MV cable termination and testing field guide for 5 to 35 kV

What an MV cable termination is, and why it is the failure point

A medium-voltage cable termination is the engineered end of a shielded cable that controls electric stress and seals the cable where its metallic shield and semiconducting layers are cut back. Anything from 5 kV to 35 kV is medium voltage in common North American practice, and a shielded cable at that class cannot just be stripped and landed on a lug. Cut the shields back without managing the field and the termination tracks, treeing across the insulation surface until it flashes over.

MV cable almost never fails in the middle of the run, and that one fact decides everything about where you spend your attention. The extruded insulation, XLPE or EPR, is a manufactured, tested dielectric, and a buried or tray run sits there for decades. The failures cluster at the terminations and the splices, the two places a person opened the cable, interrupted the shield, and rebuilt the stress control by hand. That is why the trade treats the termination as the cable, and why the splicer's workmanship and the acceptance test carry the reliability of the whole circuit.

The work splits into two halves that have to agree. You build the termination right, to the kit dimensions, with the semicon off the insulation and the shield grounded. Then you prove it with a withstand test before it ever carries load, because a termination that looks finished and a termination that holds voltage are not the same thing. The only way to tell them apart is the test.

The layers of a shielded MV cable, and why each one matters

A shielded MV cable is built in layers, and every one of them matters when you cut into it. Work from the center out: the conductor, copper or aluminum, stranded; the conductor shield, a thin semiconducting layer extruded tight against the strands; the insulation, cross-linked polyethylene (XLPE) or ethylene propylene rubber (EPR); the insulation shield, a second semiconducting layer over the insulation; the metallic shield, which is copper tape, a concentric neutral, or wire; and the outer jacket.

The two semiconducting layers, the semicon, are the reason the cable works. The conductor shield smooths the field at the stranded conductor so there are no air gaps or sharp edges where the field would concentrate and start partial discharge inside the insulation. The insulation shield does the same on the outside, holding the insulation surface at a uniform potential so the field sits radial and even through the dielectric, the same on every side. Between the two semicon layers the insulation lives in a clean, equal-stress condition. That is exactly the condition a termination has to give up.

The metallic shield over the insulation shield carries charging current and fault current and keeps the outside of the cable at ground. The jacket is mechanical and environmental protection. When you terminate, you remove the jacket, the metallic shield, and the insulation-shield semicon in stages back to the kit dimension, and you expose bare insulation that used to live inside a grounded shield. That exposed insulation is where the trouble starts.

What is stress control, and why the cutback creates the problem?

Cut the metallic shield and the insulation-shield semicon and you end the equal-stress condition the cable was designed around. Inside the cable the field is radial and even. At the point where the shield stops, the field has nowhere uniform to go, so it crowds into the small region right at the cutback edge. The electric stress there can run an order of magnitude higher than anywhere in the run, high enough to ionize the air at the surface and start tracking across the insulation.

Stress control is the part of the termination that manages that crowded field so it does not destroy the insulation. There are two common ways to do it. A stress cone, geometric stress relief, extends the grounded shield outward in a cone shape so the field lines spread over a longer distance and the peak stress drops to something the insulation handles. Capacitive or refractive stress control uses a high-permittivity material at the cutback that redistributes the field electrically instead of geometrically, which is how most heat-shrink and cold-shrink kits do it in a thinner package.

Either way the job is the same: take the concentrated field at the shield edge and grade it back down. Get the geometry wrong, leave a void, or smear semicon across the insulation where the field is highest, and you have built the tracking path into the termination. Stress relief is not a feature of the kit. It is the reason the kit exists.

Termination classes under IEEE 48

IEEE 48 is the standard that sets the test requirements and the classes for AC cable terminations on shielded cable. It sorts terminations into three classes by what they have to do, and picking the right class for where the termination lives is a design call that often gets made by habit instead of by the standard.

A Class 1 termination does the most. It controls the stress at the shield cutback, provides full external leakage insulation between the conductor and ground, and seals the cable end against the environment. That is the termination for outdoor, exposed duty, where weather, UV, and contamination are in play and you need the long external creepage and the seal. A Class 2 termination controls the stress and provides external leakage insulation but is not held to the same environmental seal. A Class 3 termination controls the stress at the shield terminus only, for an indoor, protected, dry location where weather and contamination are not a factor.

The practical version: outdoor and exposed gets a Class 1, indoor switchgear and protected dry locations can take a Class 3, and Class 2 sits between. The current edition of IEEE 48 also distinguishes by insulation type, extruded versus laminated, so match the kit to the cable construction as well as the location. Confirm the class against the project specification and the manufacturer's listing, because the kit is built and tested to a class, and substituting down is how an indoor termination ends up failing outdoors.

Termination types: heat-shrink, cold-shrink, and taped

There are four families of MV termination, and the choice is about the install conditions and the connection, not just price. Heat-shrink uses polymeric tubing and stress-control components that shrink down over the prepped cable when you heat them with a torch or gun, bonding tight as they cool. They take an open flame and a steady hand, they hold up well mechanically and against chemicals and heat, and they are slower to install.

Cold-shrink uses pre-expanded rubber tubes held open on a removable plastic core. You slide the assembly over the cable, pull the core, and the rubber contracts onto the cable under its own elastic tension, with no flame and no heat. The rubber keeps constant radial pressure on the cable for the life of the termination, so it follows thermal expansion and contraction without the voids or adhesive cracking that can open up under a heat-shrink. In rain, humidity, wind, or a confined space where an open flame is a problem, cold-shrink is the safer and faster method, and it has taken over a lot of field work for that reason.

Taped terminations build the stress control and insulation up by hand from rolls of specific tapes to a published procedure. They are the oldest method, they depend entirely on the splicer getting the tension, overlap, and dimensions right, and they have largely given way to shrink and pre-molded kits for new work, though they still show up in repairs and where a kit does not exist for the geometry. The fourth family, the pre-molded separable connector, the elbow, gets its own section because it is how MV cable lands on padmount transformers, switchgear, and sectionalizing cabinets.

Load-break and dead-break elbows

A separable insulated connector, the elbow, is a pre-molded rubber termination that plugs onto a bushing on a transformer, switchgear, or junction, so the cable connects and disconnects without being cut. The elbow carries the stress control and insulation molded in, and it lands on a matching bushing interface standardized under IEEE 386 for systems 2.5 kV to 35 kV.

The two you handle are the 200 A load-break and the 600 A dead-break. A 200 A load-break elbow is built to be connected and disconnected under load, with a hotstick, by an operator switching the circuit, and it is limited to smaller cable, commonly up to 250 kcmil. A 600 A dead-break elbow, sometimes called a hammerhead, carries more current and lands larger cable, but it is bolted together and must only be operated dead, because it is not built to make or break load. Pulling a dead-break under load is a way to get hurt.

Treat the elbow as a termination, because it is one. The cable prep behind the elbow, the semicon cutback, the shield grounding, and the cleanliness, is the same discipline as any other termination, and a sloppy prep hides under the molded rubber where no one sees it until it fails. Match the elbow rating, the bushing interface, and the cable range to the application, and follow the manufacturer's lubricant and seating instructions, because a partially seated elbow is a high-stress, high-heat connection inside the molding.

The cutback and the prep, where the termination is won or lost

Every termination starts with the cutback, and the cutback comes off the kit instructions, not memory. The kit gives a set of dimensions, jacket cutback, metallic shield cutback, semicon cutback, insulation strip, and they are matched to that kit's stress-control geometry. Move the semicon edge a quarter inch and you have moved it out from under the stress relief. The manufacturer's dimensions govern, and on a 35 kV termination they govern harder than on a 5 kV one, because the field is higher and the margin is thinner.

Three things ruin a prep, and a good splicer is fanatical about all three. The first is nicking the insulation or the conductor. A knife cut into the XLPE or EPR at the semicon edge, exactly where the field concentrates, is a future failure site, and it may be invisible. That is why the trade strips semicon with guarded tools and scores the jacket without driving the blade home. The second is leaving semicon on the insulation. The insulation-shield semicon has to come off cleanly back to the dimension, and the surface under it has to be left with no smear, no embedded particles, nothing conductive. A film of semicon you can barely see is a conductive bridge sitting in the highest-stress region, and it tracks. You remove it, then clean the surface in one direction, away from the insulation toward the cut, with the solvent the kit specifies, never wiping back toward the clean insulation.

The third is dimension and roundness. The insulation has to be smooth and round for the stress cone or shrink tube to seat without a void, because a void is an air gap and air ionizes long before the insulation does. Glass the insulation smooth where the procedure calls for it, take the high spots down, and check the diameter against the kit range.

This is the work that does not show in a photo and does not show on a megger. It shows on the VLF withstand months later, or it shows as a flashover. The semicon smear left on the insulation is the single most common reason a hand-built MV termination fails, and it is pure workmanship.

Workmanship, cleanliness, and the qualified splicer

Cleanliness on an MV termination is not housekeeping, it is the dielectric. Once you have the semicon off, the insulation surface has to stay clean and dry through the rest of the build, because anything you trap under the stress cone or the shrink tube, dirt, skin oil, metal filings, moisture, becomes part of the insulation system at the worst possible location. Splicers work clean, keep the prep out of the weather, and do not touch the bare insulation with bare hands after it is cleaned.

The silicone grease that comes in the kit goes exactly where the instructions put it and nowhere else. It lubricates the rubber so a cold-shrink or an elbow seats without dragging and without trapping air, and it fills the microscopic gaps at the interface so there is no void. Too little and the rubber drags, tears, or leaves an air path. The wrong grease, or grease where the kit does not call for it, can attack the rubber or leave a conductive film. Use the kit's grease, where the kit says.

Moisture is the slow killer. A termination built in the rain, or on a cable that took on water at an unsealed end, or in high humidity with no protection, traps moisture that drives partial discharge and water treeing in the insulation. Seal cable ends the day they are cut, build in dry conditions or under cover, and do not leave a half-finished termination open overnight.

The reason all of this rides on the person is that none of it is inspectable after the fact. A hurried termination looks identical to a careful one until it fails, usually inside a year, and by then the crew is long gone. That is the argument for a qualified, trained splicer on MV work. The skill is not exotic, but it is specific, and the cost of getting it wrong is an outage and an arc.

Grounding the metallic shield and concentric neutral

The metallic shield gets grounded at the termination, and how it is grounded is a real decision, not a formality. The shield, copper tape, wire, or concentric neutral, sits at ground in normal operation and carries charging current and fault current. At the termination you gather it, often onto a drain wire or a shield-grounding braid in the kit, and bond it to the equipment ground. A shield left floating builds up voltage and is a shock and tracking hazard. A shield bonded poorly is not there when a fault needs the return path.

Whether you ground the shield at one end or both ends is a system decision with a tradeoff. Grounding both ends gives the shield a fault-current path and the best safety, but it lets circulating current flow in the shield on long parallel runs, which heats the shield and derates the cable. Single-point bonding kills the circulating current but leaves standing voltage on the ungrounded end that has to be managed. Most distribution and data center MV runs ground the shield at both ends; long transmission-class runs are where single-point bonding and the IEEE 575 practice for shield and sheath bonding come into play. Follow the design.

At the termination itself, keep the grounded shield connection away from the bare insulation between the stress cone and the ground point, because the whole point of the stress cone is to control where the shield ends electrically. Land the ground where the kit and the design put it, and verify continuity from the shield to building ground, the same shield-to-ground proof a busway housing gets across its joints.

Field testing the cable: what actually proves a termination

Two kinds of test live on an MV cable after it is terminated, and they answer different questions. The insulation-resistance test, the megger, is the first and the cheapest. You apply a DC voltage, commonly 2500 V or 5000 V on MV cable, between conductor and shield and read the resistance in megohms or gigohms. It catches gross problems: a dead short, a flooded cable, a shield touching the conductor, a badly contaminated termination. It is a go/no-go and a continuity check, and it is the right first move every time, the same insulation-resistance discipline used to baseline busway and switchgear.

What the megger does not do is prove an MV termination good. On 5 kV and up, a 2500 V megger is a small fraction of the cable's rated stress, so it tells you the insulation is not already failed, not that it will hold operating voltage plus the margin an acceptance test demands. A termination with a semicon smear or a small void can read gigohms on the megger and still fail a withstand, because the megger never puts enough field across the defect to find it.

That is why MV acceptance needs a high-voltage withstand test on top of the megger. The withstand applies a voltage well above operating level for a set time and proves the insulation and the terminations hold it. The megger says the cable is connected and not obviously dead. The withstand says it will live under voltage. You run both, in that order, and you do not call IR alone an acceptance.

VLF vs DC hi-pot: why DC is out for extruded MV cable

Very-low-frequency (VLF) AC withstand at 0.1 Hz is the modern field acceptance test for shielded MV cable, and DC hi-pot has been largely retired for extruded XLPE and EPR for a real physical reason. VLF applies an AC voltage at a very low frequency, around 0.1 Hz, so the test set can build the high voltage a cable needs without the enormous current a 60 Hz test of the same cable capacitance would draw. It stresses the insulation with AC, the way the cable actually runs, above operating level for a set duration. IEEE 400.2 is the guide for VLF field testing of shielded power cable systems.

DC hi-pot is the old method, and on aged extruded cable it can do harm. DC drives space charge into the insulation, charge carriers that get trapped and build up localized field intensification. When the DC comes off, or worse on a polarity reversal, that trapped charge can fail the insulation. DC also accelerates water treeing in service-aged XLPE, turning slow, benign degradation into a conductive path. The result is a test that can pass a cable and then leave it more likely to fail than before you tested it. The field has moved off DC hi-pot for extruded MV cable, and IEEE 400 and 400.2 reflect that.

DC still has a place on some laminated-insulation cable, paper-insulated lead-covered (PILC) and similar, where the space-charge problem does not bite the same way. But for the XLPE and EPR that make up nearly all new MV cable, the acceptance test is a VLF withstand, often paired with a diagnostic. If a spec still calls for a DC hi-pot on extruded cable, that is a flag to raise, not a step to run blind.

Tan-delta and partial discharge: finding a termination before it fails

The VLF withstand is pass or fail. The diagnostics that ride along with it, tan-delta and partial discharge, tell you how the insulation and the terminations are aging before they fail, which is what turns a withstand into a condition assessment. Both are run on MV cable per IEEE 400.2.

Tan-delta, the dissipation factor or loss angle, measures how much of the applied voltage is lost as heat in the insulation instead of being stored. Clean, new XLPE has a very low tan-delta. As the insulation takes on water trees, contamination, or age, the losses climb. You read the value, you watch how it changes as you raise the voltage, the tip-up, and you watch the trend over time. A high or rising tan-delta says the whole cable dielectric is degrading. IEEE 400.2 publishes criteria for what new, serviceable, and degraded insulation look like, and the numbers are small, measured in units of 0.001, so the test gear and the technique have to be good.

Partial discharge finds the local defect that tan-delta averages out. PD is a tiny spark in a void, a contaminant, or a badly prepped termination, discharging across a small gap under voltage. It is the exact mechanism that erodes insulation from the inside until it punctures. A PD measurement can locate the source along the cable, and a termination or splice is a common place to find it, because that is where a void or a semicon smear lives. Tan-delta tells you the cable as a whole is tired; PD tells you which termination is the problem. Together they catch a degrading termination while you can still schedule the fix instead of finding it as a fault.

What does NETA acceptance require for MV cable?

NETA acceptance testing, the ANSI/NETA ATS standard, is the field reference for what tests to run on MV cable before it goes into service and what the results should look like. For shielded MV cable it covers the visual and mechanical inspection, the shield-continuity and grounding check, the insulation-resistance test, and the high-voltage field acceptance test, with the test method and voltages drawn from the NETA tables, IEEE 400.2, and the cable and accessory manufacturer's data.

The acceptance battery for a shielded MV cable usually runs in this order: inspect and verify the installation and the shield grounding, megger conductor-to-shield to confirm the cable is sound and connected, run the high-voltage withstand, now commonly a VLF test, at the voltage and duration the table calls for, and where the spec requires it, take tan-delta or PD diagnostics. The VLF withstand voltage is set as a multiple of the cable's phase-to-ground rating, U0, and runs for a defined time, commonly on the order of tens of minutes, with the exact multiple and duration coming from the NETA table and the edition in force.

Do not carry the VLF voltage or the duration in your head as a single number, because they depend on the cable rating, whether the test is an installation acceptance or a maintenance test, the waveform, and the edition of the standard and the guide. Pull the value from the current NETA table and IEEE 400.2 for the specific cable, and confirm it against the project specification and the manufacturer, the same way the rest of the data center power acceptance battery is run to NETA ATS. The pass is a cable that holds the withstand for the full duration without breakdown, with the IR and any diagnostics in band.

Reading the test the spec actually requires

The test you run is the test the project specification and the owner call for, and on MV cable that is where old habits and current practice collide. A modern spec written to current IEEE and NETA practice calls for a VLF withstand on extruded cable, often with tan-delta or PD. An older spec, or one copied forward without review, may still call for a DC hi-pot, and on extruded XLPE or EPR that is the test that can damage what it is supposed to prove.

When the spec and good practice disagree, raise it before you test, in writing, not after. If a spec calls for DC hi-pot on extruded MV cable, the move is to flag it to the engineer of record and propose the VLF equivalent, with the IEEE 400.2 basis, and get the change in writing. Running a DC hi-pot on aged extruded cable because the spec said so, and leaving the cable worse than you found it, is not covered by following the spec.

Two more things the spec controls that people assume away: the acceptance voltage and duration, which come from the spec's referenced edition of NETA and IEEE 400.2 and not from the last job, and whether diagnostics are required or just the pass-or-fail withstand. Read the spec, confirm the edition, and confirm it against the cable and accessory manufacturer, because the manufacturer can cap the test voltage on their termination and that cap wins over a higher number in a generic spec.

Phasing, lug torque, and the IR scan after energizing

Before the cable carries load, prove the connection and the phasing, because a perfect termination on the wrong phase or a loose lug undoes the work. Confirm phase identification and rotation end to end, at the source and at every termination, so a three-phase load downstream does not run backward and so the phases land where the drawing says. On a cable that feeds switchgear or a transformer, phasing is verified before energizing, not diagnosed after a motor spins the wrong way.

The connection at the lug is its own failure point. The conductor lands in a compression or mechanical lug, and that lug is torqued to the value stamped on it or in the manufacturer's table, with a calibrated tool, the same bolted-connection torque discipline that governs busway joints and switchgear bus. An under-torqued lug runs hot under load, relaxes further, and burns; an over-torqued one can cold-flow aluminum and loosen the same way. Torque to the number, and mark the connection so a later thermal scan knows it was verified.

After energizing and loading, scan the terminations and connections with a thermal imager. A termination or lug running hotter than its neighbors at the same load is the one with a high-resistance connection or a developing problem, and the infrared scan finds it under real current long before it fails, the same energized infrared check the rest of the power chain gets. Record the load you scanned at, because a temperature rise only means something against the current that caused it.

Splices: the in-line joint and why it is a weak point

A splice is a termination turned inward. Where a termination ends the cable and seals it, a splice joins two cable ends and rebuilds the whole insulation system across the joint, conductor connection, insulation, both semicon shields, the metallic shield, and the jacket, in the middle of a run. The stress-control physics is identical: you have cut the shields, you have to manage the field where they end, and you have to do it on two cables at once, often in a manhole or a vault with worse access and worse conditions than a termination at a panel.

Use a splice when you have to, not because it is convenient. You splice to join two reels on a long run, to repair a fault, or to tap, and every splice you add is another hand-built point that can fail. A run with no splices and good terminations is more reliable than the same run with a splice in the middle, so the design minimizes them and the field does not add them casually.

When you do splice, it gets the same discipline as a termination and then some: the kit dimensions, the semicon off the insulation clean, no nicks, dry and clean conditions, the connector crimped or bolted right, the shields reconnected for continuity, and the same VLF withstand afterward proving the joint with the rest of the run. A splice in a wet manhole, built by someone in a hurry, is the classic MV failure, and it fails the same way a bad termination does, by tracking at the semicon edge.

Safety on an MV cable test

MV cable testing puts lethal voltage on a conductor on purpose, and the cable's own capacitance stores a charge that stays dangerous after the set is off. Treat the whole job as energized work governed by NFPA 70E and the site electrical safety program, performed by qualified persons, with the test area barricaded and attended at every access point so no one walks into a cable being energized.

The grounding discipline is what keeps it safe. Before anyone touches the cable, establish an electrically safe condition: de-energize, isolate, lock out, test for absence of voltage, and ground. During the test, ground every conductor except the one under test. After the test, the cable holds a stored charge on its capacitance and has to be discharged through a properly rated discharge stick or resistor and then grounded, and on a long MV cable the discharge is not instant, so the ground stays applied for an extended period, often hours, because the charge can rebound.

This is not the place for a shortcut or an unqualified hand. The combination of high voltage, stored energy, and a conductor that can recharge itself after you walk away is exactly the hazard the grounding rules exist for. A cable you tested an hour ago and assumed was dead can still bite. Ground it, keep it grounded, and verify dead with a meter you trust before it is touched. On MV the consequence of skipping it is fatal.

What to document

Lose the termination and test record and the circuit carries no proven history the day it faults. Capture enough that someone who was never on the job can tell what cable was installed, how it was terminated, how the shield was grounded, and what the tests read, all keyed to the circuit and the termination location.

Record the cable identification and the kV rating, the termination type and IEEE 48 class, the kit and manufacturer, the splicer, the insulation-resistance reading with temperature, the VLF or withstand voltage and duration and the result, any tan-delta or PD diagnostics, the shield-grounding configuration and the continuity check, the phasing, the lug torque, and the energized infrared scan with its load. Where a termination or splice was reworked, record it as found and corrected. The table below is the minimum spine, and it is the baseline every future maintenance test trends against.

Common mistakes

• Leaving a film of semicon on the insulation at the cutback, the conductive smear that tracks in the highest-stress region.
• Nicking the insulation or conductor with an unguarded blade at the semicon edge, an invisible cut that becomes a failure site.
• Building the termination to remembered dimensions instead of the kit's, moving the semicon edge out from under the stress control.
• Running a DC hi-pot on aged extruded XLPE or EPR cable, which can drive space charge and water treeing and leave the cable worse.
• Calling an insulation-resistance megger reading an acceptance test, when a megger only proves the cable is not already dead.
• Leaving the metallic shield floating or poorly bonded, so it holds voltage and the fault path is not there when it is needed.
• Using the wrong IEEE 48 class for the location, like an indoor Class 3 termination on an outdoor, weather-exposed duty.
• Building the termination in rain or high humidity, or on a cable with an unsealed end, trapping moisture in the insulation.
• Skipping or misplacing the kit silicone grease, dragging the rubber and trapping air voids at the interface.
• Leaving the cable charged after a test, or pulling grounds too soon, when MV cable capacitance recharges and stays lethal.

Field checklist

Standards and references

The acceptance testing framework is ANSI/NETA ATS, the Standard for Acceptance Testing Specifications for Electrical Power Equipment, which carries a section for shielded MV cable covering the inspection, the shield grounding, the insulation-resistance test, and the high-voltage field acceptance test, with the test voltages and durations in the NETA tables. The field high-voltage test on extruded shielded cable follows IEEE 400 and IEEE 400.2, the guides for field testing of shielded power cable systems, with IEEE 400.2 specific to very-low-frequency (VLF) testing and the tan-delta and PD diagnostics. Confirm the editions in force, because the test values and the guidance move between cycles.

The terminations themselves are built and tested to IEEE 48, the standard for AC cable terminations on shielded cable, which sets the Class 1, 2, and 3 requirements. Separable insulated connectors, the load-break and dead-break elbows, follow IEEE 386 for the bushing interfaces. Shield and sheath bonding on single-conductor cable is covered by IEEE 575. The cable itself is built to ICEA and AEIC specifications, such as the AEIC CS series for extruded MV cable, and the cable and the termination kit manufacturer's instructions govern the cutback dimensions, the test-voltage caps, and the installation, over any generic reference.

Worker safety during the energized test follows NFPA 70E and the site electrical safety program, and the installation is built to the NEC, NFPA 70, as adopted and amended by the jurisdiction. Confirm the specific editions, the rated values, and the test parameters against the approved submittal, the project specification, and the manufacturer, because the requirement that controls is set by the project and the listing, not the rule of thumb. Where the spec and current practice disagree, the engineer of record and the AHJ resolve it.

Units, terms, and conversions

MV termination and testing crosses ratings, test units, and a stack of trade terms, so the same item reads differently across a cable reel tag, a kit instruction, and a test report.

Voltage is given as the cable class in kV, with U0 the phase-to-ground rating the test voltage is set against, not the phase-to-phase number. Insulation resistance reads in megohms or gigohms, recorded with temperature because it swings with it. VLF test voltage is stated as a multiple of U0 at 0.1 Hz for a set duration in minutes. Tan-delta is a dimensionless loss number, reported in units of 0.001, often written E-3. Lug torque is in lb-ft or in-lb in US units and N-m in metric, to the manufacturer's value. Cable size is AWG and kcmil in US practice and mm squared in metric.

Semicon

The thin semiconducting layer over the conductor and over the insulation that smooths the field and sets the equal-stress condition

Stress cone

Geometric stress relief that extends the grounded shield in a cone so the field spreads and the peak stress at the cutback drops

Stress control

The part of a termination that grades the concentrated field at the shield cutback so it does not track and fail

IEEE 48 class

The termination class, 1 through 3, set by whether it controls stress, insulates externally, and seals against the environment

VLF

Very-low-frequency AC withstand testing, around 0.1 Hz, the modern field acceptance test for shielded MV cable

Tan-delta

Dissipation factor, the share of applied voltage lost as heat in the insulation, a diagnostic of overall insulation condition

Partial discharge (PD)

A tiny localized discharge in a void or defect under voltage, the mechanism that erodes insulation and a key termination diagnostic

Cold-shrink

A pre-expanded rubber termination that contracts onto the cable under its own tension when the core is pulled, with no heat

Concentric neutral

A metallic shield of wires wrapped around the cable that also carries neutral and fault current and holds the cable at ground

Related tools

Calculators and readiness checks for this work

Compare your options

FAQ

People also ask

• What megger voltage do I use on busway?
• How long does data center electrical commissioning take?
• What is a stress cone?
• What is a TTR test?
• What tests do I run on a transformer when it arrives?