TenneT Netherlands deploys ReFlow for faster grid checks

TenneT Netherlands has deployed ReFlow, an in-house grid analysis platform built on the open source framework PowSyBl. The system has delivered at least a tenfold improvement in performance across standard grid security processes.

The platform is used by the operator of the Dutch high-voltage grid, where control room teams run continuous security studies, congestion management analysis and capacity calculations to support round-the-clock operations.

These workloads had been growing faster than previous tools could handle, as larger datasets, higher time resolutions and stricter regulatory requirements increased pressure on existing systems. One European intraday coordinated security analysis process was taking longer than the required time resolution, limiting its value for operational decision-making.

The new system combines PowSyBl, an open source framework for grid modelling and simulation, with ReFlow, an internal orchestration layer developed by TenneT to run large numbers of calculations in parallel. ReFlow takes standardised CIM/CGMES 3.0 inputs, including domestic and foreign network models and operational forecasts, then automatically launches the necessary calculations and closes them when the work is complete.

Operational pressure

Congestion has become a central issue for the Dutch grid, which TenneT describes as a major direct cost and lost opportunity for society. The operator is addressing the problem through new congestion management products, market design changes, network expansion and better use of existing assets, including weather-based dynamic line rating and machine learning-led topology optimisation.

The value of those measures increases when they are brought into a single security assessment framework. In practice, this allows operators to assess whether an action that solves one problem, such as an oscillation, could create another constraint elsewhere on the network.

More advanced integrated assessments could also change how outages are planned. Rather than only checking whether a planned outage is safe under expected conditions, the system could help identify the best future time windows for outages, improving operational flexibility and asset use.

Open source choice

PowSyBl was originally developed by French transmission operator RTE and is hosted by LF Energy. It is already used in production by RTE and by two regional coordination centres, BalticRCC and Coreso, for functions including European merging, coordinated security analysis and capacity calculation.

That existing operational use appears to have helped TenneT avoid building a new analysis engine from scratch. Instead, it focused on creating a system around PowSyBl that could meet its own throughput and automation requirements at national transmission system operator scale.

"We consider PowSyBl the most advanced open source power-system analysis engine available," said Hugo Pfister, manager of grid security applications at TenneT Netherlands. "We benefit directly from years of investment by RTE and the broader LF Energy community."

Pfister said ReFlow was designed with a modular structure to avoid dependence on a single supplier or analysis engine. While PowSyBl is the main engine for static security assessment, the architecture is intended to accommodate other engines for functions such as dynamic security assessment and short-circuit analysis.

"By embedding PowSyBl within ReFlow, we can use it in a highly automated, mission-critical context," Pfister said. "ReFlow is built on a 'single deployment, multi-customer' principle. We build once, deploy once, and scale on demand."

Five-month build

ReFlow moved from project start to first production release in five months. TenneT credited that pace to two factors: the availability of a modern on-premises platform environment based on Kubernetes, Kafka and S3-compatible storage, and a development team of software engineers embedded in the business with direct knowledge of grid security processes.

The main difficulty lay in aligning data producers and consumers on common integration patterns and standardised data products across several domains. That work took time but created a stronger foundation for scaling the system further.

TenneT also argued that the open source model was central to the project rather than incidental. In Pfister's view, using open source software reduced procurement friction while giving the operator greater freedom to adapt and extend tools for its own purposes.

"The open source model allowed us to adopt a state-of-the-art solution without a lengthy procurement process," Pfister said. "It also gives us the flexibility to adapt and extend the tooling to our specific needs. More importantly, open source promotes knowledge sharing within the industry and makes high-quality software accessible to smaller TSOs, which strengthens the collective digital capabilities needed for the energy transition."

Pfister said the approach has implications beyond software sourcing. "Beyond its technical benefits, the open source model acts as a catalyst, enabling engineers and specialists to move beyond technology constraints and think in terms of process opportunities."