Static shock from an enclosure: why millivolts still bite

Short answer: the snap off a grounded enclosure is your own static charge, picked up on a fiberglass ladder, rubber soles and gloves, and the drifting millivolts on the meter are the signature. A body charged to 6 kV carries about 0.9 µC and 2.7 mJ. It dumps the lot in under a microsecond, between the samples a meter takes, so the instrument shows nothing but noise. Fix the person, not the box: bleed the charge through a 1 MΩ wrist strap, hold the room at 40% relative humidity or higher, and keep the enclosure bonded for fault current. If a reading holds steady instead of decaying, stop calling it static and hunt the leak.

What the symptoms point to

Sort what you felt against these signatures. Three or more together and the case for static is closed.

• The snap jumps before contact. Mains voltage at 120 or 240 V cannot cross a visible gap in air. A spark that leaps is kilovolts, and kilovolts on a person means static.
• Several spots, approach only. You are the capacitor, so every grounded surface within reach discharges you. A live enclosure stings at one point and keeps stinging while you hold it.
• Tick tracer silent, meter in millivolts. There is no usable AC on the box. The drifting millivolts are the meter reporting its own pickup, not a source on the enclosure.
• The reading climbs toward the lights. A dangling lead couples capacitively to lighting and branch wiring, and a 10 MΩ meter input turns nanoamp pickup into hundreds of millivolts. Aim the lead at building steel and the reading falls; a real conductor does not care where the lead points.
• Current stuck near 0.3 mA and steady. That is a thirtieth of the 10 mA let-go floor (IEC 60479-1), so nothing in it can hold a muscle. The word to watch is steady: static never sits still on a meter.
• Seasonal and surface pattern. Dry heating-season air, a fiberglass ladder and rubber soles make up the classic triboelectric trio, and shocks that fade on humid days point the same way.

One rule outranks the whole list: prove the meter on a known live circuit, then confirm the enclosure is dead before hands go on. Static stays the diagnosis only for a box that passes that check.

Why the meter shows millivolts

Take the case a winter substation visit produces: a 6 ft fiberglass ladder, rubber-soled boots, dry air, and a body that has charged to 6 kV. Perception starts near 2-3 kV in ESD studies, so 6 kV is a firm snap that jumps a gap. Run the numbers on a body capacitance of 150 pF, between the 100 pF of the standard human-body model and the 200 pF a larger person reaches.

• Charge: Q = C × V = 150 pF × 6,000 V = 0.9 µC.
• Energy: E = ½ × C × V² = 0.5 × 150 pF × (6,000 V)² = 2.7 mJ.
• Peak current: 6,000 V ÷ 1,500 Ω of body resistance = 4 A, for under 1 µs.

The meter sees none of it. Its 10 MΩ input drains a 150 pF body with a time constant of 10 MΩ × 150 pF = 1.5 ms, while the meter integrates each reading over hundreds of milliseconds, so the pulse decays about 200 times faster than the window it lands in.

The drift toward the lights is the same physics in slow motion. A lead hanging near lighting wiring couples about 1 pF. At 120 V and 60 Hz, that coupling induces I = 2π × 60 Hz × 1 pF × 120 V ≈ 45 nA. Across the meter's 10 MΩ input, 45 nA reads about 450 mV, the same order as a 290 mV climb, with no source on the enclosure at all.

The sting is real because skin sensors respond to the 4 A edge, not to the small total energy. Lockup never enters the picture: it takes sustained current, and this pulse lasts a microsecond.

The causes, most likely first

The first three explain the snap. The last two explain any reading that sits still, and those are the rows that earn a lockout.

The fix, in order

1. Bond the enclosure for fault current. NEC 250.96(A) requires metal parts serving as equipment grounding conductors, enclosures included, to be bonded so they can safely carry the fault current imposed on them. Scrape paint and lacquer off the lug pad, as NEC 250.12 directs for grounding contact surfaces. The grounding and bonding guide covers lug placement, and the disconnect enclosure bond to the ground ring article shows the hardware. Bonding ends leakage and floating-hardware shocks; it was never the cure for the snap, because the charge is on you and the bonded box is where it jumps.
2. Bleed the charge before it can jump. Clip an ESD wrist strap with its built-in 1 MΩ resistor to the bonded box or a known ground. At 6 kV the strap bounds the discharge at 6 mA spread over about 150 µs (1 MΩ × 150 pF), low and slow enough that no spark forms. A metal ladder takes a bond wire to ground. A fiberglass ladder cannot be grounded, so the strap is the way.
3. Strike with a tool when no strap is at hand. Hold a bare screwdriver shaft or a key, touch the box with it first, and let the snap land on metal instead of a fingertip. Two touches leave the body discharged for that climb.
4. Raise the room to 40% relative humidity or higher, the working floor ESD control practice uses; below that, surfaces lose their moisture film and routine movement charges a body into the kilovolt range. Keep added moisture out of sealed enclosures, because humid air trapped inside condenses on cold walls overnight; the condensation in a sealed enclosure article covers what follows. The same band suits storing spare boards without corrosion.
5. Match the gloves to the hazard. Cotton or ESD-dissipative gloves suit static-prone cleaning work. Rubber insulating gloves belong to AC work and are chosen by class: Class 00 tops out at 500 V AC and Class 0 at 1,000 V AC in the ASTM D120 class system. The insulating glove selection page for arc-flash and shock walks that choice in detail.

When it is not static

Decide between static and leakage with one test. Set the meter to AC volts between the enclosure and a known building ground, press both leads firmly, and hold them for a minute.

• The reading decays toward zero and the snaps happen only on approach: static, and the fix list above ends it.
• The reading holds at the same value on repeat: leakage. Lock the circuit out and find the source before the next shift works the box. Start with EMI filters on nearby drives, where two or three 2.2 nF capacitors reach a steady 0.3 mA; the VFD enclosure wiring checklist covers the bonding and filter layout. Then look for damaged insulation, which the shock from a metal enclosure article walks through.
• The reading holds above a few milliamps: stop touching the box, lock out, and treat the enclosure as live until the source is cleared and the meter confirms it.

For new work, pick hardware that keeps its bond for life. Bare stainless contact surfaces need no scraping at the lug. A 16 x 12 x 6 in 304 stainless wall-mount enclosure, rated NEMA 4X / IP66 by the manufacturer and not UL listed, takes a ground lug straight onto clean metal.

Decide it in one pass. Snap on approach with drifting millivolts is static: strap up, raise the humidity, keep the bond. A reading that holds is leakage: lock out and fix the source. No snap and no reading means the box was only missing a clean bond.

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