Condensation in junction box assemblies is the failure that makes no sense until you run the numbers: a box with a valid IP67 certificate, correctly installed, undamaged, holding water a year later. Nothing leaked through a hole. The water was carried in as vapour and released onto a cold surface. This guide covers the four routes by which moisture enters a sealed enclosure, works through the arithmetic of one day-night cycle, and sets out what removes the problem rather than relocating it. It applies to any waterproof junction box installed where temperature moves.

Four routes into a sealed box

A sealed enclosure is not a dry enclosure. It holds whatever air was inside on the day it was closed, plus whatever has diffused or been pumped in since.

Each route delivers milligrams, not litres, per event. That is exactly why the problem is dismissed at commissioning and undeniable a year later.

The arithmetic of one temperature swing

Step one: the dew point of the trapped air

Air at 30 °C and 60% relative humidity holds about 18.2 g of water per cubic metre and reaches saturation at roughly 21 °C. The moment any internal surface drops below about 21 °C, the air touching it gives up water. Outdoors that is routine: the lid radiates to a clear sky and falls below air temperature.

Step two: convert it to milligrams

Cool that air to 15 °C and saturation capacity falls to 12.8 g/m³, so the 5.4 g/m³ difference has to condense somewhere. A compact 100×100×70 mm box with about half a litre of free air yields roughly 2.7 mg per cycle; a 200×200×80 mm box with around 2.4 litres yields about 13 mg. Neither sounds like much until you remember the nightly repeat.

Step three: the pressure the cycle generates

In a rigid enclosure the pressure change scales with the absolute temperature change: a 20 °C rise from a 20 °C start gives roughly 0.07 bar internally, a 30 °C rise about 0.10 bar. That explains a detail most buyers ignore. Vent plug datasheets quote airflow at 0.07 bar, which is not an arbitrary test point but approximately what one sunny morning does to a sealed box.

Pressure works both ways. On cooling the internal pressure falls below ambient, the box pulls ambient air in through the gasket line and entry threads, and that inward stroke delivers the next night’s condensate.

Why the tighter the box, the worse it looks

An IP68 box has the smallest leakage paths, so it exchanges the least air, which sounds like an advantage. The vapour that does get in stays in, and a well-sealed box with a large internal volume and metal components supplies the three things condensation needs: a trapped vapour load, a cold surface and no route out. A loosely closed box often looks drier inside precisely because it ventilates freely, at the cost of admitting liquid water and dust. The answer is not a tighter seal but a controlled exchange. The JX-AG IP67 enclosure, rated to one metre immersion and sized up to 280×190×180 mm, is that profile: a large trapped volume under a gasketed lid.

What works, and what only moves the problem

A vent plug is a small threaded body containing a microporous membrane, sealed to the wall by an O-ring. Air and vapour pass; liquid water does not. The breathable vent plug for this duty is specified at 325 ml/min airflow at 0.07 bar, water entry pressure above 5 bar, an O-ring rated IP68, and −40 °C to +120 °C with short excursions to +150 °C.

Two figures deserve interpretation. Water entry above 5 bar is roughly fifty metres of static head, so the membrane is not the weak point of an immersion-rated assembly. And the airflow is a rate, not a verdict: it is quoted at the same 0.07 bar that a 20 °C swing generates. Desiccant suits a box closed once and left alone; heaters suit large cabinets that already have power; gel suits joints that are never reopened. The pattern to avoid is fitting desiccant to a box that is still breathing, which converts a slow problem into a scheduled one.

Sizing and fitting a vent plug

Sizing uses the airflow figure rather than a rule of thumb: divide the free air volume by the rated flow at 0.07 bar. A half-litre box with a 325 ml/min vent exchanges one internal volume in about a minute and a half, far faster than any real temperature ramp. A ten-litre box takes roughly half an hour, which is where a second vent or a higher-flow body becomes worth considering.

• Mount low, membrane facing down. On a vertical wall use the lowest available position so water runs off the body instead of pooling against the membrane.
• Seal with the O-ring, not with tape, and keep paint, solvents and silicone spray off the membrane; a blocked membrane turns the vent into a plain plug.
• Do not fit one to hide a leak. A vent equalises vapour pressure, not an open knockout.

The same membrane principle is built into breathing cable glands, which combine the entry and the pressure path in one component; the trade-offs are compared in this guide to breathing cable glands.

Assembly habits that reduce internal moisture

No accessory can remove water that was built in, so three habits matter. Close the box in dry conditions, because assembly on a humid morning loads the enclosure with the maximum vapour it can hold. Never leave an enclosure open overnight through a cool, humid night and then close it in the morning with that moisture sealed inside. And seal every cable path deliberately: correct gland size for the cable’s outer diameter, a plugged spare entry, and a gasket that is compressed rather than crushed. Entry configuration and body sizing are covered in this IP68 junction box selection guide.

One structural choice matters here. The F1 mini junction box uses a silicone O-ring in a groove rather than a flat gasket under a lid, which suits a small enclosure that cycles repeatedly, because an O-ring is less sensitive to uneven screw torque.

Conclusion

Condensation inside a sealed box is not a defective seal and not a reason to buy a higher IP rating. It is a vapour and pressure problem: humid air is trapped or pumped in, a cold surface appears, and water that was always in the air becomes visible. The fix is to give that vapour a controlled way out while keeping liquid water out, which is what a membrane vent does. Size it from free volume and rated airflow, mount it low with the membrane facing down, and specify it with the enclosure rather than after the first failure.

Send the enclosure size, the temperature range and the mounting position and we will confirm whether pressure equalisation is needed and which vent body suits the environment. For the rating side of the same decision, see what an IP rating on a junction box tests.

Frequently asked questions

Can a waterproof junction box get condensation inside?

Yes, and a well-sealed one is more likely to keep it. Waterproofing stops liquid water; it does not remove the vapour in the air closed inside, or stop vapour diffusing through polymer walls.

Will a vent plug let water into the enclosure?

No. The membrane passes air and vapour but blocks liquid water, and the unit is specified with water entry pressure above 5 bar and an IP68-rated O-ring.

How many vent plugs does one box need?

Size it rather than guess: divide free internal volume by airflow at 0.07 bar and compare with the real temperature ramp. A half-litre box equalises in about a minute and a half at 325 ml/min; a ten-litre box takes roughly half an hour.

Is desiccant enough instead of a vent?

Only in a box closed once and not reopened. Desiccant has a fixed capacity and saturates, after which it holds moisture inside the enclosure.

Does condensation mean the IP rating is fake?

No. IP testing is a static immersion or jet test on a new sample and does not include pressure cycling or vapour transmission, so a genuine IP67 box can still condense water inside.