A cabinet is not sealed by one product. It is sealed by everything that closes every hole in it — and by the discipline of the person with the torque wrench. This page walks through all four kinds of opening on a typical outdoor enclosure, and the specific way each one lets water in.
They all punch through the same wall, but they do not do the same thing — and a method that seals one will destroy another. Sort your openings into these four groups before you price anything.
Where power and signal cables cross the wall. On a typical cabinet this is anywhere from a handful to several dozen entries, which makes it the largest single count of holes — and the one most often under-specified. A gland has to grip the cable jacket, seal the entry face and hold both under vibration.
Deliberately open to air, closed to water. It lets the cabinet breathe as the sun heats it and the night cools it, so the difference never gets applied to the door gasket and the glands. Usually one or two per cabinet. Without it, the seal becomes the pressure relief — which is exactly the wrong component for the job.
Closed until it is needed. A relief valve stays shut so the enclosure keeps its rating, then opens above a set pressure to give a fast, directed release. On a battery cabinet this is the opening that deals with a cell venting event rather than with a daily thermal cycle. It is a safety device, and it is specified separately from the breather.
Knockouts, spare cable entries, mounting holes, drain holes, an inspection opening. These are the ones nobody owns on a drawing, and they are how a cabinet that passed a water test in the factory arrives on site with an open hole in it. They need a blanking plug in the same thread as the opening — not tape, not a plastic cap, not a spare gland with nothing in it.
IP ratings are established by testing the assembled enclosure, not by adding up the ratings of its parts. A component rated IP68 has passed its own test; your cabinet still has to pass its own. In practice the cabinet tests to whichever opening fails first — and that is almost never the one with the worst specification on paper.
One definition worth knowing. In IEC 60529 the water code “8” means continuous immersion under conditions agreed between manufacturer and user — the depth and duration are not fixed by the standard. Two suppliers can both print IP68 and mean 1 m for 30 minutes or 10 m for a week. Always ask what the rating actually covers before you rely on it.
| Failure point | Why it leaks | How to prevent it |
|---|---|---|
| Gland not torqued | The face seal is never compressed, so water travels along the thread into the cabinet. | Torque to the supplier's figure with a wrench, not “hand tight plus a bit”. |
| NPT assembled dry | The taper seals on the thread flanks; with no sealant there is a continuous spiral path. | Wrap PTFE tape, or apply a sealant rated for the service. |
| Cable at the edge of its range | The insert cannot compress enough to close around the jacket. | Size so the cable sits mid-range, not at the minimum. |
| Panel thinner than the thread | The gland never grips the full wall, so the joint works loose under vibration. | State your panel thickness; add a locknut or order an extended thread. |
| Open knockout or spare entry | A direct hole — no seal to fail, because there is no seal. | Blank every unused opening in the matching thread. |
| Breather blocked or painted over | Pressure has nowhere to go, so it loads the gasket and the glands instead. | Keep the vent path clear; protect the membrane during painting and cleaning. |
| Sharp, un-deburred hole edge | The hole saw leaves a burr that chafes through the jacket over years of vibration. | Deburr every entry before the gland goes in. |
Failure modes are general enclosure practice, not a specification of any single part. Confirm the torque figure and thread details against the datasheet for the part number you buy.
We would rather ask five questions up front than quote the wrong thread three times. Send these and you will get a parts list per opening type, not a catalogue page.
None of these show up on a drawing review. They show up as rust streaks and failed insulation resistance two winters after commissioning — which is precisely the point at which the enclosure supplier and the component supplier start pointing at each other.
A breather port is meant to be open. Cap it to “make the cabinet sealed” and the pressure goes to the door gasket instead. The cabinet still leaks; it just leaks somewhere more expensive to fix.
Assembly discipline is free and it matters more than a material upgrade. A correctly torqued standard gland will outlast a premium one thrown in by hand every time.
Running a cable through an oversized opening with a foam collar is a temporary repair that becomes permanent. It defeats the entry completely and it is invisible on the bill of materials.
A breather mounted low on the side of a cabinet sits in splash and jet-wash range. Same part, same cabinet, wrong wall. Position is part of the specification.
Every component datasheet says IP68 and the enclosure test still fails. Ratings are a property of the assembly. Test the assembly, or at least walk the assembly against the failure table above before you ship.
If you are at the drawing stage, send the panel layout and we will mark up which opening takes what. It is usually a shorter conversation than reading four product pages.
A drawing, a photo of an existing cabinet or just a list of hole sizes — any of the three is enough to start. Where we cannot support a requirement, we will say so rather than quote around it.
