Does this system look ok?

[HappyHippyDad]

I have just finished designing the solar layout for my shed. It's a little different from standard 230V domestic electrics so I was hoping you guys would cast an eye over it and give me some opinions.

I have done all the maths and am happy with the size of the solar panel, battery, inverter and charge controller so it is more the following points I was hoping for some thoughts on:

1. Size and position of the protective devices (fuses and breakers)?
2. Earthing?
3. Types of fuses (blade or inline fuses or something else? Perhaps DC breakers?)
4. Anything else?

Cheers.

 

[HappyHippyDad]

Why are you so dead set on trying to fit an RCD into this system? Electrical separation as a method of shock protection is, if done correctly, much better than relying on an RCD


Sent from my iPad using Tapatalk

I wouldn't say 'dead set', although probably more keen than I have implied. The reason being is that I would prefer to have 2 or 3 2gang sockets in the shed therefore the likelihood of using more than one appliance is raised, meaning electrical separation is not ideal due to the increased possibility of 2 faults occurring at the same time.

Unless however I get a better understanding of creating a TN-S system I shall stick with the one socket from the inverter and so one appliance and therefore use electrical separation.

The floating one doesn't have an N-E link
But you could perhaps provide one at the DB.
Generally the ELV side will be electrically separate from the AC output but might be connected to the case, so the result of independently earthing the LV connection would be unpredictable.

Actually, after writing my post above to DS I think I will stick with one appliance and electrical seperation. If you are using words like 'perhaps' and 'unpredictable' Lucien then I'm not going to play around with it!

 

[Lucien Nunes]

What I meant was that unless the inverter data sheet specifies, or until you take a meter to it, you can't predict how it is going to be wired. Once you know this, you can decide what's best.

 

[HappyHippyDad]

The floating one doesn't have an N-E link
But you could perhaps provide one at the DB.
Generally the ELV side will be electrically separate from the AC output but might be connected to the case, so the result of independently earthing the LV connection would be unpredictable.

A few more questions!

1. So this new diagram would allow the use of an RCD IF I had an N-E link in the inverter?
2. The casing of the inverter shows no continuity with the LV side earth terminal, the LV earth terminal shows no continuity with either (ELV and LV) negative terminal on the inverter. What tests are required to determine the nature of the inverter and assist with working out earthing possibilities?
3. If I have a floating earth in the inverter the diagram would not allow the use of the RCD as there would be no neutral-Earth link and therefore no source earth?
4. If I have a floating earth in the inverter how would I provide a neutral-Earth link in the DB without tripping the RCD?
5. If I do use electrical seperation, then there is no need to have any earth conductor going from socket to socket?

Let me know when you get fed up Lucien and I'll try and stop. :smile5:

 

[HappyHippyDad]

ps.. When I say new diagram I mean the one at the bottom of the last post!

 

[Richard Burns]

At the risk of demonstrating my ignorance here!
The HSE have a very old open circular that provides some easy information about earthing that may help, but it is a bit out of date.
The MCS website also have a guide to the installation of PV systems, obviously geared to grid connected systems but this may give some helpful hints.
Section 717 of BS1671 for mobile units shows some useful diagrams for the earthing of independent generators inside the unit, even though it is also covering externally supplied units.

Your inverter not needing earthing but having an earth terminal is slightly odd.
The inverter may not provide electrical separation of the dc side from the ac side so earthing the system may give rise to unexpectedly high voltages on the ELV side.
It might be worth seeing if the inverter has any connections internally (via continuity/ IR) to that earthing terminal, it may just be a bolt on the case.

If the 230V side is electrically independent of the 12V side then one pole of the inverter output could be connected to an earth reference point ( an earth rod) and the earth for the installation (MET) separated out from this earthed point so that you create a TNS supply before the RCD.

 

[davesparks]

I wouldn't say 'dead set', although probably more keen than I have implied. The reason being is that I would prefer to have 2 or 3 2gang sockets in the shed therefore the likelihood of using more than one appliance is raised, meaning electrical separation is not ideal due to the increased possibility of 2 faults occurring at the same time.

!

There is no problem using elGood evening fair maiden, how goes the day with you?

Please allow me to introduce myself, I'm Andy and I'm very pleased to make your acquaintance.

Bonjour

Electrical separation for more than one socket, if you use a bit of common sense and precautions.

I am assuming that it is only ever going to be yourself using these sockets? In which case you should be capable of ensuring that the appliances and the installation are kept in good condition and repaired if damage occurs.
Faults are rarely spontaneous occurrences, there has to be a cause or reason for the fault occurring, in this environment I would have thought physical damage is the only real danger.

My reason for favouring the separate supply over the use of RCDs is that an RCD does not prevent an electric shock, it just limits its magnitude. From painful experience I know that the shock you receive before the RCD trips is incredibly painful and enough to knock you out cold for half an hour or so.
Electrical separation however does prevent electric shocks, far better in my opinion.

 

[HappyHippyDad]

At the risk of demonstrating my ignorance here!
The HSE have a very old open circular that provides some easy information about earthing that may help, but it is a bit out of date.
The MCS website also have a guide to the installation of PV systems, obviously geared to grid connected systems but this may give some helpful hints.
Section 717 of BS1671 for mobile units shows some useful diagrams for the earthing of independent generators inside the unit, even though it is also covering externally supplied units.

Your inverter not needing earthing but having an earth terminal is slightly odd.
The inverter may not provide electrical separation of the dc side from the ac side so earthing the system may give rise to unexpectedly high voltages on the ELV side.
It might be worth seeing if the inverter has any connections internally (via continuity/ IR) to that earthing terminal, it may just be a bolt on the case.

If the 230V side is electrically independent of the 12V side then one pole of the inverter output could be connected to an earth reference point ( an earth rod) and the earth for the installation (MET) separated out from this earthed point so that you create a TNS supply before the RCD.

I'm sorry Richard, I'm not explaining it clearly enough. When I say earth terminal, it simply has a terminal for the earth pin of a plug! This 'terminal' had no continuity with the frame of the inverter.

I'm away from home, but I did do a continuity test between the ELV terminals and the LV socket of the inverter. I remember getting around 500Ω I think between L-L and N-N. This figured changed whilst I left the probes in place, I cant remember in what fashion it changed. With the switch of the inverter either on or off it either decreased or increased respectively with the probes left in place.

I am now erring again towards electrical seperation, (as DS said above), it will just be me operating it and I will be able to keep it in good condition, however I will certainly be reading through the links you have provided as they look interesting.

 

[davesparks]

This might help in understanding
https://m.youtube.com/watch?v=fRhofcMyAyk


Sent from my iPad using Tapatalk

 

[HappyHippyDad]

This might help in understanding
https://m.youtube.com/watch?v=fRhofcMyAyk


Sent from my iPad using Tapatalk

That was interesting, thanks.

Unless any further information comes along which sways me I will stick with electrical seperation.

 

[Lucien Nunes]

Unless any further information comes along which sways me I will stick with electrical seperation.

I did do a continuity test between the ELV terminals and the LV socket of the inverter. I remember getting around 500Ω I think between L-L and N-N

Obviously there are not interconnections between both poles of both circuits otherwise there would be a constant heavy load on the AC side. Likely, DC -ve is connected somehow to AC N, and there is relatively low resistance between L & N and between + & -. The varying readings you got merely reveal the large capacitors charging to the test voltage of the DMM and you might get different readings with different meter probe polarities as you are reading not just resistors but complex electronics. But the fact that you got any reading other than infinity suggests that to achieve electrical separation of the LV side, you might need to insulate the ELV side to the same standard. I would re-check those readings carefully on the inverter.

 

[HappyHippyDad]

Obviously there are not interconnections between both poles of both circuits otherwise there would be a constant heavy load on the AC side. Likely, DC -ve is connected somehow to AC N, and there is relatively low resistance between L & N and between + & -. The varying readings you got merely reveal the large capacitors charging to the test voltage of the DMM and you might get different readings with different meter probe polarities as you are reading not just resistors but complex electronics. But the fact that you got any reading other than infinity suggests that to achieve electrical separation of the LV side, you might need to insulate the ELV side to the same standard. I would re-check those readings carefully on the inverter.

I am back home and have taken more accurate results:

Inverter switch off:

ELV (-) to LV (-) = 80.5Ω
ELV (-) to LV (+) = 80.8Ω
ELV (+) to LV (-) = 300Ω rising to >2000Ω
ELV (+) to LV (+) = 300Ω rising to >2000Ω

Similar results with inverter switch on although the resistance went from 300Ω raising to 414Ω with the test lead connected to the ELV (+)

ELV (+) to ELV (-) = 158Ω rising to 330Ω with inverter switch on

ELV (+) to ELV (-) = 71Ω with inverter switch off

No continuity between LV terminals in the provided socket.

Here is a link to the inverter 2000W(4000W peak) DC 12V to AC 240v Power Inverter Converter Electronic UK | eBay

and a picture for what it's worth

I've only put these readings up as my others were clearly inaccurate. Would you still be talking about 'insulating the ELV side to the same standard' or was that merely due to my inaccurate results?

Also, as before, no continuity between the earth terminal in the socket of the inverter and anything!

 

[davesparks]

That looks like a very faint earth symbol next to the CE mark, is there some writing to go with it?


Sent from my iPad using Tapatalk

 

[HappyHippyDad]

That looks like a very faint earth symbol next to the CE mark, is there some writing to go with it?


Sent from my iPad using Tapatalk

That is extremely well made out!

Yes, it is an earth symbol. There is some writing beneath it but it is very difficult to make out. It looks like the word 'Class' followed by a single letter or digit (unreadable), followed possibly by the word 'symbol'.

They did say that the unit does not need earthing, but it is pretty difficult to have a conversation with them over the net as English is not there first language and he is relaying my questions to the tech department, so acting as a mediator.

 

[Lucien Nunes]

Would you still be talking about 'insulating the ELV side to the same standard'

Well 80.5Ω isn't going to get in the way of you receiving a shock! I don't know what instrument you are testing with but bear in mind that these are indicative results only, you are measuring active electronics and the numbers on your display don't necessarily relate to actual resistances, only how the instrument 'sees' what's inside. But anything less than megohms indicates there's some kind of connection there; my hunch is that the electronics that drive the H-bridge (that turns the 320V DC rail into AC) run on the ELV instead of the secondary side of the step-up DC-DC converter (that makes 320V out of 12V). I suppose that's cheaper than powering them off an extra secondary. Whatever the cause, your readings (which I think I trust) suggest that a breach of the ELV insulation would defeat an IT configuration on the LV side.

My next thought is that if the hunch is right, and I accept it is only a hunch, then you can't create an external N-E link either to make it TN, or the ELV side will be flapping around on half-wave-rectified AC w.r.t earth. Perhaps the idea is you are not supposed to think about earthing or insulation, just plug in an appliance and use it like 99% of customers and try not to have any accidents with it.

Dunno what next. On my bench I'd take a look inside to see what's going on in there. ISTR this is why I only used to buy Victron or Mastervolt.

 

[davesparks]

Where did you get the inverter?
If they aren't too expensive then I might get one and 'investigate' what's inside


Sent from my iPad using Tapatalk