulose were individually and uniformly adsorbed on the hydroxyl groups by hydrogen bond, and the quantum chemistry computation results showed that the lowest unoccupied molecular orbital more appeared on the water molecule and the corresponding density of state increased. Then, experimentally, it was confirmed that the trap energy level decreased by the thermally stimulated current method. On this basis, the promotion effect of moisture on charge transport is predicted and verified by polarization and depolarization current methods. As the moisture content increased, more charge carriers escaped from the trap by hopping and participated in long-range continuous charge motion. Therefore, after dampness, the current of cellulose insulating paper increased exponentially with the increase in electric field strength, which was consistent with the hopping conductivity mechanism.

The mechanical and dielectric properties of transformer insulating paper are key factors that require close attention to ensure optimal operation. In this study, a novel approach of enhancing properties by nanocellulose modification was proposed. To investigate the effect of doping, insulating paper with different doping mass fractions of cellulose nanocrystals (CNCs) was prepared, and multiple properties of the sample were characterized. By applying a doping concentration gradient, the non-monotonic trend in performance with doping levels was revealed, and the impact of doping concentration on the data variability was investigated. The effects of CNC doping on the surface pore size of the samples were analyzed by SEM (scanning electron microscopy) and a mercury intrusion method, and the regularity of CNC dispersion uniformity with doping concentration was investigated. The Brunauer, Emmett, and Teller model was employed to fit the water vapor adsorption isotherm data, providing an analysis of the impact of doping on hydrogen bond connections and the moisture adsorption properties. Tensile strength and Young's modulus were measured and analyzed using the Page model and the Halpin–Kardos model. The tensile strength of the 6 wt. % CNC-modified paper was increased by 6.91%. 10 wt. % CNC-modified paper had a Young's modulus increase of 11.98%. 3 wt. % CNC-modified paper has a 28.53% increase in the AC breakdown field strength and an 8.72% increase in the DC breakdown field strength. The influence of CNC on the dielectric properties of the insulating paper was discussed using the Havriliak–Negami (H–N) model. The results indicate that the introduction of CNC can effectively enhance the comprehensive performance of oil-immersed insulating paper.

Dielectrically graded insulation (DGI) with spatially inhomogeneous dielectric parameter distribution can provide extraordinary insulation performance for high-voltage power equipment. The most crucial objective in the application of DGI is to identify the optimal distribution of dielectric parameters. However, the calculation efficiency and electric field optimization results of the existing methods are poor under complex insulation structures. Also, the dielectric loss of DGI calculated by these methods is relatively large. In this work, an efficient DGI parameter optimization method was developed. Firstly, a variable density topology optimization model was proposed to improve the electric field uniformity of the power module and suppress the electric field stress at the triple point. Secondly, the influence of different model parameters on the mitigation degree of the electric field was studied to obtain the optimal parameters. Finally, to reduce the dielectric loss of the insulation, the DGI micro-region was constructed by narrowing the optimized region. Numerical results indicate that the model only needs 23 iterations to obtain the optimal result. After optimization, the electric field stress at the triple point is reduced by 71.7%. DGI micro-regions reduce dielectric loss by 85.1% compared to global DGI. This study shows that the proposed method is anticipated to provide a new route for the preparation and application of DGI.

The insulating paper in oil-paper insulated transformers was subjected to significant mechanical vibration and mechanical shock, putting considerable demands on the insulating paper’s strength and resistance to deformation. Therefore, effective enhancement and accurate prediction of mechanical properties are crucial in the application of cellulose insulating paper. In this paper, modified insulating paper with different mass fractions of cellulose nanowhiskers, nanofibrillated cellulose and microfibrillated cellulose were prepared. The microscopic morphology, stress-strain curves and Young’s modulus of the modified insulating paper were measured. The modification effects of micro/nano cellulose with different aspect ratios on the mechanical properties of the insulating paper were compared. 10 wt% mass fraction of cellulose nanowhisker modified paper exhibited the highest Young’s modulus with an enhancement of 11.99%. The mechanism of micro/nano cellulose modification on insulating paper was analyzed in terms of its strength, morphology, interface, and agglomeration. The prediction of Young’s modulus of modified insulating paper under different doping concentrations was realized based on the Halpin-Tsai model for filler modification under ideal conditions, considering the effects of agglomerated and interfacial phases, on the basis of detailed experimental results and comprehensive analysis. Experimental validation showed that the error of the proposed prediction model was less than 1.57%. This study was very beneficial for nano-modification and mechanical property enhancement of cellulose paper.