Voltage drop

[JohnW2]

In real terms with 45 years in the trade that was the only time I found something would not work simply because of volt drop.

I think that probably says is all.

Kind Regards, John

 

[scousespark]

In real terms with 45 years in the trade that was the only time I found something would not work simply because of volt drop.

I think that probably says is all.

Kind Regards, John

Out of interest, what does this say to you?

 

[JohnW2]

Out of interest, what does this say to you?

Good to see you. What is says to me is that, for virtually all practical purposes (in domestic installations, which is what this site is meant to be about), VD is not something one needs to be concerned about.

To have an arbitrary guideline 'VD limit' of 5%, without any reference to the prevailing supply voltage (which is allowed to vary by about 15%) does not really make a lot of sense. For the majority of people who have a supply ~240V, a 5% (of 230V) drop within the installation will mean that equipment is being supplied with ~228.5V, which is still about 10% above the figure (204.7V, aka 230V - 6% - 5%) which, by implication, the regs would apparently deem to be acceptable.

Kind Regards, John

 

[ban-all-sheds]

And what if they start with a supply voltage of 204.7V?

 

[JohnW2]

And what if they start with a supply voltage of 204.7V?

Well, as you know, that's not allowed, so I presume you meant 216.2V.

That's obviously the potential theoretical issue, but I wonder how realistic it actually is. Given that the great majority of us have supplies around, or over, 240V, I wonder how many there really are whose supply voltages are ever even remotely as low as 216.2V? It seems to me a bit like having a national speed limit of 20 mph, on the grounds that there is a tiny minority of drivers who are not safe driving any faster than that!

All I'm really saying is that I feel that common sense should prevail. Despite the fact that many seem to think otherwise, the regs allow discretion - the 5%/3% guidance figures in the Appendix are merely ones which are 'deemed to satisfy' the regulations.

Kind Regards, John

 

[ericmark]

I look at what will go wrong or fail with too much volt drop. So an old radio power supply would have ripple on output of power supply producing a mains hum on the equipment. But today with switched mode power supplies this is not really and issue. With inductive controlled fluorescent lamps they will fail to light, with HF types they are really not affected. With tungsten lamps run on mains supply they will be dim, with switch mode extra low voltage again no problem.

So in general the problem of volt drop is becoming less and less. I suppose a fan heater could have the fan stall but heater still work but there should be a thermal fuse which would switch it off.

So lets say 7% instead of 5% volt drop starting at lowest supply voltage of 216.2 volt means 201 volt instead of 205 volt even at 10% that's 194.6 volt and most switch mode devices will run 150 ~ 250 volt so before even at lowest supply voltage one has a problem with volt drop it needs to be double the permitted limit.

As I said before I looked into the measuring of volt drop. To measure we need to measure voltage with and without a load. Also it needs to do this quickly so other items don't effect the reading. Lucky we already have a meter to do just this. I can either be set to read in amps or ohms switchable from loop impedance to prospective short circuit current. Clearly measuring live - neutral not line - earth.

When I constructed the java script program I was pleased with the results. However if one is going to condemn some ones work one needs to allow for error. This is where it all failed. Take any loop impedance tester and test same circuit twice and one will get a different reading. So to start with we need two readings, one for the supply impedance and one for the end of circuit impedance. Taking one from the other we have the impedance so looking at a ring final as an example. Incomer 0.35Ω and mid point on ring 0.94Ω will give us 106 meters of cable and a 11.4651650917593 volt drop but if really 0.34 and 0.95 volt drop 11.8 and 0.36 and 0.93 it's 11 volt so just 0.01 error in reading gives 0.8 volt error. Or 103 to 112 meters if looking at cable length.

To be frank meters tend to vary far more than that so in real terms looking at more like 5 volt error before you could be certain some one has made an error. This is why I give up with the idea of testing volt drop. It has to be really bad before one can say it is 100% installation error not a meter error.

So if you do find a loop impedance of say 1.2Ω line - neutral centre of ring final which means 16.5 volt drop instead of 11.5 volt what would you do? Likely change of owner or 10 years since last test. So do you really think the owner can get the electrician back 10 years after a re-wire and get him to split it into two circuits because he had miss calculated the volt drop? The only time you are likely to be able to show the electrician made an error is when something does not work days after the job which is unlikely

 

[JohnW2]

As I said before I looked into the measuring of volt drop. To measure we need to measure voltage with and without a load. ... So to start with we need two readings, one for the supply impedance and one for the end of circuit impedance. Taking one from the other we have the impedance so looking at a ring final as an example.

You highlight another anomaly with all this interest in VD within an installation. Consider a supply with an external L-N loop impedance of 0.35Ω and a "40A shower". The voltage at the shower will then fall by 14V (aka ~6%) when the shower is switched on, even if the final circuit introduces minimal additional VD (i.e. very short and large CSA).

Indeed, if one had a total installation load on the installation (with the shower on) of, say, 60A, an external L-N impedance of 0.35Ω and the 'permitted' 5% VD in the shower's final circuit, then, if were were unlucky enough to start with the minimum permitted supply voltage (216.2V), then the shower would only be getting about 184V.

Kind Regards, John

 

[ericmark]

Not sure I am following that, max volt drop at the head is 6% i.e. with the full permitted load. At one time I would have assumed no load at transformer voltage would be at max i.e. plus 10% however with micro generation that's really no longer an option so let us assume 251.2 volt at transformer no load giving just a little leeway, so at 100A they are allowed a 35 volt drop so the maximum impedance will be 0.35Ω. Oh where have I seen that figure before?

With TN-S earth may not be same size cable as neutral so possible the earth loop impedance may exceed 0.35Ω same within the installation so very possible mid point with a ring final that the earth loop impedance is 1.37Ω but the neutral impedance is 0.94Ω. So with both supply on max volt drop and installation on max volt drop we have 205.39 volts remember the - 5% is of the reduced voltage not the nominal supply voltage.

So for a ring final the prospective short circuit current line - neutral should be more than 218.5 amp and the loop impedance less than 0.94Ω.

But my point is at what point, considering accuracy of measurement, would you highlight it as a fault. You may be sure of load in the premises, but not the rest of the houses supplied. Even using an average of 5 reading ignoring any well out of range 200 amp or 1Ω could be within the limits it is just your readings that are out. Only real way is with a low ohm meter.

Next is the last time it was calibrated, you would want it calibrated before and after to be sure readings were spot on including the leads. Now I remember sending a PAT tester for a traceable calibration and how it had to be returned because they could not find the traceable record and how on return they admitted it could not be calibrated to the new figures for PAT testing. So I don't have much faith in calibration.

To test the mid point before extending ring final yes, see the point, but as part of an EICR can't see the point as unlikely even if out anyone would pay for it to be corrected.

The only point is of course watching ones own back. If you don't test and the next guy does are you responsible for work to correct or the original installer? Since volt drop is not a danger being over the limits would not be coded so it would only be in comments. So a comment volt drop seems to exceed laid down limits, but this is unlikely to affect operation or cause danger could be put down but really why bother.

 

[ban-all-sheds]

Not sure I am following that,

We should strap some magnets to Hr. Dr. Ohm, and wrap some copper wire around his coffin

If the supply impedance is 0.35Ω then every extra amp you draw will cause the voltage to fall by 350mV.

max volt drop at the head is 6% i.e. with the full permitted load. At one time I would have assumed no load at transformer voltage would be at max i.e. plus 10% however with micro generation that's really no longer an option so let us assume 251.2 volt at transformer no load giving just a little leeway, so at 100A they are allowed a 35 volt drop so the maximum impedance will be 0.35Ω. Oh where have I seen that figure before?

Where did the 100A come from?

I thought it was several '00A?

 

[ericmark]

Your saying a 100A supply with a 0.35Ω impedance would have a max volt drop of 35 volts, which is exactly what I also said. Assuming no load transformer is set to max then with max load it is still within limits. In real terms unlikely just one house on a transformer so if transformer can give 500A and houses even spaced then we would likely work it out at 250A at end house. So the loop impedance would be more like 0.13Ω to allow for with maximum draw still within limits. But that would be at the street cable the cable from street to house would likely mean at DNO head more like 0.2Ω. And that is volt drop is never exceeded I think they are allowed a short time with over the -6% volt drop.

Years ago there was no problem having transformer at max voltage with no load. Today however we have people trying to export power. At 253 volts the grid tie inverter should close down, also at 205 volts it should close down, and it should remain closed down for a set time. The idea is if the supply is lost either the voltage will go high or low as it has loss transformer regulation so within a very short time the grid tie inverters on that supply will all close down.

So either way high or low voltages out of range means people can't export their power. One solar panel in the street no problem. Every house with solar panels and I don't know how they will work?

But we are not really worried. All we worry about is the internal volt drop. The only reason to measure external volt drop is to work out how much is inside and how much is outside the property. But in real terms if the volt drop is less outside we can get away with more inside so.

So for a ring final the prospective short circuit current line - neutral should be more than 218.5 amp and the loop impedance less than 0.94Ω.

 

[ban-all-sheds]

Your saying a 100A supply with a 0.35Ω impedance would have a max volt drop of 35 volts, which is exactly what I also said.

Another thing you also said was that you didn't follow the argument that, at whatever voltage was present at the service head, if you whack another 40A load onto the cable that voltage will drop.

 

[ericmark]

Your saying a 100A supply with a 0.35Ω impedance would have a max volt drop of 35 volts, which is exactly what I also said.

Another thing you also said was that you didn't follow the argument that, at whatever voltage was present at the service head, if you whack another 40A load onto the cable that voltage will drop.

Did I? I was assuming potential volt drop. The volt drop that would happen if the circuit was loaded to it's maximum. So add another 40A and the potential volt drop is 230 volt. In other words the protective device has tripped so there is no voltage.

Indeed, if one had a total installation load on the installation (with the shower on) of, say, 60A, an external L-N impedance of 0.35Ω and the 'permitted' 5% VD in the shower's final circuit, then, if were were unlucky enough to start with the minimum permitted supply voltage (216.2V), then the shower would only be getting about 184V.

The voltage will not drop until we have a load so if we have 216.2V at no load then clearly the supply is out of limits. We should have 216.2V with full load. This is the whole idea of a permitted volt drop it's the difference allowed between no load and full load volts.

What John seemed to be implying was we could have a voltage of 216.2 with no load.

What I would say the 0.35 and 0.8 ohm figures we tend to use as TN-C-S and TN-S limits I have never seen written down on official documents. So it is possible where a 60A DNO fuse is used we could get a higher reading.

I remember asking for a TN-C-S supply with my mothers house and the DNO electrician measuring the loop impedance which was 0.36Ω he said it's near enough and gave us a TN-C-S supply. Not until much latter did I consider if only just within limits what does the voltage do in the street considering my mothers house was first to be built so when the supply was put in there was no street.

 

[bernardgreen]

What John seemed to be implying was we could have a voltage of 216.2 with no load.

There may be no load in your house but if there is a load next door and you at the end of along feeder then the supply to your property will drop significantly.

At one time the 230 volt feeder to a group of 13 houses also fed the village sewer pumping station, When those motors kicked in lamps dimmed, after heavy rain ( dilute sewage ) they came to speed quickly and the dimming was short duration. But when it was dry weather and the sludge was thick those pumps took a while to get up to speed. Now it has it own 11kV feed and tranny

 

[JohnW2]

Not sure I am following that, max volt drop at the head is 6% i.e. with the full permitted load. At one time I would have assumed no load at transformer voltage would be at max i.e. plus 10% however with micro generation that's really no longer an option so let us assume 251.2 volt at transformer no load giving just a little leeway, so at 100A they are allowed a 35 volt drop so the maximum impedance will be 0.35Ω. Oh where have I seen that figure before?

I'm rather confused, because there is a lot which doesn't seem seem right about all this....

For a start, IF you are right in saying that the 230V + 10% -6% limits refer to the voltage at the origin of an installation when the "full permitted load" (which I take to mean the rating of the cutout fuse) is being drawn then, as you say, if the cutout fuse is 100A and the external L-N impedance is 0.35Ω, then the permitted range of no-load voltages at the origin of the installation would be 215.2V - 253.0 - which is clearly impractically tight. Even with the much-more-common "80A supply", the permitted no-load voltage would still only be 244.2V - 253V (and, as below, even that assumes no other users connected to the local network).

Much more to the point, these discussions/calculations assume that there is only one consumer connected to the relevant part of the LV network. In reality, there are likely to be many, potentially dozens, of installations sharing much of the path back to the transformer. If just a few of them had loads even remotely approaching the "maximum permitted load", then the VD in the network would be dramatically greater than the sort of figures you are quoting. Indeed, without even going to extremes, if, say, a dozen consumers were simultaneously using, say, 20A, that would mean a network VD of 84V if the L-N impedance of the common part of the network were 0.35Ω. I realise that diversity means that even this will rarely occur, but it could. Indeed, it would only take a few houses with night storage heaters for the night-time VD to be very large.

In fact, as you calculations demonstrated, if one installation were drawing 100A, it would only require other users on the local network to be using a total of about 5A for it to become impossible for the DNO to satisfy the "230V -6% +10%) both at 'no load' and 'full load'.

Kind Regards, John

 

[ericmark]

I would think 0.35Ω is the worst reading acceptable and in the main it would be far better. As John rightly states with many houses on a supply then to have a 0.35Ω loop impedance (line - neutral) would not really be good enough. I am sure the 500A cables in the road are far heavier than the coaxial cable that feeds the house and so trying to second guess where volt drop is greatest is maybe pointless, also as to how many houses are supplied from each phase or how much current flows in the neutral. I would have to rake out all my Uni books and swat up on the two methods to calculate.

I do see the problems with large pumps. I have seen many pumping stations and also the efforts taken to ensure the volts don't dip. I know I have been told do not run more than two pumps the supply is not good enough, and seen both modern inverter and softstart and the old resistor start not uncommon to have 5 stage starters to reduce the dip on local lights as the beasts fired up. Even seen auto-transformer start which compared with star/delta is rather an expensive option. I have worked for SLD pumps and North West Water only one site did they have 11kV pumps. I was to advise on safety isolation and E stops with that site I just back healed it and said they needed some one with a high voltage license and training good by.

Oddly biggest pumps were at a power station they removed the water from the condenser so were pumping from a far lower than atmospheric pressure seem to remember again 11 kV some silly draw like 1/2 MW can't remember exact figure. But under a generator producing 750 MW so likely no problem with supply. Seem to remember 8 x 1.5 MW stand-by diesel generators to ensure supply when main turbines not running.