OpenMod workshop Freiburg 2026 - Lightning Talks and Poster Contributions

Workshop
,
44

:four: :zero:

Contribution type: lightning talk and poster

7K Challenge: Synthetic Power Injections for Open Grid Models

Presenter: Pierre-Olivier Cayetanot - CRESYM

Description:
The 7K Challenge is a new open-data initiative built around the RTE7K dataset, a detailed representation of the French transmission network released through the D-GITT project. The dataset includes the complete grid topology from 63 kV to 400 kV, with snapshots every 5 minutes over the period 2021-2023.

The first challenge focuses on generating realistic synthetic power injections, the missing component needed to run power-flow simulations on the dataset. Participants are invited to use public data, statistical methods, optimization techniques, machine learning, or hybrid approaches to reconstruct realistic load and generation profiles.

The goal is to spread awareness around the dataset, start an open data community, as well as creating a fully usable open dataset for power-flow analysis, grid optimization, and AI applications, enabling researchers and developers to work with genuine TSO topological data.

Background:
The RTE7K dataset was released through LF Energy OpenSynth and while it provides detailed information on its components and topology evolution, operational injection data is not included.

The 7K Challenge, organized by RTE, CRESYM, eRoots, and IRT SystemX, aims to fill this gap. Beyond synthetic injections, potential future challenges will explore topics such as AI-driven topology management, power-flow analysis, and advanced grid-operation use cases based on realistic open grid data.

Optional links:

Challenge link: https://www.codabench.org/competitions/16703/
RTE7K Dataset: https://huggingface.co/datasets/OpenSynth/rte7000

OpenMod workshop Freiburg 2026 - Schedule & Program
45

:four: :one:

Contribution type: lightning talk and poster

Comparing MILP formulations for non-linear hydrogen pipeline investment in PyPSA-based energy system model

Presenter: Veronika Brooks

Description: This contribution uses a PyPSA-based energy system model and extends its optimisation problem with alternative MILP formulations for hydrogen pipeline investment. The application focuses on offshore hydrogen infrastructure in the North Sea, where economies of scale can highlight the system value of hydrogen transport and shift network expansion from highly meshed low-capacity structures towards more defined corridor development.

The work compares four formulations: a linear reference case, piecewise-linear cost curves, a start-up formulation with fixed corridor activation costs, and discrete pipeline-size options. The methods are applied to a North Sea energy system model covering offshore wind, electrolysis, hydrogen pipelines, and electricity transmission optimal investment and operational decisions.

The contribution will discuss the modelling approach, preliminary results, and lessons learned from implementing mixed-integer and piecewise-linear formulations in a PyPSA workflow.

Background: This work is part of ongoing doctoral research on integrated European hydrogen and electricity infrastructure planning at Fraunhofer IEG and the Chair of Energy Economics, TU Dresden. It builds on PyPSA-Eur developments for offshore energy infrastructure in the North Sea by the TU Berlin PyPSA team and originated from previous work in the MOHN project funded by BMBF. It contributes to the open energy modelling discussion on representing economies of scale and discrete infrastructure choices in large-scale models.

OpenMod workshop Freiburg 2026 - Schedule & Program
46

:four: :two:

Dear Ramiz,

I know it is late submission and submitting with the kind hope:

Contribution type: lightening talk + poster

highRES: a high resolution open-source European electricity system model

Presenter: Muhammad Shahzad Javed.

Description: I aim to introduce our highRES electricity system model, get feedback, and explore collaboration opportunities.

Background: highRES is an open-source electricity system model for EU25+Norway, Switzerland, and the UK. It is co-developed by researchers at the University of Oslo and University College London. This will be our first time presenting our advanced highRES version to the community, integrating socio-environmental dimensions.

Repository: GitHub - highRES-model/highRES-Europe-WF · GitHub

1 Like
OpenMod workshop Freiburg 2026 - Schedule & Program
47

:four: :three:

Dear Organizers, many apologies for the late submission… I submit this through the Form, I thought this was enough. Best regards.

Contribution type: lightning talk + poster

CVXlab: a reproducible workflow for data-driven convex optimisation models

Presenter: Matteo V. Rocco

Description:
CVXlab is an open-source Python framework for building, running, maintaining integrated convex optimisation models. It separates model logic, input data and scenario definitions, combining tabular data management, symbolic expression parsing and CVXPY-based solution. The presentation will show how CVXlab supports a transparent workflow from model specification and data input to scenario execution and reproducible results, using a compact optimisation example.

Background:
Optimisation models used in research often combine mathematical formulation, data processing, solver configuration and scenario management in project-specific code. This can make models difficult to inspect, reuse and reproduce. CVXlab provides a structured workflow based on model settings, data tables and a common optimisation interface. It supports independent and integrated problem solving, Excel or CSV input data, YAML-based settings, and user-defined symbolic extensions. The lightning talk will introduce the rationale and workflow; the poster will provide a more detailed view of the software architecture and an example application.

Optional links:

This version is consistent with the documented CVXPY integration, data management, Excel/CSV inputs, YAML settings, symbolic expressions and integrated solving. CVXlab documentation (https://cvxlab.readthedocs.io/en/dev/changelog.html)

OpenMod workshop Freiburg 2026 - Schedule & Program
48

:four: :four:

late registration/waiting list for talk slot

Contribution type: lightning talk

MOSEK, a solver

Presenter: Gustaf Ehn

Description: MOSEK is a commercial optimization solver, developed by the Danish company Mosek ApS. MOSEK was first launched in 1999 and has been continuously improved ever since. It has long been used in energy applications, with users including Plexos, Gridmatic, and EDF Renewables.

The lightning talk will present MOSEK, its strengths, how to use it successfully in energy applications, and where it fits into the open-source modeling ecosystem.

Background: Optimization is at the heart of energy system modeling. And whilst great open-source solvers exist, the gap to commercial solvers is usually considerable. The choice is often presented as either commercial or open-source. However, reality is much more nuanced than that. With significant variation in different solvers’ functionality, licensing, and pricing. Choosing the right solver or solvers can improve system performance, end-user applicability, and real-world impact.

Optional links:

1 Like
OpenMod workshop Freiburg 2026 - Schedule & Program
49

:four: :five:

late registration and poster • seeking place on waiting list for a lightening talk slot

Contribution type: poster

MGA-Compass: Package for Discovery and Interactive Exploration of Near-Optimal Solutions

Presenter: Sina Hajikazemi

Description: Near-optimal solutions of energy system models are often overlooked, yet they can be more politically feasible, socially acceptable, or environmentally preferable than the strict optimum. I will present an open-source package that enables users to compute near-optimal solutions using a novel algorithm and interactively explore them in real time, even for large-scale linear energy planning models.

Background: This contribution combines two threads of work.

The first thread — presented here for the first time — is a new exact method for finding near-optimal solutions. The method simultaneously constructs an outer approximation and an inner approximation of the near-optimal solution space. At each iteration, the inner approximation expands while the outer one shrinks, until the maximum distance between the two falls below a user-defined threshold. This iterative bracketing guarantees convergence to the true near-optimal space with controllable precision, offering a principled and rigorous alternative to existing sampling-based approaches.

The second is a published IEEE paper introducing a graphical user interface for exploring near-optimal solutions interactively: starting from an initial solution, a user specifies preferences over a small number of properties and continuously traverses toward an alternative solution along an approximated cost-optimal path. After an upfront preparation phase, the exploration performance is independent of the original problem size, making it tractable even for large models.

Optional links:

https://ieeexplore.ieee.org/document/11589682

OpenMod workshop Freiburg 2026 - Schedule & Program
50

:four: :six:

late registration/waiting list for talk slot

Energy transition decision infrastructure: open foundations, governed workflows

Type: poster, plus lightning talk if a slot becomes available

Presenter: [TBC], TransitionZero

Description: Open energy modelling makes assumptions inspectable and alternative pathways testable. Turning that capability into lasting institutional practice requires more than access to software. Decision infrastructure must also preserve how evidence was produced, what changed between scenarios, and how results informed a decision, so analysis can be challenged, updated and reused as circumstances change.

Using Scenario Builder as a reference, we examine how model and scenario versioning, traceable results, controlled collaboration and clear review responsibilities connect open modelling foundations with sustained institutional use.

AI assistance adds a further responsibility: keeping the boundary between model evidence and interpretation. We outline a claim-provenance approach that distinguishes solved results, reproducible calculations, AI interpretations, questions requiring a new simulation, and human judgement. An explanation of an existing result should not be confused with evidence about a scenario that has not been solved.

Background: TransitionZero builds open tools and data for energy transition planning. This extends our OpenMod South Africa 2026 talk and July webinar on decision infrastructure for coordinated planning in sub-Saharan Africa.

• Links:

OpenMod workshop Freiburg 2026 - Schedule & Program
51

:four: :seven:

Contribution type: poster and lightning talk if possible

Comparison of Different Methodologies to Calculate GHG Emissions Related to Grid Electricity Usage

Presenter: Giulia Montanari

Description: Development of a Python package to calculate hourly GHG emissions and environmental impacts associated with electricity withdrawn from local electrical grids. The tool calculates hourly GHG emissions for each bidding zone in Europe using emission factors for energy technologies that are specific to each country’s energy system. The tool can also calculate different environmental impacts for example ecotoxicity, land use and more.

Background: I initially developed this tool to calculate GHG emissions in Northern Italy and applied it to a case study involving municipal buildings. I am now planning to extend its application to calculate the hourly environmental impacts of an eco-industrial park in Denmark (GreenLab Skive).

Optional links: Paper and package are still waiting for publication

1 Like
OpenMod workshop Freiburg 2026 - Schedule & Program
52

:four: :eight:

Contribution type: poster

Nationally Tailored Energy System Models: Design Principles and PyPSA-AT for Austria

Presenter: Max Nutz

Description: PyPSA-AT is an open-source, sector-coupled energy system model that nationally tailors the pan-European PyPSA-Eur framework to Austria, preserving high sub-national spatial resolution while dynamically embedding Austria in its international context. This poster introduces the concept of nationally tailored energy system models, presents PyPSA-AT’s design and validation, and shows through a comparison with the multi-regional PyPSA-Eur that enforcing national targets and policies, such as Austria’s 2040 climate-neutrality target, materially changes modelled renewable deployment.

Background: Planning the transition to a climate-neutral energy system requires models that combine detailed national context with a dynamic representation of international energy interdependencies. Continental-scale, multi-regional models (e.g. PyPSA-Eur) capture international cooperation and cross-border infrastructure well but lack country-specific detail, while purely national models offer high resolution with little dynamic link to the surrounding system. Nationally tailored energy system models bridge this gap: they preserve high sub-national spatial resolution, country-specific policies and local data, while endogenously embedding the focus country within its international surroundings. PyPSA-AT implements this concept for Austria and is validated against Eurostat energy balances via the model-agnostic IAMC data format. Comparing PyPSA-AT with PyPSA-Eur demonstrates that national policy constraints – such as Austria’s 2040 net-zero target – materially alter domestic renewable energy deployment and cross-border trade patterns, underscoring that nationally tailored models are essential for robust energy infrastructure planning and for understanding how national and international policy decisions co-determine transition pathways.

Links

OpenMod workshop Freiburg 2026 - Schedule & Program
53

:four: :nine:

Contribution type: poster

ATLAS: an agent-based model of electricity markets going open-source

Presenter: Florent Cogen

Description: This poster presents the electricity market model ATLAS, its main concepts, and the timeline of its ongoing open-source publication. ATLAS (for Agent-based short Term eLectricity mArkets Simulation) has been developed by RTE, the French TSO, during the last few years. It was used in several studies looking at electricity markets including day-ahead, intraday and balancing processes (mainly activation, with recent development focusing on reserve procurement processes). With ATLAS, we tried to focus on 3 main aspects that are often overlooked in usual market models: integrating forecast uncertainty (both on load and renewable), including operational complexities (technical constraints on units, process constraints such as complex bids, etc.) and modeling each market as a succession of independent steps with specific optimizations. We are now in the middle of a 3 years project that aims at recoding the whole model in Python, and publishing it open-source. Feel free to use the model and contribute to it!

Background: RTE is the French Transmission System Operator (TSO). One of our mission is to ensure the balance between supply and demand in real-time in France, which is heavily influenced by market results and actor behavior.

Optional links: Find the ATLAS repository at:

OpenMod workshop Freiburg 2026 - Schedule & Program
54

:five: :zero:

Contribution type: poster

Reproducing European Electricity Price Stability Estimates: A PyPSA-based Replication and Uncertainty Analysis

Presenter: Nick Herrmann (INATECH, University of Freiburg)

Description: This poster replicates the price estimates of Navia Simon & Diaz Anadon (2025, Nature Energy), who use GenX with 300 stochastic Monte Carlo runs to show that higher renewable shares lower both the mean 2030 European electricity price and its volatility. Using the same open ENTSO-E ERAA input data but PyPSA as an independent open-source tool and a deterministic setup, this work asks whether those price estimates hold under a different, fully open modelling approach and additionally, what makes this kind of study hard to reproduce exactly. The comparison shows that reproducing a stochastic study’s results requires matching not only its input data, but its underlying methodological design as well as full knowledge of the modelling decisions behind it.

Background: Reproducibility is a recognized challenge in energy system modelling: studies rarely share code, and even with open input data, matching a published result exactly depends on choices that are often undocumented.

This case is a practical test of that challenge: it combines the original study’s fully open input data (ENTSO-E ERAA) with an independent, widely-used open-source modelling tool (PyPSA), making an end-to-end replication feasible rather than only a partial or code-level check.

The project was carried out at the University of Freiburg as part of a Master’s program, ahead of the Master’s thesis.

OpenMod workshop Freiburg 2026 - Schedule & Program
55

:five: :one:

Contribution type: poster

Impact of PyPSA-Eur model resolution and segmentation on electricity-gas-hydrogen transmission capacity distribution

Presenter: Vuillaume J-F

Description: Model resolution is a key factor trade-off between the quality of results and the computational time. We are investigating the impact of node number (#node) and the segmentation option (SEG) on the capacity expansion at 2040 horizon for an electricity-gas-hydrogen coupled model.

We use the PyPSA-Eur sectoral modelling software package, version v2025.07.0 and adjust the number of clusters and the sector_opts options. We tested node counts of 39, 50, 150and 250 as well as SEG value of 250, 500, 1000 and 2000 respectively.

The results indicate that there is a positive correlation between total system cost and the SEG, with a ‘plateau’ at around 1000SEG. The 39-node model does not capture internal network capacity expansion potential, 50 nodes appear to be the minimal number but the cases of gas/hydrogen require further analysis. Increasing the number of nodes improves the representation of internal AC and DC capacites and, consequently, optimises of transport costs and volumes while reducing the impact of cross-border capacity.

Background: Pre-study to test JRC HPC optimisation considering computation time, model cluster resolution, and result quality.

Optional links:

OpenMod workshop Freiburg 2026 - Schedule & Program
57

:five: :two:

Contribution type: lightning talk and poster

CCS for Austria’s energy-intensive industries: a brake on rising power demand?

Presenter: Karoline Harrer, Montanuniversität Leoben

Description:
Austria, being a rather small country in population and size, has a very high density of energy intensive industries including iron- and steel, cement, pulp and paper and oil refinery. From today’s industrial energy demand of 134 TWh only 25 TWh are electric; adding biogenic fuels, environmental heat and district heat of 18 TWh leave 91 TWh that is covered by fossils. Decarbonizing these end-uses is difficult due to high capital costs and unfavorable price differentials among electricity, gas, oil, hydrogen, coal, and CO₂. As electricity and its follow-up product hydrogen are assumed to stay scarce, CO2 capture processes are considered a complement for decarbonization of industrial high temperature and process related demands. A CO2 flow methodology is implemented into an energy-scenario model. The scenario comparison of an electrification and a CCS scenario shows, that in both cases today’s total electricity demand of 72 TWh doubles to 140 TWh or more.

Background:
Optimal power flow and network flow simulations for Austria need different scenarios, so the implication of different decarbonization pathways for the energy infrastructure, most importantly the electricity infrastructure, can be investigated.