Buckling critical coefficient changes if run with another separate P-Delta analysis

[Be_S_]

I have a tall, braced steel structure model set up in Robot structure. When I run the model with buckling case only, the result of the critical coefficient differs from when I run it together with another P-delta analysis case at the same time.

Note that I defined two complete separate buckling and p-delta analysis cases. Only running them at the same run changes the result of the buckling case.

Anyone had the same issue or knows why this happens?

[Rafacascudo]

Zip and attach the model , if possible