Parametrierbare Gebäudemodelle für dynamische Energiebedarfsrechnungen von Stadtquartieren

Aachen / E.ON Energy Research Center, RWTH Aachen University (2018) [Book, Dissertation / PhD Thesis]

Page(s): 1 Online-Ressource (xxv, 158 Seiten) : Illustrationen


This thesis presents a methodology and a software framework for dynamic modeling and heat demand calculations of building stocks using basic input data. Such heat demands build the foundation for the investigation of innovative efficiency measures, such as advanced control concepts and optimal design of heat supply systems. The methodology includes simplified dynamic building performance models that fulfil the requirements regarding computational effort and model complexity on urban scale. An accompanying model parameterization routine on the basis of accessible input data allows the efficient simulation of entire building stocks. Both parts, parameterization and modelling, are embedded in a workflow automation process and implemented using the programming language Python and the modelling language Modelica, respectively. The parameterization makes use of archetypes for residential buildings, offices and institute buildings, which allows for statistical enrichment of individual datasets for the given building types. The archetype for institute buildings is developed based on statistical analyses of two research centers. The simplified building performance models are based on a reduced thermal network model, which is described in the German Guideline VDI 6007-1 (2015). As part of a characterization of different model types, the use of a second order model including two state variables proves to be the best choice concerning efficiency and complexity. This model uses of one state variable each for interior and exterior building elements in addition to one separate resistance for windows.The investigations of three use cases show the framework’s ability to simulate entire building stocks using only basic input data. The use case of one of the studied research centers results in differences between simulation and measurement of less than 3% in the annual heat demand of all buildings. The simulation can in addition capture the dynamic behavior of the heat demand, what is highlighted by a coefficient of determination of 0,894. All parts of the methodology comply with the requirements regarding verification, accuracy, transparency, stability and flexibility. They can contribute to the field of urban energy modelling, if the individual application allows the use of archetypes and reduced order models. The parameterization is implemented in the Python library TEASER, the reduced order building models are integrated into the Modelica library AixLib. Both libraries are available open source at and



Lauster, Moritz Robert


Müller, Dirk
Nytsch-Geusen, Christoph


  • ISBN: 978-3-942789-59-2
  • REPORT NUMBER: RWTH-2018-230258