Volt drop, loop impedance and micro generation?

[JohnW2]

If that current were incredibly small, it surely would not be able to trip the mechanism, would it?

The sensor has several turns of the Live and Neutral around the toroid and does give enough energy on a low differential current.

Fair enough, but my point remains that there would inevitably be a finite limit to how small a differential current would trip it.

... whereas an 'electronic' one could.

only if there was a power supply to the electronics.

True, but if that crucial, it could presumably be battery backed-up.

The requirement was to disconnect the incomer even if there was no voltage on the incomer. Leakage being due to voltages generated on the site.

Interesting, but I'm not sure I understand that. If both voltage source and 'leak' were on the load side of the RCD (with incomers either O/C or effectively connected through a 'dead' transformer winding), why/how would the device trip?

Kind Regards, John

 

[westie101]

I think that the DNOs say that the 'maximum' Ze is 0.35Ω for TN-C-S

No we say the maximum expected!

Technically a figure of 0.8 works fine as that (as with TN-S) will operate our 100A main fuse in 5 sec as required.
After that it is up to an installation designer to cope with our actual measured figure

The 0.35 figure was set by estimate not by calculation or measuring and was based on the expectation of the move to a combined neutral/earth.

In certain cases where the figure is above 0.35 it is perfectly acceptable for us to just reduce the size of the main fuse and do no other work.

The reality of all of that is that installation designers should not use other than the actual measured value in their designs or expect a figure up to 0.8 and design from that.
You would be surprised in the number of reports we often get on Friday afternoons of figures above 0.35 that prevent certificates being issued as the electrical contractor has either not measured the actual figure prior to commencing work or has just used the 0.35 as an absolute maximum for a TN-C-S supply.
Not to mention those that they only spot that don't have an earth until completion and then request we supply one. Every so often they get told to supply their own TT earth which has not been allowed for in the quote

 

[westie101]

Regarding tripping, the devices are usually designed that as you close them to on you tension a spring, such that when they trip the only power is to reduce the latch holding the spring in position.
This is common on all the circuit breakers we use up to 132kV

We have one device that uses a vacuum circuit breaker to operate to energise faulty circuits.
As we operate the manual handle to open it we tension a spring.
The spring is released by a 9V DC battery.

 

[JohnW2]

Regarding tripping, the devices are usually designed that as you close them to on you tension a spring, such that when they trip the only power is to reduce the latch holding the spring in position.

Well, yes, I'd always assumed that this was how virtually all 'breakers' worked - in fact, it's difficult to think of what alternative there would be. Indeed, even manual switches/isolators often/usually use spring action, in one way or another, to separate contacts rapidly and positively. However, as I've been discussing with bernard, it obviously must take a finite amount of mechanical effort to 'release the latch'.

Kind Regards, John

 

[westie101]

But very little to release a spring latch

Fusemate operated by a 9V battery as stated

http://www.kelvatek.com/fusemate.php

 

[JohnW2]

I think that the DNOs say that the 'maximum' Ze is 0.35Ω for TN-C-S

No we say the maximum expected!

Fair enough (whatever it exactly means!)

Technically a figure of 0.8 works fine as that (as with TN-S) will operate our 100A main fuse in 5 sec as required. After that it is up to an installation designer to cope with our actual measured figure ... The 0.35 figure was set by estimate not by calculation or measuring and was based on the expectation of the move to a combined neutral/earth.

Again, fair enough. I've never really understood why there is so much apparent interest in the exact value of Ze, given that electricians will/should usually be measuring Zs on each of their final circuits, rather than calculating/estimating from Ze and (R1+R2) of the circuit. If it transpires that Ze is clearly too high, they will obviously want to speak to you and your colleagues but, otherwise, it's really just the Zsthat matters o them.

In terms of how this bit of the discussion arose, when you estimate the 'maximum expected' Ze, does that take into account the fact that Ze will rise as loading on the relevant part of the network (hence conductor temperature) rises?

Kind Regards, John

 

[JohnW2]

But very little to release a spring latch

I don't doubt that. Albeit in a totally different field (and 'the other way around'), it's always rather amazed/impressed me that the tiny amount of electricity coming from the 'flame failure thermocouple' in a gas appliance is able to 'hold back the latch' of the mechanism of what can be a fairly substantial gas valve.

Kind Regards, John

 

[bernardgreen]

If both voltage source and 'leak' were on the load side of the RCD (with incomers either O/C or effectively connected through a 'dead' transformer winding), why/how would the device trip.

It was something to do with a damaged feeder that had been dis-connected at the sub station and may have been auto re-connected while still faulty. I am going to have to ask an old colleague who was more involved in the power provision.

 

[JohnW2]

If both voltage source and 'leak' were on the load side of the RCD (with incomers either O/C or effectively connected through a 'dead' transformer winding), why/how would the device trip.

It was something to do with a damaged feeder that had been dis-connected at the sub station and may have been auto re-connected while still faulty. I am going to have to ask an old colleague who was more involved in the power provision.

Fair enough - that sounds like an extremely unusual, and probably pretty unique, situation! Faced with the same, I think I would probably want to have the 'damage feeder' well-and-truly disconnected at my end - at least until the issue was resolved

Kind Regards, John

 

[westie101]

Of course the difficulty with these figures is that nothing is absolute!

So on a day to day basis irrespective of load and heat, DNO networks can change.
Cables can be diverted making them longer or shorter
Substations can be moved with the same effect.
One substation could be switched out and load supplied from another
A generator could be supplying the load from a particular substation
A network changed from overhead to underground
An existing small cross section cable overlaid

 

[JohnW2]

Of course the difficulty with these figures is that nothing is absolute! ... So on a day to day basis irrespective of load and heat, DNO networks can change.

True - which I suppose means that it would be 'nice' (for consumers/ electricians) if (as with supply voltage) DNOs had to comply with some 'real maximum' Ze (on services for which they provide an earth'), rather just a 'maximum expected' - so that everyone knew where they stood.

Kind Regards, John

 

[ericmark]

My understanding has been what ever the Ze measured on the incomer we should plan for it changing to TN-S 0.8 and TN-C-S 0.35 due to changes in the supply which the DNO does not have to notify one of.

So if the supply was 0.25 then on a ring final using a B32 we would want to record 1.34 Ω to allow for the supply Ze to rise to 0.35 Ω in the future.

One clearly can't have it both ways and say with one breath you should use actual readings and with other allowance should be made for a change is supply characteristics. By enquire has always been a valid method.

I will admit looking at a head it is some times hard to work out TN-S or TN-C-S that combining need not be where it is visible to the electrician. So to measure is often the only way to know and we tend to consider above 0.35 but below 0.8 is TN-S and below 0.35 is TN-C-S although one is in real terms only guessing unless written on the head.

I have in one case inquired when the supply did not exist as to what supply would be provided and was told TN-C-S so no earth rod was used and line, neutral and earth tails were provided ready for connection when DNO arrived. However it transpired an earth was not provided this was found by another electrician visiting some 2 months latter.

We were at first blamed but to give credit to DNO they did admit original they had said a TN-C-S supply would be provided but on visiting site had informed the occupant they could not provide the earth and they would need to contact their electrician before switching on the supply.

By that time my son was in a narrow boat so of no fixed abode and they had been unable to contact him. He had a major problem getting his mobile charged and we had to cycle to boat to find him.

So when you hear the story my electrician is on holiday etc they could be true. In most cases the DNO insist on the electrician being on site when power is connected and I can see why. But people seem to expect everyone to be at the end of a phone. Being relent on weekly or fortnightly visit to PO box seems to slip the minds of many.

We know all health and safety information must be in writing be it paper or electronic but we happily phone the DNO and inform them without also sending text or email to them which in theroy we should.

Had my son written to DNO and provided a return address instead of phoning they would I am sure sent him a written notification that they could only provide a TT supply. So at the end of the day it was his fault for not doing it all in writing in the first place.

 

[bernardgreen]

Reasons for not using electronic RCD

Loss of one side of the supply renders it useless

They are susceptible to being tripped by radio frequency energy. Too many false trips at hill top sites.

Robotic assembly of electronic modules removes the need for hand wound sensors but increases the number of failure modes.

This is from way back when in the 1980's The hilltop sites have powerful transmitters and field strengths are high, much higher than any domestic situation.

Accuracy of the trip point was not necessary so precise adjustment of the mechanism was not needed.

There is no doubt that automated assembly brings the cost down to a level that makes RCD economically viable but this is not always a good thing.

 

[JohnW2]

Reasons for not using electronic RCD...

I think that we're aware of the theoretical downsides but, in an attempt to translate them into anything practical, are you aware of any RCDs (at least, ones for domestic etc. use) currently being manufactured which are not electronic? ... As for a couple of the specifics you mention:

Loss of one side of the supply renders it useless

It is surely 'vanshingly improbable' that such would occur at the very moment that one needed a RCD to operate. Anyway, as previously discussed, if one were actually worried about that possibility, it could be addressed by having a 'functional earth'.

They are susceptible to being tripped by radio frequency energy. Too many false trips at hill top sites.

At worst that affects only a tiny proportion of users. In reality, truly 'spontaneous and unexplained random trips' (which might be due to RF pickup in some cases) are incredibly rare.

Kind Regards, John

 

[ericmark]

Domestic and industrial RCD's are very different beasts. The industrial was in three parts.
1) The CT
2) The electronic unit often with adjustments for time and current.
3) The moulded breaker with a trip coil.
It needed power to trip and as we pointed out many times loss of a phase could prevent some units from working. However these were never designed for personal safety as such. Often set with one minute delay and one amp differential to trip them.

However I have seen where a worker hammered in a nail to hang his coat it took out the 30mA in tea hut, the 100mA S type in sub board the 500mA with 30 seconds delay in our main board and the 1A with 60 seconds delay in suppliers board. So much for discrimination.

But tried fitting RCD's in 110 volt (63 - 0 - 63) and they failed to work. This was back in 1994 so I would assume there have been improvements back then 110 volt ones just could not be found. But as we commented at the time there was nothing to give the range of voltage.

Later we had problems the other way with 110 volt RCD's. Two major problems. One was mag mount drills falling off when the power tripped although there was always a lanyard it could swing quite dangerously when power was lost.

The other was with load sensing generators once a load was put on the generator they would rev to 3000 revs and when load removed would drop back to tick over may be 900 RPM. All the 110 volt RCD's were active types if the generator was already being used they would set but as the load was removed they would drop out and clearly you can't set without a load and you can't give it a load until set. There was around 40 volt line to line when generator ticking over. Around 15 Hz.

It was of course safe just no one could do any work.