I've been fighting the same issue. I was advised to disable alerts and enable one at a time to troubleshoot. When I've got time free to spend in the lab I can check that out. Unfortunately nobody has been able to give me any troubleshooting steps that did not involve disabling alerts--not an option in production network. I also am seeing the alert seemingly tied to interface utilization and the alerts I have tied to them. I'm unable to determine which interface utilization alert is involved.
None of the alerts tables in the database show this information when that event is getting written to the event log.