Why a Breathable Jacket Fills With Water From the Inside

This page contains affiliate links. If you buy through one we may earn a commission, at no cost to you.

Mostly you cannot, and the reason is worth understanding because it kills the instinct to solve it by buying a better jacket. The water on the inside of your shell is largely not sweat. It is condensate. It left your skin as vapour and became liquid again on the inner face of the shell, because that surface is around 20 kelvin below the dew point of the air behind it, and a waterproof membrane is specifically designed not to pass liquid water.

Once it has condensed, the only thing driving it out is the difference between the vapour pressure of a cold film of water and the vapour pressure of the cold wet air outside, which is 0.175 kilopascals. The rating on the jacket was measured across 3.55. So the shell is working on five percent of its headline figure, and the arithmetic says what you already know: it does not keep up.

On paper the jacket is exactly good enough

Start with the two numbers that make this look like a solved problem. Walking uphill under a pack produces vapour at something like 700 grams an hour, which is the middle of the usual half to one litre range. A mid grade membrane is rated at 10,000 grams per square metre per 24 hours. A jacket is about 1.8 square metres of fabric, so that is 18,000 grams per 24 hours, or 750 grams an hour.

Seven hundred against seven hundred and fifty. The specification matches the body almost exactly, which is presumably not a coincidence, and it is why the figures get printed. The trouble is entirely in the conditions those figures were measured under, because a transport rate is never a property of a material on its own. It is a property of a material and the difference applied across it.

The dew point decides everything, and it is not close

Air trapped between your skin and your insulation sits near your skin temperature and near saturation. Call it 33 degrees Celsius at 95 percent relative humidity. Water's saturation pressure at 33 degrees is 5.034 kilopascals, so that air carries 4.78, and its dew point works out at 32.1 degrees.

Now the shell. It is a thin membrane with liquid water on its outside face, in 10 degree air. Nothing is keeping it warm, and evaporation from the wetted face is actively cooling it, so its inner surface sits near ambient, perhaps 12 degrees. That is 20.1 kelvin below the dew point of the air touching it.

Which means the vapour does not reach the membrane as vapour. It reaches a surface that is far too cold to let it stay vapour, and it condenses there, every time, in any weather cold enough that you wanted a jacket. There is no version of this where it does not condense. The only question is how fast the resulting liquid leaves.

And that is now a different and much smaller calculation, because the thing driving vapour across the membrane is no longer your skin. It is the film. A film of liquid water at 12 degrees has a vapour pressure of 1.403 kilopascals. The saturated air outside at 10 degrees has 1.228. The difference is 0.175 kilopascals, which is 4.9 percent of the 3.55 the rating implies, so 750 grams an hour becomes 37.

Thirty seven grams an hour against seven hundred. Ninety five percent of what you produce has nowhere to go, and it is already liquid, on the inside of your jacket, and it reads exactly like sweat.

Stop raining and it improves, but not enough

Run the same sum on a dry cool day, which is instructive because it explains why the jacket seems to work sometimes. With no water on the outer face there is no evaporative cooling of the shell, so the inner face sits warmer, perhaps 18 degrees, whose saturation pressure is 2.065 kilopascals. The outside air at 10 degrees and 50 percent relative humidity carries 0.614. The difference is 1.451 kilopascals, which is 40.9 percent of nominal, so 307 grams an hour.

Better by a factor of eight, and still less than half of what the body is making. So the honest statement about a waterproof breathable shell is that it reduces the rate at which you get wet from the inside, and that is all it has ever done. The marketing number is a laboratory ceiling. The field number is a fraction of it set by how cold and how wet the outside of the garment is, which are exactly the conditions you bought it for.

The limiting case, which we do sell

We do not carry a rain jacket, so nothing above is read off one of our records and it is worth saying so plainly. What we do carry is the same mechanism with the dial turned all the way to one end.

LANON waterproof double latex coated winter work gloves are a complete vapour barrier. Their record describes a "double layer of fully dipped latex over an acrylic nylon knit shell" with "a brushed fleece lining" inside it. Fully dipped means no uncoated area anywhere, and a continuous latex film has no micropores at all, so its vapour transmission is not reduced, it is absent. Everything a hand produces stays inside.

Which makes one of its own stated drawbacks much more interesting than it looks. The record lists "Insulation weakens after long water immersion", and its performance text records that "LANON warns against soaking the gloves in cold water beyond ten minutes, or the warm insulation weakens". The maker is describing water getting in from the outside. But the arithmetic above says the liner will wet out from the inside too, because the insulation sits between a vapour source and a cold impermeable film, which is precisely the geometry that guarantees condensation in the loft. Lofted insulation works by trapping still air, and water in the loft collapses it and conducts heat away roughly twenty times faster than the air it replaced. The ten minute immersion warning and the thing that happens after an hour of shovelling are the same failure arriving from opposite directions, and the record only names one of them. Our LANON winter glove review has the certifications.

And it is worth correcting our own page on this, gently. Our guide to choosing insulated waterproof gloves for outdoor work recommends exactly this combination and never mentions that putting the insulation inside the barrier is what makes the insulation vulnerable. The recommendation is still right for cold wet work. The missing sentence is that a fully sealed insulated glove has a duty cycle rather than a rating, and the duty cycle is set by how long the loft takes to fill.

The partial barrier, and the honest remedy hiding in a pack size

G and F waterproof winter gloves with an HPT palm coating are the same idea built differently, and the difference is the one that matters here. Only the palm side is coated: the record specifies "a durable exterior latex coating" on "double-coated latex and foam palms" over "a 7-gauge acrylic terry lining". The back of the hand stays open knit. So there is an escape path for vapour, through an area neither record quantifies, and the glove is correspondingly less waterproof on that side. One object, two requirements, and the coated area is the dial between them.

The thing to notice is why the pack holds three pairs. The record says it directly: "A three-pair pack means spares for rotation as gloves get wet or worn." That is a concession that the waterproof glove gets wet, offered as the solution rather than as a problem, and it is the correct solution. It is also the glove version of every genuinely effective technique in this article: do not try to drive the water out through the barrier, take the barrier off and let the water out the open end. Its cons add the other half honestly, that "Prolonged submersion can eventually let moisture in". Our G and F winter glove review covers the rest.

The other end of the dial

PULI water resistant hiking cargo pants sit at the opposite extreme and are the most honest garment in this article, because of one line in their own drawbacks: "Water-resistant, not waterproof, so heavy rain will eventually soak through."

That sentence is the trade stated correctly. A woven fabric with a repellent finish and no membrane has no barrier to vapour at all beyond the air trapped in the weave, so its transmission rate is enormous and it will never condense on its inner face in the way a membrane does. The forfeit is that it has no hydrostatic resistance either, and in sustained rain the water simply arrives. For anything short of that it is the better garment to be moving in, and the arithmetic says why: it is operating on the whole 3.55 kilopascals rather than on 0.175 of it.

Worth noting that neither this listing nor any other apparel listing we carry publishes a figure for either property. The record admits as much, noting that "No exact fabric content or weight is stated in the listing". So the choice between repellent and waterproof in our catalogue is made on adjectives. Our PULI hiking trouser review has the pocket detail that is actually its selling point.

What works, and it is convection rather than diffusion

Here is the number that ends the argument. Air inside a jacket at 33 degrees and 95 percent relative humidity holds 33.9 grams of water per cubic metre. Replacement air at 10 degrees and saturated holds 9.40. So every cubic metre of jacket air you throw away and replace removes 24.5 grams of water, as vapour, without anything having to diffuse through anything.

Exchange ten litres a second, which is a modest draught from an unzipped front while walking, and that is 0.245 grams a second, or 882 grams an hour. Twenty four times what the membrane manages in the rain, and more than your body is producing. Opening the jacket is not a compromise on the jacket's function. It is between one and two orders of magnitude better than the jacket's function, and it works best in exactly the conditions where the membrane works worst, because cold outside air is dry in absolute terms no matter what its relative humidity says.

Which is also why forced air is such a good idea when you can have it. A solar fan bucket hat with two brim fans is a small and slightly comic illustration of the right physics: its record lists "Dual breathable mesh vents", "an internal sweatband", "quick-drying fabric" and two fans running "5 to 16 hours" after a two hour charge, on three speeds. Every one of those attacks the convective term rather than the diffusive one. The head is also the one part of the body where the garment is not sealed at the bottom, so air genuinely moves through it. Its honest limits are stated: "Fan output and battery life are maker-stated", and there is no air flow figure published anywhere, so the hat cannot be put on the chart above. Our solar fan hat review has the rest.

The second thing that works is buffering, and it is why wool keeps being recommended. Wool absorbs water vapour into the fibre itself rather than holding it as a film on the surface, and at high humidity its regain is around 30 percent of its dry mass. FALKE TK2 Explore merino wool hiking socks describe "a three-layer build engineered to wick moisture away", which is the surface transport story, and the fibre does something the record never mentions: a 60 gram pair can take up about 18 grams of water inside the wool without feeling wet or losing much of its loft. That is a buffer across a sharp burst of effort, not a path to the outside, and it is the difference between a damp sock and a wet one. Its cons are appropriately careful that "Exact fiber percentages are not listed", so the 30 percent figure is handbook wool rather than this sock. Our FALKE merino sock review has the sizing.

The honest limit, and the counterintuitive one

The limit is that none of the four numbers this page turns on is published by anybody we carry. No MVTR, no hydrostatic head, no membrane type, no coated area. So the 10,000 grams per square metre per 24 hours, the 1.8 square metres, the 700 grams an hour and the ten litres a second are all worked examples, chosen to be ordinary and labelled in the drawing. What is not an assumption is the vapour pressure table, the dew point and the arithmetic between them, and those are what produce the factor of twenty.

Then the free techniques, in order of how much the arithmetic says they are worth. Open the front and the pit zips, because convection is worth 882 grams an hour against the membrane's 37. Slow down, because the supply term is the only other large one and it scales with how hard you are working. Take the hood down when it is not raining hard, for the same convective reason, since the neck opening is the top of the chimney.

And the counterintuitive one, which falls straight out of the staircase in the drawing. Wear less under the shell, not more. Insulation between your skin and the membrane makes the membrane's inner face colder, which lowers the vapour pressure of the condensate, which reduces the only gradient that is actually moving water out. A thick fleece under a shell improves your comfort and makes the shell's job harder at the same time. The usual advice to layer up under a waterproof is correct for warmth and wrong for dryness, and nobody tells you that the two are in conflict.

Last, keep the outer face repellent. A wetted out outer face is cooled by its own evaporation and holds liquid in the surface fabric, both of which pull the shell's temperature down and shrink that 0.175 kilopascals further. Washing a shell properly and re-treating the outside is the one maintenance task that moves a number on this page, and it moves the right one.