RCD tripping problem

[Solarexploits]

Unusual installation, I have a solar installation (DIY) with battery backup, it powers a no. of circuits within the house separate to the rest of the mains circuits, so, effectively a lot of the house is off grid. There is an RCD on the mini consumer unit which is powered by the inverter, no mains to this CU. Everything works perfectly until overnight when the battery charger on the inverter is on, it doesn't usually happen when the charger first kicks in at 11.30pm on the cheap overnight leccy, but happens, apparently randomly sometime after the charger is on, if I leave the charger off then the RCD never kicks out. If it kicks out we get awakened by the alarm low battery warning (house alarm, not the inverter alarm) which comes on after around 20 minutes of loss of power. If I then reset the RCD it normally stays in for the rest of the night, it's only ever gone out a 2nd time in one night.
I'm totally at a loss as to what could be causing it to go out, we have various items come on overnight such as dishwasher, washing m/c, tumble drier, EV charger, but the trip doesn't appear to tie in with any of the times these come on.
Any ideas would be welcome
Also, this is going to sound like a daft Q, exactly what trips an RCD, is it that it "sees" a voltage going to earth or to the negative & are there different sensitivities of RCD?

 

[kerching]

I'm guessing there is a contactor that has a slight delay between the L and N breaking causing a momentary imbalance in the RCD sensing coils.
But it's a total unsubstantiated guess

 

[binky]

Also, this is going to sound like a daft Q, exactly what trips an RCD, is it that it "sees" a voltage going to earth or to the negative & are there different sensitivities of RCD?

earth leakage normally, although a faulty cable could also cause issues, or just a loose connection somwhere in the cct. Random tripping is frequently a fridge or freezer, which are the only appliances that randomly turn on and off at will.

So, if I'm understanding you correctly, you have the battery inverter linked to mains to charge up batteries on off peak leccy, and are feeding various appliances around the house via the UPS output of the inverter? What earthing arrangements do yuo have for the off-grid ccts, and how is the inverter connected to these, and the mains.

 

[Mike P]

I have had a similar issue although not off an inverter. As above could be a loose connection, but in my case it was a faulty RCD in the distribution board. It checked out ok with test equipment but was intermittent, always at night with little load. Changed RCD and fault went away.

 

[Ucatchmydrift]

Unusual installation, I have a solar installation (DIY) with battery backup, it powers a no. of circuits within the house separate to the rest of the mains circuits, so, effectively a lot of the house is off grid. There is an RCD on the mini consumer unit which is powered by the inverter, no mains to this CU. Everything works perfectly until overnight when the battery charger on the inverter is on, it doesn't usually happen when the charger first kicks in at 11.30pm on the cheap overnight leccy, but happens, apparently randomly sometime after the charger is on, if I leave the charger off then the RCD never kicks out. If it kicks out we get awakened by the alarm low battery warning (house alarm, not the inverter alarm) which comes on after around 20 minutes of loss of power. If I then reset the RCD it normally stays in for the rest of the night, it's only ever gone out a 2nd time in one night.
I'm totally at a loss as to what could be causing it to go out, we have various items come on overnight such as dishwasher, washing m/c, tumble drier, EV charger, but the trip doesn't appear to tie in with any of the times these come on.
Any ideas would be welcome
Also, this is going to sound like a daft Q, exactly what trips an RCD, is it that it "sees" a voltage going to earth or to the negative & are there different sensitivities of RCD?

Hi, im no expert at all, just someone that has a very similar setup to you: mine is a diy system seperate to the grid system, so i have the normal sockets etc powered by the grid, and a seperate off grid set of sockets etc powered by an all in one inverter which has a feed from the grid that either kicks in to charge the batteries when necessary, or is used as a bypass to power all the off grid sockets if necassary. My panels are mounted on my metal workshop and wooden shed, which are right next to each other. The inverter is inside the metal workshop and sends power into the house via a armoured cable. There is a consumer unit in the roof with a mains swich and 30ma rcbo, where power then goes to the sockets etc.
The metal workshop already had power which came from the house into a consumer unit in the workshop which has a breaker and an rcd to cover the workshop circuit. I took the feed to the inverter from this consumer unit. - this is where the problems can arise it seems.
Once i finished the installation of the panels and wiring to the house, my dad came round (hes an electrician) to check and test all what ive done. One thing i noticed once was when cleaning the panels, i was spraying them with a hose, i heard a click, and the rcd tripped inside the metal shed. I had earthed the panels independently directly to the metal shed and then to a grounding rod. I also earthed the inverter to a seperate grounding rod about 3.5 meters away from the panels grounding rod.- this earth is connected to the earths in the off grid system, completely seperate from any other earth.
Never after it has been raining, or during, has the rcd tripped in the shed, so i left it be and waited for some damp winter conditions to test it. I woke up the other day after a very foggy damp morning, and the rcd had tripped.

Basically, all in one inverters are not supposed to have a rcd inbetween the supply from the grid to the inverter. Only after the inverter to the house, due to earth/neutral bonding inside the inverter it seems.
Probably been a waste of time me posting this though lol as yours if im right is tripping the rcd after the inverter. But just thought it may help. Was it raining very heavy/foggy when it tripped? (as i could reset the rcd after a minute or so most times)

 

[binky]

Hi, im no expert at all, just someone that has a very similar setup to you: mine is a diy system seperate to the grid system, so i have the normal sockets etc powered by the grid, and a seperate off grid set of sockets etc powered by an all in one inverter which has a feed from the grid that either kicks in to charge the batteries when necessary, or is used as a bypass to power all the off grid sockets if necassary. My panels are mounted on my metal workshop and wooden shed, which are right next to each other. The inverter is inside the metal workshop and sends power into the house via a armoured cable. There is a consumer unit in the roof with a mains swich and 30ma rcbo, where power then goes to the sockets etc.
The metal workshop already had power which came from the house into a consumer unit in the workshop which has a breaker and an rcd to cover the workshop circuit. I took the feed to the inverter from this consumer unit. - this is where the problems can arise it seems.
Once i finished the installation of the panels and wiring to the house, my dad came round (hes an electrician) to check and test all what ive done. One thing i noticed once was when cleaning the panels, i was spraying them with a hose, i heard a click, and the rcd tripped inside the metal shed. I had earthed the panels independently directly to the metal shed and then to a grounding rod. I also earthed the inverter to a seperate grounding rod about 3.5 meters away from the panels grounding rod.- this earth is connected to the earths in the off grid system, completely seperate from any other earth.
Never after it has been raining, or during, has the rcd tripped in the shed, so i left it be and waited for some damp winter conditions to test it. I woke up the other day after a very foggy damp morning, and the rcd had tripped.

Basically, all in one inverters are not supposed to have a rcd inbetween the supply from the grid to the inverter. Only after the inverter to the house, due to earth/neutral bonding inside the inverter it seems.
Probably been a waste of time me posting this though lol as yours if im right is tripping the rcd after the inverter. But just thought it may help. Was it raining very heavy/foggy when it tripped? (as i could reset the rcd after a minute or so most times)

I have a similar setup to yourself, I had to bond the metal roof to get rid of induced voltages in it - could get a right tingle on a wet day, which was quite disconcerting! . My garage CU has been setup with the solar on the main switch and RCBOs for other ccts.

Nothing like a bit of damp to get into the DC side! I cable tie the connections to the frames to keep them off the roof itself, they are suppossed to be waterproof enough to be immersed 1m deep in a puddle, but in reality that seems to be an optimistic figure. In particular they dislike leaves, moss or other debris building up against the connectors and even sometimes the cables. A flat roof installation at a school I do maintneance work for had not been cleared of debris for years,and it's next to woodland so covered in leaves, big build of moss and plants around the cabling, net result the DC cabling has had to be replaced after only 5 years. It would seem the DC cabling isn't quite as robust as the manufacturers originally stated. I was also involved in recabling all the DC side of a solar farm, but they had buried all the DC cables direct in the ground, which wasn't a clever idea - that had lasted approximately 3 years before needing complete replacement. Like wise a gound array about 50m from the inverters had had the DC cables installed in land-drain, so again was constantly wet and had to be replaced after only a few years.

 

[Ucatchmydrift]

I have a similar setup to yourself, I had to bond the metal roof to get rid of induced voltages in it - could get a right tingle on a wet day, which was quite disconcerting! . My garage CU has been setup with the solar on the main switch and RCBOs for other ccts.

Nothing like a bit of damp to get into the DC side! I cable tie the connections to the frames to keep them off the roof itself, they are suppossed to be waterproof enough to be immersed 1m deep in a puddle, but in reality that seems to be an optimistic figure. In particular they dislike leaves, moss or other debris building up against the connectors and even sometimes the cables. A flat roof installation at a school I do maintneance work for had not been cleared of debris for years,and it's next to woodland so covered in leaves, big build of moss and plants around the cabling, net result the DC cabling has had to be replaced after only 5 years. It would seem the DC cabling isn't quite as robust as the manufacturers originally stated. I was also involved in recabling all the DC side of a solar farm, but they had buried all the DC cables direct in the ground, which wasn't a clever idea - that had lasted approximately 3 years before needing complete replacement. Like wise a gound array about 50m from the inverters had had the DC cables installed in land-drain, so again was constantly wet and had to be replaced after only a few years.

Ha ha, yes, the tingles! Nearly forgot, when testing with a hose, before i grounded the panels (silly yes i know ), if i stood about a foot or two away from the panels with my finger on the end of the hose to make a spray pattern, i would get more than just a tingle build up the closer i got! I could also feel a tingle in the dry if touching one set on one shed and my other hand on the other set on the other shed. Once i grounded them all properly, no more tingles at all. But still would trip the rcd if i persisted long enough with the hose. If it happens again in the damp, i expect i will remove the rcd to the inverter grid supply like the manufacturer suggests (epever), and just have the breaker instead.

 

[binky]

Ha ha, yes, the tingles! Nearly forgot, when testing with a hose, before i grounded the panels (silly yes i know ), if i stood about a foot or two away from the panels with my finger on the end of the hose to make a spray pattern, i would get more than just a tingle build up the closer i got! I could also feel a tingle in the dry if touching one set on one shed and my other hand on the other set on the other shed. Once i grounded them all properly, no more tingles at all. But still would trip the rcd if i persisted long enough with the hose. If it happens again in the damp, i expect i will remove the rcd to the inverter grid supply like the manufacturer suggests (epever), and just have the breaker instead.

I'm still a little intrigued by the RCD tripping, I can't quite get how that happens via the DC side?

 

[Ucatchmydrift]

I'm still a little intrigued by the RCD tripping, I can't quite get how that happens via the DC side?

I havent a clue if im honest, accept for some hybrid inverters like mine have a permanent or temporary neutral to earth bonding inside for whatever reason.
maybe in my case a stray dc current comes from the pv down to the earthing rod, and if damp enough, the current reaches the other inverter earthing rod which then goes into the inverter, then through the temporary earth neutral bond when grid power is enabled, which then allows the current down the neutral cable and is picked up by the rcd? Just a thought, but probably wrong as im no electrician or electronics wizz...
I guess what i should really say is,, i dont know lol.

 

[Solarexploits]

Thanks for the replies, I have ba strong suspicion it is almost certainly the fridge freezer. I've isolated it by running it solely from the mains overnight & not had it trip since.

However, I'm still somewhat confused by why it would only happen when the inverter is charging, I've had the freezer running 24/7 from the inverter & never had the RCD trip except overnight & during the hours the batteries are recharging.

The earth connection from the inverter does have a connection to the earth on the main consumer unit as I have a double pole switch which can switch most of the downstairs power circuits to run from either mains or inverter, so I can switch over if battery power is getting low, it switches both L & N at the same time, but the earths are connected within the switch box. I'm aware that strictly speaking the inverter earth should go to a separate earth rod (not yet investigated exactly what's required there) & shouldn't share the mains earth as that is owned by your supplier, but this has never caused the RCD to trip when the batteries aren't charging. The inverter AC supply obviously comes from the main consumer unit, so in which case are the earths automatically connected anyway via this supply - I have no idea how inverters are wired internally & if that earth is shared with the AC output earth from the inverter? But as the earths are already connected at the double pole switch why should it make any difference whether batteries are charging or not???

 

[binky]

However, I'm still somewhat confused by why it would only happen when the inverter is charging, I've had the freezer running 24/7 from the inverter & never had the RCD trip except overnight & during the hours the batteries are recharging.

I would say it's combined Earth Leakage from inverter and fridge. The RCD will trip at around 25mA, it may say 30mA on it, but that's it's max earth leakage it will tolerate, in reality they all trip at less than that. Everything has electronic controls these days, and to protect the sensitive components they having 'smoothing circuitry' that throws a bit of leccy down the earth, which is known as earth leakage. With houses having many appliances on the same circuit it can be very difficult to find exactly with one is failing, as you will get a few mA rom the TV, a few from mA from the fridge, a few mA from the wifi router etc etc, so in combination wth other appliances the earth leakage builds up until the RCD sees the 25mA it needs to trip. Now, when appliances start to fail, the earth leakage increases, so it can be hard to identify exactly which appliance is at fault, but random tripping at odd times of day or night is nearly always the fridge/ freezer, as these are one of the few appliances that turn themselves on and off at random times of day. Just to add insult to injury, even the appliance that is developing a fault, if it's on it's own RCD / RCBO, may well work happily for many years to come without tripping anything out. One of the reasons I fit RCBO consumer units rather than sharing RCDs with multiple circuits.

 

[SPECIAL LOCATION]

I would say it's combined Earth Leakage from inverter and fridge. The RCD will trip at around 25mA, it may say 30mA on it, but that's it's max earth leakage it will tolerate, in reality they all trip at less than that.

Whilst as you say, generally a typical 30ma RCD actually trips around the 25->28ma area..
Technically to comply with BS7671 is could be anywhere from 16ma upward..

i.e. Must NOT trip at half RCD rating, (15ma)..
But MUST trip at full RCD rating (30ma)..

Has a ramp test been done on the RCD to verify it is NOT oversensitive?
I have known some trip around 17->18ma which can be a problem...

Also, cumulative leakage can be a right sod to track down!

 

[Mike P]

Or is the type of RCD wrong. Is there a DC element involved?

 

[Joules]

Unusual installation, I have a solar installation (DIY) with battery backup, it powers a no. of circuits within the house separate to the rest of the mains circuits, so, effectively a lot of the house is off grid. There is an RCD on the mini consumer unit which is powered by the inverter, no mains to this CU. Everything works perfectly until overnight when the battery charger on the inverter is on, it doesn't usually happen when the charger first kicks in at 11.30pm on the cheap overnight leccy, but happens, apparently randomly sometime after the charger is on, if I leave the charger off then the RCD never kicks out. If it kicks out we get awakened by the alarm low battery warning (house alarm, not the inverter alarm) which comes on after around 20 minutes of loss of power. If I then reset the RCD it normally stays in for the rest of the night, it's only ever gone out a 2nd time in one night.
I'm totally at a loss as to what could be causing it to go out, we have various items come on overnight such as dishwasher, washing m/c, tumble drier, EV charger, but the trip doesn't appear to tie in with any of the times these come on.
Any ideas would be welcome
Also, this is going to sound like a daft Q, exactly what trips an RCD, is it that it "sees" a voltage going to earth or to the negative & are there different sensitivities of RCD?

Present consumer unit designs under 18th Edition have two options. The first is to replace all mcb with RCBO, this first appeared in 17th Edition. The second is to have multiple RCD along with their mcb. The rule here is that all circuits down stream from the RCD must not have a combined leakage greater than 9mA on a 30mA RCD. You can however benefit from a hybrid arrangement that is the combination of the two. Arrange for you electrician to create at least a two module, three is better, immediately after the main switch. Everything gets shuffled down to create the gap. The leakiest circuit(s) is the ring main. Move this to an RCBO in the gap. Next move anything solar to another RCBO in the gap and so on. YouTube's eFIXX electrician explains in EARTH LEAKAGE CURRENT - WHERE it comes from and HOW to measure it. - MEGGER DCM305E