SMA TL Inverters and type B RCDs

[paulsamuel1984]

SMA have advised me that DC injection cuts off at 100mA for the STP range. So they are certain a type B rcd is not required. However, if this is the case, is a type A rcd required, or no rcd at all? what rating should the device be if required? Where would you locate the RCD, prior to the G59 relay in your PV switchboard?

 

[SibertSolar]

Interesting, what do they mean "cuts off at 100mA"? Are they talking about an inherent DC leakage current? Or are they referring to a potential leakage current fault?

I think that whether an RCD is needed or not is down to the nature of the installation and/or the type of inverter. The DTI guidelines state (section 2.3.1) that if at least a simple method of separation (DC side to AC side) is not incorporated into the PV power supply system (for example: no transformer in the inverter) then a Type B RCD must be used. SMA's statement regarding their TL range refers to not needing a Type B. They don't mention Type A because they don't need to - in Germany they don't even use Type AC RCDs so the base/default type of RCD they would use, if the electrical installation required it, would always be a Type A. The UK is different in that we still allow the use of basic, Type AC RCDs.

In terms of ratings, the TL inverter's inherent leakage current characteristics should be considered but (in Doepke's experience) a 100mA trip limit is the typical level required to avoid nuisance tripping by the inverter itself. If there are fire hazard considerations to be taken into account then a 300mA, or even a 420mA limit may be used.

As far as location is concerned, I'm not sure I'm best qualified to advise so I'll leave that to those on here who actually work in the real-world of installations... ;-)

Hope this helps, open for correction if my understanding is wrong/incomplete

 

[paulsamuel1984]

http://files.sma.de/dl/3427/EK_RCD-SBSMC-ZEN104814.pdf

attached link for discussion

 

[SibertSolar]

@sellsolar - thanks for this. What I believe they are saying is that a minimum 100mA trip limit RCD is recommended (to avoid nuisance tripping).

Point to note is that they state that the integrated RCMU is an additional safety device to an RCD. The tests they performed to determine whether a Type B RCD is needed were done without taking the integrated RCMU into account. The only reason I highlight that point is to avoid people possibly interpreting the SMA (or Fronius, or PowerOne etc) inclusion of an "RCD-type device" such as the RCMU as being "fitted already with a Type B RCD" - which is not the case, the RCMU is not an RCD, Type B or otherwise.....

I need to check on the list of "approved RCD Type A devices" and whether that has been updated recently as the document seems to be dated 26th Nov 2010.

In case anyone needs the other SMA document regarding selection of RCDs, a copy can be obtained here: http://www.sibert.co.uk/uploads/spare_document/filename/37/RCD-UEN103120.pdf

Thanks again for this,

Andy

 

[Worcester]

This may help explain why SMA TL's don't need an RCD : http://files.sma.de/dl/3491/TECHGRID-11-AE2306.pdf Is about being more than an RCD and embodying the requirements into a standard rather than prescribing a technology.

 

[edexlab]

I've been reading this thread and Bruce B is on the money, stating exactly the same information we were given when we went to SMA last year ,they were quite clear about this and it was discussed at length that this was a factor in the design to be able to install Tls without the need for Rcds which are difficult / expensive to source