What's your average single phase measured voltage? Is 251V at point of measurement common?

[SimonH2]

Interesting. The dips to 230V are not surprising if you're that far from the transformer, but if you often get 253V athat far from the transformer, one would expect those close to the transformer to be getting well over 253V (the theoretical 'maximum')minimises I2R

Not necessarily. Don't forget that these days power flows can reversed in the local networks due to embedded generation (e.g. solar PV). So the voltage at the end of a line can be higher than at the transformer.
The DNO likes to keep the voltage up as it minimises I2R losses. Then embedded generation comes along and reduces or reverses voltage drops.

Fluorescents have a similar life issue to do with the heater filaments at the ends. (Any resistive heating element is going to have reduced lifetime by running hotter due to higher input voltages).

The heaters are only used at switch on - at least with switch start & magnetic ballast, not sure what electronic ballasts do.

Modern leds should be highly tolerant of voltage changes

Only for those with active electronics. Many only have passive electronics (rectifier & and a cap to drop the voltage) - these will be very sensitive to voltage. I have some LED lights in the garden - 80off 2W lights is "a bit much" so I run them off about 55V with a transformer.

 

[JohnW2]

Not necessarily. Don't forget that these days power flows can reversed in the local networks due to embedded generation (e.g. solar PV). So the voltage at the end of a line can be higher than at the transformer.

I think you're scraping the barrel a bit, at least at this point in time. In any event, nearly all of the 'embedded generation' is solar, so won't affect voltage anywhere in the network during the hours of darkness,

The DNO likes to keep the voltage up as it minimises I2R losses. .....

In the case of 'dumb' (usually resistive) loads (which most high-powered ones are), higher voltage will mean higher instantaneous current and hence higher instantaneous I²R losses.

What matters to the DNO is obviously 'energy loss' (I²Rt), rather than instantaneous power loss (I²R). As I see it, if the 'dumb high-powered loads are, say, thermostatically-controlled (which many/most will be), then the DNOs I²Rt losses will presumably be much the same (for the same power delivered to consumer loads), regardless of voltage, but those losses will still not be lower with a higher voltage. Isn't that the case?

Consider a simplistic and hypothetical example, with simple ('mental arithmetic'!) numbers.
... a resistive load with resistance 40Ω and a DNO network resistance of 1Ω. With a 200 V supply, the load draws 5 A, a power of 1,000 W - so, in, say, 1 hour consumes 1,000 Wh. The VD in the DNO's network is 5V, instantaneous I²R in that network is 25 W and the energy loss in 1 hour hence 25 Wh.
Now, hypothetically. assume that the supply voltage is doubled to 400 V, but that the amount of energy requires is still 1,000 Wh (e.g. to heat water in a cylinder, under thermostatic control). The current is now 10 A, now a power of 4,000 W, so to deliver the required (thermostatically-controlled) 1,000 Wh will take only 0.25 hours. However, the DNO's energy losses (I²Rt) are still 25 Wh, now over 15 mins rather than an hour.
In reality one would obviously not have voltages 'doubling';, but the arithmetic works for any increase, but my mental arithmetic can cope with the 'doubling'

.... Then embedded generation comes along and reduces or reverses voltage drops.

True but, as above, the DNO cannot rely on that (and particularly not at night ) when setting the output voltage of their transformer. An 'intelligent' transformer set-up could presumably adjust dynamically, but I'm not aware of that going on?

Kind Regards, John

 

[hovertoast]

Didn't stay up too late last night so could only get the following at 21:30:

Here is an example of a "sag" at 19:45, with a low load of 150w from my installation:

 

[hovertoast]

It's not just 'next door' the current going to every installation downstream of yours will go through the 'shared bit of cable' (between your installation and the transformer) that determines the voltage drop between the transformer and your installation.

I was referring to the cable that the neighbour and I share from the street, rather than all the way back to the transformer.

 

[JohnW2]

I was referring to the cable that the neighbour and I share from the street, rather than all the way back to the transformer.

Fair enough, but I didn't even know that you and your neighbour shared a bit of cable - most properties have their own feed from the DNO's main distribution cable.

However, what I wrote still stands, assuming that the transformer supplies many properties through a single 'main'.. With that assumption, in addition to gthe voltage drop in that bit of cable shared between you and your neighbour, both you and your neighbour will experience an additional voltage drop due to current being drawn by all downstream consumers flowing through the bit of cable upstream of you (between your shared cable and the transformer).

Of course, if the transformer supplied only you and one neighbour (through that shared cable), then that would not be true, but that would be an unusual situation in UK, other than in very rural situations.

 

[hovertoast]

All agreed @JohnW2. I did say in my original post that I was looped from next door and I was suggesting that the “lows” that I see might be due to next door using heavy loads.

Out of interest, do people agree that the 20+ volt difference I see throughout the day is somewhat unusual?

 

[ebee]

I could not make any sensible comment on that being on an urban supply and all that the peaks and troughs and numbers involved keep it more constant I should imagine.
out in the sticks and with only a few subscribers I imagine the variance shall me more marked an noticeable.
Somebody working for a supplier might have more info on that one.

 

[hovertoast]

Does anybody know whether "modern" lamps - LED, halogen etc or even fluorescent, are more or less tolerant of under or over voltage?

One halloween a few years ago we experienced a brownout. It was strange, as all the LED lights that we have remained at full brightness (as far as we could tell) but the hairdryer was running very slowly, the TV didn't work at all but the Sonos speakers did, and the one remaining filament lamp we had was very dim. The brownout event only lasted 10 mins or so, and then became a total loss of power. Very scary indeed, and in fact, the only power cut I can recall (other than the very brief ones often experienced in a thunderstorm). It's a shame I didn't get chance to measure the voltage coming in, but I suspect it was very low, as the hairdryer was unusable.

 

[JohnW2]

All agreed @JohnW2. I did say in my original post that I was looped from next door ....

You did, and I'd forgotten that - apologies.

... and I was suggesting that the “lows” that I see might be due to next door using heavy loads.

That will obviously be a factor. Do you know roughly how long the shared cable between you (and your neighbour) and the DNO's main is? One might expect that it would be fairly short, in which case your voltage drop due to the neighbour's loads might well fade into insignificance in comparison with VDs due to loads in the (possibly many) installations connected to the transformer through one main cable. Do you know roughly how many installations w're talking about?

Out of interest, do people agree that the 20+ volt difference I see throughout the day is somewhat unusual?

I personally can';t recall having seen swings that large, but I'm not sure that I've seen (m)any installations 350m from a transformer. I'm probably less than 100m from transformer, and, as I've said, have been known to see swings around 10 V (~240 - ~250), so it wouldn't surprise me if customers another 250m downstream (if any exists) saw much larger swings than I do.

I suppose it's in the DNO's interests to try to design their networks so as to maximise usage of the 'permitted range' (216.2 V - 253 V) - and that's 36.8 V wide.

 

[SimonH2]

In the case of 'dumb' (usually resistive) loads (which most high-powered ones are), higher voltage will mean higher instantaneous current and hence higher instantaneous I²R losses.

What matters to the DNO is obviously 'energy loss' (I²Rt), rather than instantaneous power loss (I²R).

As you say, a significant proportion of loads will be controlled - and a prevalence of switch mode supplies, including high power inverter drives, where the load current decreases with increasing voltage. I suspect we're still a long way off the latter cancelling the former. But however you look at it, higher voltages reduce the losses as a percentage of billable power delivered.

Incidentally, I know a few years ago, round our way the DNO (ENWL) were doing experiments with automated tap changing at the 132/33/11kV substations. I knew about it as a relative working for the DNO asked if we'd noticed anything - and I could see some step changes in the graohs I had from the UPS.
One was to drop the voltage slightly to get a short term power reduction.
Another was to unbalance loops to create circulating currents to add load (through I²T losses). E.g. instead of setting taps so that an 11kV ring was fed (roughly) equally at it's ends from two 132kV incomers, or it was split and run as two radual circuits, they'd change the taps to draw more power from one incomer, round the ring, and back out the other incomer (or more likely, just reduce the incoming power on it).
Both as ways of managing supply-demand balance. I don't know what the outcome was.

 

[hovertoast]

round our way the DNO (ENWL) were doing experiments with automated tap changing at the 132/33/11kV substations.

Interesting… ENWL are my DNO too (South Manchester/Altrincham), so maybe the drops are a result of this.

 

[plugwash]

Consider a simplistic and hypothetical example, with simple ('mental arithmetic'!) numbers.
... a resistive load with resistance 40Ω and a DNO network resistance of 1Ω. With a 200 V supply, the load draws 5 A, a power of 1,000 W - so, in, say, 1 hour consumes 1,000 Wh. The VD in the DNO's network is 5V, instantaneous I²R in that network is 25 W and the energy loss in 1 hour hence 25 Wh.
Now, hypothetically. assume that the supply voltage is doubled to 400 V, but that the amount of energy requires is still 1,000 Wh (e.g. to heat water in a cylinder, under thermostatic control). The current is now 10 A, now a power of 4,000 W, so to deliver the required (thermostatically-controlled) 1,000 Wh will take only 0.25 hours. However, the DNO's energy losses (I²Rt) are still 25 Wh, now over 15 mins rather than an hour.
In reality one would obviously not have voltages 'doubling';, but the arithmetic works for any increase, but my mental arithmetic can cope with the 'doubling'

Your math essentially assumes that all the "thermostatic loads" switch in sync.

If they switch at independent times then the DNOs energy losses will reduce.

 

[JohnW2]

As you say, a significant proportion of loads will be controlled - and a prevalence of switch mode supplies, including high power inverter drives, where the load current decreases with increasing voltage. I suspect we're still a long way off the latter cancelling the former.

Indeed, but although you slipped in "... including high power inverter drives", I would suggest that, at present, the great majority of high power loads are not supplied by SMPSUs and that the great majority of things supplied by SMPSUs are very low powered. Do you disagree with that.

But however you look at it, higher voltages reduce the losses as a percentage of billable power delivered.

That is true of (instantaneous) "power delivered", but it is energy (not power) delivered that is billable and hence what matters to the DNO - and, as I attempted to demonstrate last night, the energy losses is the DNOs network cables when supplying a particular amount of ('billable') energy would appear to be the same regardless of the supply voltage.

To recap on that argument, in relation to a simple resistive load (which, hypothetically can tolerate a doubling of voltage ), which is controlled by a thermostat or whatever, so as to limit the amount of energy consumed to a certain figure (e.g. that required to heat all the water in a DHW cylinder). If one doubles the supply voltage, hence also doubles current drawn (for as long as it IS drawn), then the I²R losses in the DNOs cable will increase by a factor of 4. However, the doubling of current and voltage means that the duration of current flow (to deliver the same amount of {'billable'{ energy) will decrease by a factor of 4, so that the energy losses in DNO cables (I²Rt) will be exactly the same, despite the change in voltage. Is that not correct?

A doubling of voltage is obviously unrealistic in the real world, but the same arithmetical argument applies to any degree of change in voltage (e.g. a 1% increase).

There definitely are advantages to DNOs in using high voltages (hence their HV distribution network), but they are primarily down to things like the decreased current facilitating the use of smaller CSA conductors (lighter, cheaper and more environmentally friendly), rather than any (I suspect incorrectly) perceived reduction in energy losses in the network (see above).

 

[JohnW2]

Your math essentially assumes that all the "thermostatic loads" switch in sync.

No it doesn't. In fact, I was considering just one load, with one thermostat switching it off when it had received the required amount if energy (e.g. when a cylinder of water was fully heated)

If they switch at independent times then the DNOs energy losses will reduce.

As above, I was considering just one load/thermostat, and appeared to demonstrate that the energy loss in DNOs cables would be the same (whilst delivering the same amount of energy) regardless of voltage - the increase in "I²R" losses due to the higher current being exactly cancelled by the corresponding reduction in time current would flow to deliver the same amount of power.

If, as in practice, there were many such independent loads/thermostats, operating 'out of synch', with one another, exactly the same arithmetic would apply to each of them - so, whether considered separately or together, DNO cable energy losses would presumably be the same regardless of voltage/ That's how I see it, anyway - am I wrong?

 

[JohnW2]

.... out in the sticks and with only a few subscribers I imagine the variance shall me more marked an noticeable.

I'm sure you'r right. We're essentially talking about a type of 'diversity' and that will always work better with 'larger numbers'.

If there is only a small number of installations, that greatly increases the probability that ('by chance') most/all of them will have many/most of their 'high loads' on simultaneously and also increase the probability that most/all will not have have many/most of their 'high loads' on simultaneously = so there is potential scope for large differences between minimum and maximum total demands of the installations put together.

With a large number of installations being supplied, things will usually 'even out' much more (but still somewhat constrained by things like 'dinner time', shower time' and 'being asleep time' being similar across most installations).

Kind Regards, John