240v AC line voltage fluctuations

[TrevorP]

Hi, so, got some alarms/warnings from our large UPSs last night, went to investigate, and our AC line voltage here in rural NSW was up to 270v. Didn't really cause me any issues (line interactive UPSs with multi-tap transformers), but I've spoken to others locally with smaller UPSs that switched to battery to maintain power quality.

Even more interestingly, the increase was a sharp rise at 10pm last night, and a sharp return at 9am. Graph from our UPS monitoring below. I also confirmed the voltage last night with a cheap plug in "Kill A Watt" style power meter - https://www.jaycar.com.au/mains-power-meter/p/MS6115

Whats going on? Did anyone else observe this? What could cause that? I've seen the voltage fluctuate in the past, but not by this much. Aren't the standards to provide between -6% to +10% of the nominal 270v for NSW?


- image from https://twitter.com/trevorp/status/917946960480321536

Take care all, thanks.

[Brumby]

The standards are 230V +10% -6% and have been since 2000.  Prior to that they were 240V +6% -10% (Do the math, it will make you smile.)

Such a large fluctuation (I would expect) typically be the result of a large load being removed at the start and then reconnected at the end.  I would normally expect a tap changer to step in and adjust the level within a few seconds.  I understand rural areas can have their own challenges - but since I'm not familiar with how things are actually structured, I won't speculate further.

270V is too high - and because of such an extended period, I would be asking your electricity supplier for a "please explain".

[wasyoungonce]

Yikes 270V AC on a 230V system supply?   Are you in the sticks (country area)? using Single Wire Earth Return (SWER)?  Any big power hungry business around?

In anycase its way outside standards!  Local authorities can put on remote measuring/monitoring  devices.  Get onto this this can damage your toys!

[richard.cs]

A lost neutral could also explain it, but the relatively constant but high voltage makes that unlikely (as it would imply a reasonably constant load-balance the whole time).

I don't suppose you have logs from any other devices (someone else's maybe) on different phases?

[digsys]

Seen this a few times over the years, but mainly in industrial parks. As commented, 1 case was a bad neutral, with a very large factory next to us likely using mainly 1-2 phases.
In domestic areas, it happens a bit if you are near a large solar array (with not many houses in the line) and some poop-hole turns his V-Feed up. An electrical inspector friend constantly investigates
these poop-holes. They do it to try to get more feed-in $$s  In our cases, many cheap P/Supplies blew up (old PCs), luckily I pulled the breakers in time.

[Ice-Tea]

Nasty. About good enough to start firing and frying a few MOVs here and there I imagine.

[Old Don]

Rural voltage in the USA can vary quite a bit too. Years ago we had a customer in Wisconsin, I believe, that was having a problem with our process controller not working. They'd remove it from the machine send it in and it tested within spec. Put it back on the machine and it would act up again. We did this several times. We replaced it with a new unit and same thing. Finally we had a tech visit and record things that might determine the problem. Turned out that the 120VAC required to run the controller was dropping down to about 87VAC and the 220/440VAC lines coming into the building were at the end of a very long feeder line and at peak power load they were dropping with the same ratio. We installed a Sola transformer and solved the problem with our controller. Don't know if te plant had ongoing problems with burnt motors or other problems, but our equipment worked correctly after the Sola was installed.

[TrevorP]

Great info guys

@Brumby thanks for confirming the standard. Are you talking about a tap changer in a substation?

@wasyoungonce definitely in the sticks, though we are close (as the crow flies at least) to a substation (few hundred meters), with a 3-phase supply. Had to look up SWER, pretty sure thats not the case. Nearest industrial sites I'm aware of are 7km away in the neighboring town.

@richard.cs tell me if I'm thinking about this incorrectly, but that would also trip the ELCBs in the distribution box? Agreed its unlikely. No other logs. I did test a few other points in the building, my understanding is the points here are split between two phases, but not sure which are which, so cant be sure I saw another phase.

@digsys there is a new large commercial array 30km away (I'd call that close for here in the sticks), though this was observed at night.

@Ice-Tea good point, didn't think of that. Getting uncomfortably close to the typical 275v threshold.

[Brumby]

@Brumby thanks for confirming the standard. Are you talking about a tap changer in a substation?

Yes.

I must also admit I missed the rather obvious possibility of a neutral failure... 

[richard.cs]

If you have a three phase supply then you are definately not on SWER.

Lost neutral wouldn't automatically trip an ELCB*, but sometimes it can happen because you can have some leakage N to E in an installation, essentially a latent fault, that does not pass enough current to cause a trip until the N to E voltage rises.

*I am assuming here that you mean a device that works by imbalance in the L and N currents, in the UK we'd call that an RCD (USA: gfci) but we have an older device that we call an ELCB or VOELCB that detects earth leakage by a different mechanism with a separate reference earth rod.

[TrevorP]

@Brumby Thanks for clarifying, I know next to nothing about equipment in substations, had never occurred to me they'd be using multi tap transformers, though fairly obvious when you think about it. Guess its a case of neural failure.

@richard.cs Interesting the differences in terminology. Wikipedia tells me ELCBs can be voltage or current operated, All the devices I've seen (that I've always called ELCBs) have been rated by trip current (in the tens of mA in the typical home/small business) so I'm assuming they've all been equivalent to an RCD. I'll have to be more precise about my terminology in the future. Thank you.

[richard.cs]

Usual UK practice has multiple taps on substation transformers but they are fixed at installation, nearly always to give 250 V open-circuit. Voltage on the 11 kV network is then regulated over a wide area with automatic tap changers giving sufficient regulation on the LV side. I have seen tap-changing autotransformers at the end of remote rural lines, the LV cable was nearly a mile long in that instance, but they are very rare here.

[G7PSK]

I have seen such voltage rise when someone uses a welder with a large power factor correction capacitor, they leave the welder switched on at the mains and either have no lod on the welder or leave the tap switch between tapping's on the welder, on one occasion I saw the voltage go over 300. Farmers are especially good at this.