RU2375718C2 - Method for high-voltage insulation diagnostics - Google Patents

Abstract

FIELD: electricity.

SUBSTANCE: invention refers to the field of electric engineering, specifically to testing methods for solid insulation materials, and can be used for predicting service life and recourse of high-voltage insulation. Essence: high voltage is applied to tested insulation sample via spike-plane electrode system. The moment of primary destruction channel origin is recorded. Time to origin of insulation destruction primary channel τ₀ is measured. High voltage is disconnected. Dendrite shaping time τ_d is determined from empiric expression:

time to insulation breakdown τ_br is determined from formula:

where

- module of complex dielectric permeability in the range of optical frequencies under condition of equality of its real ε'_∞ and imaginary ε''_∞ = ε'_∞ • tgδ components, n - refraction index of insulating material, δ - dielectric loss angle.

EFFECT: diagnostics time reduction, reduction in consumption of materials.

4 tbl, 1 ex

Description

The invention relates to the field of electrical engineering, in particular to methods for testing solid insulating materials, and can be used to predict the service life or resource of high-voltage insulation.

A well-known method for diagnosing high-voltage insulation (Koykov S.N., Tsikin A.N. Electrical aging of solid dielectrics. L .: Energia, 1968), which consists in carrying out life tests of real insulating products and calculating the time before breakdown according to the formula:

τ _CR = B · E ^-m ,

where E is the electric field strength; B and m are empirical coefficients determined for each insulating material based on the results of life tests of real insulating products.

However, the implementation of this method is possible only if the experimental values of τ _CR for different E fit on a straight line in double logarithmic coordinates log τ _CR = f (log E), which is not always done in practice, especially for polymer dielectrics that have non-linear dependence of the main electrophysical characteristics (specific volume and surface resistance, dielectric constant and dielectric loss tangent) on the electric field strength. In addition, for the implementation of this method, it is first necessary to experimentally determine the values of the coefficients B and m for various E. This procedure also requires high-voltage tests of real insulation products at given values of the working electric field E. This is associated with a significant expenditure of time and material resources, so as a result of high-voltage tests it is necessary to break through a sufficiently large number of real insulation products. The latter is especially important when determining the service life or time before the breakdown of unique expensive insulating products. Another significant drawback of this method is that it allows you to determine only the average time to breakdown for a given sample of insulation samples with a spread determined by the sample size or level of confidence.

Closest to the technical nature of the present invention is a method for the diagnosis of high-voltage polymer insulation (Gefle OS, Ushakov V.Ya. Method for determining the "life curves" of a monolithic polymer insulation, Electricity, No. 8, pp. 65-67, 1985), which that the test samples to the insulation through the point-plane electrode system alternately high voltage is applied and formed in a length register insulation dendrite l, after which the high voltage is turned off, and the time until breakdown of the insulation is determined by _direct τ eq Depending trapolyatsii lg τ _ave = f (lg K _l) to the intersection with the ordinate axis, where K _{m l} = E / E _cpl - coefficient of the local electric field nonuniformity in the length of the dendrite, which is formed in the insulation under the applied voltage; E _m - maximum electric field at the end of the tip; E _cpl is the average electric field strength, sufficient for the breakdown of the local isolation section (along the dendrite length):

where r is the radius of curvature of the tip; E _l - field strength at the point at which the development of dendrite stops:

However, this method has a number of significant disadvantages, the main of which are the following.

To plot the dependence log τ _pr = f (log K _l ) and then extrapolate it to the intersection with the ordinate axis, i.e. in the region K _l → 1 (which corresponds to a uniform field), uncertainty arises due to the fact that in practice for systems with a quasihomogeneous or uniform field (for l≤K _l <1.3) it is impossible to obtain accurate experimental data on the length of the dendrite, since in such fields the dendrite formation time is much longer than its development time, and the dendrite channel grows continuously. As a rule, the dependence log τ _pr = f (log K _l ) is constructed from the data obtained for samples with K _l ≥3. This can lead to a significant error in determining τ _pr

In addition, for the accurate construction of the dependence log τ _pr = f (log K _l ) and its subsequent extrapolation to the intersection with the ordinate axis, a large set of experimental data is needed to ensure a high level of confidence in determining the extrapolated value of τ _pr , which significantly increases the time required to implement the method.

In addition, the prototype method also determines only the average value of the time before the breakdown at a certain confidence probability.

The technical result consists in reducing the diagnostic time by eliminating the need for resource testing of samples before breakdown, as well as in reducing material consumption.

This is achieved by the fact that in the method for diagnosing high-voltage insulation, namely, a high voltage is applied to the tested insulation sample through a point-plane electrode system and the time until insulation breakdown is determined, according to the proposed solution, the moment of primary fracture channel initiation is recorded, and the time until the primary channel of destruction of insulation τ _o , after which the high voltage is turned off, the formation time of the dendrite τ _d is determined by the empirical expression:

and the time to breakdown of insulation τ _CR determined by the formula:

- модуль комплексной диэлектрической проницательности в диапазоне сверхвысоких оптических частот при условии равенства ее действительной

и мнимой

составляющих, n - показатель преломления материала изоляции, δ - угол диэлектрических потерь.Where

- the module of the complex dielectric constant in the range of ultrahigh optical frequencies, provided that its real

and imaginary

components, n is the refractive index of the insulation material, δ is the dielectric loss angle.

The method is as follows.

High voltage is applied to the test insulation samples using the tip-plane electrode system. Individually for each sample, the moment of nucleation of the primary channel of insulation destruction is recorded, and the time of appearance of the primary channel of insulation destruction τ _{o is} measured. Turn off the high voltage, then empirical expressions (1) and (2) calculate the time of dendrite formation τ _d and the time to breakdown τ _pr isolation.

The proposed method can significantly reduce the time of testing, since it eliminates the need for resource testing of samples before the dendrite nucleation, as well as significantly reduce material consumption.

In addition, the proposed diagnostic method, unlike the prototype, allows for individual diagnostics of high-voltage insulation of individual products and to determine not only the average time to breakdown τ _pr for a batch of insulation samples, but also the dendrite nucleation time τ _d from the results of measuring the primary polymer failure channel nucleation time isolation τ _о individually for each insulation sample without carrying out lengthy life tests.

An example of a specific implementation of the claimed method. To implement the claimed method, three batches of samples were made of polycarbonate (PC), which is a transparent dielectric, by injection molding. The number of samples in each batch was at least 20 pieces. Samples in three batches differed only in the direction of melt flow during molding: in samples of batch No. 1, the direction of melt flow in the molds during the manufacturing process was from plane to tip, in samples of batch No. 2, from tip to plane, and in samples of batch No. 3 the direction of the melt flow was perpendicular to the axis of the tip-plane gap.

Each sample had tip-plane electrodes. The radius of curvature of the tip was r = 7.5 ± 0.5 μm, the distance between the electrodes for all samples was d = 9.3 ± 0.3 mm.

All samples were tested at an alternating voltage of industrial frequency U = 24.5 kV, the voltage was applied to the samples stepwise. The type of test voltage can be any (variable, pulse or constant) and it does not affect the results of insulation diagnostics. Since the greatest practical interest is the diagnosis of insulation of electrical equipment, the implementation of the method was carried out on an alternating voltage of industrial frequency of 50 Hz.

The process of fracture nucleation (registration of the moment of nucleation of the primary fracture channel) was recorded in samples of polymer insulation from a PC by the method of recording partial discharges (G. Kuchinsky Partial discharges in high-voltage structures. L .: Energia, 1979). In the general case, any experimental method can be used to record the moment of fracture initiation, which allows recording the moment of formation of the primary fracture channel (microcrack or microcavity at the high-voltage electrode - tip). In particular, for transparent dielectric materials, an optical method with high spatial resolution can be applied, for example, based on an optical microscope and a fiber waveguide, or the method of acoustic emission for opaque dielectrics.

It is known that the nucleation and development of the destruction of high-voltage insulation under the action of high voltage in the tip-plane electrode system in the vast majority of cases is time-discrete. The so-called “incubation” period ends with the formation of the primary insulation failure channel in the form of a microcrack or microcavity at the tip (Shibuya Y., Zoledziowski S., Calderwood J. Void formation and electrical breakdown in epoxy resin, IEEE Trans. Power Appar. Syst., 1977 , V.PAS-96, p.198-206; Vershinin, Yu.N. Electron-thermal and detonation processes during electrical breakdown of solid dielectrics, Ekaterinburg: Ural Branch of the Russian Academy of Sciences, 2000, p. 258), from which the channel jumps further incomplete breakdown, the so-called electrical dendrite.

For the formation of the dendrite channel, a certain time τ _{d is} also necessary, during which a local electric field is formed at the end of the primary channel of destruction of the insulation and degradation of the insulating material occurs. One of the main tasks of diagnosing high-voltage insulation in a strong electric field is the task of determining the moment of the beginning of the destruction of insulation, which ends with electrical breakdown and the failure of the entire insulating product, which is associated with significant material losses. In this regard, the method that allows for individual diagnostics of individual insulating products and to determine the time of formation of the primary fracture channel τ _o , the time of formation of the dendrite τ _d and the time to breakdown τ _pr , is very relevant.

The samples were held under tension until the formation of the primary fracture channel, which is accompanied by the appearance of the first pulse of partial discharges (PD). The appearance of partial discharges is due to the occurrence of electrical microdischarges in the emerging microcracks or microcavities in the insulating material. After recording the first pulse of the PD, the test voltage was turned off and the time was fixed until the appearance of the first pulse of the PD - τ _о .

Table 1 shows the average values of the formation time of the primary fracture channel τ _о and the amplitude of the primary pulse of the CR - q _о for the three tested batches of insulation samples.

For most dielectrics, there is reference data on the real ε 'and imaginary ε' '= ε' · tgδ components of the complex permittivity ε * = ε'-jε '', where tgδ is the dielectric loss tangent for a given polymer dielectric. In the absence of such reference data, the values of ε 'and ε' 'can be easily determined experimentally by any of the known methods.

After recording the values of τ _o according to the formula (1), the values of the dendrite formation time τ _d were calculated, and according to the formula (2), the values of the time to breakdown τ _pr for each sample were calculated.

Table 1 Parameters of the initial stage of destruction of insulation samples from a PC. Parameter Sample Lot Number No. 1 Number 2 Number 3 τ _о ± Δτ _о , s 149 ± 16.5 189 ± 16.6 233 ± 23.2 q _about ± Δq _about , pC 25.9 ± 5.24 18.5 ± 4.08 14.72 ± 3.7

To compare the values of τ _d and τ _pr determined by the proposed method, with experimental values, samples of all three batches of PC were tested before breakdown. The calculated and experimental values of τ _d and τ _{ol are} presented in tables 2 and 3.

A comparison of the experimental results and the calculations given in tables 2 and 3 shows that the discrepancy between the average values of the parameters τ _d and τ _pr for the three tested batches of PC samples does not exceed 10%.

Table 4 shows the results of a comparative assessment of the time required to implement the two methods. During the implementation of both methods of determining τ _pr accepted average values of the formation time of the dendrite τ _d and τ _about for the three tested batches of samples. At the same time, the time required to determine the value of τ _CR by the prototype method is more than sixty times greater than that of the proposed method (table 4). The time for determining the value of τ _pr by the prototype method is the sum of the times of dendrite formation τ _{d of} at least four batches of samples, since to build the dependence log τ _pr = f (log K _l ) it is necessary to have a minimum of four points corresponding to the average values of τ _pri at various test field strengths. The time to determine the value of τ _CR by the proposed method is the sum of the times of formation of the primary fracture channel τ _o for a batch of samples of 20 pieces.

Table 2. Experimental values of the spatio-temporal parameters of the initial stage of destruction of insulation samples from PC (d = 9.3 ± 0.3 mm, r = 7.5 ± 0.5 μm, U = 24.5 kV). Parameter Sample Lot Number No. 1 Number 2 Number 3 l _d ± Δl _d , μm 193 ± 76 189 ± 48 165 ± 39 τ _d ± Δτ _d , s 9826 ± 1319 13532 ± 1958 16725 ± 3570 τ _{Ave Ave} ± Δτ, with 18520 ± 1736 21456 ± 2743 26303 ± 3608

Table 3. The calculated values of the spatio-temporal parameters of the PC destruction process based on the results of the registration of partial discharges. Parameter Sample Lot Number No. 1 Number 2 Number 3 l _d ± Δl _d , μm 210.8 ± 69.4 183.5 ± 46.3 168.7 ± 38.7 τ _d ± Δτ _d , s 9805 ± 1086 12437 ± 1093 15333 ± 2178 τ _{Ave Ave} ± Δτ, with 18481 ± 2046 23443 ± 2059 28900 ± 4106

For an accurate estimate of the time required to implement the two methods, it is necessary:

1. To conduct comparative tests of four batches of samples before the formation of dendrite in them. In this case, to ensure a high level of reliability, the number of samples in each batch should be at least 20 pieces. After that, measure the length of the dendrite in all tested samples in any known manner. The time required to implement the known prototype method, without taking into account the processing time of the obtained experimental data, will be equal to the sum of the dendrite formation times for 80 tested samples and the time required to measure the dendrite length in all tested samples.

Table 4. The time required to determine τ _CR Way Average time to determine the value of τ _CR , s Sample Lot Number No. 1 Number 2 Number 3 Prototype 9826 13532 16725 Proposed 149 189 233 The ratio of implementation times of two methods 66 72 72

2. To test a batch of samples of 20 pieces to determine the time until the initiation of the primary fracture channel according to the proposed method. That is, the time required to implement the proposed method, without taking into account the processing time of the obtained experimental data, will be equal to the sum of the times of the initiation of the primary fracture channel of 20 tested samples. With an accurate estimate of the implementation time of the two methods, the difference in favor of the proposed method will be at least four times greater than that shown in table 4.

Thus, the proposed method allows to determine the values of τ _d and τ _CR from the measurement results of τ _about , that is, the reduction in time to implement the proposed method is due to the lack of the need for lengthy high-voltage tests of the samples before the formation of dendrites in them. In addition, the proposed method, in contrast to the prototype method, allows for individual diagnostics (to determine the time of dendrite nucleation and time before breakdown) for each sample of high-voltage insulation.

Claims (1)

A method for diagnosing high-voltage insulation, namely, that a high voltage is applied to the tested insulation sample through the tip-plane electrode system and the time until insulation breakdown is determined, characterized in that the moment of initiation of the primary failure channel is recorded, and the time until the primary insulation failure channel appears τ _about , after which the high voltage is turned off, the dendrite formation time τ _d is determined by the empirical expression:

and the time to breakdown of insulation τ _CR determined by the formula:

Where

- the module of the complex dielectric constant in the range of optical frequencies, provided that its real

and imaginary

components; n is the refractive index of the insulation material; δ is the dielectric loss angle.