In brief
Often, the rainwater downpipe has no leak at all. The air in the shaft condenses on a wall chilled by the rain, and replacing a joint changes nothing.
- The threshold is the dew point. A rainwater (RW) pipe condenses as soon as its surface is colder than the dew point of the surrounding air.
- A low dew point does not protect. Winter rain at 2 °C leaves a bare pipe at about 3 to 5 °C (calculation assumption): below the 5.9 °C dew point of air at 18 °C and 45% humidity.
- It drips, it does not flood. About 3 to 10 g of water per hour per metre of bare 110 mm pipe in our cases: the volume collected is a criterion for telling condensation from a leak.
- Confinement does the damage. In a shaft, suspended ceiling, boxing or crawl space, the air is not renewed: condensed water does not dry and builds up.
- The remedy is well known. Continuous closed-cell insulation, supplemented where possible by ventilating the space. In our cases, 11 mm at most are enough thermally: a discontinuity can still let condensation through.
A stain on the suspended ceiling, drops running down a rainwater downpipe, and the verdict comes: a leak. Often, the pipe is watertight. This article is written for building professionals, managing agents and owners who want to settle the question with figures: what can be calculated, what must be measured, and what remains an order of magnitude.
A cold pipe in warm air
Condensation appears when the surface temperature of the pipe falls below the dew point of the surrounding air.
The dew point is the temperature at which the water vapour in the air turns to liquid on contact with a wall. Three quantities are therefore enough to predict the phenomenon:
- The temperature of the flowing water, that is, of the rain or melted snow.
- The surface temperature of the pipe. A bare PVC, zinc or cast-iron pipe stays very close to the water. We assume here 1 to 3 °C above it while water flows: this is a calculation assumption, unsourced, to be replaced by measurement (rule 3).
- The dew point of the room air, set by its temperature and relative humidity.
The phenomenon is intermittent: it starts with the rain, lasts as long as cold water flows, then stops when the pipe warms up.
Rain is colder than the air
Rain reaches the ground a few degrees below the outdoor air temperature: no difference when the air is saturated, about 3 °C less at 80% humidity.
A common estimate is the so-called wet-bulb temperature: the drops cool by evaporating as they fall. The table calculates it with Stull’s formula (2011). It is an order of magnitude, not a measurement: the real temperature of the rain depends on the height of fall and on the layer of air it crosses.
| Outdoor air | Rain, outdoor air at 80% humidity | Rain, outdoor air at 100% humidity |
|---|---|---|
| 2 °C | 0 °C | 2 °C |
| 5 °C | 3 °C | 5 °C |
| 10 °C | 8 °C | 10 °C |
| 15 °C | 12.5 °C | 15 °C |
| 20 °C | 17.5 °C | 20 °C |
| 25 °C | 22.5 °C | 25 °C |
Values rounded to the half degree. A simple rule, which stays within 0.7 °C of Stull’s formula between 2 and 25 °C outdoor air, for a humidity H between 80 and 100% (no assurance below that, as the gap grows when the air is drier):
The coldest case is not rain but melting snow or hail on the roof: the water then enters the downpipe at about 0 °C, sometimes for several hours.
The dew point: the threshold to calculate
The dew point is calculated from the temperature and relative humidity of the room air, using the Magnus formula.
With T the temperature (°C) and H the relative humidity (%), and Sonntag’s constants (1990), valid from −45 to 60 °C:
Dew point in °C according to the room air:
| Room temperature | 30% | 40% | 50% | 60% | 70% | 80% |
|---|---|---|---|---|---|---|
| 10 °C | −6.8 | −3.0 | 0.0 | 2.6 | 4.8 | 6.7 |
| 12 °C | −5.1 | −1.2 | 1.9 | 4.5 | 6.7 | 8.7 |
| 15 °C | −2.5 | 1.5 | 4.7 | 7.3 | 9.6 | 11.6 |
| 18 °C | 0.1 | 4.2 | 7.4 | 10.1 | 12.4 | 14.5 |
| 20 °C | 1.9 | 6.0 | 9.3 | 12.0 | 14.4 | 16.4 |
| 25 °C | 6.2 | 10.5 | 13.9 | 16.7 | 19.1 | 21.3 |
Simple rule above 50% humidity: the dew point drops by about 1 °C each time the humidity loses 5 points.
A low dew point does not protect
Dry air has a low dew point, but winter rainwater goes even lower.
The table gives the room humidity above which a pipe condenses, according to its surface temperature:
| Room temperature | Surface at 3 °C | Surface at 5 °C | Surface at 8 °C | Surface at 10 °C |
|---|---|---|---|---|
| 12 °C | 54% | 62% | 77% | 88% |
| 15 °C | 45% | 51% | 63% | 72% |
| 18 °C | 37% | 42% | 52% | 60% |
| 20 °C | 32% | 37% | 46% | 53% |
In a shaft at 18 °C, a pipe at 5 °C therefore condenses from 42% humidity, a common level in a heated building. The warmer the room, the lower the threshold. Confinement adds four aggravating effects:
- No drying. Without air renewal, condensed water does not evaporate between two rainfalls and builds up.
- Rising humidity. A small closed volume quickly fills with vapour: damp ground in a crawl space, a neighbouring wet room, water that has already condensed. The dew point rises.
- Air chilled around the pipe. Air at 18 °C and 40% humidity, cooled to 8 °C right next to the pipe, reaches about 77% humidity. That is close to the 80% surface relative humidity threshold, customary for mould risk, without a single drop being visible.
- Cold contact points. Clamps, metal brackets and floor penetrations take on the temperature of the pipe and condense in turn.
The risk also exists in summer, by another route: warm, humid air at 25 °C and 60% has a dew point of 16.7 °C. A thunderstorm rain at 12 °C (a case to check on site) leaves a bare pipe at about 13 to 15 °C, so below it.
Five typical situations, calculated
Of five common situations, four condense: only a dry, heated room escapes.
Common assumption: bare 110 mm pipe whose surface is 2 °C above the water flowing in it. The dew point comes from the Magnus formula; the water produced and the minimum thickness are explained in the following sections.
| Situation | Air (T / RH) | Dew point | Bare pipe surface | Margin | Condensation | Water produced (g/h per m) | Minimum thickness |
|---|---|---|---|---|---|---|---|
| Heated building shaft, winter rain at 4 °C | 18 °C / 50% | 7.4 °C | 6 °C | −1.4 °C | Yes | 2.9 | 1.4 mm |
| Same shaft, melting snow at 0 °C | 18 °C / 50% | 7.4 °C | 2 °C | −5.4 °C | Yes | 9.9 | 3.1 mm |
| Damp crawl space, rain at 5 °C | 12 °C / 80% | 8.7 °C | 7 °C | −1.7 °C | Yes | 3.6 | 4.7 mm |
| Dwelling ceiling void, summer storm at 12 °C | 25 °C / 60% | 16.7 °C | 14 °C | −2.7 °C | Yes | 9.3 | 2.5 mm |
| Dry heated room, winter rain at 5 °C | 20 °C / 30% | 1.9 °C | 7 °C | +5.1 °C | No | 0 | none |
The crawl space is not the coldest case, but it needs the thickest insulation of the five: the air there is almost saturated, and it is humidity, far more than temperature, that sets the thickness. In the dry room the surface stays 5.1 °C above the dew point: no condensation, and a surface humidity of about 70%.
Leak or condensation: the signs and the damage
Condensation is often mistaken for a leak, when the pipe is watertight.
It differs in three ways: it wets the whole exposed length of the pipe, it appears during or just after a cold rain, and drops form on the underside rather than at a joint. The two can coexist: condensation does not rule out a failing joint, which the test in rule 6 helps to separate.
- Runoff and stains on suspended ceilings, plasterboard boxing and the foot of shafts.
- Corrosion of clamps, brackets, fixings and metal pipework.
- Mould on plaster, timber, dust and insulation, with odours in neighbouring rooms.
- Soaked insulation: wet mineral wool loses its insulating power and keeps the dampness going.
- Electrical risk if cable trays or junction boxes run beneath the pipe.
How much water does condensation produce?
Condensation on a bare 110 mm pipe produces about 3 to 10 g of water per hour per metre in the cases above: enough to drip, not enough to flood.
The flow can be bounded by a simple calculation: vapour reaches the cold surface in proportion to the difference in absolute humidity between the air and air saturated at the surface temperature.
g is the condensate flow (kg/s), S the pipe surface, w the humidity ratio (kg of water per kg of dry air), hc the convective exchange coefficient and cp the specific heat of humid air (simplified Lewis analogy). Assumptions: hc = 4 W/(m²·K) in still air (an unsourced order of magnitude, higher if the air moves), saturated surface, latent heat neglected.
- A bound, not a measurement. Ten metres of bare pipe in the first case condense about 90 g in 3 hours of cold rain, and about 300 g in the melting-snow case.
- Water does not stay where it forms. It runs along the pipe and gathers at clamps, bends and floor penetrations, where it drips: condensation spread along the whole length gives one localised drip, which imitates a leaking joint.
- Volume is a test. If, during a rain, markedly more than this bound is collected at the foot of the shaft, condensation does not explain everything: look also for a joint, a crack or a roof connection. Without that comparison, a wet pipe proves neither.
- The bound assumes a bare pipe. A single uninsulated section in an otherwise insulated shaft is enough to produce the drip.
What works: insulate, seal, ventilate
The remedy is to stop humid air from touching a cold surface.
Replacing joints or redoing the watertightness of the pipe changes nothing when the water does not come from inside.
- Insulate with a closed-cell material. An elastomeric foam sleeve offers high resistance to vapour diffusion. Its outer surface stays above the dew point if the thickness is sufficient: the calculation tells.
- Seal every joint. Glued sleeves, bends and junctions covered, no open cut. A single gaping joint is enough to let vapour in.
- Avoid mineral wool without a vapour barrier. Vapour passes through it, condenses on the pipe and soaks the insulation, which then hides the problem instead of solving it.
- Deal with the singular points. Clamps with insulated supports, floor penetrations, and above all the roof outlet, which is the coldest point of the downpipe.
- Ventilate the confined space. A low grille and a high grille in the shaft allow drying between two rainfalls.
- Remove the sources of humidity. Bare ground in a crawl space, a poorly ventilated wet room, a neighbouring leak.
- Provide an access hatch. A structure that cannot be inspected deteriorates without a witness.
What insulation thickness?
The calculation looks for the thickness at which the outer surface of the insulation stays at or above the dew point. In steady state, around a 110 mm pipe, with an elastomeric foam of conductivity 0.036 W/(m·K) and a surface exchange coefficient of 8 W/(m²·K) (customary values, to be replaced by the manufacturer’s), we get:
| Room air | Water at 0 °C | Water at 3 °C | Water at 5 °C | Water at 8 °C |
|---|---|---|---|---|
| 18 °C, 40% | 1.3 mm | 0.4 mm | none | none |
| 18 °C, 50% | 3.1 mm | 1.8 mm | 1.0 mm | none |
| 18 °C, 60% | 5.5 mm | 3.9 mm | 2.8 mm | 1.2 mm |
| 20 °C, 50% | 3.8 mm | 2.6 mm | 1.8 mm | 0.5 mm |
| 12 °C, 80% | 10.7 mm | 7.2 mm | 4.7 mm | 0.9 mm |
Minimum thickness for the surface to stay at the dew point. The pipe wall and the inner water film are neglected, so the calculation is on the safe side. Four lessons:
- The thickness is small. From under 1 mm to 11 mm in this table.
- Humidity decides. At 12 °C and 80%, more insulation is needed than at 18 °C and 40%, even though the air is colder.
- Discontinuity matters more than thickness. What can still let condensation through is a bare clamp, an open joint or a forgotten section, which recreates a surface at the water temperature.
- A one-degree margin above the dew point adds 0.6 to 1.3 mm in the 18 °C cases (50 and 60% humidity).
The final thickness is calculated according to standard EN ISO 12241, with the manufacturer’s values, for the warmest, most humid room crossed; the next catalogue thickness up is chosen. Open or discontinuous insulation can be worse than none: it holds water against the pipe and hides it.
Six rules for a defensible diagnosis
A condensation diagnosis comes down to one numerical comparison: surface temperature against dew point.
Without those two numbers, “it is condensation” remains an opinion.
- Measure the air of the confined space, not that of the room. Temperature and relative humidity are taken inside the shaft or plenum, with a thermo-hygrometer.
- Calculate the dew point. With the Magnus formula or the table above, and write it into the report.
- Measure the pipe surface during the rain. Preferably with a contact thermometer. In dry weather the pipe is at room temperature and the measurement proves nothing.
- Estimate the worst case. Cold winter rain, melting snow at 0 °C, or a summer storm on a warm, humid room.
- Keep a margin. Even with no visible drop, a surface less than about 3 °C above the dew point is above 80% relative humidity, the customary threshold for mould risk (the criterion associated with the ISO 13788 method). A wall next to the pipe may therefore already be affected.
- Rule out a leak with a test. Fill the downpipe in dry weather: a leak runs, condensation does not. Caution: mains water is itself cool, and if its temperature is below the dew point of the shaft, the test creates condensation. Record the water temperature and compare it with the dew point before concluding, then compare the volume collected with the bound given in the section “How much water does condensation produce?”.
Measuring without being misled
- Contact probe or fine thermocouple, pressed against the pipe and covered with a small piece of insulation, so that the ambient air does not distort the reading.
- Infrared thermometer: usable on PVC or painted surfaces (high emissivity), misleading on zinc, aluminium or bright metal (low emissivity, wrong temperature displayed). On bright metal, stick on a matt patch or tape and measure on that.
- Log continuously throughout the rain, with the temperature and humidity of the shaft air: the curve showing the surface dropping below the dew point when the rain starts is the proof.
- Measure at the coldest point (roof outlet, first section below the gully) and at the point where the water drips.
- Record three items: the time the rain started, the quantity of water collected and the length of pipe involved. They allow the volume check.
Worked example. A shaft at 18 °C and 50% humidity has a dew point of 7.4 °C. Outside it is 5 °C and 90% humidity: the rain arrives at about 4 °C. The bare pipe settles at about 5 to 7 °C (calculation assumption), below the dew point: it condenses.
In doubt between a leak and condensation?
We are an independent building expertise firm: we measure the air in the shaft and the pipe surface during the rain, we calculate the dew point and we write a written report. We sell neither works nor treatment. Free quote within 24 hours.
Also on this site
- Leaking drain stack in a shared building: who pays? ;
- Thermal bridge condensation in Nice: temperature factor calculation ;
- “80% moisture in a wall”: a figure that does not exist, on reading moisture meters ;
- Socked drain in clay soil: should you avoid it?
Glossary
- Dew point
- temperature at which the water vapour in the air turns to liquid on contact with a wall. It depends on the temperature and relative humidity of the air.
- Relative humidity
- amount of vapour in the air, as a percentage of the maximum possible at that temperature.
- Wet-bulb temperature
- temperature reached by a wet surface evaporating into the air. It is often used to estimate the temperature of rain at ground level, with the reservations given above.
- Pipe insulation
- thermal insulation of a pipe. Against condensation it must be continuous and vapour-tight.
- Vapour barrier
- layer that resists the passage of water vapour through the insulation to the cold surface.
- Humidity ratio (w)
- mass of water contained in the air, in kg per kg of dry air. It is used to calculate how much vapour reaches a cold surface.
- Emissivity
- ability of a surface to emit infrared radiation. An infrared thermometer assumes it is high; on bright metal it is low and the reading is wrong.
- RW
- rainwater: the network that drains water from roofs and terraces.
Frequently asked questions
How do you tell a leak from condensation on a rainwater downpipe?
Condensation wets the whole exposed length of the pipe, on the underside, during or just after a cold rain. A leak is localised at a joint or crack and also runs when the pipe is filled in dry weather. Comparing the surface temperature with the dew point settles it; the two can coexist.
At what humidity does a rainwater pipe condense?
It depends on the room temperature and the pipe temperature. In a room at 18 °C, a pipe at 5 °C condenses from 42% humidity. At 20 °C the threshold drops to 37%.
Should an internal rainwater downpipe be insulated?
Yes as soon as it crosses a heated or humid volume, and even more so in a closed shaft or above a suspended ceiling. Without insulation it condenses whenever its surface falls below the dew point, typically at each cold rain.
What thickness of insulation prevents condensation?
It is calculated from the water temperature, the room temperature and its humidity. For a 110 mm pipe, the calculation gives a few millimetres up to 11 mm at most for the cases in the table (elastomeric foam at 0.036 W/(m·K), surface exchange at 8 W/(m²·K), to be replaced by the manufacturer’s values). The next catalogue thickness up is chosen and continuity is looked after, which matters more than thickness.
How much water does condensation on a rainwater downpipe produce?
For a bare 110 mm pipe, about 3 to 10 g per hour per metre in the calculated cases, that is a few tens to a few hundred grams for 10 m during a 3-hour rain: enough to drip. If the volume collected clearly exceeds this bound, it is not condensation alone.
Does the risk exist on the Mediterranean coast, where it rarely freezes?
Yes, less often but it does not disappear. The threshold is the pipe surface, not frost: in a home at 20 °C and 60% humidity the dew point is 12.0 °C, and a pipe below that temperature condenses. Rain at around ten degrees leaves a bare pipe at about 10 to 12 °C (calculation assumption). Autumn storms add a second case, with warm, very humid indoor air.
References
Titles and editions checked against catalogue entries (AFNOR, ISO, CEN) and bibliographic records on 16 November 2026. The full text of the standards and of the two papers was not re-read for this article: they are used here for their constants and methods, without verbatim quotation.
- NF EN ISO 13788 (EN ISO 13788:2012), Hygrothermal performance of building components and building elements: internal surface temperature to avoid critical surface humidity and interstitial condensation, calculation methods (French edition title: Performance hygrothermique des composants et parois de bâtiments).
- EN ISO 12241:2022, Thermal insulation for building equipment and industrial installations: calculation rules. According to its published table of contents, it includes a clause on preventing surface condensation.
- Sonntag D. (1990), Important new values of the physical constants of 1986, vapour pressure formulations based on the ITS-90, and psychrometer formulae, Zeitschrift für Meteorologie, 40(5), 340-344. Constants of the Magnus formula used here: 17.62 and 243.12 °C.
- Stull R. (2011), Wet-bulb temperature from relative humidity and air temperature, Journal of Applied Meteorology and Climatology, 50(11), 2267-2269.
Unsourced values: temperature gap between the water and the surface of a bare pipe (1 to 3 °C), thunderstorm rain temperatures, conductivity of the elastomeric foam (0.036 W/(m·K)), surface exchange coefficient (8 W/(m²·K)), convection coefficient in still air (4 W/(m²·K)). These are orders of magnitude to be replaced by measurements or by the manufacturer’s values; pipe wall and inner water film neglected. The tables are calculated with the Sonntag (Magnus) and Stull formulas and were recalculated on 16 November 2026.
Survey carried out by Frédéric Faucouneau, independent building expert, Antibes, Alpes-Maritimes (06), Var (83), Bouches-du-Rhône (13). Last updated: 16 November 2026. Find out more
Last updated 16/11/2026 - Cassini Expertise, independent building expert, departments 06, 83 and 13.