Off grid inverter issue

[Feenster]

Looking for a bit of help.
I bought a PIP4048 from MPPSolar many years ago and bought a new PIP5048 just recently.
I live off-grid and have a generator for back up.

Recently I replaced the PIP4048 with the new PIP5048.

The PIP5048 started up fine. However, when I started to turn on the MCBs on the fuse board for the house many of them kept tripping. Occasionally the RCD would also trip. For example, when I turned on the breaker for the kitchen as soon as the 'fridge turned on the breaker fell. At first it was only the circuits with a 'fridge or freezer or small motor that tripped. Then, on another circuit, someone in the house turned on a small laptop and a trip occurred on that circuit.

I have 7 socket circuits (each a ring circuit), each with its own MCB, protected by one RCD.

On the advice of MPPSolar (after waiting for a reply for a long time) I replaced two of the MCBs. I took out 20A breakers and put in 32A ones.
I put back in the PIP5048. Unlike my first attempt, the MCBs didn't trip at all. However, the RCD did.
On each circuit as soon as an electrical appliance started the RCD tripped. I tried a 100W coffee grinder. RCD tripped. I tried something without a motor - a 700W toaster - RCD tripped. The only thing that appears to work on the RCD controlled circuits is a table lamp (5W).
On one of the circuits there is a small pump. Even when the pump is on standby but not actually running the RCD trips.
The RCD is a Hager 63A 2 pole RCD 30mA (type A).


This never happened with the PIP4048. I had to put the PIP4048 back in again as I couldn't function with the 5048.

 

[Chewbacca]

Change the MCBs for RCBOs, your issue is earth leakage across shared circuits. Currently you have 30mA earth leakage max, whereas RCBOs give 30mA per circuit. It also sounds like the new inverter has higher earth leakage than the old unit. You also shouldn't be sharing RCDs with solar inverters, the shutdown time of the inverter can basically kill you after the RCD has tripped, rendering the RCD useless.

 

[binky]

changing MCBs for larger ones is a bad idea, unless you know the cable is man enough to take the higher ampage, you are risking the cable burning out and causing a fire. A quick look at manuals for these two inverters suggests you have changed from 4kw to 5 kw, 5 kw is roughly 20Amps, so it sounds like the new unit is overloading the MCB and possibly the cable. 2.5mm cable will run upto 26Amps, 4mm 32 Amps, 6mm 46Amps or there abouts, dependent on installation method. We derate cables if run through insulation, grouped together and long list of other things.

It also sounds like you possibly have an issue with the pump, which is not unusual, as per Chewies' post, I would have that on it's own RCBO.

 

[Feenster]

Thanks for that.
So, just to be clear, after the main switch on the fuse board, remove the RCD and put in 7 RCBOs, one for each ring circuit?
Should I do any thing with the circuits for the lights (not shown below). At the moment each circuit is protected by a MCB but no RCD.

 

[binky]

The lights should have RCBOs aswell, as you have a TT earthing system, unless you have some other rcd before the main switch.

 

[Feenster]

You also shouldn't be sharing RCDs with solar inverters, the shutdown time of the inverter can basically kill you after the RCD has tripped, rendering the RCD useless.

Chewbacca, can you explain this some more to me please?
On the 'house side' of the RCD, if the RCD goes am I protected?
Is it between the Inverter and RCD that the problem is, where 'the shutdown time of the inverter can basically kill you'?

 

[Feenster]

A quick look at manuals for these two inverters suggests you have changed from 4kw to 5 kw, 5 kw is roughly 20Amps, so it sounds like the new unit is overloading the MCB and possibly the cable.

Each of the ring circuits have 2.5 mm cables. However, there is very little load on any of the circuits. Because we are off grid the electric kettle is probably the most powerful appliance we have. Most of the circuits just have table lamps, etc.

 

[Feenster]

The lights should have RCBOs aswell, as you have a TT earthing system, unless you have some other rcd before the main switch.

(The circuit that I attached on Monday (there should be a '+' and '-' between the battery and inverter not 'live' and 'neutral') doesn't show that there is a neutral-ground bond inside the inverter itself. It can be manually removed.)
I was off the opinion that I had a TNS system. Perhaps I am falling between two stools. Have you any advice on where I would find good information on the TT earthing system please?

 

[michael8554]

Sounds like the RCD has coincidently become faulty, as it's unlikely that all the items you got fails with have become faulty at the same time.

But I dunno, would L and N reversed on the Inverter cause this sort of problem ?

 

[binky]

Chewbacca, can you explain this some more to me please?
On the 'house side' of the RCD, if the RCD goes am I protected?
Is it between the Inverter and RCD that the problem is, where 'the shutdown time of the inverter can basically kill you'?

The inverter and your other circuits are all connected to one side of the RCD. So, for example, you are mowing the lawn and run over the cable, the RCD trips as it should, but the inverter shut down time can be as long as 5 seconds, so with the RCD tripped the inverter can continue to supply electric for 5 seconds and kill you.

Now you can't be TN-S if you are offgrid, as that refers to 'supplied earth' from the grid. Your diagram shows you have an earth rod, so all earthing is via the rod, and this has different requirements to a mains connected (on-grid) system. Google it, it's easy enough to find information. Because earth rods can have relatively high resistances, like 200 Ohms or more, the trip times of protective devices can be affected, ie slower, so to ensure safety we RCD all circuits. Ths can be achieved in various different ways, but I'll keep it simple. If your new inverter has N-E link, this may well be upsetting your RCD, they work be looking for electric in on the L,matches electric out on the N, with a small tolerance. If you are losing some electric down the earth due to N-E link, then there will be an imbalance which the RCD sees as a fault and trips. So I would look to remove any N-E link in the inverter if possible. The fact that it's there is a bit of an oddity. With grid tied solar systems, many have a UPS function to power essential items like a freezer. To get that to work you have to disconnect the grid supply, which can mean losing the earth link aswell, hence the N-E link in the inverter to compensate for that - can't say I'm too convinced it's a safe set-up, but then it's only a temporary thing.

 

[offgridAndy]

Does your battery and inverter set up not require the battery negative to be tied to the (single common) earth point. What does the manufacturer specify that you do in their hand book. Note that the genny should also be earthed to the same common earth rod. If practical,

 

[Feenster]

Sounds like the RCD has coincidently become faulty, as it's unlikely that all the items you got fails with have become faulty at the same time.

I have tried another RCD - unfortunately the same result, it 'trips'.

 

[Feenster]

The inverter and your other circuits are all connected to one side of the RCD. So, for example, you are mowing the lawn and run over the cable, the RCD trips as it should, but the inverter shut down time can be as long as 5 seconds, so with the RCD tripped the inverter can continue to supply electric for 5 seconds and kill you.

I am still confused by this. From my diagram above, the power from the inverter goes to the fuse board. It goes through the RCD and then to the individual MCBs for each of the ring circuits. Is the RCD not between the inverter and the circuits? So if the mower cuts the wire, the RCD trips and the power from the inverter can not travel through the tripped RCD.

 

[Feenster]

If your new inverter has N-E link, this may well be upsetting your RCD, they work be looking for electric in on the L,matches electric out on the N, with a small tolerance. If you are losing some electric down the earth due to N-E link, then there will be an imbalance which the RCD sees as a fault and trips. So I would look to remove any N-E link in the inverter if possible. The fact that it's there is a bit of an oddity.

The old inverter has the same set up - it has a N-E link as well. However, it has worked fine for a number of years. Perhaps the new one, since it is more powerful, is causing a greater loss down the earth.
Removing the N-E link isn't difficult - I tried that already. However, I simply moved the N-E link to the fuse board, which I saw recommended on a number of sites. Same problem, RCD tripped (which isn't surprising really considering I only moved the link to a different point).
Now I assumed that I had a TN-S system and it was I who was providing the earth by means of an earth rod. And as with any TN-S system the neutral of the source (i.e. the inverter) is connected with the earth at one point only and if that is in the inverter so be it.
Having looked up the TT system it appears that I would need 2 earth rods. There appears to be a N-E link at the source. Does that mean using the N-E link in the inverter and connecting to one earth rod?

 

[Feenster]

Does your battery and inverter set up not require the battery negative to be tied to the (single common) earth point. What does the manufacturer specify that you do in their hand book. Note that the genny should also be earthed to the same common earth rod. If practical,

I have read different things about whether its necessary to tie the battery negative to the common earth point.
I have tied the genny to the same earth rod as the rest.
Are you completely off grid with no connections at all to the grid? What type of earthing system have you used?

 

[offgridAndy]

Yes I am completley "off grid". with 6Kw of PV panels plus 3 Kw wind turbine, 24Kw lead acid battery and 12KVA propane powered gen set. Various charge controllers and 3 Kw Outback inverter charger. The WT is AC coupled and back feeds through the inverter charger to charge the battery when the house load is satisfied. I have one earth rod to which all connectiions are made including the battery neg and the CU earth. I have found that Outback have a genuine knowledge about their product and provide full technical back up. I read their manuals and any doubts I have always asked for and got answers from their technical dept. My system has grown and altered over the years but I have not suffered from any of the tripping issues you have with your new kit. I'd take up the issue with the manufacturer.

 

[MSJ]

Is nu

The old inverter has the same set up - it has a N-E link as well. However, it has worked fine for a number of years. Perhaps the new one, since it is more powerful, is causing a greater loss down the earth.
Removing the N-E link isn't difficult - I tried that already. However, I simply moved the N-E link to the fuse board, which I saw recommended on a number of sites. Same problem, RCD tripped (which isn't surprising really considering I only moved the link to a different point).
Now I assumed that I had a TN-S system and it was I who was providing the earth by means of an earth rod. And as with any TN-S system the neutral of the source (i.e. the inverter) is connected with the earth at one point only and if that is in the inverter so be it.
Having looked up the TT system it appears that I would need 2 earth rods. There appears to be a N-E link at the source. Does that mean using the N-E link in the inverter and connecting to one earth rod? View attachment 16382

your inverter has N-E link witch is controlled i.e. with a supply to grid connection in your inverter N-E link will be open and closed in island mode i.e. as UPS

 

[binky]

I am still confused by this. From my diagram above, the power from the inverter goes to the fuse board. It goes through the RCD and then to the individual MCBs for each of the ring circuits. Is the RCD not between the inverter and the circuits? So if the mower cuts the wire, the RCD trips and the power from the inverter can not travel through the tripped RCD.

my bad, your RCD is controlling the whole supply - I've had afew things on my mind lately and didn't study your diagram properly.

 

[binky]

The old inverter has the same set up - it has a N-E link as well. However, it has worked fine for a number of years. Perhaps the new one, since it is more powerful, is causing a greater loss down the earth.
Removing the N-E link isn't difficult - I tried that already. However, I simply moved the N-E link to the fuse board, which I saw recommended on a number of sites. Same problem, RCD tripped (which isn't surprising really considering I only moved the link to a different point).
Now I assumed that I had a TN-S system and it was I who was providing the earth by means of an earth rod. And as with any TN-S system the neutral of the source (i.e. the inverter) is connected with the earth at one point only and if that is in the inverter so be it.
Having looked up the TT system it appears that I would need 2 earth rods. There appears to be a N-E link at the source. Does that mean using the N-E link in the inverter and connecting to one earth rod? View attachment 16382

source earth is the local transformer or substation, not your inverter, although there's little difference.

 

[binky]

this is an interesting read,

https://electrical.theiet.org/wiring-matters/years/2020/82-september-2020/mythbusters-6/