EGC to a plastic enclosure: run it and land it on a bushing

Run the wire. The NEC expects an equipment grounding conductor with the circuit conductors whatever the load's housing is made of, and on a 50 A, 220 V oven feed that is 10 AWG copper per NEC Table 250.122. The conductor serves the metal parts along the run, not the plastic box: the oven chassis that an element fault can put at line potential, and the 3/4 in flexible metal conduit that a nicked #8 can energize. Inside a plastic enclosure there is no ground screw to take it, so the wire lands on the lug of an insulated grounding bushing, which bonds the armor in the same move.

Is the equipment grounding conductor still required when the enclosure is plastic?

Yes: the NEC requires an equipment grounding conductor to run with the circuit conductors on a feed like this oven whip, and a plastic housing changes nothing, because the conductor serves the metal oven chassis and the metal conduit armor, not the plastic module.

The rule follows the circuit, not the box. A circuit that feeds metal equipment carries a wire-type equipment grounding conductor with it, and every conductor of that circuit, ground included, rides in the same 3/4 in flexible metal conduit. The four #8s and the #10 travel together; nothing runs outside the whip.

The plastic housing of a load-management module is genuinely nonconductive, so its shell looks after nobody. The parts the ground wire protects sit outside it: the oven's steel chassis and the armor of the conduit. The grounding and bonding guide covers the sizing and bonding chain end to end; the short version below is enough for this whip.

Keep the habit. The wire becomes optional on the day the circuit has no metal to protect, and an oven feed is never that day.

What does the ground wire actually protect on this run?

On this run the conductor protects two metal parts: the oven chassis, which an element fault puts at 220 V, and the 3/4 in flexible metal conduit armor, which an abraded #8 insulation can energize; the plastic module housing never becomes part of that path.

Walk the oven failure first. An element lets go onto the steel chassis, the chassis sits at 220 V to ground, and the fault current needs a road back to the source so the breaker can trip. With the copper ground wire in place the road exists and the breaker clears. Without it, the road is a person holding an oven door handle on a concrete floor.

The armor is the second candidate. A #8 pulled through flex with a burr under its insulation can ride against steel for years; when it finally fails, the armor goes live. Bonding both ends of the whip turns that fault into a trip as well.

NEC 250.4(A)(5) calls this road the effective ground-fault current path and states the earth is not considered part of it: soil resistance sits far too high to push a breaker past its trip point. The wire clears the fault; the dirt never will. The raceway sometimes carries that path itself, and the article on when a metal raceway serves as the equipment grounding conductor maps where that holds.

What size EGC does a 50 amp oven circuit need?

A 50 A breaker takes a 10 AWG copper equipment grounding conductor per NEC Table 250.122, which sizes the ground wire from the rating of the overcurrent device ahead of the circuit, not from the ampacity of the #8 circuit conductors.

The table lists no 50 A row, so the 60 A row governs: 10 AWG copper, or 8 AWG if the ground wire is aluminum. The #8 circuit conductors sit one size up, which is normal: 8 AWG copper carries 50 A at 75 °C, and the ground wire tracks the breaker rather than the load current it parallels.

Worked example, with the whip length as an assumption you replace: take a 15 ft [4.6 m] whip. Resistance of the copper ground wire is about 1.0 Ω per 1000 ft at 20 °C, and copper gains about 0.39 % per kelvin. The rise from 20 °C to 75 °C is 55 K:

1.0 Ω/1000 ft × (1 + 55 K × 0.0039/K) = 1.21 Ω/1000 ft at 75 °C
15 ft × 1.21 Ω/1000 ft = 0.018 Ω

So the whip's ground conductor adds about 0.02 Ω to the fault loop. Pulling the habit one size up, to #8 copper at about 0.63 Ω per 1000 ft at 20 °C, lands at 0.01 Ω over the same 15 ft: roughly 40 % lower ground-path resistance for one wire size. Both sizes satisfy NEC Table 250.122 at 50 A; the table still sets the 10 AWG.

The number that hurts is the upsize nobody caught. Move that oven to a 100 A breaker and the table's 100 A row asks 8 AWG copper for the ground wire; the #10 left in the whip is undersized from the moment the breaker changes.

When can the flexible metal conduit carry the ground itself?

The NEC lets flexible metal conduit stand in as the ground path only on runs of 6 ft [1.8 m] or less with the conductors protected at 20 A or less, so on this whip the wire-type EGC is doing the work and must be there.

The rule reads: 6 ft of conduit at most, terminated in listed fittings, with the circuit protected at 20 A or less. Meet all three and the steel flex can stand in for a separate ground wire on a short, light run. The breaker rating alone takes this whip out of that lane, whatever its length, so the wire runs.

Above 20 A the conduit still gets grounded; it simply stops doing the grounding job by itself. The wire-type EGC inside clears the faults, and bonding fittings at each end keep the steel tied to it, because the NEC grounds metal parts that can go live and abraded insulation makes the armor one of them.

That division of labor is the whole answer for the whip: the ground wire clears faults, the bushings ground the armor, and neither covers for the other.

Where do you land the EGC at a plastic enclosure?

Land the ground wire on the lug of an insulated grounding bushing where the whip enters the plastic box, because the box itself offers no ground screw, and jumper the bushing to the module's ground terminal when its instructions provide one.

You need a 3/4 in insulated grounding bushing with a solderless lug, a second grounding bushing or a bonding-type connector for the panel end, the green copper EGC, locknuts, and a driver for the bushing screw. Then:

1. De-energize the breaker, lock it off, and prove absence of voltage at the module terminals with a meter you just tested on a known live source.
2. Pull the green wire through the 3/4 in flexible metal conduit with the four #8s, leaving slack at both ends to reach the landing points.
3. At the panel end, land the whip in a listed fitting, land the wire on the ground bar, and fit the grounding bushing or bonding connector to the armor.
4. Where the panel is painted, NEC 250.12 calls for paint, lacquer and enamel removed from the threads and contact surfaces at grounding and bonding connections, so scrape the landing to bare steel before the fitting goes on.
5. At the module end, fit the insulated grounding bushing with its lug facing into the box and land the wire on the lug; where the module's instructions provide a grounding terminal, land it there and tie the bushing to it with a short green jumper.
6. Torque the bushing screw to the marking printed on the fitting and snug both locknuts so neither fitting can back out.
7. Finish with a look at the conductor: an insulated EGC of 6 AWG and smaller must be green or green with yellow stripes under NEC 250.119, and the wire in this whip follows that.

Where the whip enters through a concentric or slotted knockout rather than a plain one, the bonding rules at concentric knockouts tighten; sort the box before the fitting goes in.

What checks prove the whip is grounded before you energize it?

An ohmmeter from the EGC landing at the module end to the panel's ground bus should read roughly 0.02 Ω for the 15 ft whip, and the armor of the conduit should read nearly the same figure through its bushing bond.

Then work the list:

• Wire continuity, bushing lug to ground bar: the whip itself contributes under 0.05 Ω at 15 ft; a reading near an ohm points to a loose lug or a contact still sitting on paint.
• Armor continuity, steel at the module end to the ground bar: the same low figure, because the bushing bond ties the conduit into the wire.
• Insulation, each #8 to the armor with the conductors separated: megohms on an insulation tester; a reading near the ohm scale says the nick that motivated the bond is already there.

A plastic box gives the meter nothing else to look at; that is exactly why both readings land on the bushing.

What mistakes undo the ground path?

The failure that hides best is a whip grounded at the panel end only: the armor between the panel and the box sits unbonded, and a conductor abrading on it leaves the flex live at 220 V with nothing to trip the breaker.

The rest of the shortlist:

• Armor trusted alone above 20 A: the 6 ft and 20 A lane is closed at this breaker size, so the wire goes in and the bushings come with it.
• Green wire cut back because the box is plastic: the oven chassis loses its fault path; land it on the bushing lug instead.
• Bonding bushing set on powder coat: it bonds to the paint, not the steel; scrape to bare metal under the fitting.
• Ground rod driven at the oven and the whip's wire dropped: NEC 250.4(A)(5) keeps the earth out of the fault path, and NEC 250.54 makes the rod an auxiliary electrode you may bond to the ground wire, never a road that replaces it.

When the module lives in one of our polycarbonate or fiberglass enclosures cut to your layout, the landing shifts but the conductor stays. A non-metallic box offers no ground path of its own; the fiberglass grounding piece shows the EGC and back-panel bonding that takes its place.

The decision holds in every case: the wire runs with the circuit at the size NEC Table 250.122 sets, the armor bonds at both ends, and the plastic box contributes nothing but a clean place to mount the bushing.

Catalog dimensions to compare

Compare these catalog dimensions with your component layout. Open each enclosure for its rating, material and usable space, then check those specifications against your application.

Questions