The re-use of process models is an objective frequently discussed in the literature. The state-of-art expresses numerous efforts towards this direction, by providing a great variety of methods, models, algorithms and tools. More particularly, many approaches utilize similarity metrics, patterns, repositories and fragments in order to facilitate process model re-usability. This thesis is motivated from the need to re-use recurring parts of process models expressed in Business Process Model and Notation 2.0 (BPMN 2.0) language, and use them to synthesize representative, executable, synthetic BPMN 2.0 process models, that are utilized for benchmarking purposes. More particularly, the focus is on an already available algorithm that detects re-occurring structures in a collection of BPMN 2.0 process models. These structures are named “Relevant Process Fragments” (RPFs) and they are stored in a separate collection. In the scope of this thesis, an RPF Repository is developed, which enables the composition of the artificial, executable BPMN 2.0 process models, with respect to user-defined, benchmark-related criteria. To this edge, the contributions of this work are to design and implement a prototype with the following functionalities: a) automatically characterize the RPFs with respect to benchmark related characteristics, and store them in an RPF repository; b) retrieve the appropriate RPFs from the repository with respect to benchmark related criteria; c) compose synthetic BPMN 2.0 process models that are compliant with the BPMN 2.0 standard; d) validate the representativeness of the process models through empirical rules; and e) automatically export the executable and/or deployable forms of the synthetic process models for two different BPMN 2.0 workflow engines. Our methodology is validated through simple and complex use case scenarios.