FEED-BACK OF MEDIUM VOLTAGE POWER CABLES …

1 CC II RR EE DD17th International Conference on Electricity DistributionBarcelona, 12-15 May 2003 CES_deNigris_A1 Session 1 Paper No 63- 1 - FEED-BACK OF MEDIUM VOLTAGE POWER CABLES RELIABILITY IN ITALY :SYSTEMATIC analysis OF FAILED COMPONENTSMICHELE de NIGRIS, GIUSEPPE RIZZI,AMBROGIO TOMBACESI TRAMUTOLI, CARMINEPETRUNGAROENEL DISTRIBUZIONE SpA CABLES and accessories failures represent, on theENEL POWER distribution network, the second cause ofunavailability, after lightning induced overvoltages andtheir relative consequences. In fact POWER CABLES , after theeffective controls undergone during the productionprocesses must be laid down and, above all, jointed andterminated in the field.

2 The structure and the conditions ofthe cable network is not uniform within the Country:although, at national level, of the MEDIUM voltagenetwork is underground, this ratio ranges widely fromregion to region: from a minimum of 20% central-southernregions, to a maximum of 45%-55% of northern regionsand heavily urbanized southern locations. VOLTAGE levelsare also not uniform: northern regions are operated mainlyat 15kV (even if the equipment is very often dimensionedfor 20kV), while 20kV is mainly used in the south, with amixed situation in the central part of the Country.

3 This hasa strong influence on the failure rates of the 1 shows (data 2000) the failure rate of distributionunderground CABLES in the different parts of Italy, in termsof number of failures per 100km per year. The left column(red) refers to the total failure rate, considering all failurereasons, while the right column (blue) reports the netfailure rate, the rate of failure linked with internalreasons of CABLES and accessories (excluding thereforefailures linked with excavation works or other externalcauses). It can be observed that central and southernregions suffer a higher failure rate (also linked with thehigher margins allowed in the northern regions by thereduced distribution VOLTAGE level) and that external causesare responsible for nearly 50% of the ITALYCENTRAL ITALYSOUTH ITALY Figure1.

4 failure rate in the different parts of ENEL distribution network (data 2000) In terms of contribution of the different parts of CABLES tothe failure rate, Figure 2 shows, for the three main Italianregions the percentage of net failures (NFr) linked withdefects of the cable or its ITALYCENTRAL ITALYSOUTH ITALYCONTRIBUTION TO NFr [%]CABLEJOINTTERMINATIONF igure 2: Contribution to the net failure rate of cable partsDRIVERS AND METHODOLOGYThe stringent requirements from the National Authority interms of improvements of POWER quality, number andduration of supply interruptions has motivated toinvestigate more in details the causes of failures, especiallyof those occurring on components after a limited servicelife.

5 Moreover, in the specific case of POWER CABLES , whereseveral subjects interact between the manufacturing and theoperation ( cable and accessories manufacturers,excavation companies, laydown companies, joint andtermination operators, commissioning operator, operationand maintenance departments etc.) it is essential to pointout responsibilities in order to set up effective remedialactions. With this goal in mind, an extensive survey offailure causes was set up by ENEL Distribuzione SpA incollaboration with the Centre for failure analysis of project consisted in collecting a significant quantity offailed cable accessories from all regions of Italy andsubjecting the samples to a detailed failure analysis process,often with the presence of the manufacturer and theaccessory operator.

6 Precise methods for sample collectionand expedition were set up in order to gather essentialinformation on service stresses and failed sampleconditions at the moment of extraction and to ensure that noincorrect manipulation occurred on the way to the FailureAnalysis Centre, potentially masking failure , in order to rationalize the output of these failureCC II RR EE DD17th International Conference on Electricity DistributionBarcelona, 12-15 May 2003 CES_deNigris_A1 Session 1 Paper No 63- 2 - analysis , specific procedures were set up for the inspectionprocess and codes for an effective failure classificationwere determined.

7 To ensure the widest dissemination of thefailure analysis results an Intranet site was developed inwhich all failure reports were collected and a database wasimplemented to allow statistical MODES AND DEFECTSF ailure modes of POWER CABLES can be subdivided into threemain categories: External flashovers; Internal longitudinal discharges; Punctures In our analysis external flashovers (such as surfacepollution, surface flashovers linked with damages of thesurface insulation linked with mechanical shocks,application of an inadequate type of accessory for the localpollution conditions) were excluded because they areresponsible for short duration failures or are linked withexternal causes (shocks) and because the relevant remedialactions are to be searched locally and not at a nationallevel.

8 Internal longitudinal discharges, occurring on theinterfaces inside the accessory and whose path mainlydevelops in the direction of the cable , are very often linkedwith assembly problems: non correct finishing,imperfect sealing, inclusion of contaminants, incorrectcouplings etc. Punctures develop radially in the cableaccessory; they may be linked with ageing phenomena(reduction of dielectric strength, trees etc.), to the presenceof defects such as roughness, inclusions, protrusions in theassembling process or to the detachment of thesemiconductive layer.

9 In terms of physical phenomena leading to the failure wehave used the following classification: ( ): Presence of voids; ( ): Presence of defects leading to the increase of theelectrical stress; ( ): Presence of defects linked with the penetration ofwater or contaminants in the in detail, during the internal inspection of the failedcomponents the expert classified the defects found in thecategories shown in Table results of our investigation will be referred to the typeof defect and to the associated phenomena. Wheneverdeemed necessary, more sophisticated analysis were carriedout to determine the primary cause of the failuresconsidered: mechanical tests were carried out on plasticmaterials samples to point out degradations due to ageing,thermo-gravimetric and infrared tests were conducted todetermine the conformity of the materials composition tothe original declaration of the manufacturer, scanningelectron microscope (SEM) analysis were made to verifythe eventual presence of foreign materials in the failurepath or close to it.

10 Most of the complementary analysiswere carried out comparing the characteristics of theanalysed samples with those of similar samples taken fromnew cable 1: classification of defect( ), ( ), ( )1 Bending( )2 Damages on outer sheath( )3 Lack of adherence of outer sheath( ), ( )4 Presence of pollution on insulation( )5 Incorrect assembling of electric fieldcontrol( )6 Incorrect sealing( )7 Incorrect metallic screen assembly( )8 Incorrect conductor connectionpreparation( )9 Incorrect type of connection( )10 Damages of CABLES insulation( ), ( )11 Incorrect removal of insulationsemiconducting screen( )12 Incorrect ground connection assembly( )13 Incorrect distances and dimensions( )14 Presence of humidity( )15 Other defects( ), ( ), ( )Examples of defects.