Pump the box out, rinse the salt off, dry the conductors, and megger each one at 1,000 V DC. The acceptance floor for 600 V cable under ANSI/NETA ATS is 100 megohms corrected to 20 °C [68 °F], and XHHW-2 is built for exactly this duty: a wet-location listing at up to 90 °C [194 °F]. Cable that tests clean stays in service. What never stays is anything that trusted a sealed surface: every splice and termination that sat in the water is cut out and remade, and soaked breakers, fuses or electronics downstream are replaced. Before the cover comes off, lock the circuit out and verify absence of voltage.
What you see in a flooded box, and what each sign means
Two years of standing salt water leaves a record, and you can read most of it before the meter comes out. Open the box, photograph the water line, and work down this list.
- Water over the conductors when the cover comes off. Treat everything below the line as contaminated until a test says otherwise.
- Green or white crust on exposed copper. Chlorides have attacked the strands and are wicking under the insulation; trim back to bright metal.
- An oil sheen on the water or a film on the insulation. Hydrocarbons hold grit and moisture against the wall and spoil splice kits; clean before any tape or resin work.
- Smooth, milky insulation with no cracks or swelling. Surface wetting only; clean it, dry it, test it.
- Brown tracks or blisters in the insulation wall. A genuine insulation fault; that conductor fails no matter what the meter reads.
- A megger reading that keeps climbing. Water is still leaving a splice or the box; repeat the test the next day.
Why XHHW-2 usually survives a soaking
The cable is the strong part of this installation. XHHW-2 uses cross-linked polyethylene insulation, a dense thermoset that absorbs very little water, and the -2 suffix is the wet-location mark: 90 °C [194 °F] in wet or dry locations, at 600 V, with some runs dual-rated 1,000 V. Water resting on a clean insulation wall leaks almost no current. The leakage that fails a test lives on the surface and in the splices: salt film is hygroscopic, so it pulls moisture from the air and conducts when damp. That is how a box that passed in dry weather tracks out in the next wet spell. Oil does little electrical damage, but it holds grit against the wall and keeps heat-shrink and resin kits from bonding. Clean both films off and the cable is back to its listed performance.
What actually fails, ranked by likelihood
Insulation wall failure is the rare case; the causes below cover nearly every flooded-box failure you will meet. Work them top to bottom.
| # | Cause | How you check it | What the result means |
|---|---|---|---|
| 1 | Splices, taps or terminations under the water line | Megger with splices joined, then separated | Cable alone passes and joined it fails: remake the splice |
| 2 | Salt and dirt film on every surface | Wipe with a white cloth, then rinse with fresh water | White residue on the cloth; readings rise after rinse and dry |
| 3 | Water held in duct or conduit low points | Blow the duct out and re-test after a drying day | The reading climbs once the duct dries: the cable was sound |
| 4 | Contamination under cut or abraded insulation | Inspect exposed lengths and trim to bright copper | Corrosion under the wall: replace that length |
| 5 | Real degradation of the insulation | Corrected reading still below 100 MΩ after full drying | Rare with XLPE; replace the affected section |
Cause 1 is where flooded runs usually fail, because a splice is a hand-built seal over surfaces that were never factory-clean. A splice-free pull box skips it entirely, which is why the cable-only megger decides that job on its own. Cause 3 fools anyone who tests an hour after pumping: the box is dry and the duct is not.
The megger test and the acceptance numbers
Disconnect the conductors at both ends and test each one phase-to-ground, then phase-to-phase, at 1,000 V DC, the test voltage ANSI/NETA ATS pairs with 600 V cable. The acceptance minimum is 100 megohms corrected to 20 °C [68 °F]. An older field rule of thumb, 1 megohm plus 1 megohm per kilovolt of rating, puts the bare serviceability floor at 1.6 megohms for 600 V cable (1 + 0.6 = 1.6). Use the 100 MΩ value as the line for new or recovered work.
Temperature moves the reading: insulation resistance roughly halves for every 10 °C [18 °F] rise, so a number taken at 10 °C is halved to give its 20 °C value. Readings also fall roughly in proportion to run length, so compare like with like when one run is 300 ft [91 m] and the sample was 10 ft [3 m]. After a pass, do two more things: repeat the test the next day to confirm the reading is stable, and inspect the insulation surface under good light for cuts and abrasions. A stable pass on undamaged XLPE is a pass. For test voltages and minimums at other cable ratings, the table in minimum megger values for 600 V cable collects the values.
Worked example: one bad splice in 300 ft of #8
Take a 300 ft [91 m] run of four #8 XHHW-2 conductors between two in-ground boxes, with one taped splice per conductor in the far box. The crew pumps out, rinses, dries, and tests at a steady 20 °C [68 °F], so no temperature correction is needed. Phase-to-ground at 1,000 V DC, load end disconnected:
- Conductors A, B and C read 1,200-1,300 MΩ, 12 to 13 times the 100 MΩ floor: a clean pass.
- Conductor D reads 60 MΩ: a fail.
Separating D's splice at both ends splits the fault. The cable alone reads 1,800 MΩ, 18 times the floor, while the splice body reads 40 MΩ. That splice spent months under salt water, so it is cut out and remade with a wet-location kit. A contaminated splice is never dried and re-used; salt trapped under the seal keeps tracking long after the test passes. The cable stays in service with a logged pass. Had the same borderline splice read 90 MΩ on a 10 °C [50 °F] morning, the correction would halve it: 90 × 2^((10-20)/10) = 90 × 0.5 = 45 MΩ at 20 °C, still a fail, but an honest one. Log every reading with its test voltage and temperature; that record is what gets the installation accepted.
The fix, in order
- Pump the water out and rinse every surface and conductor with clean fresh water, then pump again; the rinse is what removes the salt.
- Dry with moving air, a fan over the open box, until megger readings stop climbing between tests.
- Trim every corroded copper end back to bright metal; strands still discolored under the insulation get cut further back.
- Cut out every splice, tap and connector that sat under the water line, remake them with wet-location kits, and strap the remakes to the side wall above the box floor.
- Megger each conductor phase-to-ground and phase-to-phase at 1,000 V DC against the 100 MΩ floor, correct to 20 °C, and log each reading with its temperature.
- Repeat the full test after another day of drying; a stable pass closes the job.
- Seat and latch the gasketed cover, fit the drainage the box allows, and add a rain check to the site walk until handover.
Keeping the box dry from now on
Water got in because covers sat open during construction, so most of the cure is site discipline: covers on and latched from the day cable is pulled. Under NEC 314.15, boxes and conduit bodies listed for damp or wet locations take approved drainage openings of 3 mm [1/8 in] to 6 mm [1/4 in], and a bottom drain at the low point turns the next intrusion into a puddle instead of a bath. A gravel base under an in-ground box keeps the floor above the standing water. It pays to read how NEMA 3R and Type 4 differ on drains before you drill.
Know what the box was bought for, too. NEMA 250 gives Type 4X everything Type 4 gets plus 200 hours of salt spray checked against a 304 stainless specimen, the extra step compared in Type 4 vs 4X vs IP66, so a sound 4X handhole shrugs off marine air. The type covers spray and hose water, not submersion: Type 6 is the temporary-submersion type, nearest to IP67. Enclosure type ratings are manufacturer ratings, not UL listings; a box carries a UL listing only when a lab has evaluated it to UL 50E. On the wider question of why sealed beats open outdoors, the sealed-box-versus-vent comparison makes the case.
When the conductors or the box come out
- Conductors: replace the run or the damaged length when the corrected reading stays below 100 MΩ after full drying, when insulation shows cuts, tracking or blisters, or when trimming to bright copper costs strands.
- Devices: anything downstream that sat under the water is replaced, not re-tested — molded-case breakers, fuses, GFCI and AFCI devices and all electronics, under NEMA's water-damaged-equipment guidance.
- The box: hot-dip galvanized cast iron usually survives wet ground; judge it on the coating, wire-brushing and zinc-rich recoating any rust bloom at chips and welds.
- Above-grade wet spots: where a box keeps taking rain or wash-down above grade, gasketed 316L wall-mount enclosures rated NEMA 4X / IP66 by the manufacturer; not UL listed keep the water out from day one. Where a box will genuinely sit under water, that is a custom design with its own cable-routing problem — see routing cables dry through a submerged enclosure.
Put the numbers together and your accept-or-replace call writes itself. Corrected megger readings above 100 MΩ with clean, undamaged insulation keep the cable; every soaked seal is remade; soaked devices are replaced; the box is judged on its coating and drained for the next storm. For the replacement itself, start from what to specify for an outdoor service conductor pull box, and for what neglect costs when a splice finally tracks, burned splices in outdoor junction boxes shows the end state.
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.
| Enclosure | Material | Outside H x W x D / display | Rating |
|---|---|---|---|
| 6 × 6 × 4 in | 304 stainless | 6 x 6 x 4 in | Rated NEMA 4X / IP66 by the manufacturer; not UL listed |
| 8 × 6 × 4 in | 304 stainless | 8 x 6 x 4 in | Rated NEMA 4X / IP66 by the manufacturer; not UL listed |
| 10 × 8 × 4 in | 304 stainless | 10 x 8 x 4 in | Rated NEMA 4X / IP66 by the manufacturer; not UL listed |
Questions
What about the streetlight heads or panel the flooded cable feeds?
The cable can pass while the devices it feeds are ruined. NEMA's water-damaged-equipment guidance replaces every molded-case breaker, fuse, GFCI, AFCI and electronic device that sat under the water; drying out and a passing insulation test do not put them back in service. Motors are the exception worth testing: bake the windings dry, megger them again, and decide on the reading.
How long should the conductors dry before a megger reading means anything?
Give the box 24 hours of moving dry air after the last free water is gone, then test. A reading that keeps climbing while you hold the 1,000 V DC test voltage means moisture is still leaving a splice or the box; repeat the next day. If you add heat, stay at or under 50 °C [122 °F], far below the insulation's 90 °C [194 °F] wet rating.
Can a splice that passes the megger stay in service?
Not when it sat in contaminated water. Heat-shrink tubes, resin and tape kits bond to clean, dry surfaces; salt and oil trapped under the layers keep leaking and tracking long after today's pass. Cut the splice out and remake it with a wet-location kit, then test and log the remake. A megger reading describes today, not the seal's future.
Is a cast iron handhole still reusable after two years of salt water?
Usually yes. Inspect the coating at chips, welds and the cover seat, wire-brush any rust bloom back to sound metal, and recoat with zinc-rich paint; two thin coats bond better than one thick one. The gasketed cover matters more than the box wall: seated and latched it keeps the next rain out, while a missing seal lets the box fill again in a single storm.
