BS 7671. Isolator at meter point

BS 7671 is non-statutory but the Electricity at Work Regulations 1989 (EAWR) are written into law.

The EAWR are general in their application and refer throughout to ‘danger’ and ‘injury’. Danger is defined as ‘risk of injury’ and ‘injury’ is defined in terms of certain classes of potential harm to persons. Injury is stated to mean death or injury to persons from:

  • electric shock;
  • electric burn;
  • electrical explosion or arcing; or
  • fire or explosion initiated by electrical energy.

Regulation 12(1)(b) of EAWR states ‘where necessary to prevent danger, suitable means shall be available for […] the isolation of any electrical equipment’, where ‘isolation’ means the disconnection and separation of the electrical equipment from every source in such a way that the disconnection and separation is secure.

The main requirements for isolation and switching are provided in Chapter 46 and Section 537 of BS 7671:2018+A2:2022.

Chapter 46 of BS 7671:2018+A2:2022 sets out the requirements for isolation and switching and Regulation 462.1 states that:

‘Each electrical installation shall have provisions for isolation from each supply.’

In my opinion the DNO/DSO/MO (Distribution Network Operator/Distribution System Operator/Meter Operator) in the UK should provide the isolator switch without question and possibly free or charge to comply with the Health and Safety act EWR 1989.  Means for cutting off the supply and for isolation – It must be possible to switch off the electricity supply to any piece of equipment in case of electrical danger. This can be by direct manual operation, circuit breakers or by stop buttons in control circuits of contactors. Precautions for work on equipment made dead – Precautions must be taken to prevent electrical equipment which has been made dead (whether temporarily so work can be done or permanently because the equipment is no longer in use) from becoming charged during that work. If an Electrician has to change a CU/DB (Consumer Unit/Distribution Board) then an Isolator should be present as they are at work, thus the DNO/DSO/MO must provide it.

  • But aren't the limits (e.g. ESQCR & G98/99) set by the generating capacity rather than export?

    Yes ... and we must also consider the fact that, at present, there are no islanding requirements in G98 which means that many DNOs require G99 for islanding.

    However, you said that islanding requires a significant amount of extra kit ... and that wouldn't be needed, if you didn't need "power backup" but wanted to set a zero export limit ... the wider issue, though, is that even if you do have islanding, it won't be long I don't think, before there will be more stringent conditions on islanding (i.e. can't "just do it" but power has to go off for a length of time first before it happens).

  • it won't be long I don't think, before there will be more stringent conditions on islanding (i.e. can't "just do it" but power has to go off for a length of time first before it happens).

    Curious- I can't immediately see any benefit for the grid for doing that (the load will have been dropped already - I would have thought that having installations present near-zero load on grid return would have been beneficial if anything*). I could see that a random delay before switching back to grid after power return might help. A delaying in islanding on loss of power would seem to undermine any UPS type requirements for the installation too.

      - Andy,

    * I recall back in the 1970s power cuts my dad following official advice and going around switching everything off after the power went off  - save one light that was left on to signal the end of the power cut.

  • Curious- I can't immediately see any benefit for the grid for doing that

    If the installation is in direct feeding mode, and it just happens, yes that can definitely potentially cause instability.

    However, let's say that we get to mid-day on a hot day, and suddenly all the parallel-connected generation is turned off because the battery is full ...

    BUT

    I think the issue will be resolved, overall (at least for those who "own" their installation ... energy poverty is a whole other discussion) by novel approaches to tariffs, which have already started with companies like Octopus.

  • However, let's say that we get to mid-day on a hot day, and suddenly all the parallel-connected generation is turned off because the battery is full

    Bear with me - I'm sure I'm missing something obvious - if the battery is full then in effect one internal load (the battery charger) is disconnected - PV continues to generate and supply the other on-site loads and any excess is exported (i.e. the system degrades to a conventional PV inverter-only system). Load on the grid is reduced by the rating of the battery charger (probably not as bad as that as most multi-stage chargers reduce the power to the battery gradually as it nears capacity) - but even in the simple case of switching directly from full power to off situation it's hardly any worse than say a heat pump or EVSE switching off (or an immersion or hot tub) and the grid seems capable to coping with that. There's likely to be a significant variance between installations (different systems, different battery capacities, differences in panel positioning and shading) not to mention sun position across the country, so I don't see a huge risk of all such system across the country acting simultaneously (unless triggered by some external event - say a tariff change - but that sort of thing is under the control of the supply industry anyway).

       - Andy.

  • Bear with me - I'm sure I'm missing something obvious - if the battery is full then in effect one internal load (the battery charger) is disconnected - PV continues to generate and supply the other on-site loads and any excess is exported (i.e. the system degrades to a conventional PV inverter-only system). Load on the grid is reduced by the rating of the battery charger (probably not as bad as that as most multi-stage chargers reduce the power to the battery gradually as it nears capacity) - but even in the simple case of switching directly from full power to off situation it's hardly any worse than say a heat pump or EVSE switching off (or an immersion or hot tub) and the grid seems capable to coping with that

    Solar PV will be doing 2 things:
    (a) Exporting excess over battery charging (if that's all that's switched on); and
    (b) Helping support the local grid frequency stability, if the installation is in a "pinch point" area.

    So the issue is, not if a single installation decides to do this, but many simultaneously. It's as simple as a small delta times by a lot equates to something.