While working on some automation to deploy a vSphere 7.0 Update 2 environment that has been configured with vSphere with Tanzu and NSX Advanced Load Balancer (NSX ALB), I noticed that when you disable Workload Management on a vSphere Cluster, the two NSX ALB Service Engine (SE) VMs were still left behind.
It turns out that this behavior is due to a default setting within NSX ALB that will NOT automatically delete the SE's in the case there is a scaled up event which would then cause a re-deploy to happen. Instead, by default it is configured to wait 120 minutes (2hrs) before cleaning up.
If you wish to change this behavior, you can login to NSX ALB UI and navigate to Infrastructure->Service Engine->Advanced and update the "Delete Unused Service Engines After" to your desired value. Please note, that the shortest time interval to wait is 1 minute and if you set it to 0, it means the SE's VMs not be deleted.