Emoji :calendar_spiral: Date

Dec 4, 2024- 15:00 - 16:00 EST

Recording: here.

Emoji :busts_in_silhouette: Participants



  • Edwin Derks (TenneT)

  • Sebastien Lepy (CRESYM)

  • Jasper van Casteren (TenneT)

  • Lampros Papangelis (CRESYM)

  • Peter Mitri (RTE)

  • Samuel Vaillancourt (Siemens Energy)

Emoji :goal: Agenda (and slides)

Emoji :speaking_head: CRESY - Colib.net - Sebastien Lepy and Lampros Papangelis

  • presented several actions that CRESYM launched

    • a web library open source, for power system component models and test cases to perform dynamic studies: colib.net

      • collect good enough different models (what kind?)

      • focussed on grid operators, but can also be academia and manufacturers

      • doesn’t see within LFenergy overlap yet, but with ESX Emoji :question_mark:, GPST Emoji :question_mark: , better grids

      • bridge simulation tools of GPST, seamless connection with portal of GPST, colib.net is a tool in GPST portal (https://opentools.globalpst.org/ )

      • Goals: host a project in LFenergy, easy for users to access / find

        • investigate what is existing, merge etc.

        • develop governance

        • decide which services on top of this tool

        • populate the library with models

        • restructure colib.net to work with more models

Emoji :speaking_head: Annual Review PowSyBl and Peter Mitri (RTE)
www.PowSyBl.org

  • start from 24:17 of meeting recording

  • presented overview, updates, (new) features, integrations with other tools

  • achieved criteria of growing the community / contributions

  • technical governance, steering community, best practice governance silver level

  • presented growth plan and help users with e.g. jupyter notebooks, new Julia binding

  • key achievements of last year (e.g. dynamic security API), new widget for jupyter notebooks

  • PowSyBl in LFE archimate

  • request for support / help (e.g. OSS logs in more languages, more feedback, keep processes practical and quality high of contributions, understand best practices score on lfenergy)

  • feedback on LFenergy: great help with e.g. promotion, organisation, badging keeps project clean and forces to be rigorous and clear list of criteria.

Questions / comments

  • Sebastien - Project took 5 years till now. Sebastien - discuss DC-bell - additional contributors to PowSyBl.

  • Jasper van Casperen

    • how do you balance you do yourself, and how much did you get contributions from OS community. Only 50% now? Or is 90% still with you?

      • RTE is main contributor, 2 other big contributors (artelys and aya), OSS, but not individual contributors, they use PowSyBl in commercial products.

      • there are community commits / request on specific topic. E.g. two weeks ago, someone of google found a bug, created pull request and it was merged.

    • Do more people in RTE contribute that are outside the core team?

      • That happens for sure. Many other within RTE contribute. Core team of 7 people, 87% is committed by RTE.

    • We miss the knowledge of other people in the TenneT that can code but currently cannot contribute. TenneT just started to look into open-source projects.

  • Thomas

    • Quantifying benefits of open-source

      • these questions apply to PowSyBl as well and need to be answered.

      • Small extra cost - with writing high quality cost, architectures in a generic way, extendable.

      • haven’t done excersize exactly

  • David

    • You mentioned developments to the model to support DSO’s, and specifically, you mentioned unbalanced load flow problems. Do you have a workplan?

      • Do you have time?

    • Can share what we had to do for Gridlab-D early on in the project in the US, with unbalanced load flow, which is a particular challenge. Conversion of models was challenging. Lots of transmission level Grid Lab D needs to do, inplemented PyPower.

No formal vote is needed to stay in a stage. Formal votes happen in the TAC.

Emoji :speaking_head: TenneT presention - ODINA - Jasper

ODINA specializes in fast linear network analysis. Speed is everything. DC loadflow is most important. Much uncertainty in dutch network, e.g. surrounding countries. Analyze billions of load-flows.

  • Work with everything that changes in network model, production / consumptions, phase shifting, topologies, splitting , adding lines, etc.

  • In order to keep it fast, things need to be linearized.

  • currently doing AC loadflow.

  • training AI models

  • brute force (2 million / s loadflow analysis)

  • where to move this library? join other OSS?

    • we have now a professional, nice robust libary in course of next year

Discussion
Interesting to hook up to API’s of PowSyBl and collaborate.

How to pick cases in future with highest priority to do dynamic analysis for. Can we only use Power Factory or is there other ways to do it?

Organize brainstorming session around this?

Emoji :white_check_mark: Action items


explore interest expressed by several to brainstorm on ODINA, Power Grid Model, Grid-Lab D, PowSyBl, CRESYM colib.net, OSS governance and possible synergies
, could you check whether 87% can be specified between RTE core team and other people inside RTE. This might be interesting for TenneT.
Is there a way @peter mitri and to find out how RTE (and bosses of Sophie/Peter) quantify benefits of OSS?
share with what they had to do for Gridlab-D early on in the project in the US, with unbalanced load flow
schedule meeting with PowSyBl for next year.
Next meeting is 5th of February 2025. register here.

Emoji :arrow_heading_up: Decisions


  1. No decisions required