High voltage meters

[Graham3232]

Hi,
I know this a DIY forum but maybe someone here can help with a metering issue…

I have an 11kV supply to a building which is fed from a HV substation nearby. The DNO meter for the supply to the building is outside the substation in a cabinet and meters the 11KV output from a transformer specifically for the building (I have been in the substation with DNO to verify this)

There is then an 11 KV transformer in the yard of the building with LV supplies coming directly into a distribution cabinet which is then privately metered at LV.

The 11 KV supply cable, transformer in the yard etc. are all privately owned.

I hope the above makes some sense, anyway the problem is the building is empty apart for security lights, cameras etc. The LV metering shows consumption of 12KWh per day. The HV meter is showing an average of 500Kwh per day! It is a half hour meter with data collector and I have data but it all seems random.

I have taken various readings on both meters, had clamp meters on LV tails and had the DNO and supplier out to confirm the setup, all appears in order apart from the readings not matching up between LV and HV. Both meters read 0 when the LV supply is off, and I have traced each DB across the building and most are off so there is nothing obvious untoward.

I have been told that HV meters do not pick up small loadings well? Anyone got any experience of this or any ideas before I go down the route of having a check meter fitted at cost.


£10K of electric in an empty building over 5 months

 

[JohnW2]

... the problem is the building is empty apart for security lights, cameras etc. The LV metering shows consumption of 12KWh per day. The HV meter is showing an average of 500Kwh per day! It is a half hour meter with data collector and I have data but it all seems random. ... I have taken various readings on both meters, had clamp meters on LV tails and had the DNO and supplier out to confirm the setup, all appears in order apart from the readings not matching up between LV and HV. Both meters read 0 when the LV supply is off, and I have traced each DB across the building and most are off so there is nothing obvious untoward.

If the HV meter reads zero when the LV installation is completely 'off' (but high readings when some of the LV DBs are 'on'), have you tried switching the LV DBs on one at a time in order to narrow down what's going on?

Kind Regards, John

 

[Graham3232]

Hi John,

Again it seems random, if a LV board is isolated the HV readings drop but when isolating individual circuits on that board one at a time the readings change down but not by the total amount as when the whole board is off and even then I can isolate the same circuits in different combinations and get readings varying by up to 6Kw on the HV meter ?!? I'm concerned with the overall accuracy of HV metering.

 

[JohnW2]

Hi John, Again it seems random, if a LV board is isolated the HV readings drop but when isolating individual circuits on that board one at a time the readings change down but not by the total amount as when the whole board is off and even then I can isolate the same circuits in different combinations and get readings varying by up to 6Kw on the HV meter ?!? I'm concerned with the overall accuracy of HV metering.

From what you're saying, it does sound as if there is something very wrong with the HV metering. If the DNO are aware of how little load there is in the building (plus the LV meter readings) they surely must at least suspect their HV metering, which implies a constant average load of about 21kW (around 90A at LV)?

If this were a domestic/commercial LV metering issue, the DNO would put in a 'check meter'. Do they not do something similar with HV metering?

Kind Regards, John.

 

[DetlefSchmitz]

Already mentioned in the OP.

I suspect the problem is that you have to pay for this if it turns out the meter is right. However probably wouldn't be significant in this case.

 

[JohnW2]

Already mentioned in the OP. I suspect the problem is that you have to pay for this if it turns out the meter is right. However probably wouldn't be significant in this case.

I think the same is true of domestic/commercial, isn't it (that there is a charge if it turns out that there was nothing wrong with the meter). However, as you say, in this case the charge would presumably be insignificant in relation to the HV-meter-based billing and, in any event, it would seem almost beyond belief that the HV metering can be right - although we don't know the OP from Adam, it is almost impossible to believe that he can somehow be 'overlooking' 500kWh per day of usage in an empty building, don't you think?!

Kind Regards, John

 

[Robin0577]

Hi,
all appears in order apart from the readings not matching up between LV and HV. Both meters read 0 when the LV supply is off, and I have traced each DB across the

Is the HV meter set up to take account of losses in the transformer? The fact that both meters drop to 0 with no load on the LV side suggest that it is but it does make me wonder if, for instance unbalanced phase load currents on the LV side could so odd things with the power factor that affect the HV meter in some way? I'm sure the designers must have considered this but then perhaps they never expected the system to cater for such a small load?

 

[Graham3232]

Thanks for the replies everyone. I have taken readings on the HV side this afternoon and they vary between 6 and 20 Kwh even though the LV side is at 0.5 to 1 kWh. The supplier says there is no issue as both meters read 0 under no load. And yes I have checked every circuit and DB and clamp meters on the LV tails to verify readings. There are other screens on the HV meter but apparently these aren't used as the meter resets ever half hour on auto billing cycle and sends the reading for that half hour.

It’s not something I’m familiar with but I was thinking along the line of ‘Robin0577’s post about transformer losses and the low usage and power factor as there was previously a high very usage with computers and air-con etc. All usage is single phase (but spread across the phases) at the moment which may make a difference.

It will cost a very lot to have a check meter done as it’s a DNO job in the substation apparently and we are already threatened with disconnection with a £1.5K reconnection cost!

 

[Robin0577]

Out of interest, what is considered to be an acceptable resolution/accuracy for metering of this kind?

How big is the supply in relation to the present loading?

My thinking being that say, a 1% error tolerance on a 500kVA supply is going to make hardly any difference when the supply is anywhere near fully loaded but with the building lying idle, +/- 5kVA would notice rather more.

 

[JohnW2]

Out of interest, what is considered to be an acceptable resolution/accuracy for metering of this kind? ... How big is the supply in relation to the present loading? ... My thinking being that say, a 1% error tolerance on a 500kVA supply is going to make hardly any difference when the supply is anywhere near fully loaded but with the building lying idle, +/- 5kVA would notice rather more.

That's a good point. However, even if the metering were designed for ('used to') a load of about 500kVA, I must say that I would not have expected it to have such poor resolution that it couldn't tell the difference between between 0.5kVA and 21kVA (which is what the OP implied). Scale that down to an 18kW domestic supply, and one would be talking about a consumer being charged for about 18kWh of electricity per day when they were only actually using about 0.4 kWh per day!

I'll be very interested to hear what those in the know have to say about this.

Kind Regards, John

 

[Lectrician]

Most HV metering reads in tens of units, not single units.

On sites we maintain, there are two meters, one operating as a constant check meter, and they take an average of both readings. They are often quite far apart.

The trouble with check meters on HV is that they still use the same CTs and VTs that are usually part of the supplying RMU, so even with a check meter, you are relying on the existing current and voltage transformers.

HV is inaccurate at very low currents, and you need to take into account the losses in the transformer. It's buzzing and getting hot, and not for free. That said, these losses are largest when loaded, and even then, around 95% or something like that.

We have sites with HV metering and LV metering, the LV either installed through choice, or as existing metering was already in place and adopted as part of an HV upgrade. In these sites, I have not seen discrepancies such as these.

I would hire an energy logger and check your LV meter, this would give you something to approach the supplier with. Not expensive, and it encourages the supplier to take action if you give them data.

 

[JohnW2]

Most HV metering reads in tens of units, not single units.

How does that work with small loads? The OP says that the LV metering suggests a usage of about 0.5 units per hour, and that the HV metering is half-hourly. If the LV metering is correct, that would mean 0.25 units per half-hour - the HV metering would surely not 'round that up' to 10 units, would it? If they did conventional up/down rounding (to nearest 10) and if usage really was 0.25 units per half hour, then all the half-hourly HV readings would be zero

I would add that even if 0.25 got rounded up to 10 every half hour, even that would only account for about half the discrepancy the OP has reported.

Kind Regards, John

 

[Lectrician]

All the recent HVs I have been involved with have been tens of units. You take a reading, multiply by ten (add a zero). The meter clearly states this with engraving or a sticker.

 

[JohnW2]

All the recent HVs I have been involved with have been tens of units. You take a reading, multiply by ten (add a zero). The meter clearly states this with engraving or a sticker.

Well, as I said, if (hopefully!) the readings were 'rounded to the nearest 10' (or 'rounded down' to the next lower 10, which is what would happen if the meter only incremented when a whole 10 units had been consumed), any half-hour load less than 5 (or 10) units would presumably be displayed/read as zero - which is quite the opposite of what the OP is observing!

Kind Regards, John

 

[westie101]

he poiint to look at with the HV taarrif is that most are not just about units used!
Most include costs for

Network demand availaability

Actual demand

Power Factor

Units used

You need to carefully investigate the tariff the site is on with the electricity supplier and see the full details of the supply.

It is not unknown for some to have a time element where network availability charges must be paid for a fixed time irrespective of load