Every regional installer has a story about a domestic battery job that sat waiting eight weeks for a piece of paperwork, or a 45kW commercial roof that got quoted, designed, funded — and then stalled because nobody checked the export capacity on the local substation first. The paperwork in question is a DNO (Distribution Network Operator) application, and whether it’s a same-day G98 notification or a multi-month G99 application decides more about your installation timeline than the panels, the inverter, or the scaffolding ever will.
For anyone quoting solar in the UK in 2026 — installer, investor, or commercial energy manager — understanding where the G98/G99 threshold sits, what determines your export limit, and why the DNO relationship matters more than most homeowners realise is the difference between a six-week install and a six-month one.
What G98 and G99 actually are
G98 and G99 are Energy Networks Association engineering recommendations that govern how small-scale generation — solar PV, batteries, wind, CHP — connects to the local electricity distribution network. They replaced the older G83/G59 framework and exist for one reason: DNOs have to keep the local grid stable, and every inverter pushing power backwards into a substation is a variable they need to account for.
G98 — notify, don’t ask. This covers the overwhelming majority of domestic and small commercial installs: single-phase systems up to 16A per phase (roughly 3.68kW) and three-phase systems up to 16A per phase (around 11kW). If your system sits inside those limits, your installer submits a notification to the DNO after installation — typically within 28 days — confirming the kit is fitted, G98-compliant, and connected. There’s no approval gate. No waiting. This is why a standard 4kW residential solar-plus-battery job can go from signed contract to switched-on in under a month, install crews permitting.
G99 — ask first, then wait. Anything larger — most notably any three-phase installation above 16A per phase, so above roughly 11kW, plus anything with an aggregated capacity at a site that tips over the G98 ceiling — needs a full G99 application before work starts. The DNO has to assess the impact on their network: fault levels, voltage rise, thermal limits on the local transformer and feeder cable, and whether other customers on the same circuit are affected. Only once they’ve approved the connection (or offered a modified one) can the installer proceed with confidence that what gets built will actually be allowed to export.
This is the crux of the keyword behind this piece — a G99 application for solar isn’t a formality, it’s a technical negotiation with the network operator, and it needs to start before you order equipment, not after.
Why the threshold catches people out
The 16A-per-phase line looks simple until you’re standing in front of a real building. A few scenarios that trip up specifiers who assume “domestic = G98, commercial = G99”:
- A large detached house with three-phase supply and a big EV charger plus battery plus solar can breach G98 limits even though it “feels” residential.
- A small industrial unit under solarpanelsforindustrialunits.co.uk’s typical client profile — say a 30-40kW rooftop array — is almost always G99 territory, and the roof size alone doesn’t tell you that; it’s the export capacity relative to the existing supply that matters.
- Phased projects catch people too: install 10kW now under G98, then add another 8kW next year, and the aggregate capacity at that connection point may now require retrospective G99 treatment. DNOs increasingly ask for full site capacity up front specifically to stop this kind of salami-slicing.
- Export limitation devices (ELDs) can sometimes keep a system under G99 review by capping what’s actually pushed to the grid, even if the installed DC capacity is larger — but that’s a design decision that has to be made and documented at quote stage, not discovered later.
None of this is exotic engineering. It’s paperwork classification. But get it wrong and you’ve either delayed a straightforward domestic job with unnecessary DNO correspondence, or worse, connected an unapproved G99-scale system that the DNO can legally require you to disconnect and redo.
Export limits: the number that actually matters
Homeowners and even some commercial clients fixate on system size in kW — understandably, since that’s what drives the quote. But the number the DNO cares about is the export capacity: how much power can flow back into their network at any moment, not how much the panels can generate.
This is where export limitation earns its keep. A site that would otherwise need a full G99 network reinforcement study — sometimes with a bill for upgrading the local transformer running into five figures — can instead install a larger array, cap the export electronically at an agreed ceiling, and self-consume the rest or store it in a battery. It’s a common enough pattern on commercial jobs discussed on hubs like commercialsolarpanelsinstallation.co.uk and increasingly baked into car park and canopy schemes profiled at solarpanelsforcarparks.co.uk, where roof or canopy area supports a much bigger array than the existing grid connection could ever export unconstrained.
The commercial reality: a 100kWp rooftop system might only be granted, say, 60kW of export headroom by the DNO given the substation it sits behind. The other 40kWp isn’t wasted — it’s consumed on-site, covering daytime industrial or refrigeration load, with the export limiter simply throttling anything above the agreed ceiling at the point of connection. Getting this number wrong at quote stage is one of the most common reasons commercial solar projects go back to the drawing board after a DNO query — and it’s exactly the kind of detail that should be modelled before a client signs, not after, which is why tools like the calculator on businesssolarcalculator.co.uk are useful for sanity-checking payback assumptions against a realistic, DNO-constrained export figure rather than nameplate capacity.
Why the DNO relationship is a regional installer’s real advantage
This is the part that rarely makes it into a sales pitch but decides real-world timelines: DNOs are regional, and so is the relationship that speeds up (or stalls) a G99 application.
The UK has six DNO groups covering fourteen licence areas — UK Power Networks, National Grid Electricity Distribution (the old Western Power Distribution patch), SP Energy Networks, Scottish and Southern Electricity Networks, Electricity North West, and Northern Powergrid. Each runs its own application portal, its own queue, its own engineering team, and — critically — its own turnaround culture. An installer who submits G99 applications to the same DNO every week knows:
- which submission format that DNO’s engineers actually want, versus what the generic template implies
- realistic turnaround times for that specific region (official target windows exist, but actual practice varies by DNO and by how loaded the local network already is)
- which substations in their patch are already near capacity, so they can flag a likely reinforcement requirement to the client at quote stage instead of six weeks into the application
- who to call when an application stalls in the queue, rather than starting from a cold generic enquiry
This is a genuine structural advantage for installers who work a defined patch rather than nationally. A firm like greenlincrenewances working consistently across Lincolnshire, or ecoaim.co.uk covering Central Scotland’s SP Energy Networks and SSEN boundary areas, builds up exactly this kind of institutional knowledge of “their” DNO that a national installer parachuting in for a single job simply can’t match. The same applies to firms like midland-solar.co.uk working the West Midlands DNO patch, or ececoenergy.com handling commercial applications across Essex and East Anglia’s UK Power Networks territory — repetition with the same regional engineering team compounds into faster, cleaner applications over time.
For homeowners this mostly shows up as “my installer said four weeks and it took four weeks” versus open-ended uncertainty. For commercial clients weighing a G99 project against a finance timeline — relevant to anyone reading proposals via commercialsolarfinance.co.uk or structuring a PPA through solarpowerpurchaseagreements.co.uk — the DNO application timeline is frequently the single largest source of schedule risk in the entire project, bigger than equipment lead times or planning permission in most cases.
What a realistic G99 timeline looks like
There’s no single published turnaround guaranteed across all fourteen licence areas, and DNOs are transparent that timelines depend on network headroom in that specific area — a rural feeder near capacity behaves very differently from an urban substation with spare headroom. As a rough planning assumption, commercial developers should budget for the application-and-approval stage alone to run from several weeks up to a few months for straightforward connections, and considerably longer if the DNO’s assessment flags a need for network reinforcement work, which then involves its own quote, timeline, and sometimes a financial contribution from the developer.
The practical implication: G99 applications need to be submitted at the start of a project’s timeline, run in parallel with detailed design and procurement, not after. Installers who wait for full technical drawings before submitting are adding weeks of pure queue time that could have been running concurrently with everything else.
MCS, SEG, and why compliance still comes first
It’s worth restating a point that gets lost in DNO conversations: none of this replaces MCS certification. Any installation intending to claim Smart Export Guarantee payments still needs an MCS-certified installer and MCS-compliant equipment — the G98/G99 process governs the physical grid connection, MCS governs scheme eligibility. Export rates under SEG vary by supplier rather than being fixed nationally, so it’s worth comparing tariffs once the DNO-approved export limit is known, since that capped figure — not the panel nameplate rating — is what actually gets paid for over the system’s life.
For homeowners the practical sequence installers like solent-solar.com or hazellelectrical.co.uk will typically run is: confirm phase supply and existing load, size the system against G98/G99 thresholds, submit whichever DNO process applies, then move to install once either the G98 notification window is understood or G99 approval is in hand. Getting that order right at the quoting stage — rather than discovering the classification mid-install — is what keeps a project on the timeline the client was originally quoted. For a broader read on how installation costs and specification choices interact with this process, thecostofsolar.co.uk’s commercial solar panel cost guide is a useful companion reference alongside this DNO breakdown.
The practical takeaway
G98 versus G99 isn’t a technicality buried in an installer’s back office — it’s the single biggest schedule variable in any solar project above domestic scale, and increasingly relevant to larger domestic jobs too as three-phase supplies, EV chargers, and batteries push aggregate capacity past the 16A threshold. Anyone speccing a system should ask their installer, at quote stage, which category the project falls into, what export limit is realistic given the local network, and how that installer’s DNO relationship in their specific region is likely to affect the timeline. The firms that answer that clearly and specifically — rather than defaulting to “we’ll sort the DNO stuff” — are the ones who’ve actually done it enough times in that patch to know.