Medium-voltage switchgear maintenance and testing field guide

What switchgear maintenance is, and why the gear fails silently

Medium-voltage switchgear maintenance and testing is the work you do, mostly de-energized, to confirm that the gear distributing and protecting the facility power will still do its job when a fault hits. It distributes power to the building and it protects every circuit downstream, and it does both while sitting energized and almost completely idle for years at a time.

That idleness is the problem. The gear that protects the whole facility fails silently until it fails catastrophically. A bolted connection works loose under thermal cycling and starts making heat. Dust and moisture settle on an insulator and start a tracking path. A vacuum bottle loses its vacuum a little at a time. A breaker that has not operated in five years stiffens in its mechanism. A protective relay that was set during construction and never tested drifts or was wrong from day one. None of that shows on a walk-by. The bus reads full voltage and the lights are on.

Then a fault hits, and the gear that should clear it in a few cycles instead fails into an arc flash. Proactive maintenance and testing find those slow problems while the gear is dead and safe to open, which is the whole point. The inspection, the electrical tests, and the protective-relay testing on a NETA schedule are what keep it reliable. This guide walks the program. The transformer-acceptance-testing guide covers the unit feeding the gear, and the arc-flash guide covers the hazard you are working inside.

Why maintain gear that has run fine for years

Because it has run fine for years is exactly the reason. Switchgear is not like a motor that tells you it is sick by running hot or loud. It sits there, energized, doing nothing visible, while the slow failures accumulate with no symptom you can see from the floor.

Walk the failure forward. A connection loosens, so it heats, so the heat oxidizes the joint, so it heats more. Contamination and moisture build a creepage path across an insulator until one humid morning it tracks over. A breaker mechanism that has not cycled goes stiff with old grease. A relay nobody injected sits with a setting that will not trip in time, or will not trip at all. Each of these is invisible right up to the fault that exposes it.

The case for maintenance is the consequence of skipping it. When a fault arrives and the protection is slow or dead, the energy that should have been cleared in a few cycles keeps pouring into the arc, and you get an explosion inside the lineup instead of a trip. People get hurt and the facility goes dark for weeks while a long-lead lineup is rebuilt. Finding the loose connection, the contaminated insulator, the stiff mechanism, and the mis-set relay first, on a schedule, while the gear is de-energized, is cheaper than any of that by a wide margin.

Metal-clad vs metal-enclosed switchgear

The construction of the lineup changes what you can maintain and how you get at it, so identify the type before you plan the work. Two families cover most medium-voltage gear in the field, and the manufacturer's drawings are the authority on which one you have.

Metal-clad switchgear is the heavy version. It uses drawout circuit breakers that rack in and out of the cell, and the design is compartmentalized: the breaker, the bus, the cable terminations, and the instrument transformers each sit behind grounded metal barriers so a fault in one compartment is contained. The drawout feature is the maintenance advantage. You rack the breaker out to a safe position and work on it, or swap it, without exposing the bus.

Metal-enclosed switchgear is a broader, lighter category that includes load-interrupter switchgear, the fused-switch lineup common ahead of a transformer. The switch and fuses are often fixed-mounted rather than drawout, and the internal barriers are fewer, so isolating a compartment to work on it usually means de-energizing more of the lineup. Knowing which you have tells you how much you can isolate, whether the breaker comes out on a truck, and how the access drives your outage plan.

The breakers: vacuum, SF6, and legacy oil

What interrupts the fault current sits at the center of the gear, and the interrupting medium decides what you test and what you cannot. Confirm the technology and the rating against the breaker nameplate, because the maintenance changes with each.

Vacuum is what you will find on most medium-voltage breakers installed in recent decades. The contacts open inside a sealed ceramic bottle under high vacuum, and the arc snuffs out as the vacuum quenches it at the first current zero. The bottle is the whole game: there is nothing to service inside it, so the maintenance question becomes whether the vacuum is still good, which you confirm with a high-potential test across the open contacts.

SF6 breakers interrupt in sulfur hexafluoride gas and show up more on the higher end of medium voltage and into the utility range. The gas pressure or density is the thing to monitor, since a leak degrades the interrupting capability, and SF6 is a potent greenhouse gas, so handling and recovery are regulated. Oil breakers are legacy now. Where they survive, the insulating and arc-quenching oil has to be sampled and tested, the contacts inspected for arc erosion, and the arc byproducts cleaned out. Whatever the medium, the breaker also has an operating mechanism that stores and releases the energy to move the contacts, and that mechanism is mechanical, so it needs its own attention.

De-energize, lock out, test for dead, and ground first

The maintenance and testing in this guide is done de-energized. Not because de-energized is convenient, but because opening a medium-voltage cubicle that is still live, or working it hot, is how people are killed and burned. Before any cover comes off, the gear is isolated, locked out, tested for absence of voltage, and grounded.

The sequence is not negotiable. Open the breaker and rack it to a safe position, isolate the source, apply lockout/tagout to every point that could re-energize the section, and only then prove it dead. Test for dead with a rated medium-voltage detector, and confirm the detector itself works on a known live source before and after, because a tester that quietly failed will tell you a live bus is dead. Then apply grounds. Grounding the conductors holds them at zero even if something upstream is switched back by mistake, and it drains any stored or induced charge before your hands are near the bus.

Understand the hazard before you ever open the door. The arc-flash energy available at the gear sets the boundary and the arc-rated PPE for the task, and the worst exposure is the moment of racking, switching, or opening a door on something that turns out to be live. Read the arc-flash study and the equipment label first, wear the rated PPE, and treat the gear as live until your own meter and your own grounds prove otherwise. The arc-flash guide covers how that hazard is calculated and labeled.

Visual inspection and cleaning

With the gear dead and grounded, the first real work is to look and to clean, and the looking finds more than most of the instruments do. A trained eye on an open cubicle catches the problems that are already underway before a single test lead goes on.

Look for the heat signature first: discolored or pitted bus joints, darkened insulation, the brown bloom around a connection that has been running hot. Look for tracking, the carbonized tree-shaped path across an insulator surface that says a creepage current has been finding its way over the dirt and moisture. Look for cracked or chipped insulators and standoffs, evidence of rodents or birds, and any sign of water entry, because moisture is the partner to almost every insulation failure in this gear.

Then clean it, because contamination is not cosmetic in medium-voltage equipment. Dust, salt, and moisture on an insulator surface lower the flashover voltage and feed the tracking that ends in a fault. Vacuum and wipe the bus, the insulators, and the compartment per the manufacturer's cleaning guidance, paying attention to the standoffs and the spaces where dust collects on a horizontal surface. A clean, dry insulator is doing the job it was built for. A dirty one is a fault waiting for the right humidity.

Check and torque the bolted connections

Loose bolted connections are the most common preventable failure in this gear, and the chain is short: loose means resistance, resistance means heat, and heat means a joint that degrades until it fails or starts a fire inside the lineup. Every bolted electrical connection on the bus and at the terminations is a candidate, and checking them is core maintenance, not an extra.

Torque to the value the manufacturer specifies for that connection, hardware, and plating. Do not guess and do not reuse a generic number off another job, because the correct value depends on the bolt size, the material, and whether the joint uses Belleville washers, and the manufacturer's table is the authority. A connection that was torqued correctly at install can still back out under years of thermal cycling, which is why the check repeats on the maintenance schedule rather than living only in the commissioning record.

Torque checking and thermography work together. The infrared survey, done energized, points you at the joints that are running hot, and the de-energized torque check is where you fix what the camera found. A connection the thermographer flagged at a 20 degree rise over its neighbors is a connection to open, inspect for oxidation or damage, clean, and re-torque to spec on the next outage. The two tests are halves of the same answer, and the termination-torque practice that governs lugs and pads applies here the same way it does in a panel.

Infrared thermography

Infrared thermography is the one survey you run energized, and it earns its place because it finds the loose, hot connection while the gear is loaded and the defect is actually producing heat. A connection that reads fine cold can be running 30 degrees over its neighbors under load, and the camera sees that the moment the door or the IR window is open to it.

The survey works by comparison. You scan the bus joints, the breaker stabs, the cable terminations, and the connections under real load, and you look for the spot that is hotter than the identical connection on the next phase or the next section. A consistent rise on one phase points at that phase. A single hot joint points at a loose or corroded connection that the next de-energized outage needs to open and re-torque. Temperature rise over a similar reference, not an absolute number, is what flags the problem, and the severity guidance comes from the thermography standards and your program, not from a single reading in isolation.

Do it safely. Opening a cubicle on live medium-voltage gear to point a camera at the bus puts you inside the arc-flash boundary, so the trade has moved toward permanently installed IR windows that let you scan through a port without exposing the energized parts. Where the gear has IR windows, use them. Where it does not, the arc-flash study and PPE govern whether the scan is done at all and how, and the infrared survey practice for energized equipment sets the method.

Insulation resistance testing

Insulation resistance testing meggers the dielectric that holds the energized parts apart from ground and from each other, and it is the baseline electrical test on the gear. With the bus de-energized and isolated, you apply a DC test voltage from each phase to ground and between phases, and you read how much current leaks across the insulation. High resistance means clean, dry, healthy insulation. Low resistance means contamination, moisture, or a path that is starting to break down.

The single reading matters less than the trend. A megger value that looks acceptable on its own can still be a problem if it has fallen by half since the last test, because the direction is what tells you the insulation is aging or wetting up. Run the test the same way each time, at the same voltage, corrected to a common temperature, so the numbers are comparable, and keep them in the record so the trend is visible across years.

The polarization index, the ratio of the 10-minute reading to the 1-minute reading, adds another read on the same insulation. A healthy, dry system keeps drawing less current as the test continues, so the ratio comes out well above one. A contaminated or wet system stays flat, and the index sits near one. Hedge the acceptable values and the test voltage to the NETA MTS table, the IEEE guidance, and the manufacturer for the specific gear, because they vary with voltage class and insulation type, and the megger is a screening tool, not the final word on dielectric strength.

Contact and ductor resistance testing

Contact resistance testing, run with a low-resistance ohmmeter the trade still calls a ductor, measures the resistance across a closed connection in micro-ohms, and it is how you find a joint or a contact that is no good before it overheats in service. You pass a known DC current through the connection and measure the tiny voltage drop, and a high reading tells you the connection is loose, corroded, or worn.

On a breaker, the test goes across the main contacts with the breaker closed, and it tells you whether the contacts have eroded or misaligned to where they no longer make a solid path. On the bus, the test goes across the bolted joints, and it confirms what the torque check and the thermography suggested. A reading that runs high against the other phases, or high against the manufacturer's value, is a connection that will make heat under load, which is the same failure the thermographer is hunting from the other direction.

The number is read by comparison and against a reference. Compare phase to phase and section to section, and compare to the manufacturer's maximum micro-ohm value for that breaker or joint where one is published. Hedge the acceptable values to NETA MTS, IEEE, and the manufacturer, because the limit depends on the device and the current rating. What a high contact resistance is telling you is simple: the connection that has to carry fault current is not solid, and a poor connection is heat and a future failure.

Power factor and tan delta testing

Insulation power-factor testing, also called dissipation factor or tan delta, reads the quality of the insulation by measuring how much of the applied AC it dissipates as loss instead of storing as a clean capacitor would. Healthy insulation has a low power factor. As the insulation ages, absorbs moisture, or contaminates, more current leaks resistively through it and the power factor climbs, so the test catches a kind of degradation the simple megger reading can miss.

On switchgear, the test is applied to the bus insulation system, the standoffs, and the breaker insulation, and the instrument the trade reaches for is a power-factor set, often referred to generically by the Doble name. As with insulation resistance, the value matters most as a trend corrected to a standard temperature: a rising power factor across successive tests is the signal that the insulation is heading the wrong way even while it still passes.

This is the same diagnostic family used on transformers, where power factor and tan delta are routine on the windings and bushings, and the transformer-acceptance-testing guide covers that application. On switchgear it is more of a condition test than a universal one, so hedge whether it is warranted, the test voltage, and the acceptable values to NETA MTS, IEEE, and the manufacturer for the gear in front of you. Used on the right insulation system, it is one of the better early reads on insulation aging and moisture.

Vacuum bottle integrity testing

A vacuum breaker interrupts the arc because the contacts open inside a sealed bottle under high vacuum, so the one question that decides whether that breaker still works is whether the vacuum is still there. You cannot see it and a slow leak gives no outward sign, which is why vacuum integrity testing is on the list for every vacuum breaker.

The standard field check is a high-potential test across the open contacts. With the breaker open and isolated, you apply an AC test voltage across the contact gap inside the bottle and watch whether it holds. Intact vacuum is an excellent insulator and the gap withstands the voltage without breaking down. A bottle that has lost its vacuum will flash over across the gap at a voltage it should easily hold, and that breaker is condemned: there is nothing to repair inside a sealed bottle, so a failed vacuum check means the interrupter or the breaker is replaced.

Run the test at the voltage and for the duration the manufacturer specifies for that bottle, and treat the manufacturer's procedure as the authority, because over-stressing a bottle during the test can itself be a hazard and the correct test voltage is specific to the rating. Hedge the value and the method to the manufacturer and NETA MTS. The result, though, is binary in the way that matters: the vacuum is good and the breaker can interrupt, or it is gone and the breaker cannot be trusted to clear a fault.

Breaker timing and travel

A breaker has to do more than open. It has to open fast enough that the protection scheme clears the fault in the time the coordination study assumed, and it has to do it on all three poles together. Breaker timing testing measures the operation itself: how long from trip signal to contact part, how long to close, and whether the three poles move in step.

A timing set captures the open and close times, the contact travel and velocity, and the pole-to-pole synchronism, and each of those tells you something. Slow opening means a fault sits on the system longer than the study allowed, which pushes the real incident energy above the number on the arc-flash label. Poles that do not move together stress the system and signal a mechanism going out of adjustment. The travel and velocity curve is where you see a mechanism stiffening or a contact wearing before it shows up as a time that misses spec.

This is a mechanical-and-electrical test of the breaker as an operating machine, and a breaker that has sat closed for years is exactly the one most likely to surprise you on the timing set. Hedge the acceptable times and velocities to the manufacturer and NETA MTS, because they are specific to the breaker design. The reason the test matters is the chain back to safety: a slow breaker is a longer-duration fault, and a longer-duration fault is more arc-flash energy than anyone planned for.

Protective relay testing

The protective relays are the brain of the gear. They watch the current and voltage, decide when a condition is a fault, and tell the breaker to trip. Every test above proves the breaker can act; protective relay testing proves the relay will tell it to, correctly and in time. A relay that was set during construction and never tested is the single most common piece of a protection scheme nobody has actually verified.

The core method is secondary injection. You inject simulated currents and voltages into the relay's input terminals with a relay test set, sweep through the conditions the relay is set to catch, and confirm that it picks up at the right magnitude and trips in the right time, all the way through to the breaker trip contact. You verify the pickup, the time-current characteristic, the instantaneous element, and any directional or differential function the relay carries, against the setting that the coordination study calls for. Then you confirm the trip actually reaches the breaker, because a relay that operates into a broken trip circuit has done nothing.

This is the test that protects everything downstream, and it is the one most worth getting right. A relay set wrong, drifted, or wired into a dead trip path means the breaker never gets the order, so the fault is cleared, if at all, by something slower and farther upstream, with far more energy released. Hedge the test method, the tolerances, and the interval to NETA MTS, IEEE relay-testing guidance, and the manufacturer. What you do not hedge is whether it gets tested: an untested relay is an unproven protection scheme.

Relay settings, calibration, and coordination

Testing a relay only means something against the right settings, and the right settings come from the coordination study, not from the relay's defaults or the last technician's memory. The study works out which device should trip for a fault at each point on the system, and how the upstream devices should hold long enough to let the closest one clear first, so a fault on one feeder does not dark the whole building.

Older electromechanical relays were calibrated against their physical taps and dials, and on those you are verifying that the disk and the tap still produce the time and pickup they are set to. Modern microprocessor relays hold their settings in a settings file, so the work shifts toward confirming the file in the relay matches the study, then injecting to prove the relay acts on those settings as intended. Either way, the setting and the test are two halves of one job: the right number, then the proof that the relay honors it.

When settings change, the coordination has to be rechecked, because moving one relay can break the selectivity with the device above or below it. The selective-coordination analysis is the engineer's domain, and the values, the settings, and the testing interval are hedged to that study, NETA MTS, IEEE, and the manufacturer. The field job is to make the relay match the study and prove it does, and to flag it to the engineer when the equipment, the loads, or the available fault current have changed enough that the study itself is stale.

Partial discharge testing

Partial discharge is a small electrical breakdown inside or across an insulation defect that does not yet bridge the whole gap, and it is the early stage of an insulation failure in medium-voltage gear. A void in solid insulation, a contaminated surface, or a sharp point at a connection can start discharging long before it flashes over, and that activity slowly carbonizes and erodes the insulation until it does. Partial discharge testing finds the defect while it is still developing.

The test comes in two forms. Offline partial discharge testing is done with the gear de-energized and an external source applying voltage, which gives a controlled, sensitive measurement. Online testing is done with the gear energized and in service, using sensors that pick up the discharge signature without an outage: transient earth voltage probes on the enclosure, ultrasonic and acoustic sensors that hear the discharge, and high-frequency current transformers clamped on the grounds. Online testing has grown because it catches a developing defect without shutting the facility down and can be repeated as a trend.

Partial discharge is the predictive end of the program, the test that points at an insulation problem before any of the pass/fail tests would flag it. It takes interpretation, so the location of the discharge, its magnitude, and its trend over time matter more than a single reading, and the methods and acceptance criteria are hedged to IEEE, NETA, and the equipment manufacturer. On critical medium-voltage lineups it is increasingly part of a condition-based program rather than a one-time check.

The operating mechanism, lubrication, and racking

A breaker is a machine before it is anything electrical, and the operating mechanism is where the stored energy that slams the contacts open or closed actually lives. It is also where a breaker that has not operated in years quietly seizes up, so the mechanical maintenance is as real as any test on the list.

Exercise the mechanism so the moving parts do not set, clean and re-lubricate the points the manufacturer specifies with the lubricant the manufacturer specifies, and confirm the charging system, whether spring or stored-energy, charges and discharges as it should. Old grease is a frequent culprit: a lubricant that has hardened or gummed over a decade slows the mechanism enough to throw off the timing test, and the fix is cleaning the old lubricant out and applying the correct new one, not piling fresh grease on top of the old.

On drawout gear, the racking mechanism is part of this work. The breaker has to rack smoothly between its connected, test, and disconnected positions, the shutters that cover the live stabs have to open and close as designed, and the interlocks that stop you from racking a closed breaker have to function. Racking is also one of the higher-hazard operations on the gear, which is why exercising it and keeping it smooth during de-energized maintenance pays off later when a remote racking tool keeps a worker out of the blast zone.

NETA MTS and how often to test

NETA publishes the standard that frames this work. The Maintenance Testing Specifications, ANSI/NETA MTS, lay out the inspections and electrical tests for power equipment in service, along with test values and the suggested condition for each. Its sibling, the Acceptance Testing Specifications, ANSI/NETA ATS, covers the tests run on new equipment before it goes into service. The difference is the point in the life cycle: ATS proves new gear at commissioning, MTS keeps it proven over the years that follow.

Use the MTS test tables as the reference for what to run and what value is acceptable, and treat them as a baseline to apply alongside the manufacturer's data, not a replacement for it. Where the manufacturer specifies a value, that value generally governs for that equipment, and where the project specification is tighter, it controls. The values for insulation resistance, contact resistance, timing, and the rest are hedged to those sources because they vary with voltage class and device.

Frequency is the question everyone wants a single number for, and there is not one. The maintenance interval depends on the condition of the gear, how critical it is to the facility, its environment, and its loading and operating history, and NFPA 70B frames maintenance intervals in that risk-and-condition way. Many programs land on a multi-year electrical-testing cycle with more frequent visual and infrared checks, and an out-of-cycle inspection after any breaker clears a fault. The right interval for a clean, lightly loaded, non-critical lineup is not the right interval for a dirty, heavily cycled main, so hedge the frequency to NETA, NFPA 70B, the manufacturer, the engineer, and the facility's own risk assessment.

Online monitoring and condition-based maintenance

The program does not have to wait for the next outage to learn something. Online monitoring puts permanent sensors on the gear, temperature probes at the connections, partial-discharge sensors on the bus, breaker-operation counters in the mechanism, and trends the readings continuously so a developing problem shows up as a curve heading the wrong way rather than a surprise at the next test.

That is the idea behind condition-based maintenance: instead of servicing everything on a fixed calendar whether it needs it or not, you watch the condition and act when the data says to. It does not replace the de-energized tests, because nothing online can hi-pot a vacuum bottle or torque a joint, but it sharpens the schedule by telling you which cubicle to open first and which can wait.

Whether the monitoring is a permanent system or a technician's periodic readings, the value is in the trend, and the trend only exists if the numbers are captured the same way every time and kept where the next person can find them. A field platform like FieldOS is where those readings, the infrared findings, and the as-found test data live as a record that builds the history instead of a stack of one-off reports that no one compares.

Run it as a program, not run-to-failure

Everything in this guide only works as a program. A single round of tests proves the gear was sound on the day someone tested it. A program proves it stays sound, because it sets a schedule, repeats the tests the same way, trends the results, and keeps the records where the next technician picks up the history instead of starting blind.

The opposite of a program is run-to-failure, and it is still the most common way medium-voltage gear is managed: leave it alone because it is working, and deal with it when it breaks. The trouble is that the way this gear breaks is the arc-flash explosion and the multi-week outage, so run-to-failure is not a cost-saving choice. It is deferring a small, scheduled, de-energized cost until it comes due as a large, unscheduled, dangerous one.

A real program has the schedule, the trend, and the records as its three working parts, plus the discipline to actually open the gear on the interval instead of pushing the outage one more quarter every quarter. The discipline is the hard part, because the gear gives no reason to act until the day it gives every reason at once.

Arc-flash safety, PPE, and remote racking

The hazard that runs underneath all of this is arc flash, and it deserves its own place in the plan rather than a line in the procedure. Medium-voltage switchgear can release enough energy in a fault to kill or severely burn a worker standing in front of it, and the operations that put a worker closest to that energy are racking a breaker, switching, and opening a door on gear that turns out to be live.

The defenses stack. Do the work de-energized whenever the task allows it, because nothing protects like an absence of voltage proven by your own meter and held down by your own grounds. When work or operations have to happen with the gear energized, the arc-flash study sets the boundary and the arc-rated PPE for the task, and you wear what the label calls for, not what is comfortable. Remote racking tools let a worker rack a breaker from outside the arc-flash boundary, which removes the body from the most dangerous routine operation on the gear. A maintenance switch or arc-energy-reduction setting, switched in before energized work, speeds the trip and lowers the incident energy for the duration of the task.

The arc-flash guide covers how the study calculates the energy, what the label has to carry, and how the two PPE methods work, so treat that guide as the companion to this one. The short version for the field is plain: prove it dead before you trust it dead, and when you cannot, put distance and arc-rated PPE between you and the gear.

What to document

A test nobody recorded is a test nobody can trend, and trending is most of the value in this whole program. The as-found and as-left readings, the relay settings, the torque values, and the NETA report are the record that turns a string of one-time tests into a history that shows whether the gear is holding or sliding.

Capture the as-found condition before you touch anything and the as-left condition when you finish, so the next crew knows what you changed and what you found. Record the insulation resistance and power factor corrected to temperature, the contact resistance per breaker and joint, the breaker timing, the vacuum check result, the relay settings injected and the trip times measured, the torque values applied, the infrared findings, and any partial-discharge data, each against the prior reading so the trend is visible. Keep it in a system, not a binder that lives in one person's truck. A field platform like FieldOS keeps the as-found and as-left data, the relay settings, and the NETA report attached to the specific cubicle so the history follows the gear instead of the technician.

Common mistakes

• Running the gear to failure with no maintenance program because it has worked for years.
• Working or inspecting the gear energized when the task could be done de-energized.
• Leaving bolted connections unchecked and never re-torquing them to the manufacturer's value.
• Setting protective relays at construction and never testing them by secondary injection again.
• Reading a single insulation-resistance value and ignoring the trend that shows it falling.
• Trusting a vacuum breaker without ever hi-potting the bottle to confirm the vacuum holds.
• Racking and switching by hand with no remote racking or arc-energy-reduction protection.
• Treating NETA MTS values as universal instead of checking the manufacturer and project spec.

Field checklist

Standards and references

NETA is the body for field acceptance and maintenance testing of this gear. ANSI/NETA MTS, the Maintenance Testing Specifications, gives the in-service inspections, tests, and suggested test values, and ANSI/NETA ATS, the Acceptance Testing Specifications, covers the tests on new equipment at commissioning. Use the MTS tables as the reference for what to run and what value is acceptable, alongside the manufacturer's published data for the specific equipment, which generally governs where it is more specific.

IEEE backs up the individual tests and the protection. IEEE guidance covers insulation testing, power-factor and partial-discharge methods, and relay and protection practice, and IEEE C37 covers switchgear and circuit breakers. On the safety side, NFPA 70B, Standard for Electrical Equipment Maintenance, frames the maintenance program and the condition-and-risk basis for intervals, and NFPA 70E governs the electrical safety work practices, the arc-flash boundary, and the PPE for any energized operation. The coordination and the relay settings come from the engineer's study for the specific system.

Hedge the tests, the values, and the frequency to those sources together. NETA MTS, IEEE, the manufacturer, the engineer, and the facility's own risk assessment set what to test, what value passes, and how often, and they shift with the voltage class, the equipment, the environment, and the criticality. Three things do not bend: maintain the gear de-energized before it arc-flashes, test the insulation, the contacts, and the protective relays, and run a NETA-schedule program instead of run-to-failure. The transformer-acceptance-testing guide covers the unit ahead of the gear, and the arc-flash guide covers the hazard you work inside.

Units and terms

Switchgear testing carries its own vocabulary, and the same idea shows up under more than one name across a NETA report, a manufacturer manual, and a relay setting sheet.

Contact resistance is read in micro-ohms, insulation resistance in megohms or gigohms, power factor and dissipation factor as a percentage, and breaker timing in milliseconds or cycles. A ductor and a DLRO, a digital low-resistance ohmmeter, are the same instrument by different names. The terms below define the words this guide leans on.

Metal-clad switchgear

MV switchgear with drawout breakers and fully barriered compartments for the breaker, bus, and terminations

Vacuum vs SF6 breaker

Vacuum interrupts the arc in a sealed vacuum bottle; SF6 interrupts in pressurized sulfur hexafluoride gas

Insulation resistance

DC megger reading of how well the insulation resists leakage to ground and between phases, trended over time

Contact resistance / ductor

Micro-ohm measurement across a closed contact or joint; high means loose, corroded, or worn

Power factor / tan delta

AC test of insulation loss; a rising value signals aging, moisture, or contamination

Protective relay testing

Secondary injection of current and voltage to verify the relay picks up and trips per the coordination study

Partial discharge

Localized insulation breakdown at a defect that erodes the insulation before it flashes over; tested online or offline

NETA MTS

ANSI/NETA Maintenance Testing Specifications, the reference for in-service test methods and values

Related tools

Calculators and readiness checks for this work

Compare your options

FAQ

People also ask

• What is a TTR test?
• What is incident energy in an arc-flash study?
• What is a stress cone?
• How do you test a UPS battery?
• What is a buck-boost transformer?