I wonder what their criteria is. I know NASA published at some point some generalized equations to gauge buckling risk for thin wall cylinders, but I didn't find it credible for use in standard piping, where buckling results in pipe bowing and not necessarily localized failure.