Documentation, comparison and processing of demonstration results on the topic of "Digital Building Twin"

In recent years, two projects for the development and initial implementation of a digital twin on real buildings in Austria have been completed. This digital twin is a detailed simulation model with the software IDA ICE, which is validated in real time with measurement data from a real building. As a result, there is a model that always represents the real condition of the building and its system technology. This model can then be used for automated error detection or to optimize the control. The aim is to reduce energy consumption and improve user comfort.

The actual work on the demonstration objects and the development of the software (the so-called Building Tracker) took place within the following two projects.

• EU project "Arrowhead Tools" (2019 - 2022). In this project, as one of many use cases, the software for the digital twin was further developed on the one hand and used on the other in the new research building of Infineon Austria in Villach. Here the focus was on the mapping of several office and meeting areas with the aim of optimizing the operation of heating, cooling and ventilation.
• "City of the Future" project "Digital Twin" (2019 - 2023). The same method was used here in the H2 office building in Vienna. The main focus here was to map the refrigeration system in order to be able to optimize operation.

In this report, these two demonstration projects of a digital building twin are compared with each other, and the first operational experiences are documented.

Different focal points were set in the two projects, so that on the one hand experience is now available in the mapping of rooms and the associated user comfort. On the other hand, knowledge was also gained when creating a digital twin of a certain part of the system technology.
The architecture of the digital twin was different. In one case, the simulation model was supplied with data from the building in a cloud, in the other case a purely local solution was used.
Weather data was identified as an important factor. Here, too, there are various solutions that have been implemented and tested in the two buildings. In addition, a study was carried out to compare different measurement methods and their accuracy in comparison with online weather data.

The most favorable case for the creation of a digital twin is when a building and system simulation has already been carried out in the planning phase of the building in order to optimize the building and building services. In this case, the existing model for the real-time simulations can simply be adopted. It may still have to be adapted to the actual implementation in a few points and validated with measurement data. A documentation of the building within the framework of BIM could also significantly simplify the creation of the model.

The decisive question is not only whether a BIM model exists for a building, but above all whether the stored data is also suitable for being transferred to a simulation model. Even if, for example, an IFC model was created by the architect, this does not mean that all the information required for the simulation model is stored there.

Furthermore, further research and development needs, which are necessary in order to help the new technology to be widely implemented, are shown.

On the one hand, this concerns a standardized creation of the simulation model for the digital twin, which could save time and money. Likewise, standardized solutions must be found for the connection to existing building management systems. This can be very expensive, especially for existing buildings.

The aim of further developments should be to derive as many phenomena as possible from the temperature data in the building, such as window openings, and thus to be able to do without additional sensors. Combination sensors, which, in addition to the temperature, also record the CO2 concentration and possibly the humidity, can be helpful, e.g., to detect the presence of people in rooms.

The next step would then be to use the real-time simulation for an optimized control strategy. For this it is necessary to develop suitable optimization algorithms, with the help of which, for example, optimized set temperatures for heating and cooling or a shading control optimized according to the weather forecast could be determined for the respective building. For this purpose, the simulation model could be used to carry out simulations with forecast data for the weather in so-called what-if studies.

AEE – Institut für Nachhaltige Technologien
Dipl.-Ing. Dagmar Jähnig, Dipl.-Ing. Franz Hengel, Dipl.-Ing. Christoph Moser, Dipl.-Ing. Reinhard Pertschy