Representative process models that satisfy specific structural criteria are requested in many different use cases. However, as process models constitute a corporate asset for the companies, they are not easily shared. More particularly, when the requestor desires a process models that satisfy specific structural characteristics, the task of gaining the process models becomes even harder. This work focuses on generating synthetic, representative, executable BPMN 2.0 process models with respect to specific user-defined structural criteria. For the generation of the BPMN 2.0 process models we are using re-curing sub-structures. The discovery of the sub structures has been introduced in previous work.The generated process models will then be utilized for benchmarking purposes. The original scientific contributions of this work are to provide: a) a method for automatically generating executable representative synthetic process models for a given set of structural criteria, b) the proof-of-concept of the proposed method through prototypical implementation and c) qualitative and quantitative evaluation of the proposed approach.