Data driven heat pump models for generation of electrical load profiles at DSO level
The present master thesis focuses on analyses of anonymised monitoring data from real heat pumps that is gathered via Web access provided by the heat pump manufacturers.
Background
Due to the transformation of the energy system, Transmission System Operators (TSOs) and Distribution System Operators (DSO)s have the increased need to adapt their networks in order to cope with the changing circumstances, e.g. higher loads caused not only by significant penetration of Electric Vehicles (EV) electrically driven heating and cooling systems. Research of the Center for Energy at AIT, Austrian Institute of Technology supports various TSOs and DSOs by simulating network loads in order to identify their impact on the network and propose the optimal solutions for congestion management, network enhancement, etc. For these dynamic network simulations, load profiles of the actual and expected future network users have to be determined. The load profiles of future network users highly depend on the assumptions made, which can be done with high accuracy. However, findings from real measurement data can also contribute to improving the techniques for load profile modelling and thereby increasing this accuracy to a higher degree.
The present master thesis focuses on analyses of anonymised monitoring data from real heat pumps that is gathered via Web access provided by the heat pump manufacturers. The work aims at development of data driven models and correlations between the energetic performance of heat pumps and different parameters such as ambient conditions, solar radiation, or the configuration of the heat source. These correlations are then used to validate and improve the simulation models that have been used to date for the generation of the electric load profiles. Another important task is the energy efficiency assessment of the investigated heat pumps regarding the configuration of the elements heat source, heat sink, and controller. Then findings will be communicated to manufacturers, planners, and other relevant stakeholders.
Thesis/Learning objectives
After the thesis has been performed the student should be able to:
- Identify and describe a gap in knowledge
- Perform a comprehensive data analysis based on measurement and monitoring data provided by industrial partners
- Apply machine learning techniques to develop data driven models and correlations between the system performance and ambient conditions and other influential parameters
- Validate the models and results against measurement
- Improve the existing model for the generation of the electric load profiles at DSO level.
Proposed time schedule, including milestones and intermediate reports
The thesis is expected to start under January 2022 (wk. 3) and be completed in June 2022 (wk. 23). Intermediate reports will be due in early March (wk. 10) and late April (wk. 17).
Contact persons
Supervisor at KTH, Department of Energy Technology
Supervisor at AIT, Center for Energy
Thomas Natiesta, Research Engineer