How long can an IP67 enclosure stay underwater

Short answer: an IP67 enclosure is proven for 1 m [3.3 ft] of still water for 30 minutes, and for nothing beyond that. The 30 minutes is a pass mark, not a countdown. Every door gasket passes a small flow of water under pressure, so a housing that stayed harmless at half an hour can be carrying a puddle by hour three. A camera filming at 0.5 m [1.6 ft] for an afternoon is the typical case: past the rating the moment you cross 30 minutes. For planned or repeated immersion you order IP68 with the depth and time agreed in writing, or a NEMA 250 Type 6P enclosure. Here is what the code proves, what it leaves open, and where the numbers put your dive.

What the second digit 7 actually tests

IEC 60529 gives the second digit 7 one test. The enclosure goes into a tank of still fresh water with its lowest point 1 m [3.3 ft] below the surface, and it stays there 30 minutes. It passes when no harmful amount of water gets in, so the pass mark is harm, not dryness. A few drops under a terminal strip can pass; the same drops across live camera contacts fail the identical box.

The first digit, the 6, is dust-tight: no dust enters at all. Here is the whole wet ladder, so you can see what each digit buys:

Digits 6 and 7 test different things, so pass one and you still know nothing about the other. A housing marked IP66 and IP67 passed both tests; order both codes when the kit takes hose-down and also goes over the side. For how the lab actually runs these, read how an IP water test checks for ingress, and keep the full ladder, first digit included, in the IP rating chart.

What the pass mark leaves out

The test answers one narrow question: did a harmful amount of water enter, once, in still water at laboratory temperature? Five things you care about sit outside it:

• Water inside is allowed. A pass can mean water parked where it does no damage, and it stays parked until you drain it.
• One exposure, one condition. A new gasket, still water, no waves, no thermal shock, the same temperature on both sides of the wall.
• The device is idle. The code grades water entry, not whether the electronics keep working once they are wet.
• No aging. Nothing in the test cycles the pressure or chills the housing first, yet that is exactly what a fifth dive does.
• Nothing about vapor. Cool the trapped air and its humidity condenses straight onto the optics; an IP67 sensor enclosure is sealed against water, not vapor.

Every seal passes water, at a rate the rating leaves unmeasured

Every mechanical seal leaks; good and bad ones differ only in how much. IPX7 tells you the leak stayed below the harm line for 30 minutes under about 9.8 kPa [1.4 psi] of water head, the pressure of 1 m of fresh water. It tells you nothing about the rate. A near-perfect O-ring could run for days; a marginal one crosses the harm line at minute 35, and both left the lab with the same label.

Depth is the smaller part of that story: that metre of water is only about 0.1 bar. Time and cycling decide the outcome. Each warm-to-cold transition and each pressure cycle works the gasket toward compression set, so the seal you trusted on the first dive is a different seal by the tenth. Door gasket materials sets out which elastomers take that punishment best.

The NEMA types that go under

US specifications speak NEMA, and the conversion is clean. NEMA 250 defines Type 6 for temporary submersion and Type 6P for prolonged submersion, which line up nearest to IP67 and IP68 respectively.

In a US installation, NEC 110.28 says an IP code alone does not replace the enclosure type marking. Order the type, and carry the IP code beside it when both audiences read the nameplate. One honest note on our own line: our stainless boxes are rated NEMA 4X / IP66 by the manufacturer; not UL listed, and IP66 is a jet rating. The catalog box is therefore the wrong tool for anything that submerges, and the NEMA 4, 4X and IP66 guide covers what that rating does buy you.

A camera housing at 0.5 m for a 3-hour shoot

Take the case this question usually hides: filming at half a metre for an afternoon. Work the loads on the seal one at a time.

Inputs:

• Depth 0.5 m [1.6 ft] and about 3 h of filming — the job itself.
• The rating on the label: tested at the rated depth for 30 min in still water (IEC 60529, second digit 7).
• Water at 15 C [59 F]; the housing was closed indoors with its internals near 35 C [95 F], typical of running electronics.
• Physics: water pressure is depth x density x gravity, fresh water at 1000 kg/m3 and g at 9.81 m/s2. Trapped air in a rigid box keeps its volume, so its pressure follows its absolute temperature.

The rule in words: two loads push water through the gasket — the static head of water above the seal, plus the partial vacuum left behind when the trapped air cools. Add them.

• Static head at 0.5 m: 1000 x 9.81 x 0.5 = 4.9 kPa [0.7 psi]
• Test head at the full rated depth, for comparison: 1000 x 9.81 x 1.0 = 9.8 kPa [1.4 psi]
• Cooling: 35 C is 308 K, 15 C is 288 K, so the air loses 20/308 = 6.5 % of its pressure: 0.065 x 101.3 kPa = 6.6 kPa [0.96 psi]
• Total across the gasket: 4.9 + 6.6 = 11.5 kPa [1.7 psi]
• Time: 3 h is 180 min, which is 6 x the 30-minute test

Half the tested depth, and the seal works harder than the lab ever made it: 11.5 kPa against 9.8 kPa, for six times as long. The vacuum does most of the damage, and it appears in every real dive where a warm box meets cold water.

You can delete the vacuum. Close the housing at water temperature with the electronics cooled, and the load falls to the 4.9 kPa [0.7 psi] static head, half the tested pressure. The time still runs 6 x past the test, so the label still does not cover the dive.

Decide it on three thresholds. Shallower than the rating, under 30 minutes, unplanned: that accidental dunk is exactly what IP67 is for. Planned immersion past 30 minutes, or any water deeper than the rating, moves you off IP67; order IP68 with the depth and time in writing — 2 m [6.6 ft] for 24 h is a typical agreement — or a NEMA 250 Type 6P enclosure. A camera you cannot afford to lose gets the same order before the first dive, because finding out in the water is the expensive way.

If the housing floats at the surface in chop, add pumping to the load: every wave flexes the gasket and moves water. Moor it at a steady depth and the job gets easier.

What to put on the order for real immersion

When the job is hours under water, put these four lines on the order:

• The rating, with numbers. IP68 for continuous immersion, with depth, time and the water named. The code leaves all three to be agreed between maker and user, so a bare IP68 on a datasheet is half a spec; why IP68 ratings carry depth and time limits explains what to demand in writing.
• The whole assembly at that rating. Every gland, window and connector drags the housing to its own level, so the weakest part sets the rating. A box rated to 10 m with an IP67 gland is an IP67 box.
• The right metal in salt water. 316L, because 304 can pit under chloride deposits even at room temperature. Rinse the housing in fresh water after every recovery and inspect the seal groove for silt.
• A body that fits the camera. Our stainless line runs 6 x 4 x 4 in to 60 x 36 x 16 in (H x W x D), cut to your drawing and delivered in 10 business days from your approval of the drawing; start in the size finder or the enclosure selector.

Between dives, dry the housing and check the gasket before it closes again. Before you open anything that has been down, isolate the supply and let it drain, because water bridging live terminals is a shock path. For a housing that lives under water for a season, what to specify for continuous immersion takes the spec from here.

Questions