array earthing / bonding

[maffa ent]

evening all, I'm after a definitive answer regarding array earthing please. links to information would be fine?

I read the 'decision tree' on the dti guide, and decided that, using a transformerless inverter, the array would need a separate 10mm bond to a dedicated rod.

The supplier then confused me by saying that, if used in conjunction with a type b rcd (presumably he meant an rcbo?), the separate bond would not be required if the array was outside the equipotential zone (which, on the roof, it will be).

Laugh at the question if you must, but I'm a newbie with pv, just getting set up for it, so it may be a simple problem to you, but I'd like to start off doing things correctly.

thanks in advance.

 

[SolarCity]

Always good to ask if you're unsure.

You will still need to bond the roof regardless of the type b rcd. A minor voltage would be present on the array frame if it is not bonded and this would exist regardless of the RCD, whether it switched on or not.

You only need to use a dedicated rod on a PME (TNCS) system.

 

[maffa ent]

thanks for the reply mate. i take it if i use an inverter containing a safety isolating transformer, then i dont need to do anything with earthing or bonding the array framework?

 

[SolarCity]

That's right. Just leave it floating if you're using a standard string inverter.

 

[malcolmsanford]

Maffa welcome to the forum and good luck with your venture, but I would be less than diligent to ask, do you feel you are ready to do a project in someone's home. On another post on the forum you are asking about what size cable would be needed when piggy backing an Enclosure onto an existing CU, without a thought to design ie loading, type of cable erection method.

On here you have rightly pointed out that for a transformless inverter a B type RCD is required. You put do they mean a B Type RCBO, do you know how a B type RCD differs from a B type RCBO ?

I have a feeling you think that the B is the inrush characteristics of the device, which is not the case of a B type RCD.

 

[SolarCity]

On here you have rightly pointed out that for a transformless inverter a B type RCD is required. You put do they mean a B Type RCBO, do you know how a B type RCD differs from a B type RCBO ?

I have a feeling you think that the B is the inrush characteristics of the device, which is not the case of a B type RCD.

To be fair, I doubt that the majority of electricians would know the difference. Judging by my correspondence with suppliers, I would say that the confusion is industry wide.

 

[VMan]

To be fair, I doubt that the majority of electricians would know the difference. Judging by my correspondence with suppliers, I would say that the confusion is industry wide.

That's quite worrying then!! ,

RCD's alone won't protect you if basic protection failed, eg, short circuit, only a fault to earth!!

RCBO's will protect against Basic and fault protection.

 

[malcolmsanford]

To be fair, I doubt that the majority of electricians would know the difference. Judging by my correspondence with suppliers, I would say that the confusion is industry wide.

That Biggs must be a sign of the woefully poor training that is now being adopted industry wide. Fingers can be pointed at 5 week wonder courses and the such, but it goes deeper IMO than this. It is OK in my opinion to have guys that may not know, and not want to know protection device characteristics, they are told to run cable here, connect to this and that and they are happy, good electrician mates or this ridiculous new scam the "Improver".

But when your designing an installation as Maffa is doing then I would expect the designer to have the expertise to know what a B type RCD is, what a B type RCBO is, especially in the PV industry. Granted an electrician not within the PV industry is unlikely to encounter any RCD other than a bog standard AC type.

 

[SolarCity]

RCD's alone won't protect you if basic protection failed, eg, short circuit, only a fault to earth!!

RCBO's will protect against Basic and fault protection.

While I've no doubt that all electricians (and I use the term tightly - I'm not referring to 5 week wonders) know the difference between an RCD and an RCBO, when you thrown in the 'type B' part it tends to confuse people. I'll be honest and admit that I'd never even heard of 'type B' RCDs before I moved into PV installations.

 

[SibertSolar]

In fairness to the likes of Maffa, it can be a bewildering array of options and it is a dynamic industry with new technologies presenting new challenges on an almost weekly basis. AC, A and B Type RCDs, B and C Type RCBOs, RCMUs, Type B, C & D MCBs, Type 1 and Type 2 AC and DC surge/lightning protection, time-delay, surge-resistant, etc the list goes on......it's difficult enough as a supplier so we empathise with engineers who are not only trying to keep up with the PV technology but also having to correctly interpret and keep to the electrical code too!

One of our primary objectives is to try to understand as much as we can and to share information with our peers and client base wherever possible. If there's anything we can do to help further then my door is always open,

Andy

 

[malcolmsanford]

I hope that this is not an avocation of design by telephone.

I have no problems with suppliers advising on technology or the advances in them, as the days are gone I think when a supplier sold someone an item of equipment and their obligation was over as soon as money exchanged.

What I'm saying is that perhaps someone that is not sufficiently experienced enough or technically informed enough to realise what a component is, and it's use, should not be designing PV, or for that matter any type of electrical. installation.

IMO it should not a matter of someone being able to call a supplier and saying

"Oh I read this inverter catalogue and I need ......hmmm a B type RCD ?"

And said supplier

" Yes we have them 50 pound+ vat (SIC)"

To me that is papering over a very large crack in someone's training and education, and is what is systematically wrong in the industry.

 

[maffa ent]

I have to admit, I have no experience regarding type b rcd's, but I'm all ears!

I do understand plenty of what you're saying, as it's been my job as a spark since my apprenticeship finished in 1990 to understand. Throughout that time it's always been my intention to do a good job, and nothing's changed.

Lack of experience is my problem with pv, and my customer understands that, and has had it reflected in the price of his job. I've done plenty of electrical work for him in the past, and he's always been happy.

I totally appreciate what you're saying - I've said the same about some inexperienced installers of various things in the past - the difference is I will get it right. It may take me lots of second guessing, and lots of time, and lots of questions on places like this!!, but the job will be right.

It's the only way to do it right?

 

[maffa ent]

by the way malcolmsanford, myself - maffa ent - and maff (who asks about piggybacking a unit) are two different people.

 

[Graeme Harrold]

When I did my PV course the instructor did mention the type B RCD and also mentioned that it was an elusive beast that none of the major wholesalers knew existed never mind carried. From a technical aspect it is quite confising for someone entering the PV market to automatically think of a type B RCD as one having B curve inrush characteristics, so I would blame the naming convention for inducing the confusion. Only recently, and Im talking weeks have Type B RCD, been pushed by suppliers (thanks to Sibert) that have seen this hole in the market. I for one was advised to fit transformered inverters UNLESS the manufacturer had stated they were not required due to their DC leakage prevention characteristics e.g. SMA.

To point out I come from a Domestic installer background, and therfore one of the 5 week wonders......I did my PV course with time served sparks who were cluelless about DC, then again I have 20 years experiance in DC and found it a piece of cake, but that did not give me the right to poke fun at the 3 day PV wonders!!! We all have to learn

 

[Ramjam]

If you use a sunny boy inverter such as a SB4000TL then you will need to earth the array frame. This I was lead to understand was because the array frame may provide a tickle to anybody working on the roof and that it would not cause injury itself it could cause them to fall...

This is only part of the story. We installed an SB4000TL and accepted an earth electrode resistance of 190ohms (I know less than 200 for regs, less than 100 for good practice!).

The inverter caused the domestic 30mA RCD to trip around every hour. It turns out that the earthing of the array frame for TL inverters is not just a matter of good housekeeping but failure to provide an escape path for the leakage means that it will travel through the inverter causing nuisance tripping of the RCD.

The solution from SMA is to use only 100mA RCDs for TL inverters as upto 80mA is possible. This is fine in Germany but not here in the UK. Apparently lots of installers simply don't RCD protect the circuit and don't earth the array frame, hard to believe but apparently quite common!

Moral of the story, fit TL inverters but put a good earth rod in or you'll get nuisance tripping.

Oh and by the way there is a 100mA type B RCD fitted internally in Sunny Boy Transformerless inverters so there is no need for an external one. I don't know about other makes because we only fit Sunny Boys!

 

[SibertSolar]

Ramjam - are you sure that the internal RCMU (residual current monitoring unit) fitted to the TL range of SMA inverters is a Type B RCD? Or are you referring to an additional device fitted internally? Do you interpret this to mean that even a Type A RCD is not required?

I'll stick my neck out here with my below comments/queries as I want to understand better but I'm simply not qualified to interpret the regulations well enough but have tried to form a simple understanding myself, I'll apologise in advance and stand corrected on anything that I've got wrong:

Why "this is fine in Germany but not in the UK" for 100mA trip limits? Surely a necessity for a 30mA trip limit (for "human" protection") should be avoided by virtue of the fact that the PV installation incorporates Class II equipment and contains non-pluggable (into sockets) devices. If other factors (fire risk, cable type/routing etc) dictate that an RCD should be used then doesn't the fact that the DTI guidelines state that the AC supply from the inverter to the DB/CU should be unique to the PV system and not have any other devices or circuits connected to it, mean that a 100mA trip limit (or more) is perfectly acceptable as it's related to a separate part of the household/property circuit from a supply/load point of view?

Bonding the array - the DTI guide suggests that whenever a TL inverter is being used, you must bond the array in some form or fashion (earth spike or direct to CU earth terminal). Does the DC leakage current you mention above from the array have any relation to DC leakage currents generated by the inverter itself? Are they the same thing? Having a TL inverter anyway means that there is no simple separation between the DC and AC sides. Hence an RCD capable of protecting against DC leakage faults/currents (pulsed or pure) as well as the fundamental protection against AC ground/leakage faults is required as a minimum. This means that either a Type A or a Type B should be selected. Type AC devices are used in the UK still but are not capable of providing protection if the internal coil is saturated with a DC leakage current. I believe that half the recent confusion with the current SMA statement(s) are that, in Germany, they don't use Type AC RCDs at all, even for their standard domestic electrical circuits. Their "base unit" if you like is the Type A. So, when they say "you don't need a Type B" the fact that they omit any mention of a Type A device is simply because that is the defacto RCD in use over there anyway so doesn't need to be clarified.

I am curious as to people's interpretations of SMA's (and Fronius', and PowerOne's) documents and whether anyone has actually spoken to SMA, or others, directly about this. It still seems like a bit of a grey area to me. I do know that discussions are still ongoing over in Germany and that the UK regulatory bodies are involved in trying to clarify the exact requirements. That's what I am told anyway.

My next area to investigate is surge/lightning protection........we're talking to Dehn now about their products.

Andy (off to don flameproof suit)

 

[SolarCity]

I may be missing something important here, but why are RCDs being fitted anyway, assuming that the cable route doesn't call for it?

 

[malcolmsanford]

I may be missing something important here, but why are RCDs being fitted anyway, assuming that the cable route doesn't call for it?

Reg 551.7.2 (iii) might be the reason people are fitting RCD protection and 30mA.

 

[BruceB]

Reg 551.7.2 (iii) might be the reason people are fitting RCD protection and 30mA.

They would be wrong to do that though because that regulation is talking about generating sets in general. The specific rules for PV in 712 do not allow a PV generator to be connected to the load side of the protective devices of a final circuit. It has to be on its own mcb in order that it is connected to the supply side of the protective devices for final circuits of the installation.

Regards
Bruce

 

[malcolmsanford]

Section 551.7 for Additional Requirements (note that ) for Installations where the generating set (PV ARRAY) may operate in Parallel with other sources including systems for distribution of electricity to the public ..........and reg 551.7.2 is upheld by reg 712.411.3.2.1.1. But some might be looking at the reg as a whole.

Regs 551.7.4, 551.7.5 and 551.7.6 all refer us to BS 50438 [h=2]
Requirements for the connection of micro-generators in parallel with public low-voltage distribution networks[/h]Status : Current Published : January 2008


So I would have thought that this section would be relevant.

 

[BruceB]

I do not think so. 551.7.2 has 2 cases: generator on the supply side and generator on the load side. With PV, the generator has to be on the supply side from 712.411.3.2.1.1, so all the load side conditions are irrelevant because the generator should not be connected in that way. Short answer only as I am away now for the day - I'll see where this has gone later.
Regards
Bruce

 

[malcolmsanford]

Section 551.7 for Additional Requirements (note that ) for Installations where the generating set (PV ARRAY) may operate in Parallel with other sources including systems for distribution of electricity to the public ..........and reg 551.7.2 is upheld by reg 712.411.3.2.1.1. But some might be looking at the reg as a whole.

Regs 551.7.4, 551.7.5 and 551.7.6 all refer us to BS 50438 [h=2]
Requirements for the connection of micro-generators in parallel with public low-voltage distribution networks[/h]Status : Current Published : January 2008


So I would have thought that this section would be relevant.

You have to remember the Regs are a minimum standard and yes your right there is no need to if the installation is fitted on the supply side to fit an RCD, but some may take the reg as an whole and fit an RCD.

Same as there is no regulation that makes me fit a double pole RCBO in a TT system if the DB I was fitting it to had a double pole 100amp main switch, but I feel IMO that the regs does not go far enough and so I fit double pole protection devices for TT.

 

[SibertSolar]

Dear all,

Just to come back to the subject of RCD type selection, I have asked Doepke UK (who are currently in Germany discussing the whole TL inverter situation) to provide me with a generic guide to RCD type selection when fitting TL inverters with built-in RCMUs (Residual Current Monitoring Units). As I hinted at in my previous post above, an RCMU is not the same as a Type B RCD. Anyway, hopefully the below will help to clarify things somewhat. I'm assuming that those reading this will understand the differences between Type AC, Type A and Type B RCDs, as well as the effect that DC currents (pulsed or pure) will have on the trip coil inside an RCD protective device (saturation causing failure of coil to energise/release etc).

From Doepke UK:

The Installer needs to verify from the inverter manufacturer the inherent d.c. leakage current in normal operation ( i.e. the RCMU will have a minimum d.c operating current – this is normally around 15mA d.c.)
If the Installer can guarantee that there will be no d.c. leakage current ( today – this requires an isolating transformer ) he can use a Type-AC RCD.

If using a TL inverter, and the inherent leakage current is less than 6mA d.c. the installer can then use a Type-A RCD ( Doepke are not aware of any TL PV inverters where the RCMU will limit the leakage current to 6mA or less)

If the TL inverter's leakage current is more than 6mA d.c. (Doepke cannot guarantee the operation of the Type-A RCD under the existing standards), the installer will then need a Type-B RCD.

The customer is responsible for the PV installation – not the inverter manufacturer

So, this is where our concerns stem from. On the one hand the RCD manufacturer is reinforcing the DTI guideline but on the other hand the TL inverter manufacturer (for example, SMA) are stating in their documentation that a Type-B is not required and a Type-A is adequate. At the end of the day, if installers are happy to defer to SMA's current documentation then that's perfectly fine, just fit Type-A units. Where doubt exists, or the manufacturer cannot clarify, then perhaps it is better to err on the side of caution and fit a Type-B.

(Don't shoot the messenger....just trying to share info)

 

[malcolmsanford]

Dear all,

Just to come back to the subject of RCD type selection, I have asked Doepke UK (who are currently in Germany discussing the whole TL inverter situation) to provide me with a generic guide to RCD type selection when fitting TL inverters with built-in RCMUs (Residual Current Monitoring Units). As I hinted at in my previous post above, an RCMU is not the same as a Type B RCD. Anyway, hopefully the below will help to clarify things somewhat. I'm assuming that those reading this will understand the differences between Type AC, Type A and Type B RCDs, as well as the effect that DC currents (pulsed or pure) will have on the trip coil inside an RCD protective device (saturation causing failure of coil to energise/release etc).

From Doepke UK:

The Installer needs to verify from the inverter manufacturer the inherent d.c. leakage current in normal operation ( i.e. the RCMU will have a minimum d.c operating current – this is normally around 15mA d.c.)
If the Installer can guarantee that there will be no d.c. leakage current ( today – this requires an isolating transformer ) he can use a Type-AC RCD.

If using a TL inverter, and the inherent leakage current is less than 6mA d.c. the installer can then use a Type-A RCD ( Doepke are not aware of any TL PV inverters where the RCMU will limit the leakage current to 6mA or less)

If the TL inverter's leakage current is more than 6mA d.c. (Doepke cannot guarantee the operation of the Type-A RCD under the existing standards), the installer will then need a Type-B RCD.

The customer is responsible for the PV installation – not the inverter manufacturer

So, this is where our concerns stem from. On the one hand the RCD manufacturer is reinforcing the DTI guideline but on the other hand the TL inverter manufacturer (for example, SMA) are stating in their documentation that a Type-B is not required and a Type-A is adequate. At the end of the day, if installers are happy to defer to SMA's current documentation then that's perfectly fine, just fit Type-A units. Where doubt exists, or the manufacturer cannot clarify, then perhaps it is better to err on the side of caution and fit a Type-B.

(Don't shoot the messenger....just trying to share info)

Very useful piece of information Thank you.

As I noted on another thread I find it worrying that Inverters designed for the European market are not geared for UK Installation methods. I know the BS 7671-2008 denotes that manufacturers instructions must take precedent but I'm concerned that sometimes these instructions do not take UK standards into consideration

 

[SibertSolar]

Exactly Malcolm, I think your concerns are quite valid to be honest. This is why I mentioned before about Germany not using Type AC RCDs anymore and potential confusion with the likes of SMA not clarifying that Type A RCDs are acceptable (not on all their documents relating to this anyway). Hopefully the ongoing discussions within the industry will eventually lead to further clarification...

 

[Graeme Harrold]

Is this not a fault of the MCS product accreditation process. Surely if these devices attain such certification then the whole installation process should be reviewed as part of signing the product off, otherwise you have an MCS certified product conflicting with BS7671 or at the very least clouding the decision process on protection and bonding.

Ive had a look at the MCS cert for the SMA 4000TL and the max permisable DC injection is 20mA with the cert denoting a max pass through result of <14mA

 

[power7171]

I am Confused, I get the point that you dont need to earth the array if you have a seperation transformer in the inverter but if you installed a inverter such as a Sunnyboy HF which has no transformer you have bond the array to the main earth or earth rod if pme.

But why would you carry out a bond if you think about the array is out of any earth zone and out of reach. Most panels are double insulated to me you are only creating problems when lighting strikes and if you use a earth rod all you are doing is creating a earth on the roof which is out of reach.

 

[flamefix]

Er the HF inverter is High Frequency and does have a transformer so no earthing is required for HF inverters, only TL on PME systems.

Second para, I agree but the DTI solar install tree states the array must be earthed.