Currently, the process IDs are generated from the activity ID and ID of the reference product of a data set. But, we should also include the activity type (1: unit process, 2: terminated system) and system model (APOS etc.) here. Then we should have different IDs for all ei process packages.