the journal
Grid & DNO
10 articles on Grid & DNO, for UK installers.
DNO applications by size: the UK guide from G98 to Type D
Which DNO application a solar or battery job needs, by size. G98, the SGI lanes, Form A1-1, the Standard Application Form and Type B, in one table.
G98 explained: solar under 16 A per phase, connect and notify
What EREC G98 covers (3.68 kW single-phase, 11.04 kW three-phase), the 28-day notification, the Connect Direct cut-out photo, and what pushes a home into G99.
SGI-1, SGI-2 and SGI-3: G99 for bigger home inverters
The three Small Generation Installation lanes in EREC G99 Issue 2. Units up to 32 A, set down or export-limited, and when a 5 kW hybrid connects and notifies.
G99 Form A1-1: the short application under 50 kW
Who uses Form A1-1 (under 50 kW three-phase or 17 kW single-phase), what it asks for, the wait, and the Form A3-1 sign-off after commissioning.
50 kW to 1 MW: the G99 Standard Application Form
From 50 kW the short forms stop. What the SAF asks for (Parts 1 to 3), the offer clock, fees, witness tests, HV connections and the sign-off, up to 1 MW.
Above 1 MW: G99 Type B, C and D connections
What changes at 1 MW under G99. The Type B, C and D thresholds, the compliance document, witnessed tests, studies, SCADA and the 5 MW transmission step.
Before the DNO application: operator, MPAN and what counts
The facts every DNO application needs before the form. The operator, the MPAN, phases, the cut-out, batteries, existing solar, set-downs and headroom.
Why Connect Direct escalates your G99 application
On one platform's data, over 60% of Connect Direct applications escalate to manual review and 0.6% auto-approve — what triggers it, and what you control.
Zero export, G100, and why the DNO still said no
An export limiter set to zero does not mean zero network impact. Why the intuition fails, and what to do when the reinforcement quote lands.
Estimated maximum demand and cut-out rating: two guessed fields
What the two DNO application fields actually ask for, why "just use the cut-out rating" is wrong, and what to do when nobody has published the method.