A Case Study on Testing LFE Seapath at National Grid Electricity Transmission
Event Recap: LF Energy Summit Europe 2025
TL;DR
At LF Energy Summit Europe 2025, Thomas Charton (National Grid Electricity Transmission) presented a proof of concept conducted with GE to evaluate LFE Seapath for hosting virtualized protection applications. The work examined deterministic real time performance, sampled value processing, system architecture, and scaling the platform for substations. In lab testing across more than 44,000 repetitions, the average round trip time was about 5.5 milliseconds, with no results exceeding 6.6 milliseconds.
Presentation Overview
Charton introduced National Grid Electricity Transmission as the transmission owner responsible for the electricity transmission network in England and Wales. The organization owns the network but does not operate it, as operation is handled by a separate national system operator.
The network consists primarily of 400 kV infrastructure with some 275 kV lines. Charton emphasized that availability is the primary design objective guiding system design and operational practices.
Protecting and controlling this infrastructure requires a large amount of secondary equipment. National Grid manages approximately 5,000 bays and around 15,000 bay devices across its substations. These systems support functions including protection of the network, control and automation, settlement metering, system monitoring such as PMU data, communications infrastructure, cyber security, and asset health monitoring.
Charton explained that traditional protection, automation, and control architectures rely on many devices performing relatively small functional groups. Substations typically include separate machines for gateway functions, monitoring, human machine interfaces, and automation controllers, along with protection IEDs deployed within each bay.
Virtualization Drivers and System Architecture
National Grid has already begun virtualizing some substation functions. In the current design approach, redundant servers host a hypervisor running several virtual machines. These virtual machines support functions such as gateway services, tooling, monitoring systems, and security domain control.
Protection systems, however, remain implemented as physical devices at the bay level.
Charton described several drivers for exploring further virtualization. A typical substation may contain around one hundred different types of devices, ranging from legacy electromechanical equipment to modern digital relays. These devices often remain in service for around twenty years or longer.
Virtualization is expected to reduce operating costs by consolidating functions and simplifying asset management. Charton also pointed to the scale of growth required to meet net zero commitments by 2050, noting that existing supply chains and internal capability cannot grow at that rate without changes in how systems are delivered and engineered.
Architecture and Testing
The proof of concept used a hardware cluster of three machines running Seapath. The system used a KVM hypervisor with QEMU for hardware emulation and virtual machines hosting application workloads. The platform build process was managed using Ansible.
Sampled value processing was identified as the primary performance bottleneck. To address this, PCI passthrough was used so sampled values could be delivered directly to the virtual machine responsible for processing them, bypassing the virtual switch.
Within the virtual machines, Docker containers implemented the sampled value processing and protection functions. Communication between containers used messaging technologies including ZeroMQ.
For testing, an Omicron device generated analog signals representing prefault and fault conditions. Merging units converted the signals into sampled values and published them onto the process bus. The protection Docker application detected the overcurrent condition and issued a trip signal.
Across more than 44,000 test repetitions, the average round trip time measured about 5.5 milliseconds, with none exceeding 6.6 milliseconds. Charton explained that the total time was roughly divided between the merging unit, the Seapath platform, and the return path back to the merging unit.
Standards Context and Future Work
Charton also discussed the implications of virtualization for standards development. He noted that virtualization will need to be reflected in IEC standards, including work related to the definition of virtual IEDs and updates to engineering processes.
He also referenced the need for additional logical nodes and continued work on engineering approaches that support virtualized systems.
National Grid plans to continue evaluating hardware platforms and clustering approaches. Charton stated that the organization intends to deploy a pilot scheme using this architecture on the network before 2030.
FAQ
What was evaluated in the proof of concept?
The project evaluated whether LFE Seapath could host virtualized protection applications while maintaining deterministic real time performance required for protection systems.
What architecture was used for testing?
The system used a hardware cluster of three machines running Seapath with a KVM hypervisor and QEMU, with virtual machines hosting Docker containers implementing protection and sampled value processing.
How was performance validated?
Lab testing used an Omicron device to generate analog signals representing fault conditions. Merging units published sampled values to the process bus, and the protection Docker application issued trip signals that were measured for round trip timing.
What performance results were observed?
Across more than 44,000 repetitions, the average round trip time was about 5.5 milliseconds and none exceeded 6.6 milliseconds.
What are the next steps?
National Grid plans further evaluation of hardware platforms and clustering approaches and aims to deploy a pilot scheme using this architecture before 2030.
About LF Energy
LF Energy is an open source foundation within the Linux Foundation focused on advancing collaboration in digital energy infrastructure.
Learn more: https://lfenergy.org
Last updated: March, 2026
