230v or 240v

[stillp]

Don't you share my view that such a situation is rather extraordinary? When a manufactured product has to comply with a Standard, doesn't one expect there to be specified tests which have to be satisfied in order to confirm that compliance?

Yes, I do expect the standard to specify tests, and it does! Just not including the one(s) you were expecting.

True, but can you blame me for 'expecting' a test of one of the most important electrical properties? In electrical terms (I accept that there are other considerations) current-carrying and voltage-withstanding properties are really all there is!

quote]

(I wish I could work out how to do multiple quotes on this forum!)
And the CCC depends on what John? The materials, which are specified, the resistance, which is specified, and the dimensions, which are specified. The voltage-withstand is tested, to identify possible manufacturing defects, but the CCC can't be affected by manufacturing defects without affecting the resistance.

The standard isn't allowing an assumption to be made - it is only applicable to certain materials.

Lord Kelvin patented his "current balance" in the 19th century, and residual current devices were around long before VOELCBs were considered unacceptable. There were probably a mixture of reasons for using them, cost and marketing considerations as well as changing attitudes to electrical safety.

You're being a little inconsistent in your final 2 paragraphs. The laws of physics haven't changed, but the perception of risk has, and perhaps CPCs in lighting circuits came about because of a changing perception of the risk associated with touching a conductive part of a light fitting that has become live due to a fault. In a few years we (or the IET's Marketing Manager) might consider we have to reduce the possible touch voltage between exposed conductive parts to 30 V instead of 50, and increase the MEB to 25mm2.

 

[JohnW2]

(I wish I could work out how to do multiple quotes on this forum!)

Just as with any 'tagged' language, just make sure that all tags are logically paired (I've had to add the asterisks to prevent the system thinking they were tags!)...

[q*uote][q*uote]...text2[/q*uote]...text1[/q*uote]
just like x = ((a+b)*c)

And the CCC depends on what John? The materials, which are specified, the resistance, which is specified, and the dimensions, which are specified. The voltage-withstand is tested, to identify possible manufacturing defects, but the CCC can't be affected by manufacturing defects without affecting the resistance. ... The standard isn't allowing an assumption to be made - it is only applicable to certain materials.

Yes, but it's assuming that actual manufactured products have been manufactured using materials which actually (allegedly) are of the correct type and have not had their properties altered by the manufacturing process. If the thermal properties of the insulation are not 'as expected', then the CCC could be seriously altered. Would you think it reasonable for it to be accepted that, say, fastners for the aeronautical industry 'must have the required strength', because they had been manufactured using a certain grade of steel as specified in a Standard, or would you prefer to see some tests of samples of actual manufactured products to be specified in the Standard?

Lord Kelvin patented his "current balance" in the 19th century, and residual current devices were around long before VOELCBs were considered unacceptable. There were probably a mixture of reasons for using them, cost and marketing considerations as well as changing attitudes to electrical safety. You're being a little inconsistent in your final 2 paragraphs. The laws of physics haven't changed, but the perception of risk has, and perhaps CPCs in lighting circuits came about because of a changing perception of the risk associated with touching a conductive part of a light fitting that has become live due to a fault. In a few years we (or the IET's Marketing Manager) might consider we have to reduce the possible touch voltage between exposed conductive parts to 30 V instead of 50, and increase the MEB to 25mm2.

If I've been inconsisent, that was not intentional - and I thought I had been clear. In nearly all the cases discussed, changing attitudes to safety (rather than any considerations of physics or engineeering) have undoubtedly been the main reason for evolving regulations.

Kind Regards, John.

 

[stillp]

'how much current can T&E really safely take'

Depends what you mean by 'safely'. OK, 70 C, but under what conditions?

However, conductor temperature itself would probably not fully answer the question of interest.

What would then, the insulation material being a constant?

the test conditions don't necessarily have to be all that 'realistic', so long as they are reproducible

Exactly my point, although I said repeatable rather than reproducible.

I'm talking about requirements to confirm that manufactured products conform to a required specification (e.g. a Standard) - sampling/testing (or, in some cases, 100% testing) of manufactured products to verify that has been part of the QC of every sort of 'serious' manufactured product I've come across.

You must have lived a very sheltered life then John.

Without that, one cannot rule out problems with the (actual, not theoretical) materials or manufacturing process.

In the far-off days of QC, you're right, but in today's world of QA such problems are reduced to an acceptable level (i.e. lower than would be achieved by any plausible sample testing) by management controls.

 

[JohnW2]

'how much current can T&E really safely take'

Depends what you mean by 'safely'. OK, 70 C, but under what conditions?

I didn't start all this. There is just an interest in a ball-park figure for what sort of currents T&E really can 'safely' carry under 'typical' conditions. You are trying to turn it into a much more precise scientific exercise.

However, conductor temperature itself would probably not fully answer the question of interest.

What would then, the insulation material being a constant?

We don't know what conductor temperature would be 'safe' for any particular insulation material.

You must have lived a very sheltered life then John.

Without that, one cannot rule out problems with the (actual, not theoretical) materials or manufacturing process.

In the far-off days of QC, you're right, but in today's world of QA such problems are reduced to an acceptable level (i.e. lower than would be achieved by any plausible sample testing) by management controls.

My goodness, we really must function in very different areas! Sure, in all walks of life, there has been an increasing emphasis on QA, but not (in my experience) to the exclusion of traditional QC, both as a 'final confirmation' and also as a source of information to feed into the QA process.

In some fields I work in, there are Statutory requirements for extensive testing of samples from every batch of finished products (to ensure that they comply with a spec) before the batch is allowed to be 'released' (in itself a very formalised process, defined by Statute) - in addition to an awful lot of QA.

Kind Regards, John.

 

[stillp]

'how much current can T&E really safely take'

Depends what you mean by 'safely'. OK, 70 C, but under what conditions?

I didn't start all this. There is just an interest in a ball-park figure for what sort of currents T&E really can 'safely' carry under 'typical' conditions. You are trying to turn it into a much more precise scientific exercise.

No, I'm just trying to point out that you can't even acheive a ball-park figure without knowing the parameters of the ball-park, and trying to explain why the standards say what they do, and don't say what they don't.
If you only want a rough figure for the CCC of T & E, why not just look at the tables in BS 7671? I thought this discussion came about because you wanted to know how much "safety margin" was built-in to those tables.

Would you think it reasonable for it to be accepted that, say, fastners for the aeronautical industry 'must have the required strength', because they had been manufactured using a certain grade of steel as specified in a Standard, or would you prefer to see some tests of samples of actual manufactured products to be specified in the Standard?

In my experience it's quite common for safety-critcal components in aerospace to be only type-tested and to depend for their integrity on tests of the material. There's little point in sample testing for small production volumes. I accept that things are very different in your pharmaceutical world.

 

[Paul_C]

Lord Kelvin patented his "current balance" in the 19th century, and residual current devices were around long before VOELCBs were considered unacceptable. There were probably a mixture of reasons for using them, cost and marketing considerations as well as changing attitudes to electrical safety.

Now that it's been considered obsolete for 25 years or so, I feel that the voltage-operated ELCB is often portrayed in an unfairly bad light. It wasn't ideal for various reasons, but so long as the integrity of the reference electrode was maintained properly, it did the job it was intended to do effectively and cheaply. There are a lot of misleading claims made these days about how any extra earth path on the installation side ("F" terminal of the ELCB) would compromise the working of the ELCB, but in fact that isn't the case - It's only anything within the resistance area of the reference electrode (on the "E" terminal) which would prevent the ELCB from doing the job it was designed to do.

 

[stillp]

Now that it's been considered obsolete for 25 years or so, I feel that the voltage-operated ELCB is often portrayed in an unfairly bad light.

I agree, this is another example of changing perception of risk. In most applications a VOELCB gave increased protection (compared to nothing at all) most of the time. They only really "became unsafe" when people started to place too much dependence on them. Also of course RCD manufacturers saw the chance of some extra sales...

 

[JohnW2]

If you only want a rough figure for the CCC of T & E, why not just look at the tables in BS 7671? I thought this discussion came about because you wanted to know how much "safety margin" was built-in to those tables.

As I understand it, the desire was for a rough estimate of the extent to which the BS7671 CCC figures are lower than the 'true' value of 'maximum safe CCC.

This question arose out of a discussion about the fact that many aspects of BS7671 (e.g. the allowance of ring final circuits, and acceptance that I2>In for an OPD) was reliant on some (unknown) 'safety margins' being built into BS7671's tabulated CCC figures. I suggested that it would be intellectually more satisfactory if we were told what the 'absolute limits' really were, so we could see what 'margins were operating - and thereby, in some cases, be able to design circuits more 'properly'.

In my experience it's quite common for safety-critcal components in aerospace to be only type-tested and to depend for their integrity on tests of the material. There's little point in sample testing for small production volumes. I accept that things are very different in your pharmaceutical world.

It's by no means only in the pharmaceutical world - but that certainly is one field which corresponds to what I described, and I think that's probably what the public expects. Would you be happy to take a medicine from a batch which had not been tested to confirm that it appeared to comply with the spec? ... and if your answer is 'yes', does that remain the case if I tell you that an appreciable (although obviously small) proportion of production batches of pharmaceutical products are rejected because batch samples fail to meet some spec/safety requirements?

Albeit a long time ago, my father was a Standards Engineer in the aeronautical (just about 'aerospace' at the time!) industry, and was involved in inspection and QC in that industry for a long time prior to that. In those days, QC was certainly paramount but Statistcal QC (i.e. sample-based) was seemingly little expected in that industry (and, anyway, as you say production volumes are generally small). hence, in those days, there was often 100% non-destructive inspection/testing, often with destructive testing on a sample, where appropriate. From what you say, that may have changed - although I have to say that I would be surprised if it had.

Kind Regards, John.

 

[stillp]

This question arose out of a discussion about the fact that many aspects of BS7671 (e.g. the allowance of ring final circuits, and acceptance that I2>In for an OPD) was reliant on some (unknown) 'safety margins' being built into BS7671's tabulated CCC figures. I suggested that it would be intellectually more satisfactory if we were told what the 'absolute limits' really were, so we could see what 'margins were operating - and thereby, in some cases, be able to design circuits more 'properly'.

That's what I thought, which implies a reasonable degree of precision in the measurement.

Would you be happy to take a medicine from a batch which had not been tested to confirm that it appeared to comply with the spec? ... and if your answer is 'yes', does that remain the case if I tell you that an appreciable (although obviously small) proportion of production batches of pharmaceutical products are rejected because batch samples fail to meet some spec/safety requirements?

You're not far wrong, but I would actually prefer my medicines to come from a manufacturer who was 100% certain that every dose is within tolerances. I understand the sampling rates can be very low for some relatively non-harmful drugs, and the variability between tablets means that some can be wildly different from the samples that were tested.

I think reliance on 100% NDT is rare these days, even in aerospace. It is sometimes argued that if a component needs that level of inspection the design must be placing too much reliance on it anyway.

Returning to CCC of cables, let's think about fuses - you can't 100% test their primary function, so you test the materials, check the dimensions, and check the resistance - just as you do for cables!

 

[bernardgreen]

There are a lot of misleading claims made these days about how any extra earth path on the installation side ("F" terminal of the ELCB) would compromise the working of the ELCB, but in fact that isn't the case - It's only anything within the resistance area of the reference electrode (on the "E" terminal) which would prevent the ELCB from doing the job it was designed to do.

Would the ELCB still work if a few yards of wire were connected between the F and the E terminals. ? In effect putting a short circuit across the coil inside the ELCB. Could the coil still detect earth leakage currents if it were shorted out by that wire between F and E ?

Now consider the CPC ( which is connected the F terminal ) being connected to a water pipe which is metallic into the ground which is also connected to the E terminal of the ELCB. It may not be a dead short as in near zero ohms but it was a shunt (bypass) to the coil.

If the voltage on the CPC relative to ground and the E terminal reaches 50 volts there will be enough current in the coil to trip the ELCB. But if the path from CPC to ground and supply star point ( the neutral ) via water pipes has a 1 ohm resistance then 50 amps of fault current will be flowing through the CPC and pipes before the ELCB can trip.

 

[JohnW2]

I suggested that it would be intellectually more satisfactory if we were told what the 'absolute limits' really were, so we could see what 'margins were operating - and thereby, in some cases, be able to design circuits more 'properly'.

That's what I thought, which implies a reasonable degree of precision in the measurement.

Oh, sure, to be able to do what I was suggesting would require a very high degree of precion, and we obviously would not be able to rely on any figures we had determined empirically ourselves. However, those who created the 'Tables with safety margins' surely must have been working from 'Tables without safety margins' - so they ought to be available to us. All that was being suggested was (I think) that it would be nice to have at lest a vague feel as to how large these safety margins might actually be (I speculated that cables may well be able to cope with 2-3 times their BS7671 CCC without harm - but that might be wildly wrong) - and that does not require much precision.

You're not far wrong, but I would actually prefer my medicines to come from a manufacturer who was 100% certain that every dose is within tolerances. I understand the sampling rates can be very low for some relatively non-harmful drugs, and the variability between tablets means that some can be wildly different from the samples that were tested.

It obviously depends upon the nature of the product. If it's a liquid (e.g. injection - when safety concerns are probably at their greatest), it's possible to mix an entire batch such that it is almost perfectly homgenous, in which case you will be very close to getting that 100% certainty, since the test results on any sample from that homogenous batch will be extremely close to reflecting the composition of each and every aliquot which is taken from that bulk batch and placed in a bottle or vial for sale.

As you say, with discrete items, like tablets (or nuts and bolts, or cables, or whatever), it's a bit different, and (unless 100% NDT is both possible and practical) one has to rely on Statistical QC and hence probabilistic criteria. However, given the vast size of production batches (whether tablets or nuts and bolts), and the fact that they can be thoroughly mixed so as to facilitate something approaching perfect random sampling, one can get extremely high degrees of confidence (obviously never 100%) that individual items (e.g. tablets) are not outside of specified limits with (in relative terms) fairly modest sample sizes - e.g. typically that the probability that any one tablet in the batch would be outside of those limits is 1 in 100,000 or 1 in 1 million. As you imply, the width of those acceptance limits will depend upon how potentially harmful the drug is, and that for relatively 'non-harmful' ones, will often be pretty wide.

I think reliance on 100% NDT is rare these days, even in aerospace. It is sometimes argued that if a component needs that level of inspection the design must be placing too much reliance on it anyway.

There's obvioulsy some truth in that argument, so long as its a situation in which safety can be increased by redundancy, but that's not always the case. Consider jet engine turbine blades (something which {along with 'screw threads'!} my father seemed to spend an awful lot of his time 'considering'!). One has no choice but to rely on it not failing in service. I would be astonished, even today, if there was not considerable 100% NDT applied to them.

Returning to CCC of cables, let's think about fuses - you can't 100% test their primary function, so you test the materials, check the dimensions, and check the resistance - just as you do for cables!

Yes, that's obviously another difficult area. When the most crucial function of an item relates to its failure (e.g. fuses, bolts designed to break at certain loads etc.), NDT is obviously of little use - it could confirm that the device's (required) failure does not occur below the desired level of 'stress', but can do nothing to confirm that it will fail at the specified stress. A lot of reliance on the sort of factors you mention is therefore required. However, harping back to my father, I recall him talking about bolts required to break under specified stresses (ISTR those used to attach jet engines to aircraft) - and telling me that there was sometimes a requirement to destructively test a very high proportion of items in a batch before it was deemed acceptable to use the few remaining in service. Very inefficient and expensive, but I suppose the closest one can get to 'safe'! ... and, again, discovery of the spread of failure characteristics of those items one subjects to such destructive testing enables one to make precise probablistic statements about the chances of any of the 'surviving' items from that batch being outside of acceptable limits.

Kind Regards, John.

 

[JohnW2]

Would the ELCB still work if a few yards of wire were connected between the F and the E terminals. ? In effect putting a short circuit across the coil inside the ELCB. Could the coil still detect earth leakage currents if it were shorted out by that wire between F and E ?

A voltage-operated ELCB will always do 'what it says on the tin', but it's important to understand that this is not directly related to 'earth leakage currents'.

Such an ELCB will operate if the voltage between its F and E terminals increases beyond the design limit - and its sole intended pupose was to limit the voltage between CPCs and true earth to that value (deemed to be 'safe' - which it did.

If you partially bypass the ELCB with a piece of wire (or parallel paths due to bonded e-c-ps etc.) such as you suggest then the voltage across its coil will obvioulsy decrease, and if it decreases to below the trip threshold, it will then not operate - but that's how it's meant to work, since the CPC-true earth pd is now less than what ws deemed to be a dangerous level. In effect, your bypass wire (or bonded e-c-ps) has actually reduced that pd to what is considered a 'safe' level, hence no need for the device to operate.

Such devices were obviously intended only to protect people who came into contact with CPCs and true earth whilst there was a significant pd between them due to an electrical fault. If the 'fault current' was flowing through the individual, then they probably didn't get much protection (I think it worked out as about 100mA required) and, as you say, any parallel paths would then work against protection of the human being (in fact, would usually increase the current through them).

Kind Regards, John.

 

[Paul_C]

Could the coil still detect earth leakage currents if it were shorted out by that wire between F and E ?

No, not unless we start talking about ridiculously high voltages, high enough to still cause enough current to flow through the coil despite the low-resistance shunt across it.

Now consider the CPC ( which is connected the F terminal ) being connected to a water pipe which is metallic into the ground which is also connected to the E terminal of the ELCB. It may not be a dead short as in near zero ohms but it was a shunt (bypass) to the coil.

That's why I said that it depends upon the integrity of the reference earth electrode. Yes, if there is anything else (connected directly to the "F" terminal) placed within the resistance area of that electrode, or if there is anything else connected to the conductor which runs from the "E" terminal to that electrode then it will affect the operation of the ELCB. But so long as that isn't the case, it doesn't matter how many parallel earth paths exist from the "F" side of the coil. It will increase the current which needs to flow through the fault, but the ELCB will still perform its intended function of tripping before the voltage exceeds the level deemed to be safe to touch.

 

[ban-all-sheds]

I accept that it would be much easier for us to work with a single insulated conductor, to get some preliminary answers.

I'd forgotten about Table 53.1.

As we know that at the rated current the fuse wire does not melt, nor does it (presumably) get extremely hot, we can learn some things of interest.

Like the csa of 80A tinned copper fuse wire is about 2.5mm²...

 

[JohnW2]

I'd forgotten about Table 53.1. As we know that at the rated current the fuse wire does not melt, nor does it (presumably) get extremely hot, we can learn some things of interest. Like the csa of 80A tinned copper fuse wire is about 2.5mm²...

Indeed, and that is broadly consistent with the 'fusing current' figures I posted - in particular, the fusing current of 197A for a 2.6mm² CSA copper conductor. If one assumes that the fusing current for an unenclosed fuse link is double or so the 'rated' current, then that is in the same ballpark as your 2.5mm² CSA for an 80A ('rated') fuse.

Kind Regards, John.