What Low Voltage Switchgear Testing Must Prove Before Energizing
Before indoor low voltage switchgear testing begins on a lineup that is about to be energized or returned to service, the gear has to be proven dead, cut off from every source, insulated, able to open and close, and protected. The hold points follow that order. A factory design-test catalog is a different job from the sheet in front of the cabinet, and a medium-voltage withstand figure stays with medium-voltage gear.
After a board is built or landed, the work people actually describe is point-to-point wiring, operating breakers and contactors, a megger test to the frame, and torque on the connections. Shops split that work by the board in front of them: a small 240 V service and a 480 V industrial board are different jobs.
Those checks sit on indoor low-voltage switchgear, the assembly that distributes, switches, and protects the circuits. Those cabinets are grouped under low-voltage switchgear.
Contents
- What Has to Pass Before the Lineup Is Energized
- Prove the Gear Dead and Isolate It
- Check Insulation Before Any Withstand Test
- Operate the Gear Without Treating Interlocks as Lockout
- Check Protection Without a Medium-Voltage Withstand Test
- Keep Design Tests Off the Field Sheet
- Read the AGCK Product Nameplate Before You Copy a Test Voltage

What Has to Pass Before the Lineup Is Energized
Before an indoor lineup is energized or put back in service, four things have to be true: it is proven dead, it is isolated from every source, the insulation is acceptable for that circuit, and the breakers and interlocks actually operate. Matching the check to the nameplate comes before any copied test voltage, and a published design-test list is a different sheet.
The low voltage switchgear testing procedure is that hold-point order: dead, insulation, function, then protection. Each stop changes what the next person may do with a meter or a breaker handle.
A passed design on a representative sample can still be undone by mounting, wiring, and torque on the unit in the room. Routine checks exist for that reason, on every assembly, in the shop or on site.
Point-to-point wiring confirms that power and control conductors land where the drawings say they land. Operating the breakers, contactors, and relays shows that those devices open and close, and that step still waits until the bus has been proven dead.
Use the stop/go sheet when the argument is whether to energize. If a row fails, the next row waits.
| Hold point | What it proves | What stops the next step |
|---|---|---|
| Dead and isolated | Every source is open, stored charge is released, and a qualified person has tested the parts | A meter stays off the bus |
| Insulation | The assembly is off, unloaded, and the reading is already above the minimum for that circuit | A withstand test stays off the schedule |
| Function | The mechanism opens, closes, and racks, and the interlock is only a mechanism check | Lockout is still the isolation |
| Protection | The trip unit is checked against its own curve, and a trip is left open until the cause is known | Reclosing waits |
Prove the Gear Dead and Isolate It
A megger stays off the bus until a qualified person has disconnected every source, released stored charge, and measured that the parts are deenergized, including induced voltage and backfeed. A selector switch or a draw-out interlock is a mechanism check, and treating it as isolation leaves the lineup without lockout.
Exposed parts are disconnected from sources, stored charge is released, and a qualified person tests for voltage. Isolation is the opened disconnect and the tag, kept separate from the interlock.
Live parts an employee may be exposed to are deenergized before work on or near them, unless deenergizing adds a hazard or is infeasible, and parts under 50 V to ground are the exception only when there is no increased burn or arc-explosion exposure.
Stored electrical energy that might endanger personnel is released, capacitors are discharged, and high-capacitance elements are short-circuited and grounded when that stored energy might endanger people. Before the lineup is energized, a qualified person confirms that tools, jumpers, shorts, and grounds have been removed.
Only a qualified person tests the circuit. The instrument and its leads are inspected before use, and they are rated for the circuit and the location.
The before-and-after tester check applies when the circuit is over 600 V nominal, and it is the boundary rather than the default step on gear whose working voltage is AC 380 V (AC 400 V).
Important: A push button, selector switch, or electrical interlock is not enough to call the gear deenergized, and it does not replace lockout and tagging. Before the circuit is treated as dead, a qualified person confirms it cannot be restarted and tests the parts for voltage, including induced voltage and backfeed. The public text of that rule is 29 CFR 1910.333.
The US work rules for this order are OSHA 1910.333 / 1910.334. A testing-community inspection guide starts from the same hold: only qualified people, and primary and secondary circuits de-energized before testing or maintenance.
Design verification of the protective circuit is described at a test current of 25 A between the protective-conductor terminal and the exposed metal, and that description carries no ohm limit. The metalwork still has to have a real path back to the equipment ground, which is what that design check is looking for.
Check Insulation Before Any Withstand Test
Insulation on an off, unloaded assembly comes before any withstand discussion, because a withstand test waits until the insulation reading is already above the minimum. Breaking devices are closed for that measurement, and loads stay disconnected.
Insulation resistance is the DC resistance between conductors and earth on an assembly that is off. In the shop it is the megger test to the frame, and it happens only after the bus is dead.
A routine-verification account of IEC 61439-1 describes that measurement at a minimum of 500 V DC, with the assembly off, loads disconnected, and breaking devices closed, and it states a floor of 1,000 ohms per volt to earth.
Sensitive devices follow the device maker's instruction, and they stay out of a test they cannot tolerate.
A US acceptance checklist holds bus insulation for one minute and withholds the withstand test until insulation is already above the minimum. Dielectric withstand on that checklist is also held for one minute, and it starts only after the insulation reading has cleared that minimum.
Control wiring on that same checklist is a separate measurement: 500 V DC on 300 V cable, 1,000 V DC on 600 V cable, held for one minute, with a floor of 2 megohms, and solid-state devices left out when they cannot take the voltage.
A failed insulation reading is one branch of low voltage switchgear repair. The lineup stays de-energized, the connected electronics stay out of the test, and the next decision is to correct the insulation fault before anyone schedules a withstand test.

Operate the Gear Without Treating Interlocks as Lockout
Racking the withdrawable device, trying a locked-open device and a locked-closed device, and opening or closing the circuit with a load-rated breaker or switch prove that the mechanism works. Those checks do not replace lockout.
A draw-out circuit breaker can be racked to a test position or a disconnected position, which is why one feeder is easier to isolate on withdrawable gear than on a fixed cabinet. Fixed gear has fewer draw-out interlocks to prove, and it still needs the same dead-circuit check.
The interlock stops a wrong racking or closing sequence. It is checked by trying to close a locked-open device and trying to open a locked-closed device, and that trial is still a mechanism check.
Load-rated switches, circuit breakers, or other devices designed as disconnecting means are what open, reverse, or close circuits under load. Cable connectors that are not the load-break type, fuses, terminal lugs, and splices stay out of that job except in an emergency.
Bolted joints more than 50 percent above the lowest similar reading are investigated, and the comparison is between similar joints on the same gear. Torque comes from the manufacturer.
A joint that runs hot after energizing belongs with that same comparison. It is another repair decision, and raising the tester voltage does not answer a hot joint.
Point-to-point checks on control wiring include operating relays and contact blocks so they open and close as the scheme requires. Where a controller is already installed, simulated inputs are used only after the power circuits are dead.

Check Protection Without a Medium-Voltage Withstand Test
A low-voltage protection check uses secondary injection at the trip-unit input or primary injection through the main circuit, set against that device's curve. The functions on that curve are long-time, short-time, instantaneous, and earth fault. The result is read against that device's curve.
Secondary injection is a low-level signal into the trip unit. Primary injection is high current through the breaker itself, and both results are read against the curve for the device in the cubicle.
Above 1 kV the design standard people cite is IEC 62271-200. Up to 1,000 V AC, the assembly standards are IEC 61439-1 / IEC 61439-2.
A medium-voltage power-frequency withstand, and a vacuum-bottle test that travels with it, stay on medium-voltage metalclad gear. The kilovolt argument that shows up on that gear stays there, and copying it onto this indoor cabinet mixes two voltage classes.
If a breaker trips during commissioning, leave it open until the equipment and the circuit are known to be safe to energize. Repetitive manual reclosing is prohibited, and an overload identified from the design is the stated exception.
Medium and low voltage switchgear maintenance & testing names two jobs. The indoor job is the hold-point work on this lineup, and medium-voltage maintenance uses the metalclad standard for gear above the low-voltage ceiling.
A breaker that will not operate is the same repair branch as a failed insulation reading or a hot joint. Correct the device or the circuit, then return to the hold point that failed.
Keep Design Tests Off the Field Sheet
Temperature rise, dielectric properties, short-circuit withstand, protective-circuit effectiveness, clearances and creepage, mechanical operation, and enclosure protection are design proofs on a representative assembly. They show that a design can meet the assembly rules, and running them as the field sheet skips the prove-dead order the lineup still needs.
Further construction checks on that same design summary cover mechanical impact, rust protection, damp heat, heat resistance of insulating materials, fire behavior, and the mechanical performance of assemblies and fixings. Those items belong to the design program, and a commissioning crew does not repeat them as the acceptance sheet.
Routine checks are done on each assembly because mounting and workmanship can undo a type-tested design. The sheet in hand is that routine and field work, including the prove-dead order the design catalog skips.
Mechanical operation of draw-out racks in the design proof is 50 operating cycles on the racks and faceplate fixings. That cycle count stays inside the design proof.
The same design summary notes an air temperature rise under 30 K, a terminal rise under 70 K, and a busbar rise not above 65 K. Those rises are design measurements on a representative assembly, and a field infrared scan still needs the manufacturer's own limit.
ANSI/NETA ATS is the field acceptance document.
Read the AGCK Product Nameplate Before You Copy a Test Voltage
The published nameplate bounds the indoor test, and the setting on a field tester is a separate decision. On the AGCK withdrawable cabinet, the published table lists AC 380 V (AC 400 V) working voltage, AC 690 V insulation, 50 Hz, and 2,500 V power-frequency withstand for one minute on both the main and secondary circuits.
Those figures bound the equipment. They are the ratings printed for the cabinet rather than the voltage to set on a field tester or a record that this unit was acceptance-tested.
A US course text describes typical 600 V class gear as 600 V and below, about 2 in clearance, and 2,200 V AC for one minute. That description belongs to the US class in the course, and it stays off the AGCK table.
AMNS lists 2,500 V for one minute on its own table as a rating, and that cell is a separate nameplate from AGCK. AGCS is another withdrawable cabinet, and its withstand cells are incomplete, so no test voltage is taken from AGCS.
The enclosure figure on the AGCK table is IP40.

The cabinet in the photo is AGCK low-voltage withdrawable switchgear. Open that table, match the hold points to the ratings below, and return to the low-voltage hub if the lineup is a different assembly.
Withdrawable gear makes one feeder easier to rack out for a test. A higher nameplate withstand or a higher enclosure figure still does not replace the dead-circuit check, and a fixed board with no draw-out interlock to prove is a different cabinet.
The published AGCK ratings below are nameplate bounds. Read them as the cabinet's limits.
| Parameter | Published value |
|---|---|
| Rated insulation voltage | AC 690 V |
| Rated working voltage | AC 380 V (AC 400 V) |
| Rated frequency | 50 Hz |
| Horizontal busbar rated current | 1250 A, 1600 A, 2000 A, 3200 A, 4000 A |
| Vertical busbar rated current | 1000 A |
| Maximum feeder-circuit current | 2500 A |
| Maximum drawer-circuit current | 630 A |
| Maximum control-motor capacity | 320 kW |
| Horizontal busbar peak withstand | 105 kA, 176 kA |
| Horizontal busbar short-time withstand (1 s) | 50 kA, 80 kA |
| Vertical busbar peak withstand | 63 kA, 105 kA |
| Vertical busbar short-time withstand (1 s) | 30 kA, 50 kA |
| Rated power-frequency withstand (1 min) | 2500 V |
| Secondary power-frequency withstand (1 min) | 2500 V |
| Enclosure protection | IP40 |
| Envelope | W600 (800, 1000) × D800 (1000) × H2200 mm |
| Effective height of the functional unit | 1800 mm |
FAQs
What has to be true before any insulation test on indoor LV switchgear?
The lineup is proven dead, every source is disconnected, stored charge is released, and a qualified person has tested for voltage, including backfeed. A selector switch or a draw-out interlock remains a mechanism check beside lockout and tagging.
What is the IEC standard for low-voltage switchgear testing?
IEC 61439-1 / IEC 61439-2 is the pair for low-voltage assemblies, with the general rules in the first and power switchgear and controlgear assemblies in the second, up to 1,000 V AC. IEC 62271-200 applies to gear above 1 kV and is not the procedure for this indoor lineup.
What is the UL standard for low-voltage switchgear?
UL 1558 / UL 891 is the US construction split people ask about: UL 1558 for metal-enclosed low-voltage switchgear and UL 891 for switchboards. The Andeli tables do not claim either mark.
Which test shows that switchgear connections are tight?
After the bus is proven dead, similar bolted joints are compared, and a reading more than 50 percent above the lowest one is investigated. Torque comes from the manufacturer.
How is a low-voltage breaker protection check different from a medium-voltage hi-pot?
Secondary injection or primary injection, read against the breaker curve, is the low-voltage protection check. A medium-voltage withstand test stays on medium-voltage gear, and the kilovolt argument that shows up on that gear is left there.
What should happen after a breaker trips during commissioning?
Leave the breaker open until the circuit is known to be safe to energize, and do not keep reclosing it. Repetitive manual reclosing is prohibited, with an overload identified from the design as the stated exception.
Do Andeli AGCK or AMNS pages publish a 2,500 V test result?
The tables list a 2,500 V power-frequency withstand for one minute as a rating on each cabinet. That cell bounds the equipment, and it is a nameplate figure rather than a factory acceptance record, a site acceptance record, or a NETA result.
Why can a shop not use one megger plan for every board?
A 240 V board and a 480 V industrial board put different voltages on the nameplate, and the connected devices change the insulation test again. Control wiring and solid-state parts do not take the bus measurement.














