Cable and Breaker Sizes

[JohnW2]

The thing is though, electrical installations typically last many decades. you might find your supply is towards the upper end f the permitted range today but it's anybody's guess if it will still be at the upper end of the range in 30 years time.

Yes, that's the theoretical problem, but it doesn't alter the fact that what matters is the (absolute) voltage supplied to the load and that (given the wide range of potential source voltages) the VD within an installation is an essentially useless indicator of that. The only certainty about the 5% recommendation is that it means that the voltage supplied to a load should never be below 205.39 V.

I would say the following:

1. I think one should question to what extent we need/should take into account things that might change in the (possibly distant) future, particularly when we don't seem to do so in relation to things much more important than within-installation VD. If, at a certain point in time, TN earth has a very low Ze, we will accept a circuit as being satisfactory/'compliant'/'safe' if the Zs at that point in time is only just over the minimum required for ADS with the OPD in question - yet, in "30 years time" (or even 'tomorrow'!) the Ze could rise and render the circuit 'non-compliant' because of its Zs. If we took into account concerns such as you mention, we would presumably require that the Zs was sufficiently low for it to remain below the permitted minimum even if/when the Ze rose to its 'permitted maximum', wouldn't we?
2. I also wonder, essentially for reasons I mentioned in my last post, whether we really need to be concerned at all about within-installation VD. Eric often talks about problems with fluorescent lighting when the voltage supplied to it is low, and I don't know how real an issue that actually is, but with that one possible exception I struggle to think of likely situations in which within-installation VD would (within reason) present any real problem at all.
3. Hence, particularly given that, for T+E >1mm², R1+R2 will always be greater than R1+Rn, I would suggest that if Zs is (currently!) 'compliant, then one doesn't, in practice, really need to think about VD.

Kind Regards, John

 

[JohnW2]

The worst case is a break in the ring, forcing all of the current down one leg, which should be picked up by regular checks

Sure, but I was obviously talking about an intact ring. Once one starts considering faults, anything is possible!

- even so the one leg on it's own is capable of maintaining around 65 to 90% of the full load of the ring. Not many people would have, or be able to afford such a load for long.

Indeed, and I often make that point myself. As I think I recently said, I think it very unlikley that, in a typical domestic setting, a 32A radial wired in 1.5mm² cable (maybe even 1mm²) would ever come to any harm. However, as some will inevitably point out, one cannot say that a problem "would never occur".

A fully functioning ring, would be even more difficult/ impossible to overload, even with all the load placed close to one end.

As above, I agree - but it remains the case that some people will remind us that we can't say that will 'never happen'!

Kind Regards, John

 

[winston1]

You missed a space. Are you stupid?

No. A space is not necessarily required. Perhaps you are ignorant.

Eric often talks about problems with fluorescent lighting when the voltage supplied to it is low, and I don't know how real an issue that actually is, but with that one possible exception I struggle to think of likely situations in which within-installation VD would (within reason) present any real problem at all.

He does, and I've done some tests with a Variac and old fittings and never encounted the problems he mentions.

 

[bernardgreen]

He does, and I've done some tests with a Variac and old fittings and never encounted the problems he mentions.

The Variac introduces inductance in the supply close to the fitting. Very likely to create voltage spikes when the starter opens. Spikes which are large enough to make the tube strike even though the RMS voltage is low. Better to test with voltage dropping resistors.

 

[Harry Bloomfield]

The Variac introduces inductance in the supply close to the fitting. Very likely to create voltage spikes when the starter opens. Spikes which are large enough to make the tube strike even though the RMS voltage is low. Better to test with voltage dropping resistors.

Dropper resistors would not be a fair test, because the voltage will depend on the load and if testing a fitting using a choke, the heater load is quite high until the starter opens.

 

[JohnW2]

He does, and I've done some tests with a Variac and old fittings and never encounted the problems he mentions.

That does not surprise me, since eric's reported experiences always have surprised me. I would have thought that anything sold for use in the UK would be 'fit for purpose' when supplied with 216.2V, if not less.

 

[JohnW2]

The Variac introduces inductance in the supply close to the fitting.

So does the 'ballast' of the fitting

Better to test with voltage dropping resistors.

That would callfor pretty meaty resistors. Incandescent bulbs might be a more realistic choice.

Kind Regards, John

 

[Harry Bloomfield]

So does the 'ballast' of the fitting

That would callfor pretty meaty resistors. Incandescent bulbs might be a more realistic choice.

Kind Regards, John

Which would be even worse than resistors, their resistance varies dramatically as they heat up.

 

[ericmark]

The case where the fluoscent lamp failed was when using 25 lamp all 58 watt on a 16 amp 110 volt supply in a long tunnel on the building of Sizewell 'B'.

Both overload and under voltage was corrected by selecting 127 volt on first 20 lamps on the auto transformer, and 110 volt on last 5, however this was before the electronic ballast.

The other was using a 65 watt ballast with 58 watt tubes, so the volt drop problem was made worse due to using the wrong tube.

207 to 253 is the voltage that solar panels and EV charging points are required to work within when using a TN-C-S supply, so that is seems to be the accepted range.

Some times one needs to consider the real world, and think of the regulations as a guide, not law.

 

[winston1]

58w T8 tubes are designed as replacements for 65w T12 types and the ballasts are often marked as for both.

 

[JohnW2]

Which would be even worse than resistors, their resistance varies dramatically as they heat up.

True. In any event, I really don't understand bernard's problem with winston's Variac. Is he really suggesting, I wonder, that a fluorescent fitting would behave differently if fed through, say, a 1:1 transformer than if it were fed directly from 'the mains' (which itself comes from the secondary of a transformer?

Kind Regards, John

 

[bernardgreen]

I really don't understand bernard's problem with winston's Variac.

The Variac affects / removes capacitive coupling that is part of the circuit for the high voltage spike that causes the tube to strike.

When the starter opens the magnetic field in the ballast collapses and this creates a high voltage spike across the ballast, This spike is the voltage that strike the tube.

The Variac will also create a spike when the starter opens (*), this spike might be in phase with the spike from the ballast ( improves striking ) or it might be anti-phase to the spike from the ballast in which case it will reduce the strike voltage across the ends of the tube.

(*) Some people may question this. The ignition coils in petrol engines create the voltage for the spark plugs when the contact breaker opens. Same with the inductive ballasts in fluorescent lighting units

 

[JohnW2]

The Variac affects / removes capacitive coupling that is part of the circuit for the high voltage spike that causes the tube to strike.

I need to think about this, but I'm not sure that I'm convinced by what you suggest. You seem to be implying that if there were no 'PF Correction Capacitor' (which would have a 'functional purpose' beyond just PF correction) and if the path to the power source was very short ()e.g. as I suggested, a nearby 1:1 transformer), hence minimal L-N 'stray capacitance', that the light would then not work. Is that your belief?

Kind Regards, John