Bulbs constantly blowing!

Do you live in an area where lots of houses have solar panels installed

What has that got to do with anything?

When the solar panels are feeding power into the local grid the supply voltage can be driven up to and sometimes beyond the upper limit of 253 volts ( 230 +10% )

Invertors in the solar panel controller should stop feeding power to the grid when the grid voltage is close to the upper limit.
 
Do you live in an area where lots of houses have solar panels installed
What has that got to do with anything?

Well, any "Solar Panel" source must produce a voltage higher than the local "Grid" voltage in order to "feed" electricity into the (local) network. (Note that this input feed is only "local".)
If there are a large number of such solar sources relative to the total number of households fed from the same transformer, it could raise the supply voltage throughout the area fed by that transformer.

Here our distribution authority has placed limits on the percentage number/wattage of "solar" inputs in any "area".
 
I presume the suggestion is that it can push up the supply voltage in the locality.
It can indeed - and does. To the point where (according to my insider knowledge) they have had to change taps in substations.
And it gets worse. When the sun isn't shining, then the losses in the distribution system reduce the users voltage - but there are times when the whole power flow is reversed because the panels are providing more power than is being used locally. Which makes me wonder about circuit protection ...*
The result has been that they've had complaints of excessive voltage at one extreme - but there's a risk of having too low a voltage at the other extreme (such as when the night storage heaters come on in the hamlet at the end of a long run).
 
Theory is if the voltage exceeds 253 the solar panels close down, and once they close down they are to stay closed down for a set time, also is below minimum voltage they must close down, idea is if they close down for a set time and only reactivate when the voltage is correct, if the supply is lost then the voltage will exceed the limits so if once closed they stay closed for set time, with a street full of solar panels they will not simply feed to each other and maintain a local supply but will fail so as not to cause a danger for some one working on the supply line.

However although the panels paid for by home owners will likely have the correct voltage set, those who hire roof space would be losing money if they close down so if set just one volt high then the home owners own panels will fail first, so it would seem some installers are setting the cut off point a volt too high.

As said in another post the transformers were set to highest voltage which kept them with in the range when under load so think allowed something like 36.8 volts to drop over the length of the run, at 100 amp that gives around 0.368Ω ELI on a TN-C-S supply, but if the transformer is set to 230 volt then the ELI would have to be better than 0.138Ω to be frank I have not seen that many houses with it that low, one solar panel in a line of houses OK, but a whole street of solar panels and there is likely a problem.

However the CFL and LED lamp are if anything more able to stand the voltage variations than the tungsten lamp, and even at 260 volt really there should not be that much of a problem, so although it could be rouge installers with solar panels, I would say more likely something else. My first thought is a poor connection on the supply neutral, this could result in 400 volt peaks, however it would need a voltage logger to see what is going on, it is quite likely that when the meter is connected the voltage is OK, but over the 24 hours you could get peaks.

Again since lights used at night and solar panels do not supply electric at night, it is unlikely to be solar panels.

Large motors can produce over voltage, specially with faulty inverters, I have seen generators over speed and voltage go really high with faults on a tower crane inverter, so if the house is close to an industrial estate that could be causing the problem.

But at end of day looking at a supply fault not some thing inside the house, so it is a phone call to the DNO and let them investigate, also talk to neighbours, are they having the same problem? If the DNO can't help then we start again, but seems likely an DNO fault not something inside the house.
 
It isn't clear what happens to the voltage in the house if the solar panel has to disconnect from the grid. Does the invertor ensure there is no transient peak on the solar supply to the house when the disconnection occurs. ?

But at end of day looking at a supply fault not some thing inside the house,
Provided the house in question does not have solar PV panels then yes Eric is correct
 
Last edited:
It isn't clear what happens to the voltage in the house if the solar panel has to disconnect from the grid. Does the invertor ensure there is no transient peak on the solar supply to the house when the disconnection occurs. ?
You seem to be implying that the solar panels remain connected to the house but 'disconnect' from the grid (incoming supply), but I thought that what they did (e.g. in response to inappropriate or absent supply voltages) was to 'shut down'.

Kind Regards, John
 
It isn't clear what happens to the voltage in the house if the solar panel has to disconnect from the grid. Does the invertor ensure there is no transient peak on the solar supply to the house when the disconnection occurs. ?
They just "switch off". I think there is more than just voltage monitoring - part of the standard they have to be designed against is that they must NOT be capable of islanded operation (providing a local island of supply that's disconnected from the grid). That gets more interesting with larger embedded generation - one of our customers at my last place installed an 89kW water turbine.
For those, it's in principle possible to supply a significant section of network that's been disconnected upstream. So AIUI the protection relay monitors not just voltage, but also frequency and rate of change of frequency (ie if the frequency is departing quickly from 50Hz then it can't be connected to the grid). G59 protection is the key thing to search for information on.

As said, it's to avoid (mostly) the risks of embedded generation backfeeding the grid with the obvious risks to personnel safety and also operations - if there's an island running at a different voltage and frequency, you need synchronisation ability to re-connect it without having some variation between "not a lot" and "very big bang" :eek: The network wasn't designed for that.

EDIT: And yes, a few people have been "a little upset" to find that their solar panels can't keep their power on during a power cut.
 
If there is a problem with the neutral connection back at the substation, then theoretically your house (along with other houses on the same 'phase') could be getting an excessively high voltage.

You say this is getting worse - if this is the problem, it could conceivably lead to other fried appliances, or in the worst case, fire.

I would get the voltage checked and see if it is within the normal range. It might not be this but it's not unheard of.
 
I’ll suggest that it could be poor contacts between the lamp and the lamp holder. Poor contacts = overheating. You don’t have issues with the chandelier because it could be newer? Before you go poking fingers/screwdrivers/chisels/hammers into lampholder, I suggest switching off at consumer unit, don’t rely on the light switch for electrical isolation.
 
I’ll suggest that it could be poor contacts between the lamp and the lamp holder. Poor contacts = overheating. You don’t have issues with the chandelier because it could be newer? ...
If we were just talking about one or two lights, then that could be a possibility. However, it sounds as if the problem exists throughout the OP's house (involving 203 lamps dying every week), everywhere other than in one chandelier. I suppose nothing's impossible, but it seems very unlikley that there's a problem with virtually every lampholder in the OP's house which is causing overheating and premature death of the lamps/bulbs.

Kind Regards, John
 
If we were just talking about one or two lights, then that could be a possibility. However, it sounds as if the problem exists throughout the OP's house (involving 203 lamps dying every week), everywhere other than in one chandelier. I suppose nothing's impossible, but it seems very unlikley that there's a problem with virtually every lampholder in the OP's house which is causing overheating and premature death of the lamps/bulbs.

Kind Regards, John

Arcing at the switch can be enough to damage the types of low cost electronic drivers in most CFLs or LED retrofits. You get broadband noise into the driver resulting in the input circuitry overheating.
 
I Notice that Rocky has not responded to this forum since posting, but regarding aptsys suggestion that "Arcing at the switch" may be causing LED light failure reminded me of this Video showing how
"Arcing contacts are really bad news for capacitive dropper based LED lamps as the dropper capacitors do their best to suppress the electrical noise and in doing so pass much higher current than normal. This stresses the rest of the circuitry and can kill LEDs and inrush/fusible resistors as has happened here".

Rocky, do your light switches need replacing?
SFK

 
Back
Top