Short answer: the "2 m [6.6 ft] for 30 min" on an IP68 rating is the condition one sample passed in a lab, not a countdown to failure. At 2 m of water the pressure is about 20 kPa [2.9 psi], and a clean o-ring squeezed into its gland holds that indefinitely; depth alone almost never kills a seal this shallow. Time shows up twice: the standard has to end its test sometime, and a real seal in the field wicks water past grit, takes a permanent set, and breathes humid air every cool night. Split the pressure problem from the time problem and the numbers on the label start making sense.
What the IP68 test actually says
IEC 60529 splits immersion into two digits. Second digit 7 means temporary immersion, and its test is fixed: the enclosure goes under water with its lowest point at least 1 m [3.3 ft] below the surface and its highest point at least 0.15 m [6 in] under, held there 30 min, with the water within 5 K [9 F] of the specimen's temperature. The pass line is not "bone dry". Water may enter as long as it does no harm, in the standard's judgment, to live parts or to operation.
Second digit 8 means continuous immersion, and here the standard fixes almost nothing. The depth and the duration are agreed between the manufacturer and the test house, and they only have to be more severe than the digit-7 test. That is why one datasheet pairs a shallower depth with a longer soak, the next goes deeper for a shorter one, and a phone case quotes 6 ft [1.8 m]: each pair is a chosen test point the maker stands behind, not a physical constant. The IP rating chart breaks the whole code down digit by digit, and IP67 vs IP68: which battery project actually needs covers choosing between the two.
Two meters of water is easy for a sound seal
Water pressure grows at 9.8 kPa per meter of depth [1.4 psi per 3.3 ft] — the familiar 1 bar [14.5 psi] per 10 m [33 ft]. At the common digit-8 datasheet point that is 19.6 kPa [2.8 psi]; at a phone case's depth, about 18 kPa [2.6 psi]. Put that next to the seal: an o-ring squeezed 20% in its gland presses on the sealing faces with roughly 1 MPa [145 psi] of contact stress, about 50 times the head above. Outside pressure also wedges the ring harder into its groove, so depth helps the seal until you reach extrusion and structural limits somewhere below 100 m [330 ft].
What depth really stresses is the shell. A flat 12 x 12 in [305 x 305 mm] door face at that depth carries 19.6 kPa × 0.093 m² ≈ 1.8 kN [410 lbf] of water. A snap-together phone shell or a wide, thin door bows under that load, the gasket unloads at the middle, and a capillary gap opens. That is a stiffness limit set by design, and it does not care whether the dunk lasts a half-hour or a month.
| Depth | Static head | What it asks of the seal |
|---|---|---|
| 1 m [3.3 ft] | 9.8 kPa [1.4 psi] | IEC 60529 digit-7 test point, 30 min |
| 1.8 m [6 ft] | 18 kPa [2.6 psi] | Typical phone-case claim |
| 2 m [6.6 ft] | 20 kPa [2.9 psi] | Common digit-8 datasheet point |
| 10 m [33 ft] | 98 kPa [14 psi] | Still a tenth of a 1 MPa squeeze |
Why time leaks when pressure does not
A clean seal either holds a given pressure or it does not; four real-world mechanisms are what make water entry depend on time instead.
-
Capillary wicking past contamination. A hair, a sand grain or a metal chip across a face seal leaves a channel a few thousandths of a millimetre wide. A wetted 20 µm [0.0008 in] gap pulls water by capillarity with about 7 kPa [1 psi] of suction — the same pull as 0.7 m [2.3 ft] of water head, and it works at zero depth. The flow is tiny, so a half-hour test shows nothing and a three-week dunk ends in a saturated interior. Battery covers and charge-port doors, the seals you open and re-close in dirty air, are the first to give trouble.
-
Compression set. An elastomer under squeeze slowly flows, and the contact stress decays without fully recovering. A good EPDM compound shows 10 to 20% set after 1000 h at 100 C [212 F] in the ASTM D395 test, and the loss is permanent. At 25 C [77 F] the same decay takes years, but every door opening adds its share. The foam vs rubber door gasket comparison shows how much ground soft foam gives up to solid rubber here.
-
Permeation through the elastomer. Water vapor diffuses straight through the seal material itself; silicone passes roughly ten times more than an FKM of the same section. The grams per week are small, but months of them matter to electronics that must never see condensation.
-
Breathing. This is the big one for enclosures. Work it through for a 20 L [5.3 gal] box sealed dry at 30 C [86 F] and 101 kPa [14.7 psi]. Overnight, box and contents fall to 10 C [50 F]; the trapped air follows P2 = 101 × 283/303 = 94 kPa, so the inside sits 6.7 kPa [1 psi] below the outside — the pull of 0.68 m [2.2 ft] of water through any leak path there is. Each cool half-cycle draws in a few cubic centimetres of humid air; each warm half-cycle lets drier air out. One cycle a day for a month pumps a measurable slug of vapor inside, where it condenses on the coldest metal. An enclosure pressure equalization vent stops the loop by balancing pressure while blocking liquid.
Where the 30 minutes comes from
A compliance test has to end. The lab dunks one new enclosure, with clean seals and clean water, at a fixed depth for a fixed time, then opens it and looks for harmful entry. Thirty minutes matches the digit-7 duration, and doubling it doubles lab cost for little new information, so the industry settled on that duration at datasheet depths. The duration is economics and warranty scope, not physics: the maker stands behind exactly what was demonstrated. How an IP water test checks for ingress walks through what the lab actually measures.
The US enclosure standards show the same trade plainly. NEMA 250 puts the Type 6 submersion test — occasional, temporary dunking — at 6 ft for a half-hour. Type 6P, intended for prolonged submersion at a limited depth, gets 6 ft for 24 h. Even "prolonged" in a standard is a one-day test. The multi-year service claim rests on design margin and on your keeping the seal in condition, which is why the mechanisms above matter more than the test duration ever will.
What this changes in a real panel
Read any immersion rating as a three-part claim: a depth, a duration, and the seal condition as tested — new, clean, unrepeated. Your panel adds grit, cleaning chemistry, sunlight and years of temperature swings. Do enclosure ratings expire when washdown chemicals change takes the chemistry side; the mechanical side is the set-and-wicking story above.
In a US installation, NEC 110.28 is blunt: an IP code alone does not satisfy a drawing that calls for an enclosure type. Write the NEMA type and carry the IP code with it; many enclosures are marked with both for exactly that reason. And treat any enclosure that has been underwater as live until you lock it out and prove the circuits dead, whatever the label says.
What to put on the spec
Match the duty, then write the numbers down:
| Your duty | What to specify |
|---|---|
| Hose washdown, rain, jets | NEMA 4X / IP66 stainless box, gasketed hinged door, drains at the low point |
| One-off accidental dunk under 1 m [3.3 ft] | The same 4X / IP66 box on a rigid flange, EPDM gasket, vent above the waterline |
| Repeated dunking to 2 m [6.6 ft] | IPX8 with the pair written in — "2 m, continuous, fresh water" — on FKM or EPDM |
| Permanent submersion | Type 6P-class construction, built and tested to that class |
For the top two duties, the stainless enclosure size finder covers it: every box is rated NEMA 4X / IP66 by the manufacturer, not UL listed, cut to your drawing and delivered in 10 business days. For the last row, continuous immersion enclosures: what to specify and verify is the deep answer, and the enclosure selector carries your conditions to a custom quote.
Decision in one line: depth picks the shell stiffness and the seal material; time picks the maintenance interval. Clean the gasket land at every opening and replace the seal on schedule, because set and grit are the only clocks actually running.
Enclosures for this job
Catalog sizes that suit the job this article answers, picked from its subject; open one for its full specifications.
| Enclosure | Material | Outside H x W x D / display | Rating |
|---|---|---|---|
| 12 × 10 × 6 in | 304 stainless | 12 x 10 x 6 in | Rated NEMA 4X / IP66 by the manufacturer; not UL listed |
| 20 × 16 × 8 in | 304 stainless | 20 x 16 x 8 in | Rated NEMA 4X / IP66 by the manufacturer; not UL listed |
| 24 × 20 × 8 in | 304 stainless | 24 x 20 x 8 in | Rated NEMA 4X / IP66 by the manufacturer; not UL listed |
Questions
Does an IP68 rating mean the enclosure leaks after 30 minutes?
No. That figure is the duration the lab tested one sample for; a sound seal carries the 20 kPa [2.9 psi] at the rated depth for years. What grows with time is wicking past a contaminated joint, permanent gasket compression set, and daily breathing cycles — effects measured in weeks, not minutes.
Can a NEMA 4X / IP66 enclosure survive being dropped in water?
Usually once, and briefly. That rating is proven with jets — a 12.5 mm nozzle delivering 100 L/min from 2.5 to 3 m [8 to 10 ft] — not submersion. A dunk to 1 m [3.3 ft] adds just 9.8 kPa [1.4 psi], which a good door gasket holds, but the rating promises nothing about it. For repeated submersion, buy the IPX8 spec with a written depth and duration.
Why do some ratings trade depth for time, like 3 ft for 60 min?
Because those are two test points the manufacturer paid to have demonstrated. IEC 60529 sets no digit-8 pair at all; it only demands conditions harsher than the digit-7 temporary-immersion test. Shallower-for-longer and deeper-for-shorter are both valid ways of being more severe, so you see both on labels.
How deep can an o-ring seal go before water pressure defeats it?
Deeper than almost any enclosure application. A ring squeezed 20% carries about 1 MPa [145 psi] of contact stress, while water adds only 9.8 kPa per meter [0.43 psi per foot]; you need roughly 100 m [330 ft] before the head nears the squeeze. The real depth limits are shell deflection — about 1.8 kN [410 lbf] on a 12 x 12 in door at the common datasheet depth — and extrusion into gaps.
What maintenance keeps a submersion rating honest?
Clean the gasket land at every opening and replace the seal on a set interval. A quality EPDM compound takes 10 to 20% permanent compression set after 1000 h at 100 C [212 F]; at ambient the same loss takes years, but it never reverses. Any grit left on the land gives water a 20 µm [0.0008 in] capillary path worth about 7 kPa [1 psi] of suction.
