Well, not even when I did the worksheet, and maybe it was 10 years ago, I was but looking for info on the matter, so really I can't answer; you see in the worksheet the averaged result came close to Ugural so I would normally use Ugural unsymmetrical value.
By the way I see I sent a quite imperfect pdf printout, then I upload a better one.
It must be said that even if local buckling and -to the purpose of design intent- overall buckling in cylindrical tank or akin shells are not as conceptually different, these evaluations deliver mainly opinion on local buckling. Hence to prevent overall effects beign unaccounted...
1. A sound nonlinear analysis should be made, including if you want initial imperfections and anisotropy in the shell if any, as is typical of tanks when spirally prestressed.
2. I would introduce some comparison stress, incorporating the effects of bending and shear stress to compression force, what essentially means, an even lower critical stress than that given by the equation should be taken.
3. One must pay attention in analysis to differential settlement if any, since this has ruined a number of tanks.
4. If the tank is big and slender, even inwards pressures from wind etc may have additional say on the critical values.