Three degrees above dew point is only the start. What the window really rules out, how to measure it on site, and why the coldest hour decides the schedule.
Condensation on steel is invisible for the first few minutes. The surface simply looks a little duller, and on a blast-cleaned plate in a shaded corner it is easy to walk past. The coating then goes on over a film of water, and the failure surfaces months later, usually as blistering or localised disbonding that traces back to the same patches of steel.
What the dew point actually tells you
The dew point is the temperature at which air, at its current moisture content, is saturated. Any surface at or below that temperature collects water. Steel does not have to look wet; it only has to be close enough that a thin film forms as the air cools against it. ISO 8502-4 is the publicly available standard that gives guidance on estimating the probability of condensation before paint application.
The rule most specifications work to — a substrate at least 3 °C above the dew point — is not an arbitrary safety margin. It is a working allowance for the fact that steel temperature and air temperature move independently, and that a reading taken at one point does not describe the whole surface.
Why steel and air are not the same temperature
- Thermal mass. A thick plate or a stiffener changes temperature far more slowly than the air around it. In spring and autumn steel can sit several degrees below the air for hours after sunrise.
- Night sky radiation. Uncovered steel loses heat to a clear sky overnight and can be colder than the surrounding air at dawn — which is exactly when the dew point margin is most often lost.
- Sun and hot work. Steel in direct sun or near hot work can be far above air temperature, which brings its own problems: premature drying, dry spray, pinholes and solvent sealed under a film that has already skinned over.
- Enclosed spaces. Tank internals, voids and box girders follow the thermal mass of the structure around them, not the weather at the yard gate.
Reading the numbers on site
A dew point meter is the practical tool: one reading gives air temperature, relative humidity and dew point. Steel temperature is then taken with a contact gauge at several points on the surface, not one convenient spot, and the comparison is made point by point.
- Measure near the surface being coated, not at the mixing station. Moisture content and temperature differ between an open deck and the inside of a tank.
- Deliberately check the coldest locations: shaded faces, the underside of stiffeners and areas near cold bulkheads.
- Record the readings with time and location. The application record is what settles a dispute later, and the only way to see a pattern across a project.
The humid end and the dry end
- Too humid. Condensation, and on amine-cured epoxies a greasy surface that repels the next coat. Adhesion loss here is often blamed on the product rather than the conditions.
- Too dry. Moisture-cure products are the exception: inorganic zinc silicate needs humidity to cure at all, which is why most data sheets state both a minimum and a maximum relative humidity.
- Too hot. Pot life falls sharply with temperature, and a film that dries quickly can seal solvent inside. Thick high-build coats are the most exposed.
Plan the hour, not just the day
The application window is usually a time window inside the day, not a whole day. On many sites the workable period runs from mid-morning to late afternoon, with the dawn hours reserved for checking rather than spraying. Where the programme does not allow that, dehumidification and controlled heating bring the surface back inside the window, and ventilation keeps an enclosed space within limits.
Exact figures — minimum substrate temperature, maximum relative humidity and the required margin above dew point — vary between products and between stages of a system. They should be taken from the latest technical data sheet for each specific product, and checked against the project specification before application starts.