B3-102 THE ROMANIAN EXPERIENCE REGARDING THE RISK OF MANAGEMENT IN THE OPERATION AND MAINTENANCE OF LARGE POWER TRANSFORMERS IN HV SUBSTATIONS

Transcription

1 21, rue d'artois, F Paris B3-102 Session 2004 CIGRÉ THE ROMANIAN EXPERIENCE REGARDING THE RISK OF MANAGEMENT IN THE OPERATION AND MAINTENANCE OF LARGE POWER TRANSFORMERS IN HV SUBSTATIONS C. Moldoveanu* C. Radu** NOVA INDUSTRIAL S.A. SMART S.A. (Romania) Abstract About 70 % of large power transformers, of rated voltage ranging from 110 to 750 kv, from Romanian power system have the life time nearest or greater than 25 years, considerated as the standard useful life time. Their replacement by the new ones is technically and economically impossible and unreasonable. The decisions regarding the real assessment of these transformers in operation is very important. This paper presents some concrete examples met by authors while diagnosing some of the large power transformers, from which results the possibility of great decisional risks regarding the real state assessment of this type of electrical equipment in operation, with important consequences regarding the costs for maintenance and even the power system security. Keywords: - Substation, power transformer, failure, life assesement, condition based evaluation, risk of management. 1. INTRODUCTION In the context of the energy system management restructuring, energy market deregulation and ever increasing competition, costs reduction and increase in the safe electricity supply have become objectives that the generation, transmission and distribution utilities have attached greater importance to as compared to the previous periods of time. Considered from the point of view of these objectives the correct operation and maintenance of the power transformers have become even more important due to the fact that transformers are among the most costly equipment in the substations and their failure influences not only the distribution and transport capacity, but also the economic efficiency of the utility. Operation and maintenance of the existing power transformers require a very good correlation between their constructive and operation characteristics, life time, operating conditions, requirements for the assessment of their actual momentary condition and monitoring policy. The risks raised by neglecting these correlations will be presented based on an international comparative analysis, by concrete examples the authors have encountered during certain diagnoses of large power transformers in the Romanian power system and based on the risk analysis model Fault Mode and Effect Analysis /1/. 2. ANALYSIS OF FALURES AND CRITICAL PARTS At present in the Romanian power system installations there are 339 power transformers (including autotransformers) the rated power of which ranges between 63 MVA and 440 MVA and rated voltage between 110 kv and 750 kv. Piata Alba Iulia Nr. 4, Bl. I3, Sc.1, Ap. 40, Sector 3, Bucuresti, ROMANIA cmoldoveanu@novaindustrialsa.ro **B-dul Gen. Gh. Magheru Nr. 33, Sector 1, Bucuresti, ROMANIA cradu@transelectrica.ro

2 The distribution of these power transformers by their useful life time, type of installation where they operate (power plants, substations), rated voltage of the high voltage winding, respectively, is presented in figures 1 and 2. Fig. 1. Transformer distribution by their operating time and the type of installation they operate Fig. 2. Transformer distribution by their operating time and rated voltage We notice that the useful life time of most of the large power transformers both in the power plants and substations is near or greater than 25 years, a life time considered as the standard useful life time. They have been designed and developed on the basis of the techniques and materials available at that time and their dimensions and weight are far greater and they are also far less efficient from the point of view of their own losses as compared to those that are produced at present. To failures, present in other power systems as well, we should add the ones that are due to the fact that at a certain time in the past the Romanian transformer designers and manufacturers could not utilize the high quality materials indicated in the technical design in the manufacturing process. The substitute materials supplementary weaknesses that influenced the transformer reliability and availability in operation. Table 1 presents the typical failures of the 1998 year, in comparison with the situation in other countries, relating almost to the same period of event analysis for power transformers rated more than 100 MVA, kv /2,3/. Table 1 Typical faults of the large power transformers in the Romanian energy system, in 1998, compared to the situation existing in the former CSI countries /1/, between 1994 and 1995 Failure mode Failure rates (%) Failure mode Failure rates (%) Romania Other countries Romania Other countries Winding failure 63,32* 30 On-load tap changer Mechanical failure of 7 11 Cooler system winding Local overheating, sparking 3 8 Other of magnetic core and the winding clamping system Bushings Total (*) the transformers whose insulation has not been within the allowable standard limits have also been considered. The main weak parts (critical parts) of the large power transformers in the Romanian power system are the following: a. - Windings (the most typical defects are characterized by decreasing of the insulation 2

3 parameters under the minimum allowable limits, thus creating the possibility of winding insulation breakdown due to overvoltages and diminishing resistance to electro-dynamic stress, respectively, as a result of the paper insulation settling). Favoring factors, especially in the case of old transformers: insufficient quality of the used materials; the constructive solutions less resistant to electro-dynamic stresses, (constructive solutions characteristic of the level of knowledge of the respective time); lack of oil conservators with oil protection against direct contact with the ambient. Direct consequences relating to the transformer availability: relatively low level of insulation, low mechanical resistance, especially in the tertiary winding, moisture penetration inside the transformer (thus diminishing the insulating resistance of the windings up to values under the allowable limits), accelerated oxidation (ageing) of oil (quicker than in the case of transformers having oil conservators with rubber bag oil preservation system). b. - Bushings (the most typical defects are characterized by modifying the internal state, cause insulation breakdown followed by explosion and transformer fire). Favoring factors: low quality of some of the bushings, especially the 145 kv multi-conductor insulated bushings initially reinforced with a paste with high sulfur content, impossibility to control the level and quality of oil of many of the bushings during operation. c. - The winding clamping system in certain cases some clamping pieces have been made of magnetic materials (a deficiency resulting in the overheating of the pressing part - bolt/washer- up to their thermal deformation, thermal degradation/decomposition of the insulating materials -transformerboard, oil- they come into contact with, and even electric discharges between clamping metallic pieces). The adopted clamping solutions allow the recovery of coil pre-compression of the power transformers only through untanking of their active part, in a specialized shop. d. - On-load tap changers (the most typical defects are characterized by loosing of the tightness, especially between the diverter tank and the transformer tank, contact damages, mechanical malfunctions of the driving elements). Favoring factors: old constructive solutions with low tightness to vacuum, lack of spare parts as a result of taking out of fabrication of old type switches, low reliability of the drives. e. - Magnetic core (diminishing of sheets, stacks of sheets, scaffolds insulation) Favoring factors: the relatively high amount of mechanical impurities and oil humidity that has determined the sheets and scaffolds short-circuiting, in several cases. f - Cooling system (diminishing of the cooling capacity, etc.) Favoring factors: improper construction solutions at the old coolers (tubes for the circulation of oil made of iron, with their inner part chemically unprotected); clogging of the external cooling channels with the industrial dust, undersized fan driving motors, etc. Among these the most important ones are the ones listed at points a, b, and c because their faults may cause serious failures, they are much more difficult to repair, may take the transformer out of operation for a relatively long period of time and their repair is costly. 3. MANAGEMENT OF TRANSFORMER OPERATION AND MAINTENANCE, BASED ON THEIR STATE DATA By carefully analyzing the data in Fig. 1 and 2 it becomes obvious that the management of the existing number of transformers belonging to the Romanian power system and their keeping in operation is a difficult task, especially due to the transformer age. This situation is met in other countries, as well. When a transformer surpasses its useful life time the evaluation of its momentary state and of its actual working capacity requires specific supplementary investigations and well documented conclusions due to the fact the same type of fault may be caused by several causes and it may take different forms whose scope may also depend on the age of the transformer. At present in Romania there are the main types of efficient test equipment and installations for carrying out these objectives (such as efficient laboratory gas chromatographs for the analysis of free and dissolved gases, liquid chromatographs for determining the content of the furan compounds, automatic particle counters for determining the level of oil pollution, devices for determining the water content in oil by automatic coulometric Karl Fisher titration, portable measurement systems for 3

4 diagnosis of the insulation condition performing automatically the polarization spectrum test, installations for detecting the mechanical deformation of the windings through the low voltage pulses method and the frequency response analysis method, respectively, etc.). Mention should also be made of the test installations and equipment based on the traditional verification methods included in the Romanian operation norm PE 116/4/ (methods included in the national norms/standards of other countries, too). At present there is only a limited number of these highly efficient investigation equipment and installations, distributed in several enterprises and at relatively long distances one from the other. This has the following consequences in the case of investigating a transformer in operation: - increases the investigation time; - increases the level of subjectivity of the results. The maintenance types applied to these large power transformers represent a combination of preventive, predictive and reliability based maintenance. Until now the advanced age of the high power transformers in the Romanian national power system has not been the main cause for their deterioration, but it has represented the main cause favoring it. The cost of replacing the old large power transformers is very high, and even if the funds necessary for their replacement were available this replacement operation would take relatively long (necessary for the development of the new technical specifications corresponding to the actual technological and operation requirements, for the manufacturing of the new transformers and replacement of the old ones). Without ruling out the necessity of replacing the old large power transformers, considering the economic and functional points of view, the efficiency of the existing transformer management is very important as it is directed towards their maximum utilization, to making the most efficient measures for preventing the major faults (that have a major impact on the economic and functional performances of the owner s installations and commercial operator), and applying the most adequate measures for repairing the registered faults, respectively. It is not enough to simply compare the results of the investigations with the technical requirements imposed by the operation norms or standards. As a wrong evaluation may lead to incorrect decisions in operation and consequently even to important damages (from among which only the transformer replacement cost amounts to about USD 750,000-1,500,000) the investigation result interpretation and establishment of the measures to be taken in operation in correlation with the momentary, actual state of the power transformer in operation is an issue implying great responsibility, which has important technical and economic consequences. That is why in Romania the analysis and interpretation of the results is carried out by some well-known experts in this field (selected to carry out consultancy on the basis of strict criteria) and by means of the TRANSPOWER specialized data base /5/, HUMIDITY Expert System /6/ and based on the information from the efficient on or off-line monitoring systems, Romanian or foreign ones /7/. The lack of expert participation in establishing the maintenance conditions and the analysis of the investigation results or the automatic application of standard requirements (no matter of which one) to any power transformer in operation may lead to wrong conclusions and decisions in operation and maintenance, that have unfavorable technical and economic consequences for the power companies. In order to exemplify the mentioned ideas, we will present the following cases we found interesting: 1. The 250/250/80 MVA transformer 400±8x1.56%121/20kV no /82 from the Constanta Nord substation has been placed under special surveillance in 1996 on the basis of the chromatographic analysis results. Finally, in 1998 it was taken out of operation only on the basis of the acetylene concentration value and its evolution with time (in comparison with the concentration of the other gases, table 2). It should be underlined that the concentration of all the analyzed gases (including acetylene) were well under the maximum limits allowed by the CEI or the Romanian equivalent standard, table 3/8/. The alarm state resulting from the chromatographic analysis was confirmed afterwards by the partial discharge measurements (in the case of the traditional measurements, in agreement 4

5 with the operation standard, the results obtained corresponded to the standard), as well. Table 2 Evolution of concentration of some of the gases dissolved in oil of the 250 MVA 400/121/120 kv no transformers in the Constanta Nord 400kV substation Date Concentration of the analyzed gases, in ppm Comment H 2 CH 4 C 2 H 6 C 2 H 4 C 2 H 2 CO CO ,0 13,8 6,3 36,8 4,8 475,8 3304, ,0 22,0 8,0 61,0 9,0 564,0 3304, ,1 24,1 9,4 61,8 10,2 454,1 2810, ,3 31,3 12,3 88,5 15,6 624,1 3743, ,9 13,7 97,1 17,8 637,0 3666, ,8 32,0 12,7 91,5 16,7 568,5 3380, ,8 8,8 4,8 28,1 4,1 90,8 1371,8 Oil degassed ,6 18,4 14,4 50,1 9,8 132,4 1666, ,3 13,7 10,6 44,6 13,4 142,6 2029, ,0 13,9 7,3 46,4 13,9 173,7 2098,0 Table 3 Normal range of concentration values for the 90% percentage of the power transformers (any type of transformer) Category of Analyzed gas concentration, in ppm transformers H 2 CH 4 C 2 H 6 C 2 H 4 C 2 H 2 CO CO 2 No OLTC Communicating OLTC When it was untanked the transformer at the factory it was found that the capacitive protection ring was broken and the solid insulation between the guard ring and the first two flat coils of the high voltage (400 kv) coil was local breakdowned, fig. 4 and 5. Fig. 4. Location of the fault within the assembly capacitive protection ring plus high voltage coil, phase A, corresponding to the 250MVA transformer Fig. 5. The guard ring in the upper part of the coil, with charring traces there where the conductive tape was interrupted (broken) By the timely taking out of operation of the mentioned transformer its severe damage was avoided. In normal conditions, this damage would have been accompanied by important failure (the losses could have surpassed the amount necessary for the transformer replacement, of about USD 1,350,000). 2. At the 200/200/60MVA autotransformer 231 ± 12x1,25%//121/10,5 kv no /82 from the 5

6 Gheorghieni substation there occurred the signaling of gas protection and flammable gas accumulation in the Buchholz relay. The chromatographic analysis results shown very high acetylene and hydrogen concentrations, characteristic of certain local over-temperatures surpassing 800 ºC and electric discharges of high energy inside the autotransformer tank. In the technical report interpreting these results of the chromatographic analyses the expert recommended the immediate taking out of operation of the autotransformer for checking, establishing and repairing the fault. After the first chromatographic analysis, the user carried out further investigations (including partial discharge measurements) and on the basis of the results of these measurements, carried out autotransformer fault identification actions on site (by taking out the oil from the tank and visually checking the condition of the active part, through the sight holes on the tank). As the visual inspection did not point out anything, the oil was partial degassed and reintroduced in the tank, the measurements required by the PE 116 norm repeated and the autotransformer was put back into operation. Nevertheless, it was decided to monitor its operation by means of chromatographic analyses at relatively short periods of time, table 4. Table 4 Evolution of some of the gases dissolved in the oil of the 200MVA 231±12x1.25%/121/10.5kV autotransformer no /82 from the 220kV Gheorghieni substation Analysis date Concentration of the analyzed gases in ppm Comment H 2 CH 4 C 2 H 6 C 2 H 4 C 2 H 2 CO CO Partial oil degassing The fault that had been initially identified continued to occur, increasing its negative impact. After about 2 months the autotransformer was finally taken out of operation and sent to a specialized repearing shop for establishing and repairing the fault. When the autotransformer was untanked, a charred coil compression set (bolt/washer) was found displaying clear overheating and metal thermal deformation traces in the area where the two parts came into contact (fig. 6, 7). Fig. 6. Metallic coil compression parts damaged by local very high temperatures and electric discharges of high energy Fig. 7. Metallic parts for coil compression of the same autotransformer, in good operating condition, which are not affected by any fault It was also noticed that the respective parts had magnetic properties. As they were in the way of the magnetic leakage flux, these parts got overheated and changed their form locally. The high temperature of the parts also damaged the solid 6

7 insulating materials and thermally decomposed the oil in the contact area or in its close vicinity. This set of parts was located in an area where the mentioned visual inspection, carried out in the substation, could not be achieved. It was obvious that even if the autotransformer had been maintained in operation it would have been severely damaged in a short period of time. 3. In many technical operation/maintenance norms/standards limit values have been established for certain parameters of the transformer insulation among which the water content in oil or the solid insulation. In the great majority of cases the reference conditions are not given for these limit values (for example: the transformer temperature, the conditions for bringing back the transformer in a balanced state of water distribution in oil and in the solid insulation, respectively, the mathematical relations for comparing the humidity level at the measurement temperature with that at the reference temperature, etc.). The issue is very important as an incorrect evaluation may lead to the taking out of operation of the investigated transformer and/or to useless maintenance costs (for on site drying) which are high (between 5 and 20% of the transformer cost). To see how difficult it is to find an answer to this rather simple issue the results of some measurements of the water content in oil carried out on two 200MVA 231/121/10.5 kv transformers of the same constructive type are given in table 5. Table 5 Evolution of the water content in oil (determined by means of the Karl-Fisher method) with the oil temperature, oil sampled from the lower part of the tank Type of investigated transformer Water content in ppm 23 C 28 C 46 C 60 C 75 C 85 C AT 200 MVA nr /73 from 400 kv Lacu Sarat substation 17,60 25,63 27,85 35,43 AT 200 MVA nr /67 from 400 kv Brazi substation 14,79 35,43 The tank level where the oil sampling was carried out may also influence the results of the analysis of the water content dissolved in oil at the same autotransformer. For example, in the case of the 200 MVA 231/121/10.5 kv no /73 autotransformer different results relating to the water content dissolved in the tank oil were obtained in agreement with the oil sampling place, for the same transformer temperature (23 C) : - Upper Level: 12,48 ppm; - Medium Level: ppm; - Bottom Level:17,60 ppm. 4. THE RISK OF INCORRECT EVALUATION The advanced life of the large power transformers in operation and their accessories may cause faults and breakdowns, which may pass unnoticed during current investigations or even during special investigations. Experience in operation has pointed out that severe breakdowns have occurred in large transformer units, shortly after their latest preventive investigation (with positive results), a fact that proves that the necessary diagnosis methods and equipment are still not available for the identification in due time of all the existing types of faults of the power transformers in operation. In this respect mention should be made of the case of the 250/250/80 MVA transformer no /85 from the Cluj Est 400kV substation. At the preventive investigation made in agreement with the PE 116/94 norm and the standard technical procedures, they have been found the values for the insulation resistance of the windings and the dielectric dissipation factor of the insulating oil, close to the permissible limits in the mentioned norm. On the basis of these investigations the user decided to carry out the refurbishment of the transformer insulation on site. On this occasion a visual inspection of the transformer active part was also carried out through the sight holes on the tank. Surprisingly, one of the 400 kv bushing was severely damaged in its lower part, fig. 8. This could have led to the pass insulation breakdown and explosion accompanied by the transformer setting on fire and its severe damage. The fault could not have been identified neither by 7

8 chromatographic analysis of the gases dissolved in the tank oil, nor by the traditional measurements carried out at the bushing (insulation resistance R C1, R C2, capacities C1, C2, dielectric loss factor tgδ C1 and tgδ C2 ). Fig. 8 The fault in the lower part of the 400 kv bushing of the 250 MVA 400/121/20 kv transformer no /85, identified during visual investigation through the sight holes on the tank This type of the fault can be probable monitoried by the measurement of the bushing leakage current. 5. CONCLUSIONS In the case of high power transformers whose life time is close to or even greater than the standard, designed life time, the diagnosis/monitoring policy and the maintenance programmes should be correlated with the real state of each transformer and its history. The constructive characteristics of the power transformers introduce supplementary elements which should be taken into consideration when evaluating the results of the investigations and making operation decisions. Lack of expert participation in the evaluation of the investigation/diagnoses results and of the information received from the on-line power transformer monitoring equipment, respectively, as well as the automatic application of the maintenance criteria/requirements from standards (norms) (any of them) to every power transformer in operation, may lead to wrong management decisions with unfavorable technical and economic consequences for the user. In order to make decisions in cases of power transformers breakdown, including the already mentioned ones, the authors of the paper have considered only the technical and economic aspects without taking into consideration the social aspects. 5. BIBLIOGRAPHY [1] C. Booth, et. al.: Information strategy to support utility asset management. Electra, n0. 207, apr. 2003, pg [2] V. Sokolov: Preferential objectives of power transformers on-line monitoring techniques. Proceeding of the CIGRE International Seminar 31 May 1996, ref.1.1. [3] V. Sokolov: Evaluation of effectiveness in-service diagnostic tests for large transformers. Proceeding of the CIGRE International Seminar 31 May 1996, ref.1.1. [4] PE 116 Standard for testing and measuring the electric equipment and installations (Romanian standard). [5] Transpower - system for the computer assisted maintenance activity of primary electrical equipment, license NOVA INDUSTRIAL. [6] Humidity Expert System for assessment of the humidity in power transformers using four methods, license NOVA INDUSTRIAL. [7] Trafomon - complex system for on-line monitoring of power transformers in operation and SMT2 - system for the on-line monitoring of transformer bushings, NOVA INDUSTRIAL catalogue or [8] CEI Mineral oil - impregnated equipment in service. Interpretation of dissolved and free gases analysis. 8