hosted by RLI, Berlin, October 16th, 2015
- short presentation of participants and introduction (objectives of the platform)
- presenting the first proposal of configuration
- discussing parts of the configuration
- conclusions and discussion of next steps
Martin Glauer (IWS), Eva Wiechers (ZNES), Wolf-Dieter Bunke (ZNES), Ilka Cußmann (ZNES), Mathis Buddeke (Wuppertal Institut für Klima, Energie, Umwelt), Cord Kaldemeyer (ZNES), Stefan Pfenniger (ETHZ), Jonas Hörsch (FIAS), Pedro Crespo Del Granado (ETH Zürich), Berit Müller (RLI), Ludwig Schneider (RLI)
In August 2015 the four institutes RLI, ZNES, Next Energy and IWS (Uni Magdeburg) started a modelling project called "open_eGo". Apart from developing an open source model for the German electricity grid a main objective of the project is to develop an OpenEnergy modelling platform that provides the possibility for modelers to work together on models, to exchange themselves about models, to find open source models, to define quality criteria and plausibility checks, and much more which is constructive for a scientific exchange about energy system modeling.
Guido guided a breakout group about that at the openmod workshop in London to start the discussion about the composition, the requirements and the organisation of the platform. The results of the breakout group are here. One main result was, that it should be closely connected with the openmod website or even better be a part of it.
This input was considered for the first suggestion for the configuration of the platform which was the basis for the discussion of this workshop. The different elements can be summariesed as:
- User information and communication
- Data and model presentation incl. model integration
- Code and interface documentation
- Usability and understandability for inexperienced users
See also figure "Concept of the OpenEnergy platform (October 14th, 2015)".
(The options in the green boxes with the thin edges have been discussed, but are not planned to be implemented.)
The following text is the summary of the results, not of the whole discussion! The discussion was along the whole scheme which was later adjusted to the outcomes.
It was pointed out, that it is necessary to make the benefits of the platform clear (e.g. introduction into energy system modelling, common coding rules / guidelines, an "energy system modelling"- glossary (that was discussed in several breakout groups in the last openmod workshops), model results in a certain shape, place for discussions and central data storage and visualisation, common (benchmark) dataset / library, Exchange of code snippets / how tos).
With regard to the integration tool the depth of the integration was enquired. Leaving the integration to the modellers was recommended.
The three main discussion points were the representation of the models, the availability and form of data and the form of communication.
Representation of Models
The extension of the already existing presentation of the models in the openmod wiki (factsheets) is intended.
- The models themselves (and their documentation) are mostly on GitHub.
- In future a user can easily see, if the model, he/she is interested in, is worked on and how many developers are active (e.g. current version and date, last commit, milestones). A dashboard will show the activity on GitHub. -> API to draw model activities (commits or versions).
- The desired glossary (part of the openmod wiki) was considered to be the base for the ontology. The factsheets in the openmod wikishould be harmonized with the glossary: the terms used in the factsheets should be explained in the glossary (or even be glossary entries). The glossary needs a search interface. All data should be in the wiki (semantik wiki?).
Good communication with the project "open power system data" (OPSD) is necessary to avoid double work. Wolf will keep an eye on that as he is involved in both projects.
The discussion on how to save data on the platform couldn't be finished. The four discussed possibilities were
- git or github.com? not recommended for big data volumes
- datdata? disadvantages have to be carved out again
- set up an postgresql database and use CKAN?
- use an existing CKAN like "datahub.io"?
Continuous integration of new datasets, meaning automated tests and comparisons of their results with the results of a simple reference dataset, is planned.
Communication including Q&A site
The majority was against inplementing a forum into the platform for the discussion of various subjects. Reason: Nobody knew a good working forum; It needs a lot of time for moderation to make work well.
Options for a Q&A site:
- setting up an ask bot on own server (redundant and responsibility for security patch)
- open a new Stackexchange board "energy system modeling" (subdomain)
- use an existing board with specific "tags" (like e.g. openmod)
Closed questions could be summariesed and written into the openmod wiki.
A discussion concerning a subdomain on Stackexchange was started on the openmod mailing list by Martin Glauer on October 20th.
Channels of communication and information can grow organically and might stay empty when they are designed and not developed by the active community. The glossary will probably not grow organically. Therefore, the process could be supported.
Name of the Platform
Is the planned OpenEnergy platform a platform or a portal? Also the name is not yet fixed … until the final decision we call it openEnery Platform (OEP).
Summary of changes
The scheme of the platform has been updated according to the discussions of the workshop (see figure "Concept of the OpenEnergy platform (October 20th, 2015)").
A reduction of the scheme has been made later on.
The first version of the platform will be developed until January / February 2016. Then it will be tested, discussed and adopted. The release of the first official version is schedule for April / May.
You like to be involved in further discussions?
Get in contact with us by mailing to:
Berit Müller firstname.lastname@example.org Eva Wiechers email@example.com
Partners of the project open_eGo are:
- ZNES, Flensburg
- NEXT ENERGY, Oldenburg
- Reiner Lemoine Institut (RLI), Berlin
- IWS, Otto von Guericke University, Magdeburg
The project is supported by:
Federal Ministry for Economic Affairs and Energy on the basis of a decision by the German Bundestag