Short answer: white chalky dust on your terminals is zinc plating that died protecting the copper beside it, and water has been living in that joint for weeks. Stranded conductors wick moisture a metre [3 ft] or more between their strands, so the entry is often a wet gland at the far end of the cable rather than the joint that corroded. Cut the conductor back to bright copper, re-terminate on stainless hardware, and seal the entries. Where the corrosion is box-wide or the box itself is rusting, replace it with a 316L stainless enclosure rated NEMA 4X / IP66 by the manufacturer, not UL listed. The corrosion itself tells you where the water came from — this article reads it.
What the colour of the corrosion tells you
Clean nothing until you have read the damage. The colour names the metal that lost, and the metal that lost names the water's path.
| What you see | Metal that died | What it points to |
|---|---|---|
| White chalky powder, copper bright and shiny | Zinc plating | A galvanic pair with the copper; water sat in the joint for weeks |
| Green or blue-green crust | Copper | Standing water in the box, or copper paired with passive stainless in a wet joint |
| Red-brown flaking and stains on the body | Iron and steel | A plain-steel box or a field-drilled hole; that box is done |
| Dull black on the joint itself | Copper oxide | The joint ran hot as well as wet; treat it as a burned connection |
| Fluffy white on the box body | Aluminium | Aluminium paired with copper lugs in a wet box |
Most boxes show the first row: copper shining, hardware white. Zinc and copper sit about 1.1 V apart in the standard electrochemical series (-0.76 V for zinc, +0.34 V for copper), so with any electrolyte between them the zinc corrodes while the copper stays protected — your stud has been a sacrificial anode. The area ratio piles it on: a small zinc-plated stud clamped under several large copper lugs corrodes many times faster than the same two metals the other way round. A black joint that also smells of heat is a different chain of events; burned connector in a sealed enclosure covers it. For aluminium bodies, the detail is in aluminium enclosure earthing: studs, oxide and corrosion.
How the water got in: wicking and breathing
Water rarely enters where you find the corrosion. The gaps between the strands of a 10 AWG conductor act as capillaries, and a wet gland at the far end feeds moisture along them: a metre [3 ft] is typical on a level run, and 2 m [6 ft] happens on a downhill one. If the strands at your terminal are damp and dark while the box itself is dry, follow the cable, because the fault is at the other end. The cure at that end is in stop water wicking through cable into an enclosure.
A gasketed box also breathes with no hole at all. Warm afternoon air inside expands and escapes past the gasket; the cool night inhales humid air, and any metal below the dew point takes the water out of it.
One year of breathing, in numbers
Take a 10 x 8 x 6 in 316L wall-mount enclosure: about 463 in3 of usable inside volume, from the CAD, which is 7.6 L — 0.0076 m3. Saturated air holds about 23 g of water per cubic metre at 25 C and about 13 g/m3 at 15 C, so a 10 K overnight drop condenses roughly 10 g/m3. Per breathing cycle: 0.0076 m3 x 10 g/m3 = 0.076 g of water. Per year: 0.076 g x 365 = 27.7 g, call it 28 g [1 oz], all of it settling on your lowest hardware overnight. That is how a sealed box with intact glands still grows white dust; to split condensation from a real leak, read moisture inside an enclosure: failed capacitors or water ingress.
Find the entry, in order of likelihood
Work down this list before you order anything; every check takes minutes.
- Bottom glands and cable entries. Undo the gland nut and look for a water track or a green streak below the entry. A gland clamped on the wrong cable OD leaves a capillary gap — reseal on the measured OD.
- The far end of the cable. Damp strands at the terminal with dry glands mean the water travelled; inspect the low point of the run and the opposite termination.
- The door gasket. Pinch it. A gasket that has gone hard and no longer springs back needs replacing, with the latch pressure checked against the closed door.
- Field-drilled holes. A rust trail below an unglanded hole is the whole story.
- Nothing wet anywhere. Corrosion on the floor and the lowest terminals, with dry entries above, points to nightly condensation — the breathing case.
For the full hunt with the door shut, water inside a sealed electrical enclosure: find the cause walks the same box in test order.
The fix, joint by joint
- Lock the circuit out and prove it dead before you touch a conductor.
- Cut each affected conductor back to bright copper — 25-50 mm [1-2 in] past the last darkened strand. Oxide grows between the strands inside an old crimp. A joint made over dark strands runs hot again within months.
- Crimp new tinned copper lugs with the die that matches the lug.
- Replace zinc-plated studs and screws with 316 stainless hardware. Tinned copper against stainless is a far closer pair than zinc against copper, and a film of joint compound on the mating faces keeps the electrolyte out.
- Replace any component with crusted legs. A diode or terminal whose plating has turned to powder has lost conductor cross-section, not just finish.
- Re-gland everything you disturbed: IP66 minimum, which means the gland held a 12.5 mm nozzle at 100 L/min, and glands rated IP68 where the box sits in wash spray or standing water. Form a drip loop below every entry so water runs off the cable instead of into it.
Prove the repair under load: run the panel for an hour and compare the repaired joints with their neighbours. A joint about 5 C hotter than its twin, or a resistance a few milliohms higher, comes back apart.
Stop it coming back
A rusting or plain-steel body leaves with the repair. 316L wall-mount enclosures start at 6 x 6 x 4 in, delivered in 10 business days counted from your order; a box with factory-made cutouts runs the same promise counted from your approval of the drawing. Order the holes with the box, on a gland plate if the wiring is dense, instead of drilling on site: stainless work-hardens, and a burr under a gland is a leak.
Keep the assembly genuinely sealed. An open drain hole drops a Type 4X box to about a Type 3R, because the weakest part sets the rating; if the box must breathe, use a vent rated for jets and call the assembly by that vent's mark.
Choose the grade on chlorides. NEMA 250 / UL 50E put Type 4X through another 200 hours of salt spray, judged against a type 304 stainless specimen — that is the corrosion baseline the rating promises, and 304 covers general outdoor duty. Where coastal air, road salt or chemical washdown reaches the box, take 316L with its about 2 % molybdenum, which is what resists chloride pitting. NEMA 4X versus IP66 sets out the tests behind both marks, and on the purchase spec write both, since NEC 110.28 says a type rating cannot be replaced by an IP code; the IP rating chart lists what each digit tests.
The decision by what you found
| What you found | Decision |
|---|---|
| White dust on the hardware, copper bright, entries dry | Re-terminate: stainless studs, tinned lugs, sealed glands — the box stays |
| Damp, dark strands at the terminals, dry box | The cable wicks from the far end; seal or replace the run, then re-terminate |
| Green copper crust box-wide, standing water | Replace the enclosure: sealed 316L, IP66 minimum glands, and lift the cable's low point |
| Rust on the body, field-drilled holes | Replace with stainless and order the cutouts made with the box |
| Corrosion back within a year in a sealed box | Gasket or far cable end: new gasket, glands up to IP68, and re-walk the run |
For size, a handful of terminal blocks fits the 8 x 8 x 4 in 316L enclosure; a dozen blocks and a small power supply sit in the 12 x 10 x 6 in size. The enclosure selector filters the rest by inside dimensions and rating.
Enclosures in this article
| Enclosure | Material | Outside H x W x D / display | Rating |
|---|---|---|---|
| 6 × 6 × 4 in | 316L stainless | 6 x 6 x 4 in | Rated NEMA 4X / IP66 by the manufacturer; not UL listed |
| 8 × 8 × 4 in | 316L stainless | 8 x 8 x 4 in | Rated NEMA 4X / IP66 by the manufacturer; not UL listed |
| 10 × 8 × 6 in | 316L stainless | 10 x 8 x 6 in | Rated NEMA 4X / IP66 by the manufacturer; not UL listed |
| 12 × 10 × 6 in | 316L stainless | 12 x 10 x 6 in | Rated NEMA 4X / IP66 by the manufacturer; not UL listed |
Questions
How far can water wick along a stranded cable into a junction box?
Along the gaps between the strands, a metre [3 ft] is common on a level run, and 2 m [6 ft] happens on a downhill one where gravity helps. Solid wire does not wick — the strands are the channel. Wet strands at the terminal with dry glands mean the water entered at the other end of the run.
Is white powder on terminals dangerous or just cosmetic?
It is real damage. The zinc plating is gone and bare steel follows, so joint resistance climbs. At 20 A, a corroded joint at 5 mOhm dissipates 2 W (I squared R = 400 x 0.005), which runs hot and oxidizes faster still. A white-dusted stud has already lost its plating, so re-terminate on stainless hardware.
Do I need 316L or is 304 enough for the replacement box?
304 carries the 200-hour salt-spray baseline behind a Type 4X rating and covers general outdoor duty. Take 316L wherever chlorides reach the box — coastal air, road salt, washdown — because its about 2 % molybdenum resists pitting. The L itself is low carbon, about 0.03 %, so welds and cut edges do not sensitize.
Can I clean the terminals instead of cutting the cable back?
Clean solid metal, replace stranded. A busbar or lug face comes back with an abrasive pad and joint compound. Oxide also grows between the strands inside an old crimp, so cut stranded conductor back to bright copper, typically 25-50 mm [1-2 in]; a new crimp over dark strands runs hot again within months.
