TPM bachelor student Niels Goedegebure presents innovative research on price forecasting uncertainty at international conference
The increase in electric cars is causing peak loads on the existing, and sometimes outdated, regional energy network, with the result that the power 'blows out' when overloaded. Expanding the network requires a substantial investment and is therefore not ideal. TPM bachelor student Niels Goedegebure is working on a solution.
The research he did for his Bachelor End Project was so innovative and thorough that he wrote a conference paper about it together with PhD student Roman Hennig. The paper, 'Generating Electricity Price Forecasting Scenarios to Analyze Price Uncertainty Impact on Tariffs', was recently accepted at the 17th International Conference on Probabilistic Methods Applied to Power Systems to be held in June 2022.
New tax structure
‘Instead of radical and costly extensions of the energy network to cope with the demand for power, a new tariff structure is now being considered to smooth out peaks in energy demand', Goedegebure explains. Such a new charging structure, which is likely to be in place in the near future, will consider usage per kilowatt-hour, for example, instead of a fixed annual amount. This is then collected by your regional network manager, such as Stedin or Enexis. This charge should ensure that people do not all charge their cars at the same time, when there is a lot of electricity and the price is cheaper.
Flexible price predictions
TPM PhD student Roman Hennig has made a model to calculate the effects of different taxes. It is a model that assumes fixed prices, whereby an electricity user can see the electricity price four days in advance and, based on that, can plan when to charge his car. However, in practice the price is not fixed but uncertain. Goedegebure therefore made a modification to this model that assumes flexible price predictions over a number of days. The expectation was that this uncertainty would lead to worse results and higher peaks. But this adjusted model with price forecast uncertainty actually shows peaks that are smaller. This built-in price forecast uncertainty in combination with such a model has not been applied before and makes it very innovative.
Solving societal problems
For his Bachelor End Project Goedegebure wanted to work with probability and statistics as part of the minor in Finance he took at the faculty of EEMCS. At the same time, he wanted to contribute to solutions for social problems. With his background knowledge of the domain Energy and Industry it was obvious that he would do 'something' with energy. After contacting a number of teachers, Laurens de Vries came up with this great assignment. Goedegebure found the combination of theory and practice ideal. Nevertheless, it was sometimes quite tough: 'In the beginning of the project it took some familiarising with the theory behind the tariff structure and the theoretical models. But when looking back at the end of the project, it is very satisfying to have mastered the subject matter from scratch. And the fact that you can also make a social impact is very important to me. It is precisely the application of theory in practical situations that you learn so well at TPM. In this way you can really make a difference and that appeals to me very much’.