The DNO process (G98/G99)

Any generation or storage connected to the grid in Great Britain needs the distribution network operator’s blessing. The engineering recommendations that govern this are G98, for small installations up to sixteen amps per phase, and G99, for everything larger. Almost every commercial solar or battery project falls under G99, which means a formal application to the DNO and approval before the system connects. This process is not paperwork at the end of a project. It is a gating decision that belongs at the start.

The reason is that the DNO’s answer can change the project. The local network has finite capacity to accept exported power, and where it is constrained the DNO may take longer to assess, attach conditions, require reinforcement, or cap export. Each of those outcomes affects system economics. A project designed before the connection position is known is a design built on an assumption the DNO may not share.

How the process runs

For G98-scale equipment, the installer connects and notifies the DNO afterwards. For G99, the sequence is: application with the technical details of the proposed system, DNO assessment of the network impact, a connection offer with any conditions, then installation, commissioning and, for larger schemes, witness testing before final acceptance. The application is normally made by the installer, but the connection agreement and its conditions belong to the site, and the site owner should read them.

Timeline expectations should be set by size and by network, not by a single rule of thumb. Small commercial systems that fall within a fast-track category can sometimes receive a straightforward approval in a matter of weeks, because the network impact is small enough to assess against standard limits without a full study. Larger systems, and anything approaching the export capacity of the local substation, move into a slower category: the DNO has to model the impact on the network in more detail, which can extend the process to several months, longer again if reinforcement work is triggered. Region matters as much as size. Some parts of the network have absorbed years of distributed solar and battery connections already and have little headroom left, so even a modest project can trigger a constrained assessment there, while the same system elsewhere clears quickly. The only reliable planning assumption is to apply early and build the rest of the project timeline around the DNO’s answer, not the other way round.

Export limits are a design input, not a defeat

Where the network cannot take full export, the DNO may offer a connection with an export limitation scheme, which caps or eliminates what the site can send out. Because self-consumption carries most of the value in a well-sized system, an export-limited connection often costs the business case little, provided the system is sized with the limit in view. The expensive version of this story is the one where the system was sized for export that the connection never permitted.

Batteries go through the same G99 gate as generation, though the assessment looks at a different number. A DNO cares about the export capability the connection agreement will authorise, which for a battery is set by the inverter’s rated output, not the size of the battery in kWh. A large battery behind a small inverter is a small connection as far as the network is concerned. Where solar and a battery share one connection point, the DNO typically assesses them together against the combined export capacity at that point, which is why an export limitation scheme, capping what the combined system can send out, is a common condition on hybrid systems rather than a sign anything has gone wrong. Charging a battery from the grid also counts, because it adds import demand at the site, and the DNO reviews that alongside export when the network is already tight. None of this is a reason to avoid batteries. It is a reason to model the connection with the battery included from the start, rather than adding storage to a design that was approved without it.

Sequencing the project correctly

The defensible order is: model the load, size the system provisionally, apply to the DNO, then finalise design and procurement against the connection offer. Projects that commit capital before the DNO responds are gambling the design on someone else’s network. An independent view across the modelling and the connection position keeps the sequence honest, because no party in it is waiting on hardware margin.

If you are planning a system and want the connection position handled in the right order, book a review at /book or request a benchmark at /benchmark.