Digital Substation is an engineering publication focused on IEC 61850 automation, protection systems, and digital grid intelligence.
A new and important mechanism for validating SCL files is emerging in the IEC 61850 engineering process: OCL (Object Constraint Language), a language for formally describing constraints and rules for the objects of an information model.
A report from the tutorial of CIGRE Study Committee B5 at the 2026 Paris session. Three renowned experts built their talks into a single historical sequence: process bus → centralized protection (CPC) → virtualization. The protection function is becoming a software application, and the server a computing platform that can be replaced.
A report from the exhibition floor of CIGRE 2026 in Paris. The main theme for us was the virtualisation of protection — and it is no longer a single experimental product: we met it at the stands of ABB, Siemens, Moxa, Welotec, Kalkitech and others. Some offer virtualised protection functions (ABB SSC600 SW, Siemens SIPROTEC V, Kalkitech VPR), others the substation servers to run them on (Moxa DA-920E, Welotec RSAPC), others still assemble multi-vendor vPAC architectures on the open SEAPATH platform. The key question has shifted: no longer «can protection run in a virtual machine», but who defines the hardware architecture and answers for the reliability of the computing platform. Plus a contrast: while the secondary systems move towards a software-defined substation, Arteche's optical current transformers never went mainstream, and Hopf showed a GNSS-antenna signal carried over fibre for 30 km.
One run of Tekvel Magic — and the device chosen from the SCL comes alive: an MMS server is started from its data model, and a control panel opens in the browser. You change signal values and quality — and the upper-level client receives real dchg/qchg reports, exactly as from a real IED. The module works without a single physical device on the bench: an SCL file is enough. The fourth case in the series tackles the task from the other side of the table: for the first time Magic does not watch the traffic but generates it — it becomes an IEC 61850 server to test how the receiving side (SCADA, telecontrol server) receives and displays reports, including invalid quality, test data and rare scenarios. Plus an auto-sweep of values and quality for bulk testing in one click.
The Yamal LNG plant is powered by an islanded on-site power station rated at 376 MW and built around eight gas-turbine units; its distribution network spans 110, 35, 10 and 0.4 kV and more than five hundred feeders. Islanded operation places tough demands on stability, short-circuit and protection-selectivity calculations — and therefore on how up-to-date the digital model is. The article describes a cycle that automatically updates a DIgSILENT PowerFactory network model from DMS telemetry: the load-flow calculation runs as a single process in external-control (Engine) mode, without opening the GUI, and starts automatically whenever a new export appears. All control, power-balance diagnostics and analysis of the accumulated case history live in a Windows notification-area icon. The solution was built in-house and, in operational effect, is comparable to commercial real-time modules whose licences for a power system of this scale run into tens of millions of roubles.
In digital-substation projects most attention goes to GOOSE, SV, PTP and the network. But there is another important layer — MMS exchange and IEC 61850 reports, through which SCADA, HMI, gateways and monitoring systems receive telemetry from IEDs. In practice the problems here are deceptively simple: the MMS connection is up, the device is reachable, the client «sees» the server — yet data does not update, some signals disappear, events are duplicated, or the report control block does not activate at all. This article covers the sharpest problems with report control blocks (RptEna, ConfRev, a block already taken by another client, TrgOps, a DataSet mismatch, duplication from the buffer via entryID and too much faith in the SCL file) and a practical principle: cross-check the SCL file, the device's actual MMS model and a PCAP capture together.
On a digital substation, the end devices kept reporting a communication fault that just as suddenly cleared itself, and the cause stayed unclear — the alarm was a shared one. The only source of information was a pcap file about 11 minutes long — almost 186,000 GOOSE packets and more than 200 sources. A manual analysis in Wireshark at that scale would have taken hours. We show how, in a few minutes, the Tekvel Magic software turned the network capture into a structured Excel report of an unscheduled inspection.
Precise time is a mandatory part of a modern substation: it is needed for event recording and oscillography, for Sampled Values, for phasor measurements and for other applications. But relying on a local receiver alone is becoming increasingly dangerous: jamming, interference and spoofing turn a single satellite source into a systemic risk. An architecture built around an ePRTC (enhanced Primary Reference Time Clock) in the telecom network and a TDG (Time Distribution Gateway) on site offers a heterogeneous approach: one precise time comes «from the sky», another «via the ground». The article — what ePRTC and TDG are, how they fit IEC 61850-9-3 PTP, and why a single redundant GNSS receiver is not enough.
"Shall support IEC 61850" is too general a wording. We look at where to check IEC 61850 conformance certificates today, what a conformance certificate really confirms, how a registry of 1,650 test records is structured by editions and roles, the seven parameters to check in a certificate, and how conformance testing differs from interoperability testing — and how to phrase a requirement in a specification instead of a vague phrase about "standard support".
GOOSE is convenient within the substation but is not routable. R-GOOSE and R-SV extend the IEC 61850 publisher/subscriber exchange onto routable IP networks and open up scenarios for inter-site functions — direct trip, special protection schemes, synchrophasor measurements, etc. We look at the L2 vs L3 difference, the requirements for latency and quality of service (QoS), the model of secure group communication (Security Group, KDC, PKI), the PUSH / PULL key delivery modes and the KDA mechanism.
In a pcap from a digital substation attention usually goes to GOOSE, SV, MMS and PTP. But useful diagnostics often live lower — in the MAC addresses. The article explains what an OUI is and how MA-L, MA-M, MA-S and CID differ, how to guess a vendor from the first octets, why a shared Source MAC on two IEDs breaks MMS and triggers MAC flapping, and what a digital-substation engineer should check on a first pass over a pcap.
CIGRE's first dedicated framework for virtualised IEDs will not publish until Q1 2028 — but the vPAC pilots that have to align to it are being commissioned now. Its scope is already public, through the working group's Terms of Reference and a framework article refeatured in ELECTRA 345 (April 2026): what is mandatory, what is deliberately out of scope, and what every 2026 vPAC pilot should already be doing in response.
What IEC 61869-9 actually requires, and where the boundary lies between "still synchronized" and "no longer synchronized". The article unpacks the "2 → 1 → 0" misconception about SAMU/NCIT behavior on PTP loss, the link between clockClass, timeTraceable and SmpSynch, and how a complex protection function should react when one of its source streams loses synchronization.