As an Amazon Associate, Akermin earns from qualifying purchases. Product links on this page are affiliate links.
Read a humidity record by converting it to dew point first, because relative humidity is a ratio whose denominator changes with temperature and therefore moves on its own without any water being added or removed. A trace that climbs through the night in an unheated room usually contains exactly the same amount of vapour at the peak as it did at the trough, and reporting that climb as a damp problem is the single most common error in this work. The questions a logged trace can honestly settle are whether a surface was cold enough for condensation, whether the pattern follows occupancy or weather, and how long a condition persisted. Which kind of damp you are looking at is not one of them.
Relative humidity is a ratio, and the denominator moves
The percentage on the display is the amount of water vapour actually present divided by the maximum the air could hold at its present temperature. Only the numerator describes the moisture. The denominator, the saturation capacity, rises steeply with temperature, roughly doubling for every eleven degrees C or so across the range buildings live in. That single fact explains most of the shape of a typical indoor trace.
Work an example. Air at 20 degrees C can hold vapour up to a pressure of about 23.4 hectopascals. At 65 percent relative humidity the vapour actually present exerts about 15.2 hectopascals. Cool that same air to 14 degrees C without adding or removing a thing and the capacity drops to about 16.0 hectopascals, so the ratio becomes about 95 percent. The record now shows a thirty point rise and the room contains precisely as much water as before. The explanation of why humidity changes through the day walks through the daily version of this effect.
Why the cold wall reads wetter than the room
The same arithmetic applies in space as well as in time. A thin layer of air in contact with a surface takes on roughly the temperature of that surface, so an external wall at 14 degrees C in a room whose bulk air is at 20 produces a local layer sitting near 95 percent while the instrument on the shelf reports 65. Mould and surface condensation respond to the local condition, not to the room average, which is why a room can read entirely acceptably and still grow mould in a corner.
The variable that stays put through all of this is dew point, the temperature to which the air would have to be cooled before it reached saturation. Air at 20 degrees C and 65 percent has a dew point of about 13.2 degrees C, and that figure is unchanged whether you measure it at the ceiling, at the window reveal or inside a cupboard, as long as the air is the same air. Any surface colder than the dew point will collect liquid water. Any surface above it will not. That makes a surface temperature measurement, taken alongside the logged air condition, the single most useful addition to a damp investigation.
The same air against four surface temperatures
| Surface temperature | Local relative humidity at that surface | Condensation | What it means for the surface |
|---|---|---|---|
| 20 degrees C, the same as the room air | 65 percent | No | Comfortable, no surface risk |
| 18 degrees C | About 74 percent | No | Below the commonly cited threshold for surface mould risk |
| 16 degrees C | About 84 percent | No | Above that threshold, a sustained risk even with no visible water |
| 14 degrees C | About 95 percent | No, but marginal | Effectively saturated at the surface |
| 13.2 degrees C, the dew point | 100 percent | Yes, from here downwards | Liquid water forms on the surface |
Notice that the mould risk threshold is crossed long before any water appears. A surface sitting steadily at 16 degrees C in this room never drips, never looks wet, and is nonetheless held at a condition where growth is supported. That is why reports that only record visible condensation miss the majority of cases, and why the pair of numbers worth logging is air dew point alongside surface temperature rather than room humidity alone. The guidance on measuring relative humidity correctly covers the probe placement side of getting those numbers.
What a record can and cannot prove about a damp source
A logged trace can establish four things with confidence. Whether the air dew point ever exceeded a measured surface temperature, and for how long. Whether the pattern repeats daily, which points at occupancy or heating control, or follows the weather, which points at the external envelope. Whether moisture is being added to the space, which shows as a dew point rising independently of temperature. And whether conditions persisted long enough to matter, since almost every damage mechanism is about duration rather than peaks.
It cannot establish where water is entering, because every mechanism ends with damp air. It cannot measure moisture held inside plaster, timber or masonry, which needs an instrument that interrogates the material rather than the air. And it cannot distinguish a genuine step change from an instrument that was moved, which is why the deployment notes matter as much as the file. Pairing an air record with a carbon dioxide channel helps here, because carbon dioxide tracks occupancy closely and occupancy is the main indoor moisture source. The AccUnni carbon dioxide and humidity monitor publishes a carbon dioxide range of 400 to 5000 ppm with a stated accuracy of plus or minus 5 percent plus 100 ppm, humidity across 1 to 99 percent at plus or minus 3 percent, and a manual calibration procedure against outdoor air with an automatic baseline routine over a rolling week.
Differences smaller than the accuracy band are not findings
Every comparison in a damp report is a comparison between two uncertain numbers, and the uncertainties add. The HOBO USB temperature and humidity logger states an accuracy of plus or minus 3.5 percent relative humidity across the portion of its scale from 25 to 85 percent, with a measuring range running from 15 to 95 percent, an interval selectable from one second to eighteen hours, and an enclosure rated for limited ingress protection only. Set two such instruments side by side in an identical condition and a gap of several points between them is entirely expected behaviour.
| Comparison | Stated accuracy of each | Difference that means nothing | Difference worth investigating |
|---|---|---|---|
| Two loggers each at plus or minus 3.5 percent | 3.5 and 3.5 | Up to about 5 points | More than about 7 points |
| A logger at 3.5 percent against a monitor at 3 percent | 3.5 and 3.0 | Up to about 4.6 points | More than about 6.5 points |
| Two readings from the same instrument at different times | Drift and repeatability only | Up to about 1 to 2 points | More than about 3 points |
That third row assumes an instrument whose condition you have recently confirmed, because a long term drift figure is Not stated in the listing bullets for any of the four units here. It is the useful row all the same, and it is why a single instrument moved between locations often beats two instruments left in place. Most of the error in a humidity measurement is a fixed offset rather than random noise, so the same unit comparing two rooms cancels the offset out, while two different units comparing the same room do not. Settling time has to be respected if you use that approach. The Protmex YT606 is unusually candid about this, stating four measurement modes covering ambient temperature, relative humidity, wet bulb temperature and dew point, and instructing the user to wait ten minutes after changing environments before taking a reading.
Writing up a humidity record without overreaching
Separate the report into observation and inference, and keep the boundary visible. Observation covers the instrument used, its stated accuracy, where exactly it sat, how high, how far from the nearest external wall, the interval, the start and end of the deployment, and the measured trace itself expressed as both relative humidity and dew point. Inference covers what you think the pattern means, and every sentence of it should be traceable to something in the observation section.
Three habits keep a report defensible. State surface temperatures wherever you claim a condensation risk, since the claim is meaningless without them. State the accuracy band next to any comparison, so that a reader can see which differences you treat as real. And state plainly what the record does not show, including the periods when the instrument was not deployed and the questions the data cannot answer. For choosing an instrument to run this kind of work, the overview of temperature and humidity monitors covers the deployment options, and the definition of relative humidity is worth putting in front of any client who reads the percentage as a direct measure of wetness.
Frequently Asked Questions
Why does humidity rise overnight even in an empty house?
Because the air cools and the ratio changes. Relative humidity compares the water vapour present against the maximum that air could hold at its current temperature, and that maximum falls steeply as the air cools. The same unchanged quantity of vapour therefore produces a higher percentage at four in the morning than at four in the afternoon. Nothing has got wetter. Only the denominator has moved.
Which number should I actually record, humidity or dew point?
Record both and reason with dew point. Dew point is a direct measure of how much vapour is present and it does not move when the air temperature does, so it is the number that reveals whether moisture is genuinely being added to or removed from a space. Relative humidity is still the figure most guidance is written against and the figure that matters at a surface, so you need it too.
Can a logger prove that a wall has rising damp?
No. A logger measures air, and every form of damp produces damp air eventually, so the trace cannot distinguish a leak from groundwater from condensation from a flooded appliance. What a record can do is rule condensation in or out by comparing the air dew point against measured surface temperatures, and show whether the pattern follows rainfall, occupancy or heating. Identifying a source needs fabric measurements and inspection alongside the trace.
How far apart do two readings have to be before the difference is real?
Combine the two accuracy figures rather than using one of them. Two instruments each stated at plus or minus 3 percent produce a combined uncertainty of roughly 4 percent when the errors are treated as independent, and closer to 6 percent if you assume the worst case that they err in opposite directions. A four point gap between two such instruments is therefore not evidence of anything about the building.
Does it matter where in the room the logger sits?
Enormously, and it is the detail most often left out of a report. A unit on an internal shelf, a unit taped to an external wall and a unit in a corner behind furniture will produce three different traces from the same room, because the air near a cold surface is colder and therefore reads a higher percentage. Position, height and distance from the nearest wall should be written down at deployment, because they cannot be reconstructed afterwards.
How long does a hygrometer need to settle after being moved?
Longer than most people allow. Moving an instrument from a warm car into a cold room, or from a dry hallway into a damp cellar, starts a process in which both the body of the instrument and the sensing element have to come to the new condition, and the element cannot read correctly until the air immediately around it has stopped being influenced by the instrument own temperature. Ten minutes is a reasonable minimum and longer is better for a large temperature step.