Register now After registration you will be able to apply for this opportunity online.
This opportunity is not published. No applications will be accepted.
Physics-based building models
Buildings are large energy consumers and we need new control strategies to decrease it. As it would take years to find controllers in reality, simulation tools (models) are needed to accelerate the process. One key is for these models to be physically consistent to then derive plausible controllers.
Over the past years, buildings consumed 30% and 40% of the end-use energy in Europe and worldwide, respectively. Besides retroffiting old buildings, designing new efficient control algorithms will be key to decrease the energy intensity of the sector.
However, to create and test those algorithms, we have to rely on simulations, for two main reasons. Firstly, it is not always possible to access an existing building and deploy a controller. This is emphasized if the building is occupied, as the algorithm could have negative impacts on the inhabitants if it is not well tuned. Furthermore and more importantly, we face a big problem concerning the running time of experiments needed to tune the controller. Indeed, one needs to let the controller run for several days to assess its performance and several weeks to make sure it is working as expected. When searching for the right parameters of the algorithms or testing different solutions, one would thus need years to find the right option. Luckily for us, years can easily be simulated in a matter of hours with modern-day computers.
Using a satisfactory model, one can thus run experiments quickly to find interesting controllers. Consequently, there is currently a need to develop accurate models of building dynamics to develop the next generation of controllers.
Over the past years, buildings consumed 30% and 40% of the end-use energy in Europe and worldwide, respectively. Besides retroffiting old buildings, designing new efficient control algorithms will be key to decrease the energy intensity of the sector. However, to create and test those algorithms, we have to rely on simulations, for two main reasons. Firstly, it is not always possible to access an existing building and deploy a controller. This is emphasized if the building is occupied, as the algorithm could have negative impacts on the inhabitants if it is not well tuned. Furthermore and more importantly, we face a big problem concerning the running time of experiments needed to tune the controller. Indeed, one needs to let the controller run for several days to assess its performance and several weeks to make sure it is working as expected. When searching for the right parameters of the algorithms or testing different solutions, one would thus need years to find the right option. Luckily for us, years can easily be simulated in a matter of hours with modern-day computers. Using a satisfactory model, one can thus run experiments quickly to find interesting controllers. Consequently, there is currently a need to develop accurate models of building dynamics to develop the next generation of controllers.
The goal of this master thesis is to develop physics-based thermal building models, with a practical application on NEST (https://www.empa.ch/web/nest). By building models, we mean here room temperature models, which is the main component to satisfy occupant comfort. Starting from physics equations (thermal heat transfer equations) and taking advantage of existing data, we want to develop models of the thermal behavior of buildings as accurately as possible. These models are usually referred to as Resistance-Capacitance (RC) models.
We know the main drivers of the room temperature evolution: the heating/cooling system, the ambient weather conditions and the occupants (opening/closing doors, windows, blinds, …). The main objective will thus be to understand and capture all these elements and create models that are then consistent with the known law of physics.
The student is expected to analyze different modelling strategies and adapt them to NEST to create accurate simulations. Importantly, these models have to stay grounded in the underlying physics to ensure that the predictions make sense. Characterizing the impact of the weather and the occupants and then taking them into account in the modelling phase will be key.
The goal of this master thesis is to develop physics-based thermal building models, with a practical application on NEST (https://www.empa.ch/web/nest). By building models, we mean here room temperature models, which is the main component to satisfy occupant comfort. Starting from physics equations (thermal heat transfer equations) and taking advantage of existing data, we want to develop models of the thermal behavior of buildings as accurately as possible. These models are usually referred to as Resistance-Capacitance (RC) models. We know the main drivers of the room temperature evolution: the heating/cooling system, the ambient weather conditions and the occupants (opening/closing doors, windows, blinds, …). The main objective will thus be to understand and capture all these elements and create models that are then consistent with the known law of physics. The student is expected to analyze different modelling strategies and adapt them to NEST to create accurate simulations. Importantly, these models have to stay grounded in the underlying physics to ensure that the predictions make sense. Characterizing the impact of the weather and the occupants and then taking them into account in the modelling phase will be key.
Required qualifications of the eligible student for this MSc thesis are a broad knowledge of the physics behind heat transfers or building physics and good programming skills, preferably in Python for compliance with other projects, as well as a general interest in the modelling or the understanding of complex physical system. The candidate should be proficient in English.
This project will take place in the ehub team [2] at Empa, Dübendorf (https://www.empa.ch/web/energy-hub/overview). For further enquiries, please contact Loris Di Natale, loris.dinatale@empa.ch.
Required qualifications of the eligible student for this MSc thesis are a broad knowledge of the physics behind heat transfers or building physics and good programming skills, preferably in Python for compliance with other projects, as well as a general interest in the modelling or the understanding of complex physical system. The candidate should be proficient in English.
This project will take place in the ehub team [2] at Empa, Dübendorf (https://www.empa.ch/web/energy-hub/overview). For further enquiries, please contact Loris Di Natale, loris.dinatale@empa.ch.