RCD tripping on new Consumer Unit

[billywolfe]

Not sure id be happy with that solar circuit feeding back into a RCD bank like that.

Hi Jono - what problems do you see with the above?

 

[JonoPashley]

Hi Billy,

The issue I see with this is your RCD will cut the power in the event of a fault in a maximum of 0.4 seconds. However a Inverter can take upto 5 seconds to shut down following power loss. This means for upto 5 seconds your Inverter could be back feeding the circuits connected to the RCD bank.

 

[billywolfe]

Hi Billy,

The issue I see with this is your RCD will cut the power in the event of a fault in a maximum of 0.4 seconds. However a Inverter can take upto 5 seconds to shut down following power loss. This means for upto 5 seconds your Inverter could be back feeding the circuits connected to the RCD bank.

That's interesting Jono. The Solar company have been speaking to Power-One who manufacture the inverter and (as far as I know) this has not been mentioned.
The first solution suggested to me by the solar company is that the solar circuit is taken out of the mcb bank in the CU and instead put in a separate box still on a 20amp MCB, but protected by a 300ma RCD. (which according to Power One is still within the regulations). They then found out that the 300ma RCD would cost about £380, so have ditched that idea due to cost and suggested that the MCB could be unprotected by an RCD but they would put trunking or conduit over the cable from the inverter....(again I was told that this was within regs). I then suggested using RCBOs instead of the MCB/RCD combination in the CU and I think that is what is going to happen....
As I said, I am not an electrician - I just ask you guys for help and read all I can about this subject. So I don't really see it should be me making suggestions to the Solar company.
So no electricians have been back so far. No tests have been made on existing circuits or appliances....they are just assuming that the solar feed is the problem. At least if RCBOs are used it should identify the circuit at fault (although an earlier poster said that RCBOs on a different circuit could trip...)
Jono, do you think RCBOs in the existing CU would be OK? Or would it be better to have a separate box for the solar feed with an MCB/RCD protection?

thanks again to all ...

 

[bernardgreen]

They then found out that the 300ma RCD would cost about £380, so have ditched that idea due to cost

Strikes me you need to get a competent and experienced electrician to sort this out. That price is far more than a 300 mA RCD would cost to a commercial operation.

If they are trading safety against costs then they should not be selling and/or installing equipment about which they appear to lack essential knowledge as they cannot make informed decisions about how far safety is compromised.

Electric shock and/or fire that are not the only hazards, loss of power at night could lead to accidents in the darkness.

 

[bernardgreen]

However a Inverter can take upto 5 seconds to shut down following power loss.

There would be no reason for the invertor to shut down as long as it has power coming from the solar array. It would continue to run feeding power to the house even though the safety devices in the CU had detected a fault that required power to be removed. That is extremely dangerous.

The invertor power should feed into the system BEFORE the CU so that alll the safety devices have control over ALL power to the circuits in the house.

 

[JohnW2]

They then found out that the 300ma RCD would cost about £380, so have ditched that idea due to cost

Strikes me you need to get a competent and experienced electrician to sort this out. That price is far more than a 300 mA RCD would cost to a commercial operation.

Not only that, but I don't really understand why they were talking about a 300mA RCD (for the arrangement they were postulating), rather than a bog standard 30mA RCD or RCBO? Am I missing something?

If they are trading safety against costs then they should not be selling and/or installing equipment about which they appear to lack essential knowledge as they cannot make informed decisions about how far safety is compromised.

Quite so. From what little we know, it sounds quite frightening.

Kind Regards, John

 

[Greenie9657]

Hi Billy

Have you tried tripping all the MCBs that the suspect RCD is protecting. Reset the RCD then in turn reset each of the MCBs and leave it a few minutes. That should at least tell you which circuit the fault is on.

 

[plugwash]

There would be no reason for the invertor to shut down as long as it has power coming from the solar array.

AIUI grid tie inverters are deliberately designed not to "free run" in the event that they lose the connection to the grid to protect electricity company engineers.

 

[Peter1919]

As a solar installer, although admittedly not on the electric side (I fit the panels). I would say the cause of your nuisance tripping is indeed likely to be the solar PV circuit as there can be a small amount of current leakage from an inverter which means anything else such as a fridge switching on etc which adds to the problem can take you over the tripping point for the RCD, especially if you happen to have got one which is on the slightly more sensitive side (i.e. actually trips at say 20mA).

We usually install our solar circuits in their own small consumer unit with a dedicated RCD and MCB so that it is not influenced by any of the rest of the house electrics nor has any influence on them.

I think that the plan of switching the Solar PV Circuit on to an RCBO is probably the best solution at this stage.

 

[billywolfe]

Well I seem to have opened a can of worms here...

I have just quickly done a couple of google searches for 300ma RCD's and it seems the price can vary from around £72 to over £400 depending upon the type. I don't know what type is required (I think Power One recommend a type B) . I have found another forum which discusses this exact point with reference to a PowerOne Aurora TLS inverter (which is what I have got) see here :- http://www.electriciansforums.co.uk...forum/43974-power-one-ac-side-protection.html

Thanks Bernard for your comments - I will certainly check on the installation re inverter feed path before the CU.

 

[billywolfe]

Correction to above. The RCD Type should be AC according to document below published by Power One :-


RCD PROTECTION INTEGRATED INTO
POWER ONE AURORA PHOTOVOLTAIC AND WIND INVERTERS
Power One Aurora PV and Wind Inverters (*) integrate an RCD protection device accordingly with the VDE
V 0126-1-1:2006-02 German Standard (refer to par. 4.7). All the Aurora models/versions sold in Europe
include this protection.
Power One Aurora PV and Wind Inverters have build in redundancy for current leakage readings. Detection
has sensitivity with respect to all leakage current components, including the DC component. The leakage
current measurement is performed simultaneously, and independently, by two different microprocessors. It
is sufficient that one of the two devices observes a fault condition to enable the protection and so disconnect
the inverter from the grid.
Aurora PV and Wind Inverters have an absolute threshold for the total leakage current (AC+DC
component), which has a maximum tripping time of 300msec.
In addition three other tripping levels are present, dependent on the rapidity of the leakage current. For
these levels the threshold is set respectively at 30mA/sec, 60mA/sec and 150mA/sec to cover the quick
changes of the failure current due to accidental contacts with live parts. In these cases, the higher the rate
of current change, the shorter the protection tripping time. In particular the tripping time is (at a maximum)
300msec for 30mA/sec variation, 150msec for 60mA/sec and 40msec for 150mA/sec.
It is critical to note that the integrated RCD protection device will protect the system just from the ground
fault that could occur in part of the plant between the PV modules or Wind generator and the output terminal
blocks of the inverter (i.e. the AC lines aren’t protected against ground fault by the RCD protection
integrated into the Aurora PV Inverters).
The detection of leakage currents that could be present in AC lines requires an additional external device.
For the AC line protection, with reference to the German standard VDE100-712, if the PV or Wind inverter
isn’t able to inject direct current to ground (i.e. isn’t able to leak direct current), it is not required to install a
B-type RCD. For this reason, it is not required to install a B-type RCD at the output of an Aurora Inverter.
Power-One suggest using a magneto-thermic automatic switch with an RCD module (with adequate voltage
and current rating basing on the grid characteristics and on the output current of the inverter) with a
sensitivity of 300mA, AC-type. This will aim to avoid wrong protection trips, due to the capacitive leakage
current related to the PV modules and that will be present because of the transformer less inverter’s
topology.
Date: February 4th, 2010
(*) All string models.


So I have searched again and can find a Lifestar 100A 300mA DP RCCB for £101 at London & Kent Electrical. This is a Type AC RCD but I don't know what rating I would require (63A is a bit cheaper)

As I said in the first instance, I had the CU fitted at the same time as the Solar install as it made sense with the electricians here anyway.
I have asked the installer why (or if) none of his other installations have the same problem. The answer is that all their other installations have been wired (from inverter) to a separate box which just has the Solar on a 20 amp MCB protected by a 32ma RCD. This is the first one where they have put the Solar in a bank of MCBs in a CU. (Also the other installations have had possibly smaller PV output (my sysyem is 16 panel 4 Kw) and different inverters.

Of course I didn't know that Inverters seem to have a possibility for earth leakage so the setup seemed fine to me - the technical chap I spoke to at Napit also said that the wiring into the CU from the inverter was OK according to regulations (but he strongly recommended the use of RCBOs).

To be fair to the Solar Company, I think they are doing their best to find the best solution to a tricky problem. The Power One publication does say that the Inverter has built in RCD protection, so maybe the best solution is to have the solar on a 20A MCB in a separate box protected by a 300ma Type AC RCD.

 

[bernardgreen]

AIUI grid tie inverters are deliberately designed not to "free run" in the event that they lose the connection to the grid to protect electricity company engineers.

In an ideal world the connection between invertor and the incoming supply should be disconnected by a FIT connection unit when loss of network supply is detected. The FIT connection unit may be in the same "box" as the invertot but it's function is separate from that of the invertor.

The invertor may still run to provide power to the house while the network supply is not available . When network supply is restored the connection unit must synchronise the invertor's sine wave to that of the network sine wave before reconnecting invertor to the incoming supply.

It does seems that cost constraints and marketing pressures have led to installations that may not be ideal.

 

[PrenticeBoyofDerry]

Thanks for all for the replies. I am trying to get the electricians back but not succeeded so far!

If you have his details, input them here http://www.competentperson.co.uk/, thst will tell you whether he or the company is a member of a scheme provider and if so, I would inform the electrician, that you are making a formal complaint about his standard of work.

To remove all loads from live and neutral - does this mean unplugging all appliances?

Anything plugged in is best unplugged, as you need to remove both neutral and live from appliances, the switch on a socket outlet is likely only to take the live conductor of the appliance, but some socket outlet are double pole but most are not.

Is switching off the cooker at the isolating switch ok? (The cooker is about 2 years old).

Generally cooker isolators are double pole and as you have a dedicated circuit I would guess yours was also.

Is there any other way of testing appliances for the sort of problem which could cause an RCD trip? (I guess not...or I would need specialist equipment...)

You would require special equipment, portable appliance testing would help diagnose an appliance fault.

In-Service testing involves
Preliminary visual inspection
Earth continuity tests (for Class 1 equipment)
Insulation testing (Which may sometimes be substituted by touch current measurement)
Functional checks.

Although the cooker is only two year old, they still can develop faults.
I would check the warranty date, it may still be within expiry.

 

[JonoPashley]

AIUI grid tie inverters are deliberately designed not to "free run" in the event that they lose the connection to the grid to protect electricity company engineers.

In an ideal world the connection between invertor and the incoming supply should be disconnected by a FIT connection unit when loss of network supply is detected. The FIT connection unit may be in the same "box" as the invertot but it's function is separate from that of the invertor.

The invertor may still run to provide power to the house while the network supply is not available . When network supply is restored the connection unit must synchronise the invertor's sine wave to that of the network sine wave before reconnecting invertor to the incoming supply.

It does seems that cost constraints and marketing pressures have led to installations that may not be ideal.

Ideal world or not. Any inverter manufactured and certified for G83 use must shut down within 5 seconds of mains power loss. I must agree with the standard of solar plummeting although its been on a downward trend for awhile now.

 

[billywolfe]

The solar company have contacted me today and this is the proposed solution.

In the CU, the solar circuit on a 20a MCB will be taken out of the bank of MCBs protected by a 30ma RCD. So the RCD will then protect 5 "house" circuits - 2 lights, 2 sockets and 1 cooker. The 20ma MCB to which the solar is connected will be slotted into the CU as an unprotected MCB i.e. it will have NO RCD PROTECTION.

Of course I questioned this as I have not seen this solution either documented OR talked about on this forum. I have been assured that both PowerOne and NAPIT have said that this is a safe solution. I have not been able to contact either.

My a/c cabling from the inverter in the loft down to a cupboard where the CU is housed (I live in a bungalow) is attached to the surface of the wall in the loft i.e not in trunking (if this makes a difference).

Any comments? Does this sound OK?