In the project NeqModPlus we developed models and tools to simulate energy demand and different energy generation systems of buildings and city quarters that can help reach the goal of zero carbon emissions. One of the case studies is a new building of the inner-city campus of the University of Applied Sciences in Stuttgart, Germany. Here we have measured data and can compare this with our models. In this visualization, we compare the measured heating demand with the results of two different simulation models that vary in their complexity. The complex model is a dynamic calibrated white box model that is described in [1]; the other is a steady-state simulation [2] according to German norms DIN V 18599 and VDI 4710. Both simulations include detailed building physics data as well as locally measured weather data; the steady-state simulation also uses the calibration results of the dynamic model for temperature setpoints and schedules.
[1] P. Monsalvete Alvarez de Uribari, and V. Coors: A dynamic model for district-scale building demand simulation, Proceedings Dynastee Symposium The Building as the Cornerstone of our Future Energy Infrastructure, 10-11.4.2019 Bilbao, Spain.
[2] https://simstadt.hft-stuttgart.de/
Measured/Simulated Heat Demand
Select Sampling Rate:
Chart Type:
Dataset Explanation
Measured Heat Demand [kW]: Heating power demand from four heat meters in HFT building 8.
Predicted (Dynamic White Box Model) [kW]: Simulated heating power demand based on a complex simulation model.
Predicted (Simplified Steady State Model) [kW]: Simulated heating power demand based on a simulation model with reduced complexity.
Result Interpretation
According to the measured data and simulated result, it could be summarized as follows:
The demand on some days is underestimated by both, the Simstadt and the detailed INSEL model
The INSEL model follows the demand fluctuations more closely
The results are as estimated, the more detailed INSEL model shows a better prediction
quality
compared to the Simstadt model. However it also demands a significant greater effort to be
created and has a higher computational demand