Robustness Analysis for Multi-stage Adaptive Optimization with Application to Strategic Energy Planning



We would like to invite you to the EI-ERIM-OR seminar. It will take place on campus, on Friday, November 11, 12:00-13:00 in room Polak 2-09. Lunch will be provided. The event is in-person only (no live streaming). You can find the information about the talk below.

If you would like to have a separate meeting with the speaker on the day of the talk, please email one of the organizers.

Abstract: In making the transition to a more sustainable energy system, certain strategic investment decisions have to be made in regards to the energy system. Fortunately, we can use mathematical modeling techniques to inform and improve these decisions. The HyChain project ( has developed a dataset and optimization model to serve this purpose for development of the Hydrogen economy in the Netherlands. 

Yet, in its current form, the modeling approach is somewhat inadequate. The existing model is fully deterministic and assumes perfect foresight, i.e., results in optimal decisions for a (known) expected/nominal future. However, a large-scale long-term energy transition brings wide uncertainties regarding future developments in technology, economics and policymaking. The goal of my PhD research project is to extend the existing model to account for the relevant uncertainties.

In the first paper we perform the first steps towards achieving this goal, which is to characterize the uncertainty and analyze the impact. In our analysis of impact we extend the existing "robustness analysis" (RA) framework to a multistage adaptive optimization setting. Furthermore we argue why RA is more interesting than conventional "sensitivity analysis", which is often encountered in scientific literature.


About: Justin Starreveld is a PhD student at the University of Amsterdam. In his research he is developing robust optimization methodology to improve investment decisions regarding the production, storage and transportation of hydrogen. Justin works under the supervision of prof. Dick den Hertog.