EFFECTIVE METHODS OF ASSESSMENT OF INSULATION SYSTEM CONDITIONS IN POWER TRANSFORMERS: A VIEW BASED ON PRACTICAL EXPERIENCE V. Sokolov Z. Berler, V. Rashkes ZTZ-Service, Ukraine Cutler-Hammer Predictive Diagnostics, USA 5421 Feltl Road, Suite 190, Minnetonka, MN 55343 Abstract: Condition-based monitoring of power transformer The following trends can be extracted from the statistics available: insulation should center on the prediction of a substantial drop • Over 70% of transformer failures, particularly in 1996-97, in the dielectric safety margin under the impact of moisture, oil have occurred after 20 years of service due to aging diseases. by-products, contaminating particles, paper insulation aging • Transformer life is limited predominantly by accelerated and partial discharge activity. A functional failure model of deterioration of two components: bushings and LTC (over power transformer insulation and possible effective methods of 50% of failures in both surveys). the insulation condition assessment are discussed based on • A stable high rate (15-20%) of failures is attributed to the practical experience. impairment of the conditions of major and minor insulation due to a particle contamination or the ingress of moisture INTRODUCTION reducing the impulse withstands strength. It’s worth During recent years the technical policy of power utilities is emphasizing here that most of the problems have been changing under the pressure of economic considerations. There is a associated with the degradation of HV winding insulation that common tendency of moving from the time-based to the condition- is typically more sensitive to the deterioration. based maintenance to reduce maintenance costs, to extend • Decomposition of cellulose can be attributed only to a low significantly the life span of equipment and, at the same time, to portion of failures (3-5 %) caused by an excessive aging. A prevent possible catastrophic failures. Thorough knowledge of deficiency of old designs is basically involved (bad cooling, actual condition of the transformer and effective diagnostic overinsulated coils, non-uniform current distribution in the methods are necessary to meet these requirements. conductors, unforeseen stray losses, etc.), . • Mechanical weakness and winding distortion are the cause of This paper presents some considerations on condition-based 10-15% of failures. monitoring of power transformer insulation centered on a • Recurrent cases with localized overheating in the magnetic prediction of a substantial drop in the dielectric safety margin circuit due to the deficiency of old designs appear. Some of under impact of such practically inevitable agents of degradation, the defects do not affect directly the transformer as moisture, particles, oil aging products and partial discharge serviceability, but cause gassing and contamination of activity. The paper discusses typical defects and the functional insulation system with by-products. failure model of HV transformer insulation as well as effective Thus, dielectric and mechanical weaknesses seem to be the main methods of detection and identification of its defects. life shortening factors for transformers. The degradation of the dielectric safety margin appears as a really vital problem in large CRITICAL DISEASES OF AGED TRANSFORMERS power transformers. Approximately 80% of transformer failures There is a large population of aged equipment approaching the end could be predicted and prevented if an effective diagnostics system of the original design life. Questions arise: What’s happening with is used. It is important to note that most of defects caused failures aged transformers? What kind of problems may they have? What are of reversible mode and they could have been not only detected are the critical aging diseases? but also corrected in the field. The best means to answer these questions is the failure analysis. DEFECTS IN COMPOSITE INSULATION SYSTEMS Such analyses based on periodic reliability surveys were performed by ZTZ-Service Co (the population of about 5000 units, above 100 Major Insulation MVA, 110-750 kV), by the CIGRE WG12.18, and by Doble • Excessive moisture in the cellulose insulation. This defect is Engineering which distributed Technical Questionnaires in 1994- inherent basically to the transformers with open-breathing 1998 the results of which have been partially published [1-3]. The preservation system or to those which have an insufficient comparison of the statistical data from the ZTZ-Service and Doble sealing. Distribution of the moisture in the course of the Engineering investigations is given in Table 1 based on . transformer life is kept quite non-uniform. Most of the water is stored in so called “cold thin structures”, namely in the thin Table 1. The causes of power transformer failures pressboard barriers that operate at bulk oil temperature [5, 6]. in 1996-1997 (above 100 MVA) • Oil contamination with water, particles and oil aging Defective element Rate, % products. The most problems caused by these defects occur Doble clients ZTZ-Service clients typically in the space “HV winding (HV bushing) –Tank”. Bushings 35 45 • Insulation surface contamination in the forms of the LTC 16 9 adsorption of oil aging products on a cellulose surface or Major insulation 9 17 deposits of conducting particles and insoluble aging products Winding (turn, coil) aging 16 12 in areas of high electrical stresses. The surface contamination Winding distortion 12 10 can cause a distortion of electrical field and a reduction in the Core 7 7 impulse strength of the insulation system. Leads 5 - • Partial discharges in weakened insulation spots. The presence of water and impurities changes dielectric parameters This makes it difficult to detect a defective condition of a of deteriorated components, namely, their conductivity, winding’s minor insulation until a critical partial discharge or permittivity and dissipation factor, particularly with temperature. noticeable gas generation occurs. The temperature effect of water on insulation characteristics of cellulose and oil, as well the temperature effect of contaminated FUNCTIONAL FAILURE MODEL oil, are well known, e.g.. Changes in the dielectric parameters of A failure model has to be based on possible defects in the the defective component(s) result in related changes in the particular transformer component and a possible scenario of dielectric characteristics of the whole transformer integrity, and progression of the defective condition into a failure (the breakdown requires further investigation to detect the “guilty” component . or flashover). Table 2 summarizes typical problems with transformer insulation. Defects related to excessive moisture, oil contamination or surface contamination usually fall into the category of reversible defects. Table 2. Typical defects and developing faults in the The damage created by partial discharge activities is usually components of the dielectric system of power transformers irreversible. This type of damage usually results in carbonized Component Defect Fault/Failure mode tracks (creeping trees) that extend between the electrodes along the Major insulation Excessive water Destructive PDs surface. Oil contamination Localized tracking Surface contamina- Creeping discharge Creeping Discharges tion lead to: Creeping discharge is likely the most dangerous failure mode that Abnormally aged oil Breakdown or flashover typically results in catastrophic failures at normal operating Low energy PDs conditions. The phenomenon occurs in the composite oil-barrier Static electrification insulation and progresses in four steps: discharges Winding distortion • Partial breakdown of a gap, typically between the winding and the nearest pressboard barrier. Minor insulation Same as for the Destructive PDs • Surface discharge in oil across the barrier (an appearance of a major insulation + Localized tracking black carbonized mark on the barrier). Abnormally lead to: • Forcing oil and water out of the pressboard pores in the aged cellulose Flashover or short- vicinity of sliding discharge that results in microscopic Bubble evolution circuits between turns sparking within the pressboard. The presence of some excessive moisture stimulates this process. Lead insulation Excessive water Destructive PDs Overheating Gassing/bubble evolution • Splitting oil molecules under effect of sparking; the formation Abnormally aged lead to: of hydrocarbons followed with the formation of carbonized Mechanical destruc- Major insulation failure traces in the pressboard. This process can continue from tion or circuit opening minutes to months or even years, until the treeing conductive Displacement path cause shunting of an essential part in the transformer Electrostatic shields Mechanical disrup- Destructive PDs insulation resulting in a powerful arc. tion or displace- Gassing/bubble evolution ment lead to: Faulty conditions can be characterized in terms of PD activity, Low energy PDs Major insulation failure fault gas generation, and changes in the conductance, capacitance and dielectric loss factor of a defective area. The critical level of relevant diagnostic parameters may be evaluated using Typical Scenarios of Developing Insulation Failure experimental data from [9, 10]: The following typical scenarios of an insulation failure model are • Oil destruction PD intensity q >300 pC; presented on the basis of historical cases analysis (summary of PD power P< 0.4 W; over 200 failures): rate of gas generation 5-170 μlitres/Joule (different for different oils). • Critical contamination of oil (typically presence of free • Oil-breakdown PD intensity-q>100000 pC. water) + rapid change of temperature PD appearance at • Cellulose destruction (creeping discharge) rated voltage breakdown. PD intensity q>100 –1000 pC; PD power P=0.1-1 W; • Surface contamination + water + rapid change of rate of gas generation 40-45 μl/J. temperature PD appearance flashover. Minor (Turn & Coils) Insulation • Particles contamination + switching surge critical PD • Overheating leading to accelerated insulation aging. • Excessive moisture content leading to bubbles appearing at breakdown. some zones with elevated temperatures. • Insulation surface contamination with conducting particles • Water + particles contamination (or bubbles present in oil) and oil aging products. critical PD Creeping discharge progressing breakdown. Above mentioned defects may cause a sharp change in electric parameters of the minor insulation, e.g.. However, it has only a • Surface contamination + lightning impulse Surface minor impact on overall dielectric characteristics of the whole discharge Flashover. transformer, due to relatively high capacitance of turn insulation. • Distortion of winding geometry PD appearance • Levels of water content: the condition of a transformer, Creeping discharge progressing Breakdown. which may result in an increase of the relative saturation of water in oil over 50% in the range of operating temperatures. Total water content is considered. • Distortion of winding geometry + Switching surge Flashover between coils (sometimes with restoring • Water in solid insulation: the compliance with the above withstand strength) Gas evolution. stated level of water content. Corresponding water content in the barriers exceeds 1.5-2% (the estimation is made through dielectric characteristics). Objectives of the Diagnostic Technique Three main diagnostics objectives can be formulated on the basis • Particles in oil: the level of contamination is as per NAS of the failure analysis: class 7 and higher; the presence of visible and conducting • Detection and identification of defective conditions caused by (metals, carbon) particles. an accumulation of degradation agents. • Detection and identification of an irreversible damage to the • Oil aging: insulation (critical PD and creeping discharge). • The appearance of a sludge in the period between the • Detection of a pre-failure state of PD activity (prior to the tests. breakdown). • The end of the induction period (trend of accelerated degradation). INDICATORS OF DANGEROUS INSULATION • The presence of acids and non-acid polars that accelerate CONDITION cellulose decomposition. A traditional approach to the identification of critical transformer conditions is based generally on the establishment of some value • Presence of bubbles in oil: symptoms of possible bubbles for each tested, agreed-upon diagnostic parameter: fault gas evolution, including C2H2 generation due to high temperatures concentration, water-in oil content, insulation power-factor, etc. (>8000C) when the bubbles evolution is practically an However, they are rather symptoms than characteristics of the inevitable phenomenon. defective condition to be evaluated. A defective condition defines a threat to the equipment serviceability; however, sometimes there is • Partial discharges: presence of PDs with q> 100-500 pC in no any direct correlation between particular test results and the the oil-barrier structure; symptoms of PDs revealed through defective condition. The effective diagnostics system for DGA analysis. transformer insulation has rather to reveal the threat of a critical reduction in the dielectric safety margin of insulation. • Mechanical integrity: a radial mode winding distortion that may change the insulation geometry. The assessment is made Studying the models of the transformer insulation [1, 12, 13] has through the relevant change of the leakage reactance. shown that the dielectric safety margin of both major and minor insulation contaminated with water is still determined by the • Insulation surface contamination: the symptoms assessment dielectric withstand strength of the oil. The dangerous effect of the is made through the change of insulation dielectric dissolved water is determined by a sharp reduction in oil dielectric characteristics with temperature. strength with increase in its relative saturation, due to increase in the conductivity of the particles available or emulsion formation in the vicinity of a surface-active substance. EFFECTIVE METHODS TO ASSESS THE For example, 30 ppm of dissolved water in oil can be a problem at INSULATION CONDITIONS 20oC (relative saturation ~70%) but does not affect the dielectric strength of oil at temperature >40oC (relative saturation <30%). A Oil Tests low particles content improves the situation. On the other hand, Practical experience has shown that more than 60% of the latent polar oil-aging products may store a significant amount of water defects have been revealed through oil tests . However, the that can transfer in a dissolved state at some elevated temperature variations in different oil characteristics make it difficult to identify and make the situation worse. One may emphasize that the modern the type of the problem revealed. We have found that diagnostic Karl Fisher method measures practically only dissolved water, effectiveness for oil parameters could be improved by separating of leaving the full amount of water unknown. oil tests into four groups: • Identification – parameters that specify the oil and remain The evaluation of the permissible dielectric state of oil needs to be practically unchanged during its life. further studied. The following factors must be considered: the level • Aging status - parameters relevant to the aging process. of particle contamination; presence of polar and surface-active • Dielectric status – parameters affecting the dielectric safety products; value of interfacial tension of oil; limitation of percent oil margin (water, particles, etc.); saturation and water content in paper. • Diagnostic tests – utilization of oil as a diagnostic medium. Based on its wide experience in designing, testing, maintenance A test program and the protocol “Dielectric Status” includes and refurbishing of power transformers 110-750 kV, ZTZ-Service determining water content (before and after transformer heating); established the following criteria for the evaluation of a dangerous particles counting; particle identification (with a microscope); insulation state: measuring the breakdown voltage; measuring oil power-factors at determined water contamination. In most of the cases the defective 20, 70, and 900C; DGA (symptoms of PD); condition was confirmed by the following estimation of water resistivity measurements at 20, 70, 900C; polarization index content through dielectric characteristics tests  or by the direct measurement; IR-scanning (non-acid polars). measurement of water content in pressboard samples removed from the transformers. The most effective diagnostic components are by-products related exclusively to the degradation process – DGA, furans, dissolved Fig.1 demonstrates the case of 180 MVA, 220/18 kV, GSU metals, et al. DGA is indisputably the best detector of transformer, with open-breathing preservation system after 18 abnormalities. However, there are still a number of unsolved years in operation. During its Water Heat-Run Test the build-up problems: water content in oil practically followed the rise in temperature. • Difference in the rate of gas generation in different oils. The presence of free water in the transformer was determined and • Migration of gases between the oil and cellulose. later confirmed by additional tests. • Unusual sources of gas generation. T oC, • Location of the source of gas generation. One can expect some new benefits from advances in the DGA W ppm technique, namely: % to C switching • Detection of low temperature faults (150-400 °C) using C3 – off C5 hydrocarbons, particularly, C4H8 buten-1. • Determining the temperature signature of overheated oil through C3-C5 hydrocarbons response. W • Determining the correlation between amount of gases and dissipated energy. Water Heat-Run Test for In-Service Assessment of the Level of Water Contamination This method was described in [3, 14] .The objectives of the method are: • Assessment of the transformer health under rated conditions – the maximum permissible temperature. • Assessment of the level of water contamination using the build-up of water content in oil with time and temperature. • Assessment of possible state of water and distribution of water within a transformer using the rate of building-up of water in oil. Figure 1. An example of Water Heat-Run Test on an old A loaded transformer is heated, by reducing its cooling, up to the 180 MVA, 220 kV GSU transformer (time in hours) maximum possible temperature, to lower oil per cent saturation and W-absolute water content, ppm, relative saturation,% to obtain a “moisture potential” in the vicinity of insulation. The test duration must allow “discharging” the insulation and building Table 3 shows another case of the assessment of the level of water up a significant amount of dissolved water in the oil. To detect contamination in a 200 MVA, 347 kV, GSU transformer with an water content over 1.5-2.0%, the temperature 60-75°C and test open-breathing preservation system, after 27 years in operation. time of 3 days has been recommended. However, practical Following the heating of the transformer from 40 to 700C the water experience has shown that water contamination over 2% or the content in oil increased from 24 ppm to 44 and further to 49 ppm. presence of free water can be assessed even using a one day test. Water contamination > 2.5% in the thin structure was estimated. The further direct test on pressboard samples after draining the oil Assuming that the main source of water contamination is a “thin” shown a water content of 2.75 %. During drying process 36 kg of structure of transformer insulation, the level of water content W water was extracted can be estimated using the equation Table 3. Water Heat-Run test of 200 MVA, 347 kV , transformer time t, Woil We Estim. Real where W is the amount of water “discharged” from the solid hrs °C ppm % % water water insulation into oil related to the mass of the thin structure; content in content thin We is the equilibrium moisture, which can be determined from the structure absorption equation for given relative oil saturation; Initi 40 24.2 15 3.9 In the 2- [1-F(Z)] is the diffusion function . al mm 24 70 44.2 5.5 1.38 > 2.5% PB- In 1992 – 1998 the condition of over 150 transformers, rated 25 – 10.1 2.1 in 1000 2.75%. 1250 MVA, that came under suspicion was assessed using this kg of 2- Extracted “Water Heat-Run Test”. Fifty-three (53) units, predominantly of mm PB 36 kg open-breathing design, were recognized as defective due to their 48 70 47.2 10.8 2.2 of 72 70 49 11.2 2.24 water . Experience has shown a poor correlation between the water content Table 4. Typical defects and diagnostic possibilities in the oil and solid insulation at temperatures below 60-70 °C. for the main transformer insulation Tests at 25-40 °C have typically led to the overestimation of the moisture content in insulation. A special “two-steps” test at 50 and 70 °C was carried out on the 200 MVA, 347 kV, open-breathing transformer after 27 years in service. After heating the unit from 30 to 50°C and maintaining for 24 hours, the water content increased from 10 to 12.7 ppm only. However, a similar test at 70°C has shown a significant rise in water up to 34 ppm. Contamination of “thin structure” of over 2.5% was concluded. Use of Insulation Characteristics to Assess the Condition of Oil-Barrier Insulation Practical experience has shown  that 30-35% of the problems still can be detected through off-line tests only. Those are winding distortions, insulation surface contamination, some bushing problems, etc. The experience has also demonstrated that water content in the pressboard barriers, the surface contamination level and oil contamination can be effectively estimated using the temperature responses of insulation power-factor and DC insulation resistance for interwinding and winding-to-tank insulating spaces, when taking into account the value of relative portions of oil and solid insulation within the space. The The model of the oil-barrier structure can be presented in transformer insulation is a composite dielectric system, located accordance with Figure 2,A . The total current through the between the electrodes, i.e., winding conductors, and grounded model may be expressed as a sum of three components: the current parts of the transformer. Dielectric measurements allow you to through the solid insulation IP, the current through oil I0, the determine the partial conductance of the dielectric system between current along the surface IS: each accessible pair of electrodes. Sometimes the measured value I = Ip + I0 + Is is equal to the conductance of the insulation zone between electrodes. For instance, in the zone between the high-voltage It means that the full current through the composite insulation (HV) winding (outer) and the tank, all the current from the HV space and any of the insulation characteristics derived from it winding flows to ground. Sometimes the measured value is not depend not only on the solid insulation condition, but also on the equal to the conductance of the insulation zone. For example, in conditions of the oil and the surface. Therefore, the sensitivity of the interwinding space, in case of a severe contamination of the dielectric parameters to deterioration of the barrier depends on the barrier surface, the portion of current between the HV winding and share of current flowing through the barrier and consequently on the low-voltage (LV) winding flows down to the ground along the the relative amount of cellulose insulation in this space, i.e. on the surface of the barrier, resulting in a decrease of the measured insulation structure design. Consequently all the insulation dissipation factor. characteristics must be analyzed according to their interrelationship. The most important components of the main transformer insulation are: This model allows us to determine the dissipation-factor at the • Insulation between the HV winding and the tank, including power frequency (or at any other frequency) as well as the dc the HV bushings; insulation resistance . The third (surface) component is of • Insulation between the HV and the LV windings; practical importance only when surfaces of all the barriers are • Interphase insulation. severely contaminated. In the majority of practical cases another, simplified model is valid (Figure 2 B). These components usually have the smallest margins in the dielectric strength, and, as a result, are the most sensitive to the The above approach allows us to estimate the condition of the oil insulation deterioration. The monitoring of the solid and liquid within the space as well as the condition of the solid insulation insulation in these components, i.e., the monitoring of their considering the relative amount of the solid insulation. For dielectric characteristics, is a subject of great importance and one example, the dissipation-factor of the interwinding space at power of the main objectives of transformer diagnostic tests. In other frequency can be expressed by the following simple equation: areas of the insulation, specifically the insulation between the LV tan HV-LV = Kp x tan P + K0 x tan 0 winding and the core, the margin of dielectric strength is usually significantly higher than in the spaces that include HV winding. Therefore, here only a high degree of deterioration can be the The design parameters Kp and K0 can be evaluated using , but cause for concern. typically K0 =0.4-0.6 and Kp =1- Ko=0.6-0.4. The possible effectiveness in the detection of typical defects through dielectric characteristics is characterized with Table 4. • Evaluation of insulation surface contamination (including traces of creeping discharges) using temperature dependence of tan values for the interwinding space . On-Line Partial Discharge Measurements The recent progress in the PD measuring technique  has opened really new opportunities in an effective rejection of external interference, detecting weak PD signals and on-site IP IS IO IP IS IO diagnosis of the condition of transformer insulation, quite similarly A B to well established laboratory tests at the transformer factories. The practical experience with the application of PD Analyzers UPDA Fig 2. The model of interwinding oil-barrier structure (A) & in 1998 has confirmed that this test technique provides the simplified model of oil-barrier space “Winding-Tank” (B) sensitivity to PDs in field conditions of about 20 pC in power plants and 50 pC in the 500…750 kV substations. Over 10 years of experience with the evaluation of insulation Particularly, the following problems have been identified in the conditions through the measurement of tan and DC insulation transformers already tested: resistances of relevant insulation spaces has shown that the • Defective busbar isolators were found on 13.8- kV side of a following problems can be successfully solved: GSU transformer; • Source of PD was located in a 500-kV bushing of a 300 MVA Estimation of the average water content (over 1-1.5%) in the autotransformer; barriers using test data for interwinding space and the following • The source of critical PD was detected in a 750 kV algorithm: autotransformer (Figure 3). These PDs were caused by a • Measure tan HV-LV at the elevated temperature. progressing creeping discharge across the 750 kV bushing • Determine tan 0 of oil at the same temperature. insulation that was confirmed by internal inspection. The • Define the design parameters Kp and K0 . problem was associated with water penetration through • Calculate the value of the dissipation factor of the loosed top sealing of the 750 kV lead. Internal inspection has pressboard tan P. shown that a really catastrophic failure had been prevented. • Define the water content using well-known dependence The sketch of defects found at inspection is presented in for pressboards from . Figure 4. The examples of such water content estimation using the measured dissipation-factor are presented in Table 5. A good correlation between the estimated values and the results directly measured in the samples of the pressboard, after draining the oil, has been found. Table 5. Examples of estimations of the water content in barrier insulation through dissipation-factor measurements Fig 3. Results of PD measurements on the defective 750 kV transformer at two voltage levels (0.8-1 and 1.05 per-unit) • Estimation of Oil Contamination in Insulating Spaces. The experience has shown that in some cases the oil in the Figure 5 demonstrates how effective was the UPDA in the rejection insulation spaces can be more contaminated than the oil in the outside corona noise during PD measurements performed on 22 sample taken from the bottom of the tank. The condition of transformers and autotransformers rated 500 kV . Another the oil can be better evaluated through the dissipation-factor feature of this PD analyzer is the capability of analyzing the PD and dc insulation resistance in “HV-Tank” space, using the signatures, particularly, the power dissipated in PD, and in this way difference in tan of the space “HV winding–Tank” measured to utilize the diagnostic technique developed by CIGRE WG 15.01 at two different temperatures (e.g. 60 and 300C) . (TF “PD signatures”) [16, 17]. the bushing insulation is used in many cases. The increase in the resistive or capacitive currents associated with the bushing insulation are signs of insulation deterioration: power-factor being more sensitive to the initial steps in insulation deterioration and the capacitance – to the developed defect. In the majority of cases the monitoring devices sum up the currents through bushing insulation of all three phases in the set [18, 19]. Initially the currents are balanced, so the output signal is close to zero. The increase in any one or two of the currents sets the balance off and produces an output signal proportional to the current increase. These devices are sensitive and are able to detect small increases in the current. The increase about 1-2% corresponds to a developing defect and to 5-6%- to a critical defect. In Russia where simplified monitoring devices were widely used for about 30 years, they prevented about 75% of catastrophic failures in non-hermetically sealed bushings. The experience of similar monitoring hermetical bushings in the USA and Canada is more limited, but also is positive [18, 19]. The method can be further improved if the dependence of the imbalanced current on the bushing temperature is taken into account. When all three bushings are in an identical insulation Figure 4. Sketch of the defects found in the 750-kV bushing state, their reaction to the temperature rise or reduction is also identical and does not create an output signal from the device. But when the condition of the insulation in bushings are different their reaction on the temperature changes will be different as well. That makes the output signal dependent on the deviation of top-oil temperature from its value during the initial balancing. This dependence, together with the absolute value of the imbalance can be used for diagnostics. If the given imbalance is reached at the initial top-oil temperature, the state of the insulation is worse than if it is reached at a changed temperature. As an example, Figure 6 shows the dependence of the imbalance current (current in per cent of its initial value) versus top-oil temperature observed during two years of monitoring on a set of Signals prior to the noise rejection 138 kV bushings. With worsening the insulation state the absolute values of rise, and the dependence (t) becomes somewhat steeper. Off-line tests confirmed worsening defects in two bushings of the group. Signals left after the noise rejection Figure 5. Plane projections of 3-dimensional distributions of pulse repetition rates obtained on a 500 kV transformer prior and after the noise cancellation Monitoring Power Frequency Current through Bushing Insulation As oil-filled high voltage bushings are often the weakest insulating Figure 6. Gamma versus top-oil temperature in a set of component in transformers, their monitoring is of a special 138 kV bushings during three consecutive terms of observation practical importance. Above-mentioned PD monitoring provided since September 1997 till February 1999 periodically is able to detect developing defects. Also the periodic or permanent monitoring of the power frequency current through CONCLUSIONS Vibro-Acoustic Monitoring of Winding & Core Clamping Force Condition-based monitoring of power transformer insulation has to A reduction in clamping forces is frequently accompanied with be based on an extensive understanding of the processes of winding distortions and short circuits between elements of the insulation deterioration. The diagnostics have to be centered on the magnetic circuit. These defects can lead to PDs and combustible prediction of the substantial drop in the dielectric safety margin gas generation, therefore monitoring the clamping forces can be of under impact of moisture, oil by-products, contaminating particles, help in determining if the problem is associated with the partial discharge activity, tracking and creeping discharges, and transformer insulation and in its localization. On-line vibro- paper aging. The functional failure model chosen highlights the acoustic monitoring of the residual clamping forces in the core and most dangerous processes where attention has to be concentrated. windings of a transformer was described in . The method is based on the measurement of the steady-state vibration at several The application of several, practically proven monitoring methods points of the transformer tank and further analyzing the energy and multi-step test techniques provides the expert with relevant distribution between different frequencies in this vibration. To interrelated information that increases the reliability of the differentiate vibration from the windings and from the core, two diagnosis and allows corrective actions to be timely implemented. load modes are used: the maximum possible load close to the rated In many cases corrective measures can be applied to the defective one and no-load (or small load). The system is able to discriminate transformers; in other cases its catastrophic failure can be still the separate portions attributable to the core or windings. The prevented, or plans made to minimize the effects of such diagnosis is provided in the form of coefficients: when the catastrophic failures. coefficient is between 1.0 and 0.9 the clamping force is good (60- 100% of the initial one), between 0.8 and 0.9 the clamping force is satisfactory (20- 60% of the initial one), below 0.8 the condition is critical (the clamping force is below 15-25% of the initial one). REFERENCES Figure 7 shows the diagnostic window of the computerized diagnostics system during the test of a 345 kV step-up transformer. 1. V. Sokolov "Experience with the Refurbishment and Life The parts of the core and windings are colored: for three gradation Extension of Large Power Transformers," Minutes of the mentioned above colors are green, yellow and red, respectively. Sixty-First Annual International Conference of Doble Clients, 1994, Sec. 6-4. 2. V. Sokolov “Transformer Life Management Considerations”, Proceedings of the 1997 CIGRE Regional Meeting, 1997, Melbourne, Australia. 3. V. Sokolov “Consideration in Transformer Life Management - a View from Abroad”, Proceedings of the Techcon ‘99 Annual Conference, TJ/H2b, February 18-19, 1999, New Orleans, LA. 4. P. Grestad, “Life Management of Transformers. Case Story”, Proceedings of the CIGRE SC12 Colloquium, 1997, Sydney. 5. V. Sokolov, B. Vanin “Experience with In-Field Assessment of Water Contamination of Large Power Transformers,” Proceedings of the EPRI Substation Equipment Diagnostic Conference VII, February 20-24, 1999, New Orleans, LA. Figure 7. The diagnosis of the core and winding residual clamping forces for a 345 kV, 880 MVA step-up transformer 6. P.Griffin, V.Sokolov, B.Vanin “Consideration on Moisture Distribution in Transformers”, Proceedings of the 66th Annual International Conference of Doble Clients, April 12- Table 6 demonstrates the effectiveness of the vibro-acoustic 16,1999 Boston, MA., technology in the revealing internal defects in windings and cores. Totally 290 power transformers rated 2-880 MVA (mainly over 50 7. I. Gussenbauer "Examination of Humidity Distribution in MVA), 10-500 kV were tested. 34 of them were found to have Transformer Models by Means of Dielectric Measurements", problems in the core or windings. In more than 80% of cases the CIGRE, 1980, # 15-02. diagnosis was confirmed during an outage, and the compression was restored. 8. V. Sokolov, B. Vanin. “Evaluation of Power Transformer Insulation through Measurement of Dielectric Table 6. Results of vibro-acoustic method application to power Characteristics”, Proceedings of the 63rd Annual International transformers Conference of Doble Clients, 1996, sec. 8-7. Country Transformers Diagnosed Defect in Defect in tested as defective the core the windings 9. P. Marutchenko, T. Morozova “Voltage versus Time Russia 260 25 13 15 Characteristics of Surface Discharge in Transformer Oil under USA 12 4 4 2 Long –Time Voltage Action”, Elektrotechnika,1978, #4, pp. Canada 18 5 5 - 25-28 (in Russian). 10. B. Arakelyan, E. Senkevich “Early Diagnostics of Oil- BIOGRAPHIES Immersed HV Equipment”, Electricheskie stantsii, 1985, #6 (in Russian). Dr. Victor Sokolov, a CIGRE Member and the Convenor of CIGRE WG 12-18, obtained his MSEE in 1962 from the Kharkov 11. T. K. Saha, M. Darveniza “The Application of Interfacial Polytechnic University, and Ph. D. in High Voltage Technology in Polarization Spectra for Assessing Insulation Condition in the 1982 from the Kiev Polytechic University. Dr. Sokolov possess a Power Transformers”, Proceedings of the CIGRE SC12 world level expertise in all questions of designing, producing, Colloquium,1997, Sydney, Australia. testing and maintaining of power transformers. He is an author of numerous publications. Now he is the Technical Director of the 12. V. Ryzhenko, V. Sokolov “Effect of Moisture on the Scientific-Engineering Center “ZTZ-Service” (Ukraine). Dielectric Strength of Winding Insulation in Power Transformers”, Electricheskie stantsii, 1981, # 9 (in Russian). Mr. Zalya Berler, a Member of the IEEE and CIGRE, is the General Manager of Cutler-Hammer Predictive Diagnostics 13. Yu. Kalentyev “Investigation of Short –Term and Long-Term Division. He obtained his MSEE from the Leningrad State Behavior of Oil-Barrier Insulation of HV Power transformers University (Russia) in 1973 and held various engineering and in Real Operation Conditions”, Dissertation, Sant-Petersburg, managing positions in electric utilities, including 10 years with Russia, 1985. Northern States Power Co, and power engineering construction companies. He has over 10 publications. 14. V. Sokolov, V. Vanin "In-Service Assessment of Water Content in Power Transformers," Proceedings of the 62nd Dr. Viktor S. Rashkes, PE, Senior Member of the IEEE, Member Annual International Conference of Doble Clients, 1995, Sec. of CIGRE, is the Manager of the Transformer Diagnostics 8-6. Department of Cutler-Hammer Predictive Diagnostics. He obtained his MSEE and Ph. D. in High Voltage Technology from the 15. A. Golubev, A. Romashkov, V. Tsvetkov et al. “On-Line Moscow Power Engineering University in 1956 and 1966, Vibro-Acoustic Alternative to the Frequency Response respectively. State certified as a Senior Researcher in High Analysis and On-Line Partial Discharge Measurements on Voltage Technology in Russia, a Registered Professional Electrical Large Power Transformers”, Proceedings of the Techcon ‘99 Engineer in the State of Massachusetts. Prior to CHPD he worked Annual Conference, TJ/H2b, February 18-19, 1999, New for the Electric Power Research Institute VNIIE in Moscow where Orleans, LA. contributed to the creation of 787 and 1200 kV equipment and . transmissions, then for General Electric Co. at EPRI High Voltage 16. S. Lindgren, H. Moore. “Diagnostic and Monitoring Transmission Research Center in Lenox, MA. His primary field of Techniques for Life Extension of Transformers”, Proceedings activity covers HV/EHV transmissions and equipment, transients, of the CIGRE SC12 Colloquium,1997, Sydney, Australia. overvoltages and insulation, field and laboratory high voltage tests and measurements. He holds 33 patents, and has published over 17. S. Cesari, C. Hantouche, T. Muraoka, B. Pouliquen “Partial 140 papers. Discharge Measurement as a Diagnostic Tool”, Electra, December 1998, # 181. 18. Z. Berler, L. Letitskaya, V. Rashkes, P. Svy “Experience in the Application of the On-Line Monitoring System Using Power Frequency and Partial Discharges to High Voltage Transformer and Bushing Insulation”, EPRI Substation Equipment Diagnostic Conference VI, February 16-18, 1998, New Orleans, LA. 19. M.F. Lachman, W. Walter, S. Skinner “Experience with On- Line Diagnostics and Life Management of High Voltage Bushings”, Proceedings of the 66th Annual International Conference of Doble Clients, April 12-16,1999, Boston, MA.
Pages to are hidden for
"EFFECTIVE METHODS OF ASSESSMENT OF INSULATION SYSTEM CONDITIONS IN"Please download to view full document