LF Energy Summit Europe 2026

openSTEF is launching a major new version at LF Energy this year: 4.0. In this presentation, co-hosted by SIA and Sigholm, we'll cover:

1. What and for whom is openSTEF, and what's new in 4.0 (me)
2. Creating the 4.0 documentation through genAI pipeline built in AWS (SIA)
3. Experiences from a community member who switched from 3.0 to 4.0 (Sigholm)

For reliable and cost-efficient distribution grid operation and redispatch management, accurate short-term forecasting is a core prerequisite. At E.ON, we originally relied on vendor-built “black box” forecasting solutions and experienced the familiar trade-offs: slow iteration cycles, limited transparency, and dependence on external roadmaps. In this talk, I’ll share our journey to an open-source target picture for transformer load forecasting in a redispatch context. After a quick intro to the redispatch problem and our system setup, I’ll walk through our OpenSTEF-based approach. I’ll then present results from a real benchmark against an external model and the practical lessons learned; what mattered, what didn’t, and where open source helped us move faster. Finally, I’ll cover the organizational path: how we navigated compliance and governance with our Open Source Office to turn from user into contributor.

Redispatch volumes have increased enormously over the past year, creating new challenges for existing technologies, tools, and processes. These challenges become particularly visible in settlement and billing, which sit at the end of a long chain of operational and data-processing steps.

The use of multiple commercial tools from different vendors can make data access, process transparency, and adaptation to changing requirements difficult. To address this, Bayernwerk has started using open-source tools to improve data transparency and support error detection, correction, and analysis in the Redispatch 2.0 process.

The initial experiment was successful and, beyond the original use case, has attracted new users across the organization. Making data more openly available internally has enabled new applications across departments and contributed to a broader shift in how data is used.

The next challenge is to scale and industrialize these tools to meet growing internal demand, while integrating more advanced features where LF Energy and other open source tools can play an increasingly relevant role.

LF Energy has a growing ecosystem of open-source projects for forecasting, digital substations, data-driven operation, and more. However, for many users, one question remains difficult: how do these projects fit together in a realistic digital grid workflow?
Over the past year, we explored many LF Energy projects and analyzed how they can work as parts of a larger grid ecosystem. Based on this, we're building a miniature city-grid prototype in our research lab that connects different projects across a shared operational chain. For example, one workflow starts with load and renewable forecasting, uses grid-model calculations to assess network impact, and surfaces the resulting event to an operator workflow.
The session will include a demo walkthrough using recorded lab segments, architecture diagrams, and data-flow examples. It will show how projects such as OpenSTEF, SEAPATH, PowSyBl, OpenGridFM, and OperatorFabric can fit within the same architecture. Attendees understand where to start, what role each project can play individually, and how they can be connected to solve practical digital grid problems. The session will also highlight integration gaps for community discussion.

PowSyBl runs load flow in production at TSOs; its short-circuit side is younger. We do not need a perfect fault engine to make the next leap. Treat PowSyBl as a *grid engine* — it gives the short-circuit power Scc, the fault current Icc, line parameters and source strength, the boundary conditions a protection study needs since relays are set on Icc min and max. Feed those into the *real* protection, virtualized on a SEAPATH base, close it into a software loop over one zone, and you test selectivity the way software teams test code: deterministically, repeatably, in the open.

Modern power systems are undergoing a structural transformation that planning tools were never designed to handle. System operators today manage this complexity with tools built for a different era, decisions made on intuition shaped by experience rather than probabilistic foresight.
The Grid Enhancement Platform (GEP) solves this software architecture challenge as an open-source reference architecture. Built on a modular, loosely coupled stack of four mature projects LF Energy’s PowSyBl, Power Grid Model, OpenSTEF and Pandapower. At day-ahead clearing, GEP ingests the cleared positions and runs a forward-looking network state simulation. Using the probabilistic outputs at the asset level, propagated upward through aggregation engine and evaluated against the network model in Power Grid Model and PowSyBl, the platform generates a time-series of nodal state distributions. At the aggregation level, flexibility assets-batteries, demand response contracts are modelled as degrees of freedom that can be co-optimised against both the energy market objective and the network constraint objective simultaneously. The result is a genuinely localised, network-aware dispatch signal .

We need the capability to thermally model all components in the chain of assets. The global energy transition is rapidly reshaping electricity systems, driven by decarbonization, electrification, and the integration of renewables. These developments introduce new operational challenges for grid operators, including increased congestion and the need to maximize existing grid capacity while ensuring reliability.

Building on our Transformer Thermal Model, we add two new members to the family of thermal models: Cable Thermal Model (CTM) and Switchgear Thermal Model (SGTM). These models estimate asset temperatures based on load, ambient conditions, and technical specifications. Together, they allow operators to identify bottlenecks, assess hidden capacity, and mitigate congestion risks. As this now enables operators to model most primary components in the grid, the applicability and value increase exponentially!

In this talk, we demonstrate how we apply these models in practice across multiple asset types, enabling higher asset utilization, optimal grid planning, improved monitoring, and more efficient system operation.

Understanding, simulating and predicting the behavior of overhead lines and their supporting structures (including electrical towers) is essential for securing grid maintenance and extending asset life. Phlowers provides tools to achieve this in a simple way, despite the complexity of the underlying algorithms.

Phlowers is an Open Source ecosystem that brings mechanical and thermal modeling of overhead lines to engineers, researchers and operators, including those working offline. In the future, it will integrate the modeling of supporting structures (pylons/towers) to simulate mechanical constraints and interactions within the grid.

The ecosystem currently includes Python libraries for simulations and an offline web application called Stellar. It enables users to customize models, integrate their own data and adapt tools to operational needs. It will also support large-scale batch computations to assess maintenance policies, simulate multiple scenarios and optimize long-term asset management strategies.

In this session, we will explore its architecture, use cases and contribution to innovation in grid asset management.