The deprovisioning step in PowerVS IPI uses openshift-install destroy command to destroy the resources it created. And this seems to fail more often with some timeout which leads to the stale entries. This blocks the consecutive runs on PowerVS on that particular zone, as before cluster creation it tries to delete the stale entries and fails stating that its unable to destroy due to some error. 

Need to enhance code to have some kind of force delete option to destroy the resources in case we hit such issues