MODEX framework comparison study for Germany

Release | 02

This posting is based solely on what I read on the web — so please note any mistakes or misconceptions you encounter or alternatively contact me for corrections.

This describes a model framework comparison exercise.

This suite of projects are variously tagged MODEX (energy system models), open_MODEX (open source models), MODEX‑Net (electricity network models). The acronym MODEX itself expands and translates as "model experiments for the energiewende. This package of work was commissioned by the German Federal Ministry for Economic Affairs and Climate Action. The project landing page is here (in German):

Some 40 modeling frameworks were involved in the overall investigation, with eight being open source and the remainder closed source. Much of the work is captured in this special issue (subject to stepwise additions, it would seem):

The studies themselves typically include the European power system where appropriate, but the focus is on analysis that is able to assist national policy formation in Germany.

Regarding terminology, a “modeling framework” or just “framework” is the phrase that energy system analysts seem to have settled upon to describe a core codebase and associated workflows, tooling, documentation, and technical support — roughly what one would find on GitHub or GitLab. And a “model”, when used more specifically, refers to a framework that has been downloaded and populated with data to investigate some particular research question. The research question and the selected framework need to align, of course.

Comparison methodology

The following paper covers the methodology and tooling that supported some of the comparison studies:

The overarching project also sought to develop and use harmonized data to improve the validity of the comparisons:

A five framework comparison

The most significant publication for this community would doubtless be this study that compares five open source frameworks:

Those five frameworks selected are:

And for completeness, the following three open frameworks are covered in the literature below:

Articles that cover at least one open source framework

The following publications also covers at least one open source framework as indicated:

Generally pertinent articles

And here are some further pertinent publications involving only closed source frameworks:

Regarding van Ouwerkerk et al (2022), this April 2022 openmod google groups thread is also related to power system modeling using open source frameworks in a German context: The AC/DC question?.

Finally, several of the authors cited above are active in this community too, so nice to see some social overlap between the open and closed worlds. :paperclips:

1 Like

Nice overview @robbie.morrison

Please add this publication to the list. It is the result of the “Datenbeauftragte”, a cross-project initiative to handle input and output data using the OEMetadata and the OEP.

1 Like

Thank you Robbie for the nice summary! The last paper under “Generally pertinent articles” does not ony involve closed source frameworks as open_eGo is involved.

1 Like

For a little history, here is my attempt to provide a harmonized structure for high‑resolution energy system models some 20 years back:

  • Morrison, Robbie (6 September 2002). The energy-services supply systems (ESSS) common model initiative — Release A. Berlin, Germany: Technical University of Berlin. doi:10.5281/zenodo.6619604. Previously published on a now defunct website from that era. :open_access:

The abstract reads:

This document proposes a framework to facilitate the sharing of models between various energy-services supply systems modeling projects. The material presented is intended as a starting point for discussion.

1 Like

Thanks! A number of those articles look interesting for when I’ll write up my thesis in the coming months.

I had a look at “Systematic comparison of high-resolution electricity system modeling approaches focusing on investment, dispatch and generation adequacy”, and I remember reading it now because I remember that the LOLE they report is insane - 300 hours per year minimum across all models for 2025. I realise this isn’t the point of the paper, but still…

admin edit: LOLE = loss of load expectation

Text and images licensed under CC BY 4.0Data licensed under CC0 1.0Code licensed under MITSite terms of serviceOpenmod mailing listOpenmod wiki. Openmod YouTube.