Novel Method for Estimating Furfural Content in Transformer Insulating Oil Using Spectroscopic Analysis and Pattern Recognition

Abstract

The transformer is the most essential component in electrical power transmission. Once a transformer is installed, it is generally used for decades. When failure occurs in a transformer, serious problems can arise, and numerous electrical components that depend on the transformer may be affected. Thus, deterioration diagnosis of transformers is important. The furfural content in the insulating oil of a transformer is widely used as an indicator of transformer deterioration, as furfural is generated by decomposition of the insulating paper. Although the degree of transformer deterioration can be estimated by measuring the furfural content, this measurement is time-consuming. This paper proposes a novel method for estimating the amount of furfural contained in insulating oil within a short time. The method is realized by combining spectroscopic analysis in the visible region with a pattern recognition technique. The effectiveness of the proposed method is verified through a series of experiments.

Development of portable deterioration diagnosis device for oil filled transformers based on near‐infrared spectroscopic information

“…Diagonal plots for patterns ①-⑤ are presented in Figure 7. In addition, Figure 7 shows the results of furfural content estimation in sampled oil using mid-infrared spectroscopy 26 (marked as Literature ( 26): IR). Here the horizontal axis plots furfural content measured by HPLC, and the vertical axis plots estimated values.…”

“…18 When estimating furfural content in the mid-infrared range, 44 wavenumbers were distributed among 4 bands. 26 As mentioned above, overtones and combination tones of a molecule's fundamental vibrations are observed in near-infrared spectroscopy, and correlation is difficult to examine when too many wavenumbers are used; thus, in this study, the correlation coefficient threshold was adjusted so that the number of wavenumbers was around 20. Now, further examination is necessary to make sure that no significant wavenumbers were excluded in an attempt to narrow down the explanatory variables.…”

Evaluation method for insulation degradation of power transformer windings based on incomplete internet of things sensing data

“…Degradation evaluation methods relying on traditional indicators [4] Proposed a method based on spectral analysis and pattern recognition using furfural content as an index Cannot be applied to online deterioration evaluation [5] Investigated a method based on the methanol content [6] Explored the correlation between water content in oil data and transformer degradation [7] Developed a model based on feedforward neural networks using the degree of polymerization as an indicator [8] Studied the impact of electrical conductivity on insulation aging Degradation evaluation methods relying on a single type of IoT data [10] Proposed a method based on statistical indices of partial discharge Lack of consideration for multiple degradation factors and data incompleteness [11,12] Proposed two methods to eliminate partial discharge in transformers by preparing nanofluid to absorb gases such as acetylene in oil [13] Proposed a degradation prediction model based on temperature data [14] Investigated the method based on leakage current [15] Studied the electrical damage trend by incrementally increasing the voltage Consider multiple degradation factors but rely on traditional indicators [17] Constructed a dynamic model under the influence of electrical and thermal stress Cannot be applied to online deterioration evaluation [18] Explored the changing trends of indicators under thermal and mechanical stresses [19] Investigated the trends of the indicators under electrical, thermal and mechanical stresses [20] Studied the method based on tensile strength and dielectric constant under thermal-mechanical stresses Data completion for a single type of IoT sensing data [21] Completed voltage data using deep learning and unscented Kalman filtering Lack of consideration of spatiotemporal correlation between multiple IoT sensing data [22] Investigated data filling method in photovoltaic power using recursive long short-term memory network [23] Used the normal distribution method for filling power data of smart meters [24] Proposed a filling method for household load data based on noisy interpolation Abbreviation: IoT, Internet of Things.…”

“…Traditional approaches for transformer winding insulation degradation evaluation have primarily focused on utilizing the inherent trend-type degradation parameters specific to oil-paper insulation. For example, Oshima et al proposed a winding insulation degradation evaluation method based on spectral analysis and pattern recognition, using furfural content as a degradation index [4]. Similarly, Chen et al investigated the degradation eval-uation method of oil-paper insulation by analyzing the methanol content in the oil [5].…”

“…For example, Oshima et al. proposed a winding insulation degradation evaluation method based on spectral analysis and pattern recognition, using furfural content as a degradation index [4]. Similarly, Chen et al.…”