- \n
- Une connexion qui assure une \u00e9tanch\u00e9it\u00e9 parfaite peut tout de m\u00eame pr\u00e9senter un d\u00e9faut \u00e9lectrique : l'\u00e9tanch\u00e9it\u00e9 et le contr\u00f4le des contraintes r\u00e9pondent \u00e0 des probl\u00e8mes diff\u00e9rents.<\/li>\n\n\n\n
- Voltage class on the kit label isn’t a formality; it maps directly to the stress-control geometry molded inside.<\/li>\n\n\n\n
- Never substitute a kit rated for one voltage class “because the diameter fits” \u2014 diameter match doesn’t imply field-grading adequacy.<\/li>\n<\/ul>\n\n\n\n
Les principes physiques des contraintes \u00e9lectriques au point de raccordement du c\u00e2ble<\/h2>\n\n\n\n
![\"Concentration](\"https:\/\/zeeyielec.com\/wp-content\/uploads\/2026\/08\/zeeyielec-cold-shrink-stress-control-figure-01.webp-1024x559.webp\")
Vue en coupe transversale de la concentration du champ \u00e9lectrique au niveau d'un point de d\u00e9coupe non prot\u00e9g\u00e9 d'un \u00e9cran semi-conducteur sur un c\u00e2ble moyenne tension.<\/figcaption><\/figure>\n\n\n\n At the semi-conductive screen edge, the electric field is no longer distributed radially \u2014 it becomes tangential and highly non-uniform, peaking sharply right at the point where the screen terminates. This is a geometric consequence of removing a grounded conductive surface from immediately adjacent to an insulation system still carrying the full phase-to-ground potential. The field doesn’t weaken gradually moving away from the conductor the way it does under an intact screen; instead it spikes locally at the discontinuity, often reaching several times the field strength present under the undisturbed cable insulation a short distance away.<\/p>\n\n\n\n
Pourquoi un point de coupure non contr\u00f4l\u00e9 provoque-t-il une d\u00e9charge partielle ?<\/h3>\n\n\n\n
Partial discharge \u2014 a small, localized electrical breakdown that doesn’t fully bridge the insulation but still erodes material with each occurrence \u2014 begins once local field strength at the screen edge exceeds the discharge inception threshold of the surrounding insulation or the air\/insulation interface. On a typical 15 kV class distribution cable operating at normal system voltage, an unmanaged cut point can drive localized field strength into a range where partial discharge activity initiates at levels that would never occur under the intact, shielded portion of the cable \u2014 this is the core reason stress control isn’t optional at MV voltage classes. Each discharge event erodes a microscopic amount of material, and repeated cycling over months or years develops into an electrical tree that eventually bridges enough insulation thickness to cause flashover.<\/p>\n\n\n\n
Pourquoi cela rev\u00eat une importance accrue \u00e0 mesure que la classe de tension augmente<\/h3>\n\n\n\n
The relationship between voltage class and required stress-control performance isn’t linear. A termination on a 35 kV system doesn’t just need “more” grading than one on an 8.7 kV system \u2014 the geometry and permittivity of the element have to be engineered specifically for that higher operating field, since the margin between normal system voltage and discharge inception voltage narrows as voltage class rises. This is the physical basis for why cold shrink kits are sold as voltage-class-specific products: the geometry that adequately manages field concentration at 8.7\/15 kV would be undersized for a 26\/35 kV application.<\/p>\n\n\n\n
M\u00e9thodes de contr\u00f4le des contraintes g\u00e9om\u00e9triques et r\u00e9fractives (capacitives)<\/h2>\n\n\n\n
![\"Sch\u00e9ma](\"https:\/\/zeeyielec.com\/wp-content\/uploads\/2026\/08\/zeeyielec-cold-shrink-stress-control-figure-02.webp-1024x559.webp\")
Comparaison c\u00f4te \u00e0 c\u00f4te, sous forme de coupe transversale, entre la g\u00e9om\u00e9trie conique d'un c\u00f4ne de contrainte g\u00e9om\u00e9trique et la structure en couches d'un tube de contrainte r\u00e9fractif \u00e0 haute permittivit\u00e9.<\/figcaption><\/figure>\n\n\n\n Deux approches techniques permettent de r\u00e9soudre le m\u00eame probl\u00e8me de concentration de champ au niveau du point de raccordement du c\u00e2ble, et la plupart des terminaisons \u00e0 r\u00e9tr\u00e9cissement \u00e0 froid utilisent l'une d'entre elles ou une combinaison des deux, en fonction de la classe de tension et de la conception de la terminaison.<\/p>\n\n\n\n
C\u00f4nes de contrainte g\u00e9om\u00e9triques : propagation du champ m\u00e9canique<\/h3>\n\n\n\n
A geometric stress cone works by physically extending the effective transition distance between the grounded screen and the exposed insulation. Rather than a sharp cut, the cone’s molded profile creates a gradually widening conductive boundary that spreads the field over a longer physical path, lowering peak field intensity at any single point. This approach relies purely on shape \u2014 no special dielectric properties are required beyond the base elastomer (typically silicone rubber or EPDM) already used in the cold shrink body. Geometric control is common on 8.7\/15 kV class terminations where the required field-spreading distance stays within a practical cone length.<\/p>\n\n\n\n
Compos\u00e9s r\u00e9fractifs (\u00e0 haute permittivit\u00e9) : gradation capacitive<\/h3>\n\n\n\n
Le contr\u00f4le des contraintes par r\u00e9fraction, ou capacitif, repose quant \u00e0 lui sur un compos\u00e9 dont la permittivit\u00e9 est nettement sup\u00e9rieure \u00e0 celle de l\u2019isolation du c\u00e2ble environnante \u2014 souvent dans une plage de permittivit\u00e9 relative plusieurs fois sup\u00e9rieure \u00e0 celle du XLPE. Cette diff\u00e9rence de permittivit\u00e9 redistribue le champ de mani\u00e8re \u00e9lectrique plut\u00f4t que m\u00e9canique : le mat\u00e9riau \u00e0 haute permittivit\u00e9 attire les lignes de champ vers lui et les lib\u00e8re plus progressivement sur toute sa longueur, att\u00e9nuant ainsi le pic qui se formerait autrement au niveau du bord du blindage. Cette m\u00e9thode permet d\u2019obtenir un profil de terminaison plus compact, ce qui est important \u00e0 mesure que la classe de tension augmente et qu\u2019il devient plus difficile de trouver l\u2019espace physique n\u00e9cessaire \u00e0 un long r\u00e9tr\u00e9cissement g\u00e9om\u00e9trique.<\/p>\n\n\n\n
Recommandations relatives aux classes de tension pour d\u00e9terminer la m\u00e9thode \u00e0 appliquer<\/h3>\n\n\n\n
Param\u00e8tres<\/th> C\u00f4ne de contrainte g\u00e9om\u00e9trique<\/th> R\u00e9fractif (capacitif)<\/th><\/tr> Classe de tension type<\/td> 8,7\/15 kV<\/td> 15 kV\u201335 kV et plus<\/td><\/tr> M\u00e9canisme<\/td> Allongement physique du trajet<\/td> Permittivit\u00e9 (\u03b5r<\/sub>) \u00e9valuation des discordances<\/td><\/tr> Profil de r\u00e9siliation<\/td> Longueur de conicit\u00e9 plus importante<\/td> Longueur axiale plus compacte<\/td><\/tr> Sensibilit\u00e9 aux erreurs d'installation<\/td> Mod\u00e9r\u00e9e \u2014 tol\u00e9rance de position du c\u00f4ne<\/td> Plus haut \u2014 le positionnement du compos\u00e9 et un contact sans vide sont essentiels<\/td><\/tr><\/table>\n\n\n\n In field practice, some 26\/35 kV kits combine both mechanisms \u2014 a shorter geometric taper paired with a refractive layer \u2014 to hold the termination’s overall length within a practical range while still meeting the field-grading performance the higher voltage class demands. Neither method is inherently superior; the correct choice depends on the cable’s insulation thickness, the available installation space, and the kit manufacturer’s validated design for that specific voltage class rather than a generic rule of thumb.<\/p>\n\n\n\n
Que se passe-t-il lorsque la gestion du stress \u00e9choue ou est mal appliqu\u00e9e ?<\/h2>\n\n\n\n
![\"Propagation](\"https:\/\/zeeyielec.com\/wp-content\/uploads\/2026\/08\/zeeyielec-cold-shrink-stress-control-figure-03.webp-1024x559.png\")
formation progressive d'arborescences \u00e9lectriques prenant naissance au niveau d'un c\u00f4ne de contrainte mal positionn\u00e9, et se propageant progressivement du bord de la grille vers le conducteur.<\/figcaption><\/figure>\n\n\n\n Stress control failures rarely announce themselves at commissioning. A termination with a misseated cone or an incompletely bonded refractive layer will typically pass initial hi-pot or insulation resistance testing without issue, because the defect is a localized field concentration, not a bulk insulation problem \u2014 and bulk-focused tests aren’t always sensitive enough to catch it. The failure develops afterward, under normal operating voltage, over a timeline that can run from a few months to several years depending on how far off the installation was from the correct geometry.<\/p>\n\n\n\n
\u00c9tude de cas sur le terrain : positionnement du c\u00f4ne de contrainte d\u00e9plac\u00e9<\/h3>\n\n\n\n
In one recurring field pattern, a 15 kV cold shrink termination fails well ahead of its expected service life \u2014 sometimes within the first 12\u201318 months \u2014 and post-failure inspection traces the origin point to a stress cone that was seated a few millimeters short of its specified position over the screen cut-back. That small positioning error leaves a sliver of the screen edge outside the cone’s field-grading zone, recreating the exact concentration point the cone was designed to eliminate. The failure signature \u2014 a carbonized tracking path originating at one specific point rather than diffuse insulation degradation \u2014 is a strong diagnostic indicator that stress control geometry, not bulk cable insulation, was the root cause.<\/p>\n\n\n\n
Le processus de d\u00e9faillance progressive<\/h3>\n\n\n\n
The pathway from a misapplied stress control element to a full failure generally moves through three stages. First, partial discharge activity begins at the unmanaged concentration point, often at voltage levels the discharge wouldn’t reach anywhere else on a correctly terminated cable. Second, each discharge cycle carbonizes a microscopic amount of insulation material, extending a visible tracking path \u2014 commonly called electrical treeing \u2014 deeper into the insulation thickness. Third, once the tree bridges a sufficient fraction of the remaining insulation path, the reduced withstand capability leads to flashover, typically during a voltage transient such as switching or a lightning-induced surge rather than under steady-state load.<\/p>\n\n\n\n
Pourquoi les sympt\u00f4mes restent-ils cach\u00e9s jusqu\u2019\u00e0 un stade avanc\u00e9 ?<\/h3>\n\n\n\n
\u00c9tant donn\u00e9 que les d\u00e9charges partielles \u00e0 cette \u00e9chelle ne produisent aucun signe ext\u00e9rieur visible et n\u2019entra\u00eenent qu\u2019une variation marginale et difficile \u00e0 d\u00e9tecter du courant de fuite, les \u00e9quipes de terrain et les propri\u00e9taires d\u2019installations ne re\u00e7oivent g\u00e9n\u00e9ralement aucun avertissement avant la d\u00e9faillance, \u00e0 moins que le site ne soit \u00e9quip\u00e9 d\u2019un syst\u00e8me de surveillance continue des d\u00e9charges partielles \u2014 ce qui est rare sur les installations de distribution situ\u00e9es en dehors des grandes sous-stations. C'est l\u00e0 l'argument pratique en faveur d'un positionnement correct des dispositifs de contr\u00f4le des contraintes d\u00e8s l'installation, plut\u00f4t que de se fier \u00e0 une d\u00e9tection en aval.<\/p>\n\n\n\n
Les r\u00e9alit\u00e9s de la mise en \u0153uvre sur site : mise en place et v\u00e9rification du contr\u00f4le des contraintes<\/h2>\n\n\n\n
Le contr\u00f4le des contraintes de r\u00e9tr\u00e9cissement \u00e0 froid ne fonctionne exactement comme pr\u00e9vu que lorsque la g\u00e9om\u00e9trie de l'installation correspond \u00e0 celle valid\u00e9e par le fabricant lors des essais de type \u2014 ce qui fait des dimensions de pr\u00e9paration du c\u00e2ble et du positionnement du c\u00f4ne les deux variables qui d\u00e9terminent le plus directement si une terminaison tiendra toute sa dur\u00e9e de vie pr\u00e9vue de 20 \u00e0 30 ans.<\/p>\n\n\n\n
Tol\u00e9rances de pr\u00e9paration des c\u00e2bles ayant une incidence sur la position du c\u00f4ne<\/h3>\n\n\n\n
Every cold shrink kit specifies a cable preparation dimension sheet: how far to strip the outer jacket, where to cut the semi-conductive screen, and how far to bevel or taper that cut edge. These dimensions aren’t arbitrary \u2014 the stress cone or refractive tube is molded to sit at a specific distance from the screen cut, and a preparation error of even 5\u201310 mm can shift the cone off its designed position relative to the field concentration point it’s meant to manage. In field practice, the most common preparation error is cutting the screen too close to the connector, which leaves insufficient exposed insulation for the cone to seat correctly and forces installers to compensate in ways the kit wasn’t designed to accommodate.<\/p>\n\n\n\n
Contr\u00f4les visuels et dimensionnels avant le retrait du noyau<\/h3>\n\n\n\n
Before the removable core is unwound and the silicone body is allowed to collapse onto the cable, installers verify cone position against the manufacturer’s marked reference lines \u2014 most cold shrink kits print alignment marks directly on the cable prep or provide a separate marking tool for this purpose. A cone that sits even slightly forward or back of its reference mark at this stage is still correctable; once the core is removed and the body has collapsed under its own elastic memory, repositioning generally isn’t possible without cutting the kit off and starting over with a new one.<\/p>\n\n\n\n
Common Crew-Level Mistakes and How They’re Caught in QC<\/h3>\n\n\n\n