The behavior of complex production automation systems is hard to predict, therefore simulation is used to study the likely system behavior. However, in a real-world system many parameter variants need to be tested with limited re-sources. Therefore, test cases need to be generated in a syste-matic way to find suitable scenarios efficiently. This paper investigates the effort of two approaches for providing test cases based on available testing knowledge. The traditional approach uses a static generator script based on implicit test-ing knowledge, which takes significant effort to add new pa-rameters. The innovative approach uses a dynamic generic generator script based on an ontology data model of the testing knowledge. We empirically evaluate these approaches with a use case from the production automation domain. Major result is that the high-level test description of the ontology-based approach takes more initial effort for setup, but increases the usability and reduces the risk of errors during the test case generation process.