Ofgem approves structured network model exchange

Ofgem has approved Grid Code modification GC0139, introducing a structured two-way exchange of planning models between the National Energy System Operator and electricity network operators.

Taking effect on 6 August 2026, the modification replaces relevant elements of existing annual data submissions with routine exchange of solved power-system models in the Common Information Model format. Data Registration Code schedules and associated documentation will define the information and processes used by participating organisations.

Electricity North West raised the proposal in February 2020 to address limitations in planning-data exchange across transmission and distribution boundaries. Following industry development and consultation, the Grid Code Review Panel unanimously recommended implementation before Ofgem issued its final direction.

Existing arrangements can leave organisations working with incomplete representations of adjacent networks or datasets assembled at different times. GC0139 is intended to provide NESO and network operators with a more consistent electrical model for power-flow, fault-level, voltage, and other planning studies.

A solved model contains network topology, asset parameters, generation and demand assumptions, and a calculated operating state. Exchanging that model rather than a collection of disconnected fields can reduce interpretation differences, provided the effective date, scenario, boundary conditions, and modelling assumptions remain clear.

The Common Information Model provides a standardised structure for representing electricity assets and their relationships. Interoperability between planning platforms should improve, although implementation still requires disciplined naming conventions, boundary management, version control, and validation.

Data quality becomes an engineering dependency

Distribution networks now contain increasing volumes of generation, storage, electric-vehicle charging, and flexible demand, whose effects extend beyond local circuits. Active and reactive power flows, fault contribution, voltage behaviour, and network constraints can cross grid-supply points and influence the transmission system.

Changes to the transmission network similarly alter distribution planning assumptions. Boundary voltage, fault level, topology, reinforcement, and planned outages affect the conditions against which distribution assets and customer connections are assessed, so delayed or inconsistent information can produce divergent study results.

The National Grid DSO ten-year development plan provides a forward view of distribution requirements, while GC0139 establishes a route for aligning such work with the transmission models used by NESO and neighbouring operators.

Implementation will require considerably more than transferring files. Each organisation must establish responsibility for correcting errors, resolving duplicated equipment, reconciling connectivity, and confirming that exchanged models produce credible results. Automated validation can identify missing fields or impossible values, but engineering review remains necessary.

Detailed power-system models also contain sensitive infrastructure information, bringing cybersecurity and information governance into the exchange process. Secure transfer, access control, retention policies, audit trails, and incident procedures must protect the data without making routine use unnecessarily cumbersome.

Model timing presents another difficulty because networks change continually through outages, new connections, asset replacement, and reconfiguration. Planning studies do not require a real-time representation, but each exchanged dataset must have a defined effective date and a clear record of the operating assumptions applied.

Other distribution digitalisation programmes are moving in the same direction, including UK Power Networks’ open-source Python package for network data. Standardised, machine-readable information is becoming increasingly important for planning, flexibility assessment, and automated analysis.

Improved models cannot remove physical constraints or shorten equipment manufacturing times, although they can reduce repeated data requests and improve the order in which reinforcement, flexibility, and operational options are assessed. Consistent network representations should also make study results easier to reproduce and challenge.

After more than six years of code development, the focus now moves to model quality, system compatibility, and routine operational discipline. The modification will be judged by whether planning studies become more consistent and whether network changes are reflected across organisational boundaries quickly enough to support investment decisions.