Guidance for plug in solar for Electricians

The Department for Energy Security and Net Zero (2026) have produced a product specification (Plug-in Solar Device Interim Product Specification Version 2.0, July 2026) which outlines the requirements for the PV panels and the respective electrical installation. It's probably a good starting place to have a review of that document.

https://assets.publishing.service.gov.uk/media/6a60a03c47af652afa0c89f5/plug-in-solar-final-_interim-product-specification.pdf for those that haven't stumbled on it already.

   - Andy.

 Page 25: Anti-islanding and termination of power supply
An external resistance of greater than 2000 Ω must be applied across the plug pins of the
mains plug within 40 s after disconnection - this delay allows the inverter to detect loss of
mains supply. 

Is that really 40 seconds, or should it be 40ms?

That does not make sense. Why a resistor? Surely with a loss of supply the anti islanding should kick in and turn off the inverter? 

JP

When I was at college doing the C&G2360 Part Two we had a workshop lecturer who did his apprenticeship in a carpet factory and then worked for the MEB and the companies that took over as the local DNO until he retired and was persuaded to come and teach us practical skills in the college workshop.

He advised us that having carried out the safe isolation procedures to prove an installation or circuit “dead” that we should then put our insulated long nose pliers or something similar across the terminals as a final test, as to quote him “it's better to sacrifice your long nose pliers than your fingers”.

The resistor in the plug in PV to prove dead is a test as simple and as unsophisticated as shoving your long nose pliers across the pins of the plug in PV 13-amp plug when you pull it out of the socket, it should not be necessary, but it’s a very good idea as it may stop you or someone else getting injured.

Forty seconds.

assets.publishing.service.gov.uk/.../plug-in-solar-final-_interim-product-specification.pdf

Yes, that's the document I quoted from. Where did 40 seconds come from? It seems an awfully long time when we typically think of disconnection within 5 seconds.

But , reading it again, it requires the resistor to be in place 40 s AFTER disconnection. So it will disconnect, then within 40 s a resistor will bridge the terminals so there is no chance of any charged components backfeeding the plug.

But, presumably the resistor will still be in place when the mains supply is restored and the system is reenergised?

But, presumably the resistor will still be in place when the mains supply is restored and the system is reenergised?

I fear that is partly a German to English translation issue.

In VDE 0126 -95: 2025 as part of the testing, after removal from the supply, a  resistor of any value at least 2000 ohms is connected  back across the exposed pins of the plug, and this is listed as one of a set of actions to test that the inverter does not try to restart when there is no grid, but there may or may not be connected low current resistive loads or  worse a person in parallel.

It does seems a bit odd to cherry pick parts of a test in this way without the rest of the explanation and other tests.

The VDE standard stipulates that, following a grid disconnection 
(e.g., due to a power outage or voltage faults), a photovoltaic inverter 
must continuously monitor the public grid for a period of at least 
20 to 60 seconds (with the standard setting being 40 seconds). 
The inverter is permitted to automatically reconnect 
only once the grid has remained 
(imagine this says within tolerance and - MAPJ1) 
absolutely stable throughout this period.

Mike

That is all very nice but what I would like to know what time it takes for the inverter output voltage to fall to zero on not seeing the grid?

Then I can add that time to the operating time of the RCD in the  event of a circuit fault to earth to verify that is an acceptable time for protection against shock?

JP

what I would like to know what time it takes for the inverter output voltage to fall to zero on not seeing the grid?

Then I think we can all agree, you don't want the 40 seconds figure - that's all about not being fooled into automatically restarting too soon..

And of course you don't really just want 'fall to zero', you need  'fall below, and stay below, a safe level'  Much as you would for oltages  generated by a washing machine motor spinning down or an SMPS with capacitors on the input. And here depending on what product standard you want to compare  to , there is a choice. 
 
Probably the most useful in your case is 

"The voltage at accessible plug pins shall decrease to below 34 V within 100 ms following
disconnection from mains supply"
 
Really though to align with BS7671, you need 50V as the touch voltage limit, but as 34 is less than 50, we can say 'faster than 100msec, but by an indeterminate amount '

the other alternative, more use if opening covers,  is

"Following disconnection from mains supply, capacitors larger than 100nF shall discharge to a voltage not exceeding 34 V within 1 s"


Both of these are directly lifted from existing product standards for equipment that plugs in .

So in the very worst case you may observe the voltage on an RCD that only just opens at  the maximum permitted 0,4 seconds after 30mA fault, with a worst case inverter plugged in, fall away to less than 50V in something under 0,5 seconds instead.
(and indeed disconnecting a worst case washing machine mid spin cycle would be similar)

Generally though most RCDs are a lot faster than that to open their contacts, so the real ADS situation is not as bad as that suggests.

Mike.