At 5 C the epoxy does not cure slowly, it barely cures at all. Substrate temperature, condensation risk and how to verify a cold cure without guessing.
The foreman reads 8 C on the thermometer, decides the day is too cold, and sends the crew home. The next morning a competitor's crew coats the same steel at 6 C and the paint turns out fine. Neither decision was wrong. What separates them is whether the temperature was measured in the right place, and whether anyone checked what the coating actually did over the following week.
Why the reaction slows down this much
- Amine-epoxy chemistry is temperature-driven. The opening of the epoxy ring by the amine hardener follows an Arrhenius relationship. As a working rule, reaction rate roughly halves for every 10 C drop. At 5 C a system that gels in 6 hours at 23 C may need two or three days to reach the same degree of conversion.
- Touch dry does not mean cured. In a solvent-borne coating the solvent leaves first. The film feels dry and can be walked on while crosslinking has barely started. This is the single most common misreading on a cold site, because the evidence available to the applicator (finger, thumb twist) only registers solvent loss.
- Thin films suffer more than thick ones. A 60 micron coat loses heat to the substrate almost instantly and has little thermal mass of its own. A 200 micron build holds reaction heat slightly better, which is one reason thick-film systems are sometimes specified for winter work.
Measure the steel, not the air
Air temperature is the number that gets written in the daily report and the number that matters least. What drives cure is the temperature of the surface the paint is sitting on.
- Mass matters. A 6 mm plate tracks ambient air within minutes. A 40 mm flange, a pump casing or a tank shell lags by several hours and can still be at 2 C when the air has warmed to 12 C.
- Orientation matters. In low winter sun a south-facing web can sit 10 C above a north-facing one on the same structure, at the same hour. The shaded face is the one that fails.
- Measure the shaded face. Contact probe or infrared gun on the back side of the member, in three or four spots, recorded in the report. If the specification names a minimum substrate temperature, that reading is the compliance evidence, not the air thermometer.
Cold weather makes condensation more likely, not less
It is intuitive to think dry winter air is safe. Absolute humidity is indeed lower, but the saturation capacity falls faster than the absolute content does, so relative humidity at the steel surface frequently climbs above 80 percent in the early morning.
- The 3 C rule still applies. The standard guidance to keep steel at least 3 C above dew point does not relax in winter. It becomes harder to satisfy, because both numbers move quickly through the day.
- Thermal lag creates hidden condensation. Steel that was cold overnight and is warming in morning sun can pass through the dew point from below. Water flashes onto the surface while the air reading looks acceptable.
What actually helps
- Heat the substrate, not the paint. Space heaters, induction pads or heating blankets under the steel raise the surface temperature and keep it there. Heating the coating itself only lowers viscosity and shortens pot life; it does nothing for the reaction once the film is on cold steel.
- Build an enclosure and dehumidify. A tent with desiccant or refrigeration dehumidification addresses both temperature and humidity at once. On tanks and vessels this is often the only route to a compliant application in January.
- Extend the recoat window deliberately. If the specified interval is 16 hours at 23 C, treat it as a variable rather than a constant and confirm the surface state before overcoating. A second coat applied over an under-cured first coat traps solvent and produces intercoat failure months later.
- Accept the schedule cost. Cold work is slower work. Compressing the programme by ignoring cure time moves the failure into the warranty period, where it is far more expensive.
What does not help
- Adding extra hardener. A common field habit, and a damaging one. Excess amine leaves unreacted hardener in the film, plasticises it, promotes amine bloom and can reduce adhesion. The mix ratio is a stoichiometric number, not a throttle.
- Thinning to improve flow. Cold paint is viscous, and adding thinner makes it flow. It also reduces film build per pass, raises solvent retention and can push volatile content beyond what the specification permits.
- Judging cure by touch. On a cold site the thumbnail test will pass long before the film is fit to overcoat. It measures the wrong thing.
Verifying a cold cure
When the application has happened below the recommended temperature, the film should be treated as unproven until it is tested.
- Solvent rub according to the relevant standard gives a quick, repeatable indication of conversion on the shop floor.
- Pencil hardness provides a numerical result that can be compared against a reference panel cured at 23 C.
- Pull-off adhesion is the decisive test before handover, and it should be scheduled after the film has had its full cure allowance at the actual site temperature, not at the specified one.
- Delay immersion and chemical exposure. A coating that will eventually sit in water or a chemical should be given extra time before service. Seven days is a reasonable minimum for a thin film that cured at 5 C, and longer for high-build systems.
None of this argues against winter work. Solvent-borne systems are more forgiving in cold conditions than water-borne ones, because there is no film-formation step that depends on water coalescing above a minimum temperature. What they are not is indifferent to temperature. The material will do what the chemistry says it will do; the only variable under control is whether anyone measured the steel before the first coat went on.