ANALYSIS OF FAILURE IN POWER CABLES FOR PREVENTING POWER OUTAGE IN ALEXANDRIA ELECTRICITY DISTRIBUTION COMPANY IN EGYPT

Transcription

1 24 th International Conference on Electricity Distribution Glasgow, 2- June 27 Paper 89 ANALYSIS OF FAILURE IN POWER CABLES FOR PREVENTING POWER OUTAGE IN ALEXANDRIA ELECTRICITY DISTRIBUTION COMPANY IN EGYPT Amani ATTIA Alexandria Electricity Distribution Company (AEDC) Egypt ABSTRACT Underground cables are fundamental elements in any energy system, as they allow energy to be transported and distributed to consumer points. Power cables are usually subjected to electrical, thermal, mechanical, and environmental stresses on a constant basis when they are in service. These stresses, will lead to insulation degradation or other defects, such as, excessive sheath circulating current, partial discharge activity and increasing dielectric loss which can result in premature cable failure causing unplanned outages. This paper will show a better understanding of MV underground cable failure in Alexandria Electricity Distribution Company (AEDC) by analyzing data in connection with cable failures and by identifying the stress factors leading to cable defects or degradation. This will be supported by several diagnostic field examples in AEDC. In addition, the techniques adopted for preventing and solving MV underground cable failure problems are evaluated. These techniques mainly depend on the predictive maintenance (PM) program which is cost effective providing minimum downtime on one hand and improving the system reliability on the other hand, in turns reducing the amount of outages that was 926 in 27/ 28 and reaches 9 In 2/26 as a sign of improvement in the underground cable network's performance. Also, the causes of cable failure are decreased during the same period. Consequently, AEDC enhances and preserves the system reliability which will lead to a high level of continuity and quality of supply to customers specially that MV underground cable failure have been contributing 6% of annualized System Average Interruption Duration Index (SAIDI) for a number of years. INTRODUCTION In the last few decades the main focus of utility companies has been on developing diagnostic, location and pin pointing techniques for medium voltage cables. As cables and their accessories age, their propensity to fail in service increases. Experience obtained while conducting predictive diagnostic evaluations of cable demonstrates that cable deterioration manifests itself through discrete defects. Some examples of discrete cable insulation defects are electrical trees, water trees eventually leading to electrical trees, impurities, delamination of semi-conducting screens, protrusions in extruded insulations and carbonized tracking in laminated insulations. Accessories, on the other hand, typically fail because of manufacturing defects, poor workmanship, impurities, or moisture ingress along interfaces with the cable. AEDC network is very complex network and has large lengths of underground cables which mostly oil paper cables and XLPE cables. MV underground cable in AEDC suffers from an increasingly faults which occurs at the cable sealing box, the cable joints and the cable connection box. Another very common reason is the insulation degradation and ruptures in conductors. In addition, the cable may also get damaged due to vibration fatigue or overheating. Moreover, such cable failure will cause many undesirable effects and will lead to a high cost to the society as a whole specially for institutions like hospitals, airports, and train stations, power outages can be disastrous. One drawback of underground cables is that the procedure for finding the exact place of failure is harder, since no visual inspection can be performed. In addition, even when a fault is localized, the process of digging the ground to reach the cable, and also repairing the cable, is very difficult. Consequently, there is a great need for better methods to determine the condition of the in-service underground cables and their remaining useful life. In this paper, the testing devices used by AEDC in order to determine MV fault location are described. Also, the flow chart of a cable test is shown in section II. Then, AEDC experiences are analyzed in section III. In addition, the maintenance strategy for MV cables adopted by AEDC and its role in improving the amount of outages during the period 27-2 is evaluated in section IV. Finally, the concluding remarks are summarized. TESTING DEVICE The testing device used in this study is the cable test van which consists of a high voltage device, very low frequency (VLF) device with testing connection output, an impulse voltage device, Teleflex device, arc stabilizing unit, audio frequency generator, burning and testing unit, audio frequency receiver, measurement device and cable identifying generator. This is shown in the fig. () []. The testing voltage applied should be less than the allowable voltage according to the equation () to avoid more insulation failure [2]. Testing Voltage 2 or operating voltage CIRED 27 /

2 24 th International Conference on Electricity Distribution Glasgow, 2- June 27 Paper 89 Start system:. Establish the power supply 2. Set system to the Ready for operation state 3. Switch on industrial PC Start BAUR software for cable testing Enter cable data or load a project HV switch 2 Time domain reflectometer IRG 3 incl. monitor 3 Audio frequency transmitter TG 6 4 HV tester PGK / Low voltage selection unit LP3 6 Arc burning unit EB26. 7 Current limiter SB6 8 Burn down transformer. 9 Mains supply unit MS 23. Surge voltage generator. System coupling SA 32 2 Control Unit MGS 32 Fig. () The cable testing and fault location system transcable in AEDC Manual test Select manual or automatic test Start test Automatic test Set system to the Ready to switch on state. (Press Ready to switch on button) The cable testing uses BAUR software in order to determine the fault location which is shown in the following flow chart []. End manual test End automatic test Prepare test object. Activate 2. Secure against re-energisation 3. Ensure that there is no voltage 4. Short circuit the test object to earth. Protect or isolate the test object from adjoining live HV plant. Set system to the Ready for operation state Discharge, earth and short- circuit test object and all live parts Earth and connect system View and edit report Fig. (2) Flow Chart adopted by AEDC for determining the fault location in the cable test van CIRED 27 2/

3 24 th International Conference on Electricity Distribution Glasgow, 2- June 27 Paper 89 CASE STUDIES AEDC medium voltage network consists of two main types of cable design. The first one is the oil impregnated paper cable which usually suffers from the high leakage current that occurs in the non- homogeneous points and weakness points [2]. This cable type is rarely used nowadays in AEDC network. The second type is the cross linked polyethylene (XLPE) cable which may fail due to the moisture ingress into the insulation, combined with imperfections that created high electrical stress points leading to water tree growth [3]. Time Domain Reflection (TDR) technique is applied to determine the total length of a cable, the location of low resistive cable faults, the location of cable interruptions and the location of the joints along the cable. This is used to determine the cable fault type (e.g. earth fault, cable sheath fault, open conductor and failure of insulation between conductors without earth). CASE STUDY NO. The cable under test is (x 4) XLPE and exists in the transformer point number 364 in El Agamy substation in West Zone. Its length is 34m. In this case study, when technicians prepare the lighting earthing network, this cable is failed due to iron which is inserted in the cable. Figures (3 and 4) show that the phases and 3 are healthy. Figure () shows that phase 2 has a fault at a distance of about 8 m. Fig. () The wave propagation on phase 2 (faulted phase). This fault is solved by creating connection box. CASE STUDY NO.2 The cable under test is (x 4) XLPE and exists in the distribution point El Souk El Hora El Gedid and feeds El Nasreya area in El Amreya in West Zone. Its length is 6 m. In this case study, constructions which occurred during repairing the railway station lead to a fault in this cable. This fault point happened at 26m from El Nasreya which is shown in fig. (6). A short circuit exists in one of the three phases and this is solved through a connection box. Fig. (6) The wave propagation of the tested cable in the distribution point El Souk El Hora El Gedid. Fig. (3) The wave propagation on phase (healthy phase). Fig. (4) The wave propagation on phase 3 ( healthy phase). The figure shows the wave propagation of the two waves, the main wave (wave ) and the secondary wave or fault wave (wave 2). The two waves are corresponding with each other until fault point, at which the main wave is reflected in a positive reflection and the faulted wave is reflected in a negative reflection. The reflection of the two waves will still opposite to each other until the end of the cable. The two wave's reflections are returned to corresponding and change with each other after the end of the cable (after cable length), means that distance of two opposite reflection equal to cable length. CASE STUDY NO.3 The cable under test is (x 4) XLPE and exists in the distribution point El Tawasoate in Industrial zone 3 in Borg El Arab in West Zone. In this case study, a bridge is constructed which lead to a fault in this cable. The only solution in this case study is changing the cable path under the bridge in a bus bar created on the form of U in order to prevent its fail another time. CIRED 27 3/

4 24 th International Conference on Electricity Distribution Glasgow, 2- June 27 Paper 89 CASE STUDY NO.4 The cable under test is (x 4) XLPE and exists in the distribution point 6 October which is fed from El Agamy Substation I, II. Its length 86 m. This cable failed due to the insulation breakdown which lead to welding between the earthing and the conductor which is illustrated in fig. (7). The fault point occurred at 2m. Fig. (7) The insulation breakdown This insulation failure in a cable network may be caused by: lower dielectric strength due to aging processes. by internal defects in the insulation system. an external factor like e.g. poor workmanship installation or assembling works (missing insulation screen at the cable/joint transition or cable splice improperly prepared (cable splice not round) especially XLPE cables [4]. This is solved by creating a connection box. AEDC CABLE MAINTENANCE STRATEGY AND ITS EFFECTS ON INTERRUPTIONS The maintenance policy of the distribution network is a complex action, made up of a large number of activities, moving different equipment. It is carried out within a strict regulation framework and must satisfy with the requirements of quality and traceability. Even, if these activities can be grouped into recurring phases (removal, maintenance, etc.). The Maintenance consists of inspecting in a periodic way the network in order to its control state and its performance level. The network maintenance can be preventive (conditional and systematic) on the basis of a periodic visit in order to ensure a good performance and to reach a technical life span beforehand established, or curative showing itself by repairs with the occurrence of breakdowns and unforeseen events []. AEDC adopted a preventive maintenance program for MV cables. Electrical engineers disconnect feeders and sub-feeders according to a schedule which is repeated each six months for all feeders and sub- feeders in Alexandria in order to check up all cables to predict faults before it occurs. This is can be done using the cable test van. The testing voltage applied according to table (2) which is approved by Egyptian Electric Holding Company (EEHC): Cable Type Oil Impregnated paper XLPE voltage network. k.v XLPE voltage network 22k.v Cable Status Tested Voltage Test Time new 26 k.v min. being 2 k.v min. new 2 k.v min. being 8 k.v min. new k.v min. being 3 k.v min. AEDC cable preventive maintenance program goal is to supply reliable power to supply reliable power to customers at low cost, prevention of power system failures is of paramount importance during the design and operation of the system. Consequently, cable preventive maintenance program in AEDC can have a significant effect on the achievement of high reliability which can be illustrated in table () that shows the interruptions number that decreases from 926 in 27/ 28 to 9 in 2/ 26. Table (2) The interruptions number during the period 27/28 2 /26 Year Outage Number 27/ / / / / / / /2 96 2/26 9 In addition, the preventive maintenance policies adopted by AEDC are aimed to detect cable deterioration of the equipment before it fails; in turn the major causes of interruptions to the system can be determined and can be prevented. The classification of interruptions by cause is shown in figure (8) Table () The tested voltage applied in AEDC cable preventive maintenance program Cable Failure System Failure Percent (%) Constructions Cable Connection Box Cable End Box Fig. (8) The relation between the percent of system failure and the interruptions main causes during 2/ 26. CIRED 27 4/

5 24 th International Conference on Electricity Distribution Glasgow, 2- June 27 Paper 89 This figure illustrates that the percent of system failure during 2/26 is affected by a series of factors can be summarized as follows:. Cable Failure 2. Cable Connection Box 3. Cable End Box 4. Constructions AEDC created a system for collection, processing and reporting interruptions main causes data during the period 27/28 2/26.This leads to the percent of system failure improvement which will be shown in the following figures that will show each factor amelioration during the period 27/28 2/ /28 System Failure Percent due to Cable Failure (%) / / /28 29/2 2/2 2/22 22/ /24 System Failure Percent due to Constructions (%) 28/ / /2 2/ /23 23/24 System Failure Percent due to Connection Box (%) / /2.7 2/2.2 2/ /23 24/ / Fig. (9) The system failure percent due to cable failure during the period 27/28-2/ 26. Fig. () The system failure percent due to constructions during the period 27/28-2/ 26. 2/ / Fig. () The system failure percent due to connection box during the period 27/28-2/ /24 24/2 2/26 CONCLUSION AEDC distribution networks faced great problems as they constructed from many years and there is not any accurate mapping for cables routing. Also, the networks contain different types of cables. The life time of a great number of cables is finished and becomes invalid. The above problems generate a large number of failures in the distribution network so that AEDC has adopted a preventive maintenance program in order to minimize the cable failure and to improve the system efficiency and power quality. This is proved through the improvement in: the number of interruptions which was 926 in 27/ 28 and now is 9 in 2/26.. the percent of system failure due to cable failure which is ameliorated from % in 27/28 to 38.9 % in 2/26. the percent of system failure due to cable connection box that becomes 3.6% in 2/26. the percent of system failure due to cable end box that becomes.3% in 2/26. the percent of system failure due to constructions that decreased to 8.%. Concluding, AEDC enhances and preserves the system reliability and reduces the maintenance costs which leads to a high level of continuity and quality of supply to customers REFERENCES [] BAUR Cable Testing and Diagnostics User Manuel. [2] M.S. El-Bages, M.A.Abd - Allah, and M.Z.A.El hawary, 23, "Long Term Analysis of Cables Problems in Egyptian Distribution Network", International Journal on Electrical Engineering and Informatics, vol., [3] J.Tanaka, 23, " History of Underground Power Cables", IEEE Electrical Insulation Magazine, vol. 29, no.4, 2-7. [4] P.Cichecki, R.A.Jongen, E.Gulski and J.J.Smith, 28, " Statistical Approach in Power Cables Diagnostic Data Analysis", IEEE Transactions on Dielectrics and Electrical Insulation, vol., no.6, [] M.Morad, E.Abdellah and E.Ahmed, 23, " MV Electrical Network Maintenance Strategy: A New Management Approach", ARPN Journal of Engineering and Applied Sciences, vol.8, no.2, System Failure Percent due to End Box (%) /28 28/29 29/2 2/2 2/22 22/23 23/24 24/2 2/26 Fig. (2) The system failure percent due to end box during the period 27/28-2/ 26. CIRED 27 /