RCD's not tripping with TL transformers

[grandfortune]

HI all,

getting some info though that there is problems with Transformer less Inverters and RCD's not tripping.

thought id check one of mine and low and behold it didnt trip.

It is the first time ive tested it with the PV on though so i would never normally pick it up. Looks like the the Inverter is injecting DC into the RCD toroid and saturating it.


ill try and get some more info as this now has gone above my head lol

Cheers
Crofty

 

[jason121]

the trip lock injects DC into the RCD to stop it tripping when testing for earth loop on an RCD protected circuit

 

[SolarCity]

I don't think he is talking about earth loop tests.

I'm actually surprised about this as it isn't something that has happened to us on the few installs we have carried out on RCD protected installations. What inverter are you talking about?

I'd be interested to hear if any other installers have had this issue. This is going to open the 'What RCD type do I need for TL inverter' debate once again I would imagine. The general consensus is that a type B RCD is only required if the inverter can inject DC currents into the AC side of the inverter. The reason being that it would not operate with standard RCD types in this scenario. But surely this is only under fault conditions and I've yet to see an inverter that isn't protected from this kind of issue anyway.

 

[jason121]

I know that
if you inject DC into a rcd it woont trip
thats why when you have to test for earth loop on an rcd circuit you use trip lock.
trip lock injects DC current to prevent rcd from tripping
:32::32::32:

 

[grandfortune]

Im not doing a loop test, im testing the RCD tripping times.

 

[BruceB]

What inverter are you having a problem with? I tried connecting my 8000TL to the wrong side of the rcd recently and the rcd still worked fine on the trip time test at x1/2, x1 and x5 (0 & 180 in each case) whilst the inverter was outputting.

 

[grandfortune]

It was a Eversolar 4000TL, it seems to be tripping now, i didnt test it the apprentice did, seems ok for me, now im wondering if he did it properly which he said he did, and it is set to do all tests automatically. i thought id try it as this problem arose on another forum.

Cheers
Grand

 

[BruceB]

Yes, it is being discussed on the Navitron and IET forums. Do Eversolar give the same assurance as SMA and Power One that their TL inverters cannot by design inject dc into the grid? I have never used Eversolar.

 

[grandfortune]

Not sure bud, i never really asked the question, seems strange its now tripping, TBH ive never tested one with the load on the RCD/RCBO before.

 

[BruceB]

To answer my own question, the Everolar instructions for the TL3000-TL6000 I found online (albeit canadian version) specifical state that if an rcd is connected on the supply side then it must be a type b. So it looks like they can leak dc which would impede the operation of a normal type ac rcd found in most consumer units in UK.

Edit: Also found the UK instructions for the TL1700 and TL2000 inverters. Surprisingly the word "rcd" does not appear anywhere in the installation instructions. What they do say is:
"5.6.7 DC current injection monitoring
The Eversol is integrated with DC current injection monitoring unit. The DC output current at the AC terminals shall not exceed 0.5% of its rated output current of 5 mA, whichever is the greater."
I think the second 'of' is a typo and should read 'or'. At 2kW output, then 0.5% comes to 43mA of dc I believe. I doubt that is good for a type ac rcd, but do not know their resilience.

Makes me glad I have always used European inverters!

 

[MorganPVI]

I believe i've been told 6ma can swamp a type AC RCD and can cause it not to trip.

 

[SolarCity]

Is this DC to grid leakage something which happens under running conditions or only under fault conditions?

 

[BruceB]

Normal running for Eversolar. SMA and Power One say theirs can't by design inject dc to grid. Have not looked up other manufacturers.

 

[SolarCity]

You learn something new every day - I always thought it was something that could happen under fault conditions only.

 

[grandfortune]

Darn, ill look more into it but it looks like ill have to fork out £200+ for type B rcd's.

 

[BruceB]

Or install in way that does not require an rcd. Normally relatively easy to do unless TT.

 

[grandfortune]

As a spark i find it almost impossible to install anything with a depth of 50mm, even in trunking i usually goes though some sort of wall or floor and the cable becomes less than 50mm somewhere along the line.

 

[BruceB]

A cable just going through a wall or floor from one side to another (say a wall from outside to trunking inside) does not bring in the 50mm rule. What I was suggesting was surface run swa, or solar cable or T&E in surface trunking. The only times I have 'had' to install rcds has been TT.

 

[SibertSolar]

Hmmm.....

TypeAC trip coil will saturate with any DC component, pulsed or pure
TypeA trip coil will saturate if the pulsed DC component is greater than 6mA but will saturate with any pure DC component
TypeB trip coil(s) will not saturate with either pulsed or pure DC component, regardless

Interesting thread. I shall monitor and try to offer assistance if required. For anyone needing pricing/stock of TypeA or TypeB RCDs then feel free to get in touch,

Andy

 

[SibertSolar]

Btw....if anyone's planning on visiting the SolarPowerUK show in B'ham in October then we'll have representatives from Doepke with us along with a demo unit to show how TypeA and TypeB RCDs operate.

 

[sparkatus12]

Can somebody explain why dc prevents rcd tripping. Also is it true that capacitors in appliances can cause rcd to go over range with trip times. If so why? Thanks.

 

[SibertSolar]

Basically, the DC component of the AC waveform "magnetises" the trip coil of the RCD and prevents it from functioning normally under AC leakage current fault conditions (magnetisation of the trip coil stops its ability to sense changes from phase to ground)

Your second question relates more to "nuisance tripping" which is a separate issue to DC leakage/sensitivity. PV array frames can generate parasitic capacitive AC leakage currents (especially on moist mornings...) which, when combined with the inverter's inherent AC leakage current, can exceed the minimum tripping level of the installed RCD on the AC side, even when a ground fault doesn't exist.

 

[SibertSolar]

Apologies, just re-reading and perhaps I didn't answer your 2nd question actually! Not entirely sure about capacitors in electrical appliances other than perhaps the same principal applies regarding excessive AC leakage current being generated by the capacitors.....

 

[sparkatus12]

No I was told capacitors released dc back into the system preventing rcd from tripping in correct times.

 

[SibertSolar]

Hmm, not sure about that on typical AC electrical appliances to be honest. Anywhere there's frequency conversion or DC/AC inversion, and no galvanic isolation, tends to be an area where DC onto AC is a concern but I've no experience with standard electrical appliances incorporating capacitors causing issues.

 

[danesol]

HI all,

getting some info though that there is problems with Transformer less Inverters and RCD's not tripping.

thought id check one of mine and low and behold it didnt trip.

It is the first time ive tested it with the PV on though so i would never normally pick it up. Looks like the the Inverter is injecting DC into the RCD toroid and saturating it.


ill try and get some more info as this now has gone above my head lol

Cheers
Crofty

Did you not know this at installation stage when you did the electrical tests for customer documentation ???

 

[grandfortune]

Yes, i dont test RCBO's or RCD's with the load connected.

 

[grandfortune]

No I was told capacitors released dc back into the system preventing rcd from tripping in correct times.

you dont test a trips timings with the load on so it should never affect the times on a test meter

Cheers
Grand

 

[grandfortune]

A cable just going through a wall or floor from one side to another (say a wall from outside to trunking inside) does not bring in the 50mm rule. What I was suggesting was surface run swa, or solar cable or T&E in surface trunking. The only times I have 'had' to install rcds has been TT.

True, im probably speaking from a sparky point of view as working with mostly T&E, so basically without worrying about the 50mm rule you just RCD it and be done with it