Ring extension load

[DetlefSchmitz]

I suspect that the drafting group thought that 20A was the greatest current likely to be required in any one place. Remember that in the 40s and 50s any load over 1 kW was quite unusual in a domestic situation.

It's of course possible that you are right. I was certainly not there at the time.

But Occam's razor would suggest that when deciding how to test a 2x13A socket:
if double sockets could actually supply 26A, why not choose that value, unless
there was a particular reason to choose a different lower value.

I know what I believe...

 

[JohnW2]

I suspect that the drafting group thought that 20A was the greatest current likely to be required in any one place. Remember that in the 40s and 50s any load over 1 kW was quite unusual in a domestic situation.

Indeed, that might well have been the initial thinking, but that does not mean that the requirement of the Standard could/should not have been re-visited in the ensuing 60-70 years, during which the nature of loads evolved dramatically.

I think we are probably all agreed that the sort of extended periods of carrying currents >20A which the MK tests looked at are incredibly unlikely to be experienced in normal domestic situations. However, I would say that of much more concern are the not uncommon reports we hear of thermal damage to plugs/sockets when, say, a WM and dryer are run from the same double socket. The durations of high (>20A) total currents will then be very modest, but it does seem that the plugs/sockets don't necessarily tolerate it well. I could be a co-incidence/red herring (due to loose connections etc.), but it might not.

Kind Regards, John

 

[EFLImpudence]

I don't understand why (as an above comment) - with two higher loads the faceplate may show signs of overheating.

This is not a part that would carry or be susceptible to the combined current.

 

[JohnW2]

But Occam's razor would suggest that when deciding how to test a 2x13A socket: if double sockets could actually supply 26A, why not choose that value, unless there was a particular reason to choose a different lower value.

Quite so. Furthermore, if, as stillp suggested, the situation arose because domestic loads >1kW (~4A) were very uncommon at the time, why would the BS1363 test have utilised a current anything like as high as 20A?

Kind Regards, John

 

[AndyPRK]

I don't understand why (as an above comment) - with two higher loads the faceplate may show signs of overheating.

This is not a part that would carry or be susceptible to the combined current.

All metal items getting fing hot (inc plug pins) and the white bits can't cope.

We're the MK tests done decades ago?

 

[EFLImpudence]

Yes, but the plug pins would not be affected by two 13A loads.

 

[Echo the husky]

I would have though the very old MK double sockets with the 4 plate screws would be able to take a full 26Amps without problems. They are so robust and reliable.

 

[AndyPRK]

Yes, but the plug pins would not be affected by two 13A loads.

But the sockets would be hot and not be able to dissipate the heat like a single socket would.

 

[JohnW2]

Yes, but the plug pins would not be affected by two 13A loads.

I would have thought that all the metal parts of the socket and plugs, including the plug pins and receptacles, would be in pretty good thermal contact with one another - so that if anything was overheating, everything would get hot.

Kind Regards, John

 

[stillp]

No different from allowing the concept of diversity in other aspects of circuit design.

I think it actually is rather different from almost all other domestic applications of the concept of diversity. Application of diversity does not usually compromise 'safety' in any way. It allows one to use cable of smaller CSA than would be possible without application of diversity, but that cable still has to be protected by an OPD appropriate to its CSA. Hence, the worse that happens 'if the concept of diversity fails' (which, being a statistical concept, it sometimes will) is that the OPD operates.

[/QUOTE]Hmm. A fair point, provided that the OCPD is chosen correctly.

Don't forget that, in the absence of any required tests at currents >20A, a double socket could theoretically be BS1363-compliant and yet 'fail catastrophically' at currents above 20A carried for relatively short periods of time. Very unlikely - but not impossible.

I would say vanishingly improbable. Remember each socket-outlet in the pair has to be rated at 13 A, and is tested at 14 A. In the heat rise test, the twin socket-outlet is mounted in a wooden enclosure 25mm thick and each outlet is tested at 14A while the adjacent one carries 6A. There is no outgoing cable, i.e. the configuration simulates the end socket on a radial or spur, so there is only one 2.5 mm² cable acting as a heat sink. The test duration is a minimum of 4 hours, or until a stable temperature is reached, with a maximum of 8 hours. The permitted temperature rise for accessible external surfaces is 52 K. Given that the materials used by some manufacturers can withstand temperatures in excess of 90 ° C, what are the chances of a "catastrophic failure"? Also, the constructional and performance requirements of the standard are not just those that are necessary to pass the tests.

 

[stillp]

But Occam's razor would suggest that when deciding how to test a 2x13A socket: if double sockets could actually supply 26A, why not choose that value, unless there was a particular reason to choose a different lower value.

Quite so. Furthermore, if, as stillp suggested, the situation arose because domestic loads >1kW (~4A) were very uncommon at the time, why would the BS1363 test have utilised a current anything like as high as 20A?

Kind Regards, John

At the time, it would have represented a considerable safety margin.

 

[EFLImpudence]

Ah. You can see on this old socket that the current path does not, as I would have though, split and go to each socket but one part (L and N) would indeed carry the 26A.

 

[AndyPRK]

Humm. Now I can understand why it's always the neutral that overheats lol

 

[JohnW2]

Hmm. A fair point, provided that the OCPD is chosen correctly.

Indeed, but there is nothing about applying the concept of diversity which allows one to choose an OPD with an inappropriate In in relation to the cable

Don't forget that, in the absence of any required tests at currents >20A, a double socket could theoretically be BS1363-compliant and yet 'fail catastrophically' at currents above 20A carried for relatively short periods of time. Very unlikely - but not impossible.

I would say vanishingly improbable. ....

We could debate the choice of words for ever, but can presumably agree that it's "very improbable"! However, in scientific/engineering terms, it really is not 'good practice' (to say the least!) to use common sense (no matter how reasonably based) to extrapolate (some would say 'guess') beyond the conditions of a 'safety test', is it?

Similar common sense says that it is probably 'vanishingly improbable' that any harm would result from designing a circuit with a Zs fractionally above what was required for compliance with BS7671, or which used a cable with a CCC fractionally below the In of the OPD protecting it ... but try telling that to 'an inspector'

Kind Regards, John

 

[Echo the husky]

Those bits of copper appear chunkier than the copper and brass parts of the MK Logic Plus socket I've just dismantled.

I have just been messing about with a thermal imaging camera and some heavy loads on an MK socket and did some very non scientific tests. The load was a 3.03KW downflow heater in one socket and a 2.2KW and 1KW Kettles in the other socket. Total measured current was 26.5A. Measured voltage was 240V. Load was applied for 10 minutes, by which point my house was getting very warm and I was tried of getting splashed with boiling water from the 2 kettles. Temperatures are all approximate as they were measured using the camera. In the pictures the maximum and minimum temperatures in that current picture is shown at the bottom of each image, the big temperature reading at the top left is the centre of the crosshair.

Observations:
Current flowing through leg 1 of the ring: 5.5A
Current flowing through leg 2 of the ring: 21A
Leg 2 was getting warm, insulation of line conductor was about 45°C.
I broke the ring on the line side and ran the test again for long enough to boil the larger kettle. temperature of the conductors insulation was around 45°C.
The 13A fuses were the main source of heating in all tests, closely followed buy the areas in the socket where the conductor bars carried the current for both outlets. The temperature shown was on the plastic back of the socket, the metal parts inside would be hotter. I should have drilled a hole in the back of the socket and jammed a little temperature probe in there.
Even the fuse in the 2.2KW kettles plug quickly reached around 55°C in the time it took to boil 1.7L of water.
The fuse for the 3.03KW heater quickly reached 65°C in about 2 minutes, but the temperature had not increased after 10 minutes. Tests were run with the lids removed from all plugs which would have provided some cooling. All plugs are quality MK Safetyplugs.

I took the max KW rating of a random TD and a WM and the figure came to about 20A. For a wash and dry cycle run together connected to an MK socket I would expect it to be fine, provided care was taken in relation to the CCC of the spur cable and the refernce method it is installed. With both loads run through a 13A switch fuse, or connected together with a short 2gang extension lead I would expect the fuse not to blow, but to reach temperatures high enough to cause serious damage to the FCU/plug. In the last image I have connected the heater and large kettle (measured current 20.3A) to an extension lead and run it for another 10 minutes. The 13A fuse reached 160°C after around 3 minutes, after 10 minutes the temperature was the same. I had to stop the test as the 0.75mm² flex on the short extension lead was becoming very soft and starting to deform!

The 2 Plugs at the end of the test, shown with their lids removed

The socket face after the test after the plugs were removed.

The rear of the socket after the test (socket upside down with switches at the bottom of the image). The hot areas are above the switches and the 2 bars going down to the wiring terminals.

The top surface at the rear of the socket.

Fuse of extension lead with approx 20A load.

So in conclusion, this experiment has proved fark all, and I'm not running it for 75 hours to see what happens.