Zero export, G100, and why the DNO still said no

An export limiter set to zero, and a bill for thousands

Posted to ElectriciansForums on 7 November 2025:

"I've applied for G99 for 16kW solar PV and battery storage all behind a G100(2) export limitation device set to 0kW (zero export)… DNO has come back asking for a large payment (multiple thousands) for network and supply upgrades as they say this amount of generation will lead to unacceptable voltage rise (over 12%) in MY property due to the length and age of my supply cable… I don't understand. I thought significant voltage rise would only occur when exporting… Is DNO trying to pull a fast one?"

Every part of that is reasonable. The limiter is set to zero. Nothing leaves the property. So what business does the network have objecting, and why is the bill in thousands of pounds?

The last question is the one to be careful with. This post explains why the intuition fails. It does not tell you whether that DNO was right, because nobody reading a forum thread can tell you that.

Where the intuition breaks

"No export, no problem" rests on an assumption: that voltage rise is something that happens out there, on the network, caused by power flowing outwards from the property.

It isn't — or at least, not only. A generator raises the voltage at its own point of connection while it is generating. It does this whether or not any power reaches the network. Self-consumption is not a free pass. The inverter is still injecting current at the property, and that current is flowing against the impedance of the service cable that ties the property to the transformer. A longer run gives it more to push against. The DNO's reply names the length and the age of that cable, which tells you what they were looking at.

Put crudely — and this is crude — the concern being described is not about power escaping past the meter. It is about what the voltage does at the customer's end of a long piece of cable while a 16kW array is working hard, regardless of where the power finally ends up. The inverter is inside the property. The cable it pushes current through is not.

That is the shape of the concern. It is not the calculation.

The part this post will not do for you

The number in that reply — the "over 12%" — is the output of a model. It comes from the DNO's picture of that street: what the cable actually is, how long the run is, what else sits on the same transformer, what worst-case conditions they assume. You cannot check it from the roof. You cannot check it from a forum. You certainly cannot check it from a blog post written by a software company.

So there is no formula here, no tolerance band, and no worked example. Be wary of anyone who offers one for a network they have never modelled. The right response to a reinforcement quote is to ask the DNO for the assessment behind it — not to assume bad faith, and not to press on regardless.

Zero export is a commercial decision, not a permission slip

Setting the limiter to zero decides what happens to the electricity. It does not decide whether the installation is connected to the distribution network. It plainly is. The customer's board is tied to the same transformer as the neighbours', through the same service cable, and the generator sits at the far end of it.

This gets treated as an administrative nicety. It isn't. The Only Panels installer podcast, in an episode published on 24 June 2026 under the title "Are DNOs Forcing Installers to Break the Law?", flags:

"the massive liability and potential criminal offense of installing grid-tied 12kW systems with export limiters without prior DNO permission"

That is the episode's own wording and its own characterisation. It is not legal advice from us, and we are not going to quote you a statute we have not read. Take it as a flag. Installers do postpone or skip the application on zero-export jobs, on exactly the reasoning in the forum post above, and the people who talk to installers for a living think that is a serious exposure.

The application is a project schedule item

Underneath the engineering problem is a commercial one. Reinforcement costs land after the customer has signed. The survey is done, the deposit is taken, the kit may be on order — and then a letter arrives with a number on it that nobody priced into the job. Now the choice is: absorb it, go back to a customer who thought the price was the price, or unwind the deal and hand back a deposit you have already spent time against.

The wait makes it worse. The same podcast episode describes "the crippling 45-working-day wait times", and calls out "the DNO double standard: why heat pump approvals are highly automated and practically instant, while solar PV applications are stuck in the dark ages."

Nor do the applications sail through. Renbee, who sell DNO application software, report the following about applications submitted through Connect Direct — the shared connection-application route run by the Energy Networks Association:

Those are their numbers, from a company with a commercial interest in the problem being real. Even discounted, they say something you can act on. The DNO application is a scheduling dependency with a long and highly variable tail. It belongs on the programme next to scaffolding, not in the pile of paperwork you do once the deposit clears. The same job in a UKPN area and a Last Mile area is not the same job.

When the reinforcement quote arrives

Four things, in order.

Ask for the assessment. Request the reasoning behind the figure. You are being asked to pay to reinforce a connection on the strength of a model; you are entitled to understand what that model was told. If the answer comes back thin, that is information too.

Check whether the design changes the answer. A lower generation capacity, a different connection arrangement — these may change what the DNO's model returns. Ask the question explicitly rather than treating the first quote as the only available outcome.

Get an engineer on it, not a thread. If you want the arithmetic checked, whoever checks it needs the network data and the competence to use it. That is not a supplier's sales engineer, and it is not us.

Talk to the customer before, not after. A zero-export job on a long, old service cable is a job with a known unknown in it. Say so at survey. A customer told at quotation that the DNO may require reinforcement has a decision to make. The same customer told after paying a deposit has a complaint.

Then be honest about what you have and have not established. The forum poster asked whether his DNO was pulling a fast one. We do not know, and neither did anyone who answered him. What can be said is that his reasoning — no export, therefore no network impact — does not hold, and that the thing his DNO described is a real category of concern rather than an invented one. Whether it applies at the magnitude claimed, on his street, on his cable, is a question with exactly one competent audience.

This is the post that tells you why the intuition was wrong and who to ask. It is not the post that does the sums. Anyone doing the sums for you, for a network they have never modelled, is guessing.