How facades leak: rain penetration
Water does not need a hole to get into a building. It needs a gap, a film of moisture, and one of six forces to push it through — and a monsoon supplies all of them at once.
Every facade leak is the same equation: water, a gap, and a force. Remove any one and it stays dry.
A facade leak is almost never a hole. It is a hairline gap at a joint, a film of water sitting on a sill, and one of six quiet forces nudging that water the wrong way. In a Mumbai monsoon, wind-driven rain hits a tower at the energy of a thrown bucket, the wind builds a pressure difference across the skin that sucks water inward, and gravity, capillarity and surface tension finish the job. **The water does not need to be invited in; it only needs a path and a push.** This is why face-sealing fails and why a senior facade engineer never asks 'is it sealed?' but 'if water gets to this joint, what force is acting on it, and where does it go?' Name the six mechanisms and you can read any leak.
Water, a gap, a force — the anatomy of every leak
Five forces drive water through a gap; remove the force and the gap stops leaking
Water at a facade joint sits there harmlessly until a force moves it inward. There are five, and a wind-driven rainstorm runs all of them at once.
Kinetic energy — raindrops carry momentum. Wind-driven rain at 20-30 m/s arrives with enough energy to splash and drive straight through an open joint, like water through a sieve. Gravity pulls any film of water downward and inward across a sloped or flat ledge. Capillarity (capillary action) wicks water against gravity through any narrow gap below roughly 0.5 mm — the tighter the gap, the stronger the pull, which is the cruel paradox of a 'tight' joint. Surface tension lets water cling to and run along the underside of a horizontal surface, defeating the intuition that water only falls. And the big one on tall buildings: an air-pressure differential — wind pressurises the outer face and the building interior is often at lower pressure, so air rushes inward through any gap and carries the water with it.
The lesson is structural, not incidental: the cure for each mechanism is different, so 'just seal it' addresses none of them reliably. You break the kinetic path with a baffle, beat capillarity with a wide gap, beat gravity with an upstand, and beat the pressure differential by equalising it — which is the whole of the next lesson.
Water + a gap + a force = a leak. You rarely control the water or close every gap. So a facade engineer's real job is removing the FORCE.
The six classic rain-penetration mechanisms — every leak is one of these
Classical building science names six mechanisms by which rain penetrates a joint, and every real leak is one (or several) of them. Kinetic energy / momentum: the drop is simply thrown through the opening. Gravity flow: water runs down an inward-sloping surface and drips inside. Surface tension: water clings to and tracks along an underside, then drops off inboard of the seal. Capillary action: water wicks through a fine gap by surface forces, even upward. Air-pressure differential: a pressure drop across the joint pumps water through with the in-flowing air — the dominant mechanism on tall buildings in wind. Gaps and momentum of moving air combine so that even a sheltered joint leaks when the wind finds a path.
The defence is geometry, not just sealant. Slope every ledge outward to defeat gravity. Add a drip groove on every overhang to break surface tension. Leave a gap wider than capillary range (an air gap, not a tight crack) to kill wicking. Set a baffle or upstand in the path to stop kinetic splash. And — the master move — equalise the pressure across the rain barrier so the air-pressure mechanism has nothing to pump with. A well-detailed joint quietly disarms five of the six with shape alone.
A single perfect seal is a bet you always lose; defence-in-depth is the only honest strategy
Face-sealing — relying on a continuous outer bead of sealant to be the one and only line of defence — is the most failure-prone facade strategy, and the reason is physics plus time. Every sealant degrades under UV and movement. Every material expands, contracts and fatigues the joint. A facade has kilometres of joint, and the probability that all of it stays perfect for forty years is zero. The day one bead cracks, water gets behind the seal with nowhere to drain and no second line to catch it, and it travels — often emerging metres away and a floor below, which is why facade leaks are so hard to trace.
Modern weatherproofing flips the assumption. It assumes water will get past the outer face and designs a drained, ventilated, pressure-equalised cavity to catch that water and shed it back out — the rainscreen principle. The outer skin becomes a screen that breaks most of the rain; a second concealed line does the real waterproofing; and the cavity between drains whatever gets through. Defence in depth beats a single perfect seal every time. In the Indian monsoon, where a facade can take 2,000-3,000 mm of driving rain in three months, this is not a refinement — it is the difference between a dry building and a dispute.
Trace any leak UPWARD: water always enters above where it shows. The stain is the exit, not the entry. That is facade forensics.
Your joints are where leaks live, so design the joint, not just the panel. A flush, minimal, 'invisible' joint usually means a face-sealed joint — beautiful and fragile. Favour expressed, drained, shadow-gap joints that give water a path out; they read as a deliberate line, not a failure waiting to happen. When you draw a sill, a coping or a reveal, ask one question: if a film of water sits here, which way does my geometry send it? Outward is a detail; inward is a callback in three monsoons.
Diagnose every critical joint against all six mechanisms before you specify a sealant. Slope ledges outward (a minimum fall, never flat or back-pitched), put a drip groove under every overhang, keep drainage gaps above capillary range, and locate the true waterproof line _behind_ a screen wherever the architecture allows. Treat the air-pressure differential as the dominant driver on anything tall, and design to neutralise it (Lesson 5.2) rather than out-sealing it. Specify static AND dynamic water testing so the pressure mechanism is actually proven, not assumed.
On site, leaks are forensics. Water enters _above_ and _windward_ of where the stain shows, then tracks down and across hidden surfaces, so never trust the symptom's location. Learn to spot the four geometry cures and check they survived installation: is the sill still sloping outward (not back-pitched by a settling bracket)? Is the drip groove clean, not bridged by squeezed-out sealant? Is the drainage gap open, not blocked by debris or mortar? Most field leaks are a correct detail defeated by one careless centimetre on site.
ASTM E331 / E547
Static water penetration test
Standard lab method spraying water at a fixed pressure to find penetration. Honest about a limit: static pressure under-represents the gusting, fluctuating pressure of a real monsoon — which is why dynamic testing exists.
CWCT Standard (UK)
Facade water-tightness performance
The de-facto international benchmark for water-penetration test pressures and pass criteria; widely written into Indian premium-project specs. Sets the bar, but the engineer must still pick a pressure suited to the local wind.
NBC 2016 / IS 875 (Part 3): 2015
Wind that drives the rain (India)
IS 875-3 gives the design wind speeds and pressures that, with rainfall, set the driving-rain load. It governs the wind, not the water-tightness pass mark — you still inherit that from a facade standard.
“If a joint is sealed tight and there are no visible gaps, water cannot get in — a good silicone bead is waterproofing.”
A 'tight' gap is the worst kind for capillary action, which wicks water _harder_ the narrower the gap, even upward against gravity. And a single outer bead is a face-seal: when it inevitably cracks, water gets behind it with nowhere to drain. Real weatherproofing assumes water gets past the outer face and gives it a drained path out — geometry and drainage do most of the work, sealant only some of it.
Worked example — does this joint leak by capillary action?
Capillarity is the most counter-intuitive force: a tighter joint wicks water harder. Let's put a number on it for a typical aluminium-to-glass gap and see where the danger zone is.
A calculator and the capillary-rise relation.
GIVEN a narrow water-filled gap between two wettable surfaces: capillary rise h = 2 * T / (rho * g * d) surface tension T = 0.073 N/m (water, ~20 C) density rho = 1000 kg/m3 gravity g = 9.81 m/s2 gap width d = the joint gap (m) Question: how high can water climb in a 0.2 mm joint versus a 6 mm drained gap?
- 1Take the tight joint first, d = 0.2 mm = 0.0002 m. Compute the numerator: 2 * 0.073 = 0.146. Compute the denominator: 1000 * 9.81 * 0.0002 = 1.962.
- 2Divide: h = 0.146 / 1.962 = 0.074 m = 74 mm. Water can wick upward 74 mm through that hairline gap, against gravity, with no wind at all. A 'tight' joint is a capillary pump.
- 3Now the drained gap, d = 6 mm = 0.006 m. Denominator: 1000 * 9.81 * 0.006 = 58.86. Then h = 0.146 / 58.86 = 0.0025 m = 2.5 mm. Effectively nothing.
- 4Read the result: widening the gap from 0.2 mm to 6 mm cut capillary rise from 74 mm to 2.5 mm — a factor of ~30. The capillary force falls as 1/d, so an air gap defeats wicking that a tight seal invites.
- 5Draw the design rule: never rely on a tight joint to keep water out, because tightness strengthens capillarity. Either fully seal AND drain, or — better — leave a deliberate drained air gap wider than the capillary range. This is the seed of the pressure-equalised rainscreen in Lesson 5.2.
- 6Sanity-check the units: N/m divided by (kg/m3 * m/s2 * m) = (kg/s2) / (kg/m2/s2) = m. Units resolve to metres. Good.
You’ll walk away with
A number you can defend: capillary rise of ~74 mm in a 0.2 mm joint versus ~2.5 mm in a 6 mm drained gap — proof that for capillarity, a wide air gap beats a tight seal, and the quantitative basis for choosing drained joints.
Two quick reads to anchor the forces in the real world.
- 01Find a stained or leaking facade near you. Trace the stain upward and toward the prevailing monsoon wind — name which of the six mechanisms most likely drove the water in, given the geometry you can see.
- 02Look at a window sill or a coping nearby. Is it sloping outward with a drip groove underneath, or flat / back-pitched? Predict which one will leak first in a heavy monsoon, and why.
Every facade leak is water, a gap and one of six forces — kinetic energy, gravity, surface tension, capillarity, and above all the air-pressure differential. You rarely stop the water or close every gap, so the engineer's real lever is removing the force, mostly with geometry: slope, drip, air gap, baffle, and pressure equalisation. Face-sealing fights all six with one fragile bead and loses; defence-in-depth disarms them by design.
Water needs a gap, a film and a force. Five forces (kinetic, gravity, capillarity, surface tension, air-pressure differential) drive the six rain-penetration mechanisms. Capillarity wicks harder in tighter gaps — so an air gap beats a tight seal. Face-sealing relies on one perfect bead and fails; modern design assumes water gets past the face and drains it. Trace any leak upward and windward.
What are the mechanisms of rain penetration through a facade?
Classical building science names six: kinetic energy / momentum (the drop is thrown through the gap), gravity flow (water runs down an inward-sloping surface), surface tension (water clings to and tracks along an underside), capillary action (water wicks through a fine gap, even upward), and the air-pressure differential (a pressure drop across the joint pumps water through with in-flowing air) — plus the combined effect of gaps and moving air. The pressure differential dominates on tall buildings in wind.
Why does a tighter joint sometimes leak more, not less?
Because of capillary action. The capillary rise of water in a gap is inversely proportional to the gap width — the narrower the gap, the higher water can wick, even upward against gravity. A 0.2 mm joint can draw water up about 74 mm; a 6 mm drained gap, only about 2.5 mm. So a 'tight' joint can act as a capillary pump, while a wider deliberate air gap defeats wicking. This is why drained, ventilated joints outperform tightly sealed ones.
Why does face-sealing a facade fail?
Face-sealing relies on a single continuous outer bead of sealant being the one and only barrier. But sealant degrades under UV and movement, materials fatigue, and a facade has kilometres of joint — so the probability all of it stays perfect for forty years is zero. When one bead cracks, water gets behind it with nowhere to drain and travels far from the entry point. Modern design assumes water gets past the face and provides a drained, pressure-equalised path out.
Peer-reviewed journals & authoritative standards
- 01Ventilated facade system: A review (water control, drainage and the rainscreen principle). — ScienceDirect (Elsevier), 2025.
- 02Squadroni, F., De Michele, G., Mazzucchelli, E.S. et al. Analysis of condensation and ventilation phenomena for double skin façade units. — Journal of Building Physics (SAGE), 2022.
- 03IEA EBC Annex 43/44. Double Skin Facades: A Literature Review. — International Energy Agency (IEA-EBC), 2008.
If the air-pressure differential is the force that pumps water through tall facades, the smartest cure is not to seal harder but to remove the pressure difference itself. That single idea — pressure equalisation — is the most powerful move in weatherproofing, and it is next.