Wasn’t there a bit of discussion not too long ago about how Noack values were getting harder to estimate? Some of the large oil companies stopped publishing their estimates.
Maybe keeping the sample at exactly 250°F and the "constant air flow" (kind of vague) is where it can cause some variations. also how big is the sample and what size & material container, where you measure the temp, test chamber environment and many other variables ...
Unless the test is fully defined and universal, the one line test from Wikipedia won't do it:
"In this test, a sample is heated at 250 °C for 60 minutes with a constant flow of air over it. The weight fraction lost is the result for the Noack volatility test."
It will be interesting to know the variation in test results with different interpretation and execution of the noack test when not fully defined.