Photo Sonja Iskov

Statistical modelling for energy system planning

The Centre for IT-Intelligent Energy Systems in Cities (CITIES) and the organizing committee of the DTU-Compute Seminar Series “Mathematics and Informatics for Intelligent Energy Systems” are happy to announce that Dr. Chris Dent from the School of Engineering and Computing Sciences at Durham University, UK, will be giving the talk: "Statistical modelling for energy system planning".

Computer models are widely used for decision support in energy system capital planning. A key feature of most capital planning decisions is that they must be taken under uncertainty regarding the future background against which the system is planned, as well as the statistical fluctuations of outturn in system operation. This talk will discuss the different origins of uncertainty encountered when using computer models as decision support tools, using generation planning as an exemplar (in which case generation adequacy assessment models are very simple and transparent). It will then introduce the idea of using a statistical emulator of the full model to encode uncertainty about the model’s output for inputs where it has not been evaluated, which is important for more computationally expensive models where the number of possible evaluations may be very limited. The use of emulators will be illustrated using examples from network capital planning and from economic projection of generation investment and market prices.

Everybody is welcome. Feel free to forward this email to anyone that could be interested.


Mon 11 May 15
15:00 - 16:00


DTU Compute


DTU Compute, building 303b, room 130