Great logic gunman. Everyone knows proof is a risk, as it s a test. When significant numbers of apparently sound barrels start failing in a manner not usually observed, even RFDs have a good reason to wonder if all is not right.