Where to start diagnosing high voltage spikes?

[von555]

Just looking for some advice.

I bought a UPS to protect a computer from occasional outages. It has been going to battery despite there being no power problems. After speaking to the manufacturer it seems the problem is one of high voltage input. Here's a nice graph I made from the readings of the UPS. Resolution is 2 seconds, so it's about 2 hours or so.

Untitled

• von555
• 5 Dec 2012
• 1

How do I go about working out what the cause of the high voltage spikes are. The UPS has a default shutdown of 266 volts and you can see on the graph that there are quite a few spikes of that magnitude.

I literally have no idea where to start with this one. Funnily enough I had a sparky come visit today (install some sockets). He didn't mention noticing anything dodgy. I hadn't got this far then, so didn't ask him about this at the time.

If the problem is in the house, what are the likely candidates?

 

[Monkeh]

Nothing to do with the house.

Your voltage is unusually high (well out of spec, actually). Verify the results with a suitable meter (uhh, on second thought, ask the spark if he wouldn't mind dropping in for two minutes to check for you).

If it's really sitting around 255-260V that much, you need to get the DNO to fix it. If the spark's meter shows it to be more normal, your UPS is faulty.

 

[von555]

Already contacted the DNO. They'll be round tomorrow. This is the second UPS I've got in as many months. I assumed the first one was faulty. Seems unlikely that the second one is too.

Thanks for the info.

 

[westie101]

Yes you need to raise a Voltage Complaint with them for high volts as this looks to be above the 253 maximum for too much of the time.

ESQCR sets the supply voltage at 230V +10%/-6%. What it does not state is that this must be maintained for at least 95% of the time when monitored over a 7 day period (it's in a BS, but as I'm not at work I can't quote it). This does allow for volts to rise above that from time to time.

Have you or your neighbours recently fitted any solar panels? As these have been a cause of excess volts all over the network

 

[Monkeh]

Have you or your neighbours recently fitted any solar panels? As these have been a cause of excess volts all over the network

Lots of crappy inverters I guess. Mine don't cause any troubles.

 

[westie101]

Lots of crappy inverters I guess.

Not really, if the volts are already towards their upper limits the addition of generation can just tip them over the limit, in most cases it only affects the property that has the panels but could affect neighbours as well.

Generally we can reduce the voltage from the local substation but if that fails we may have to consider other actions which could include

Telling the customer, if only they are affected, that it is their problem as they are causing it.
or
If they are affecting other customers, requiring them to switch off the panels or face disconnection!

 

[von555]

Yes you need to raise a Voltage Complaint with them for high volts as this looks to be above the 253 maximum for too much of the time.

Have you or your neighbours recently fitted any solar panels? As these have been a cause of excess volts all over the network

I explained the situation to them. They said there were no issues with the area but an engineer would be round to check. No solar panels near me. There is a huge construction site not far (the new forth road bridge). Perhaps it's the cause?

Wouldn't the sparky have noticed anything when he was fitting new sockets?

 

[Monkeh]

There is a huge construction site not far (the new forth road bridge). Perhaps it's the cause?

Could be.

Wouldn't the sparky have noticed anything when he was fitting new sockets?

He might have dismissed it as simply a little high, if it's below 250V it's fine.

 

[westie101]

They said there were no issues with the area but an engineer would be round to check

I like that, there are no means of centrally monitoring voltage on the LV network that I have come across - a nice bit of reading from a script.

The electrician should have noticed it if the volts were above limits when he checked them (he did didn't he?) but they could well have been OK at the time. The DNO should install a voltage recorder and monitor the volts over a week.

 

[Monkeh]

They said there were no issues with the area but an engineer would be round to check

I like that, there are no means of centrally monitoring voltage on the LV network that I have come across - a nice bit of reading from a script.

No reported issues.

Wouldn't it make sense to have all these fancy smart meters which are rolling out report back voltage stats for DNO consumption? Hell, maybe they already do.

My solar inverter or meter could, too, both have provision for that.

 

[von555]

They said there were no issues with the area but an engineer would be round to check

I like that, there are no means of centrally monitoring voltage on the LV network that I have come across - a nice bit of reading from a script.

The electrician should have noticed it if the volts were above limits when he checked them (he did didn't he?) but they could well have been OK at the time. The DNO should install a voltage recorder and monitor the volts over a week.

When the sparky was here I only knew the UPS was playing up. No idea that it was voltage related. So didn't really occur to me to ask.

I will keep monitoring it and see what the DNO says. Least I feel a bit more confident that it is a voltage problem and not a faulty UPS. Thanks.

 

[westie101]

At least you've got a UPS that's been properly set up, the schools round here had a batch that would alarm at 250V that caused some fun until we realised they were not set up for the UK

 

[JohnW2]

Least I feel a bit more confident that it is a voltage problem and not a faulty UPS.

I fear that you can't really be all that confident yet. It could be that your supply voltage is 'normal' and that the only thing wrong is the calibration of the voltage measurement going on in the UPS - leading to incorrectly high reported voltages and hence fooling the UPS into thinking that the voltage often goes over the 266V shut-off threshold.

Kind Regards, John

 

[Monkeh]

Least I feel a bit more confident that it is a voltage problem and not a faulty UPS.

I fear that you can't really be all that confident yet. It could be that your supply voltage is 'normal' and that the only thing wrong is the calibration of the voltage measurement going on in the UPS - leading to incorrectly high reported voltages and hence fooling the UPS into thinking that the voltage often goes over the 266V shut-off threshold.

Kind Regards, John

He's already had it replaced once..

 

[JohnW2]

He's already had it replaced once..

True, but the first one may have (coincidentally) had entirely different problems. The voltages being reported by the UPS are so high so it only needs someone with a meter to determine if thsoe figures are in the correct ballpark or not - I was just wondering whether the DNO would/will charge for attending if it transpires that the UPS's voltage measurement is the problem.

Kind Regards, John