Master switch for a circuit of sockets?

[akirk]

we have already had a full inspection on the house, new main CU and new CU in the garage run from the main CU so we are all set!

 

[sometimewoodworker]

How? If the circuit protection is adequately sized for the conductor size and installation specifics, the cable can’t become overloaded?

I have no issues with ring finals but a break in the circuit creates an undersized and inadequately protected radial and even if the ring is complete, plugging appliances in close to the start/end of the ring can cause current imbalance, particularly in a long circuit. However... it’s a lot easier to pull 2.5mm cable!

The difference between pulling a single 2.5mm and a single 4mm isn’t that great but some of the fault conditions that can occur on a ring final can (assuming 2.5mm cable) give you a ring with 1 of the conductors split while the others are not so you have a 32A MCB that will not trip on steady state load until probably over 40A draw for quite a long period 30 minutes plus so potentially a 2.5mm cable passing 40A for that amount of time. It will get toasty to say the least and quite easily more than just toasty, the MCB can’t do its job.
However with 2 X 2.5mm radials each will have a 20A MCB so allowing a greater total draw than the ring final and if you do overload a single circuit to 40A the time to trip on the MCB is going to be minutes or seconds so the MCB does it job of protecting the conductors. This is one of the reasons why very few new ring finals are ever installed as the ONLY benefit is using a single 32A MCB in the CU vs 2 20A MCBs the supposed savings in copper are tiny if they exist at all.
It is more than likely that in a future revision of the regulations they will be totally deprecated for new installs.

Unlike the single dose vaccine split by 3 times the tested interval before the second booster, forced by lack of supply and hugely rising hospitalisation, that is proving to be a very good idea as the protection is much greater. The use of ring finals while they may have been useful in the 40s and early 50s. Were not a good idea and should have been phased out many decades ago.

 

[NetBlindPaul]

But a switch is not an isolator and vica versa.

Yes plastic CUs do burn and spread the resulting fire, often not helped by being under the stairs and burried in allsorts of stuff someone has put there out of the way, similar to stacking a bonfire! The idea of metal units is to contain the thermal event, but regular test and inspection would probably be more effective, just like people who never test their RCD's on a regular basis.

Actually the idea of metal consumer units is that the metal, which it doesn’t have to be, it can be any non combustible material, hence the consumer unit can be boxed up, is to remove the fuel.
The shame is that the root cause of incompetent installers that caused the hot connections and the profiteering manufacturers who allowed extremely flammable plastics to be used for the housings was not addressed properly.
It was the members of the trade body BEAMA that went the steel route together. So every one followed.

 

[NetBlindPaul]

The difference between pulling a single 2.5mm and a single 4mm isn’t that great but some of the fault conditions that can occur on a ring final can (assuming 2.5mm cable) give you a ring with 1 of the conductors split while the others are not so you have a 32A MCB that will not trip on steady state load until probably over 40A draw for quite a long period 30 minutes plus so potentially a 2.5mm cable passing 40A for that amount of time. It will get toasty to say the least and quite easily more than just toasty, the MCB can’t do its job.
However with 2 X 2.5mm radials each will have a 20A MCB so allowing a greater total draw than the ring final and if you do overload a single circuit to 40A the time to trip on the MCB is going to be minutes or seconds so the MCB does it job of protecting the conductors. This is one of the reasons why very few new ring finals are ever installed as the ONLY benefit is using a single 32A MCB in the CU vs 2 20A MCBs the supposed savings in copper are tiny if they exist at all.
It is more than likely that in a future revision of the regulations they will be totally deprecated for new installs.

Unlike the single dose vaccine split by 3 times the tested interval before the second booster, forced by lack of supply and hugely rising hospitalisation, that is proving to be a very good idea as the protection is much greater. The use of ring finals while they may have been useful in the 40s and early 50s. Were not a good idea and should have been phased out many decades ago.

However if the circuit is correctly designed, small overloads of long duration shall be unlikely to occur. (BS7661, 433.1) So if having the work done by a scheme member & in compliance with Approved documents then the work must comply with this, that is the designer must design the circuit such that small overload labof a long duration are unlikely. They would need to be able to justify and design in compliance with that requirement.

 

[NetBlindPaul]

That can’t be done as you have to isolate the ring final that you are testing from all other circuits, so must be done at the CU wherever it is, I have sympathy for sparkles and the contortions they have to go through.

Not true as the ring must he electrically separated from other circuits when designed and constructed.

 

[sometimewoodworker]

Not true as the ring must he electrically separated from other circuits when designed and constructed.

You should look at the current testing regulations. When testing a ring final you have to do the same test as when installing which involves detaching from the MCB so you have to do this part along with the loop tests at the CU.
That is why a Domestic Electrical Installation Condition Report costs upwards of £200 with repairs on top of that

 

[sometimewoodworker]

However if the circuit is correctly designed, small overloads of long duration shall be unlikely to occur. (BS7661, 433.1) So if having the work done by a scheme member & in compliance with Approved documents then the work must comply with this, that is the designer must design the circuit such that small overload labof a long duration are unlikely. They would need to be able to justify and design in compliance with that requirement.

A load of 32A is considerably greater than a 2.5mm cable is considered safe for. While the circuit is in good condition it’s safe however there are fault conditions that can run that through a single not double length of 2.5mm. Those fault conditions can be completely symptomless as all outlets can function normally but the ring is faulty.

So ring finals are safe if in good condition but much worse than a radial in some fault conditions

 

[Spectric]

The reason for disconnecting at the distribution board is so that your ring is completely independant, you need the two CPC's to be able to prove the ring is continous, it is no good having them both connected along with the other CPC's on the common earth rail.

 

[NetBlindPaul]

You should look at the current testing regulations. When testing a ring final you have to do the same test as when installing which involves detaching from the MCB so you have to do this part along with the loop tests at the CU.
That is why a Domestic Electrical Installation Condition Report costs upwards of £200 with repairs on top of that

I am more than familiar with the current version of BS7671, before I was hit by the cancer (less than 6 months ago), part of what I did was tutoring and practical examination for the inspection and testing qualifications related to BS7671.
The integrity of a ring final circuit as part of an EICR is subtly different than what is required for an initial verification.
I have done enough to know what I am doing.
I have been a PDH, QS & sole operator under the NICEIC AC scheme for many years and was assessed, personally annually without flaw.
I only left the NICEIC as I did not get value for the costs of membership.
For what I needed NAPIT was cheaper so I transferred schemes. Again being assessed by them.
I have contributed to books on the subject for both NAPIT & the IET.
I can assure you that you have got an EICR cheap at £200, and that to do the ring integrity checks at a socket outlet is fine.
This does not preclude needing access to the consumer unit, but it does mean that the person undertaking the work can potentially work with greater safety, by doing some work away from the CU.

 

[NetBlindPaul]

A load of 32A is considerably greater than a 2.5mm cable is considered safe for. While the circuit is in good condition it’s safe however there are fault conditions that can run that through a single not double length of 2.5mm. Those fault conditions can be completely symptomless as all outlets can function normally but the ring is faulty.

So ring finals are safe if in good condition but much worse than a radial in some fault conditions
[/QUOT

A load of 32A is considerably greater than a 2.5mm cable is considered safe for. While the circuit is in good condition it’s safe however there are fault conditions that can run that through a single not double length of 2.5mm. Those fault conditions can be completely symptomless as all outlets can function normally but the ring is faulty.

So ring finals are safe if in good condition but much worse than a radial in some fault conditions

Again I an fully aware of the CCC of a 2.5/1.5 flat twin & cpc cable.
If you read and understand my post you will see that it relates to the original design of the circuit which is required to be such that small overload shall not occur.
Nothing was mentioned regarding fault conditions.
If the circuit is designed, constructed and initially verified correctly in accordance with BS7671 then mo faults can be present. Additionally the design will be such that small overload cannot happen.
Now if the work has been done correctly a fault is extremely unlikely to occur in said circuit unless it is disturbed.
It is then down to the person “disturbing” the circuit to ensure that they di not compromise its integrity.
No one has a crystal ball, and as I have already stated, most ring final circuit faults are due to incompetence by DIY operatives interfering with and not understanding the construction of such a circuit.

 

[NetBlindPaul]

The reason for disconnecting at the distribution board is so that your ring is completely independant, you need the two CPC's to be able to prove the ring is continous, it is no good having them both connected along with the other CPC's on the common earth rail.

BS7671 requirements are that every final circuit shall be electrically separate from every other.
This is quite easy to achieve with an “all insulated” installation such as a domestic.
It does become a slight problem with industrial and perhaps commercial when metallic containment such as conduit is used especially when the conduit forms the cpc.
For a ring final in flat twin & cpc or singles in pvc conduit, to do the ring integrity checks to derive r1, r2 & rn, along with R1 & R2, IF these are even essential for an EICR can be done without disconnecting from the CU.
Other tests such as IR may well be influenced by the connections between the neutral and earthing in the supply or other circuits.
However the guidance from the IET (GN 3) on pg 81 is “A detailed visual examination of the installation is required, together with appropriate tests. The tests are mainly to confirm that the disconnection times stated in Chapter 41 are met.”
GN 3 goes on to state on pg 89, “... a detailed inspection comprising an examination of the installation without dismantling, or with partial dismantling as required, together with the tests of Chapter 64 considered appropriate by the person carrying out the inspection and testing.”
Perhaps here is not the best place to get into a competence debate on the operatives entering the electrical industry and undertaking periodic I&T with little training and experience.

 

[Spectric]

For a ring final in flat twin & cpc or singles in pvc conduit, to do the ring integrity checks to derive r1, r2 & rn, along with R1 & R2, IF these are even essential for an EICR can be done without disconnecting from the CU.

I agree that to find the prospective fault currents using a modern tester that the circuit can be tested without any disconnection simply by pluging/connecting the tester to the circuit and either noting the values and or recording within the tester.

A ring circuit requires additional test when compared to a radial because you have to confirm that A) The ring is infact a ring, consisting of L,N & CPC and B) It has not been bridged by any additional circuits or addons. To this end the only way to prove the ring is a ring is by disconnecting at the board to perform a continuarity test as shown. Note this applies for 1 to n sockets.

 

[NetBlindPaul]

I agree that to find the prospective fault currents using a modern tester that the circuit can be tested without any disconnection simply by pluging/connecting the tester to the circuit and either noting the values and or recording within the tester.

A ring circuit requires additional test when compared to a radial because you have to confirm that A) The ring is infact a ring, consisting of L,N & CPC and B) It has not been bridged by any additional circuits or addons. To this end the only way to prove the ring is a ring is by disconnecting at the board to perform a continuarity test as shown. Note this applies for 1 to n sockets.


View attachment 105622

I am fully aware of the requirements for initial verification of a ring final circuit.
A periodic inspection does not require the full suite of testing that is required for an initial verification.
If you read & understand IET GN 3 you will see that it is explained in that.
Additionally the ring can be split at a socket and the tests for the ring done there to check if it is a ring.
The requirements for full initial verification, i.e. the 3 step ring verification tests as per GN 3 2.6.6 are not essential for an EICR unless inspection and other tests suggest that it is needed.

 

[guineafowl21]

My view is that the ring tests will only return correct results if the ring is intact, without bridges or breaks. This applies whether done at the CU or at a socket.

Any errant result from either approach would require investigation - full disconnection and sockets off the wall, etc.

@Spectric you’re arguing that testing at a socket is invalid, which implies either false passes or false fails. Could you show us a situation where this could happen? A diagram would be useful.

I am genuinely interested, not just being argumentative, but that’s hard to convey by text alone.

 

[Spectric]

you’re arguing that testing at a socket is invalid, which implies either false passes or false fails. Could you show us a situation where this could happen? A diagram would be useful.

No I am not saying it is invalid but I always tested at the board, my reasoning was simple: The Neutrals are all connected to a common neutral bar unless there are RCBO's and the same for the CPC's. Testing at a socket could give incorrect readings because someone may have picked up a neutral or CPC from another circuit.

Again periodic test and inspections may not require the same full testing done following an installation but I always gave a full complete job and with rings I always liked to be over safe rather than have any doubt. Regulation 621.1 states as far as I can remember that you need to confirm the installation is safe and in such a condition as to be ok for further service. How can you state this when doing the EICR form without testing things you know are more liable to give rise to issues. Simple logic, broken conductor in a radial something is no longer working, broken conductor in a ring it still works but no longer safe. Ok I always tried to avoid domestic work, much prefering industrial but you always get far more dodgy work in the domestic setting so approach with that thought in mind.

 

[akirk]

A little update... workshop now has two sets of plugs:

white (2.5mm cable) for ‘normal stuff’ such as charging batteries / wifi / tv / sonos / etc... some of which also have usb ports...

red (4mm cable) are separate and all run through an isolator switch, like the stock photo below... works really well, turn it and all machinery is dead...

(the storage shed / barn / was a carport... next to the garage has a separate white circuit and then separate blue plugs for the chest freezer and fridge-freezer)

 

[Dr AndyB]

Afraid I don't know the electrical details but, when I had my workshop wired, my wife insisted on a "panic switch" being installed by the door. This is a button that you hit if something goes wrong, but I also use it when I leave the workshop. All the sockets are switched off by this switch but not the lights. You just pull it out again when you want to switch the sockets on again.

 

[Spectric]

Afraid I don't know the electrical details but, when I had my workshop wired, my wife insisted on a "panic switch" being installed by the door

That is normally refered to as an E-Stop or an emergency stop and in most industrial situations there is one located near / on every machine. In some circumstances such as training shops and the like that E-Stop will operate to stop all machines, remember back to a) when there were such classes in schools and b) there were STOP buttons everywhere so the teacher could kill all the power if the class started to behave like idiots.

 

[sometimewoodworker]

A little update... workshop now has two sets of plugs:

white (2.5mm cable) for ‘normal stuff’ such as charging batteries / wifi / tv / sonos / etc... some of which also have usb ports...

red (4mm cable) are separate and all run through an isolator switch, like the stock photo below... works really well, turn it and all machinery is dead...

(the storage shed / barn / was a carport... next to the garage has a separate white circuit and then separate blue plugs for the chest freezer and fridge-freezer)

View attachment 108004

View attachment 108005

The switch is good but I consider this better

I have 4 of these that go to a contactor in a sub CU this kills all the power circuits in the workshop, they are easier to use in an emergency and latch off

 

[akirk]

I think that adds a slightly different value (and could be added) I am concerned less by emergency stop and more by temporary isolation to avoid children playing with machinery