CryoPaint after 1,500 hours of accelerated weathering
Swift Coat ran CryoPaint through ISO 4892-3 accelerated weathering and measured its reflectance 44 times over 1,500 hours.
What the paint was put through
The QUV chamber runs eight hours of UVA-340 light with the black panel at 140°F, then four hours of condensing humidity at 122°F, and repeats that cycle without a break for 1,500 hours. UVA-340 lamps reproduce the short-wave end of sunlight, which is the part that breaks coatings down.
Reflectance was read on a spectrophotometer at three spots on the specimen, 44 times over the run.
| Standard | ISO 4892-3 Cycle 1 (fluorescent UV and humidity), ISO 4892-1 |
|---|---|
| Chamber | Q-Lab QUV accelerated weathering tester, UVA-340 lamps at 0.76 W/m²/nm at 340 nm |
| Cycle | 8 h UV at 140°F black-panel temperature, then 4 h condensation at 122°F |
| Total exposure | 1,500 hours |
| Instrument | Hunter Labs UltraScan Pro, 350 to 1050 nm, 2 nm resolution, diffuse d/8° |
| Sampling | Three locations on the specimen per reading, 44 readings |
| Specimen | CryoPaint white outdoor paint |
Only a 1.4% reduction in reflectivity, and holding steady
The only drop happens in the first 300 hours. After that the readings sit inside a band about one point wide, with no trend in either direction, all the way out to hour 1,500.
From the first reading to the last, the change is 1.4%, 93.2% down to 91.9%. The panel found its level in twelve days of chamber time and sat there for the next fifty.
Solar-weighted reflectance over 1,500 hours
Each point is the mean of three measurements, weighted across the solar spectrum from 350 to 1050 nm
Regression over the readings from hour 292 onward finds no decline. The fitted slope is slightly positive and not distinguishable from zero (p = 0.24).
What's with the discrepancy between the 93.2% here and 94% elsewhere?
CryoPaint’s published 94% is weighted across the full solar spectrum. The instrument used for this test reads 350 to 1050 nm, and that range holds only 76% of the energy in sunlight.
Score this same panel over four different ranges and it reads anywhere from 93.2% to 95.8%. All four numbers are correct.
Reflectivity versus measured spectrum
Where the reflectance actually sits
Full reflectance curve at hour 0 (blue) and hour 1,500 (orange)
The curve climbs steeply below 400 nm and runs flat above it. Any range that takes in more of that ultraviolet foot averages lower. Plotted 360 to 1030 nm; readings at the very edges of the instrument’s range are not reliable.
CryoPaint did not chalk
Chalking is how a white roof usually loses its reflectance. UV breaks down the binder at the surface, the pigment underneath works loose as a fine white powder, and rain carries it off. It is the normal aging path for water-based acrylic coatings and for other white radiative cooling paints.
The industry already plans around it. The Department of Energy’s cool roof guide tells building owners to estimate an aged roof by keeping only 70% of whatever sits above 0.20, which puts a coating starting at 0.94 near 0.72. Chalking is a large part of why.
Swift Coat looked for chalking at every inspection through hour 1,500 and never found any.
What this means if you are buying it
Cool coatings are sold on a first-day reflectance number. That is why aged ratings exist and why they sit so far below the number on the can.
CryoPaint drops 1.4%, levels off and holds. It does not chalk, so the mechanism that pulls most white coatings down over time is not running here. There is no powder coming off the surface to take the reflectance with it. CryoPaint is self cleaning, so dust gets washed off naturally with rainfall. If you do need to clean CryoPaint, use a hose and/or apply a thin layer CryoPaint to refresh the top coat.
More on how CryoPaint behaves in the field is on the product page, and measuring performance covers how to check it yourself.
Testing by Swift Coat. Swift Coat is an ASU spinout in Arizona building nanoparticle thin-film coatings for solar and aerospace. They ran the exposure, took the measurements and wrote the report. The original full report is available on request through our contact form.
Solar-weighted reflectance calculated from the raw spectra against the AM1.5 global reference spectrum across the instrument’s 350 to 1050 nm range. The specimen was cast thick and had surface cracks before exposure began; Swift Coat recorded them at hour 0 and logged no surface changes at any reading after that.
Check it yourself.
Put a sample on a square foot of metal and take a reading.