I'd assume that the engineers would have performed some form of stress tests for these installations that were actually meant to afford wind resistance and that the noise should have been noticed. Does anyone know how these conditions are tested in a simulated large structural engineering lab?