Alliant powder has a sample of one of the first runs of Unique from 1899. I read, but have never confirmed that they keep it submerged in water and periodically dry out and test a sample to see if it is still good, and so far, after 122 years, it still is.
But that sort of longevity is dependent on processes that can vary in their effectiveness. How perfectly acid-neutralized were the lots of nitrocellulose and nitroglycerin that were used? How much stabilizer was added (assuming water hasn't taken that role over)?
Additionally, why, contrary to much of the long life reported for so many powders, did so many lots of IMR 4007 SC start to break down early and have to be recalled starting in 2015 until, finally,
all past lots were finally recalled in 2019? Why did
Accurate have recalls on their 2495, 4064, and 4350 made before 2016 in that same year? The explanation of blending with over-age old lots for burn rate control wouldn't seem likely to cross three different numbers in Accurate's case, so something in their original compounding had to have been off; too little stabilizer or overly acidic components or some such thing. And there have been
other powders in other years.
Whatever causes a powder to go bad, it usually seems to be lot-related. Some lots are going to be great for an extremely long period of time, and others are going to drop dead prematurely. The question is, how do you know which one you have of which until you get there? You don't. So you keep in the habit of rechecking them whenever you load.