Number of DC isolators required?

Post Reply
Posts: 40
Joined: Thu Apr 16, 2020 12:53 pm
Has thanked: 11 times
Been thanked: 1 time

Number of DC isolators required?

Post by TPower »

If I have a PV system with two or more PV arrays connected to an inverter, what is the requirement for the number of switch-disconnectors?

I’ve only ever seen one installed at the inverter in accordance with 5033: But should there be a separate one for each array? I that what clause 2.1.4 is getting at?
Posts: 4
Joined: Sat Dec 05, 2020 11:39 am
Has thanked: 3 times
Been thanked: 0

Re: Number of DC isolators required?

Post by RikN »

I initially thought there was a contradiction, and it took me a while to get my head around it.
None if you read the first line of

The way I read the rest of the reg is that a 'means of isolation' at each PV array location shall be provided.
This could be an isolator or fuses or an MCB.
I suppose it all depends on how much you are paying for parts and the gear can handle the right switching conditions.
And its labelled accordingly.
Posts: 476
Joined: Thu Apr 16, 2020 11:24 am
Answers: 2
Has thanked: 2 times
Been thanked: 160 times

Re: Number of DC isolators required?

Post by AlecK »

Short answer: one for each array.
And while 2.1.4 deals with related matters; it isn't directly relevant to the question.

There are 2 separate isolation requirements to be considered.
Unfortunately "5033" (both versions) is not very well put together. It doesn't follow the normal Standards style. In particular it jumbles some things together that should be stated separately; while for other matters it splits related stuff up.
So it takes careful reading to extract the meaning.
And unfortunately it also means this explanation will be longer than most .

Any discussion of what's required - as against what may be considered good practice - has to start with looking at the officially correct version of the Standard. There's no clause in the cited version (2012 +A1+A2; as modified by the citation in Schedule 2 of ESRs).
We can use the later (2014) edition; but only as long as it doesn't conflict with the cited edition.

Both isolation requirements relate to means of isolation of the array from the PCE, as required by Clause 4.4.1. directs us to table 4.3 to find out what the specific requirements are for the particular type of array;
generally one for each array or sub-array .
Note also that some of these isolation devices are specified as "disconnection device";
but the array cable in all cases (except where the PV modules have micro-inverters) must have a "readily available load-breaking switch disconnector". specifies that the isolation of array from PCE must operate in all poles; and gives details of the 2 permitted options.
Neither option is remote from the PCE, one is adjacent to, and the other is part of, the PCE. covers how the means if isolating the PCE is to be installed.
1st 3 paragraphs: what types of device may be used where the Table just says "disconnection device"
para 4: location of devices relative to PV modules
para 5: no array switch disconnector needed if all sub-arrays have switch disconnectors and they are all located "close" to the PCE.
(this should have been written as an Exception)
para 6: additional requirements where isolation (of array from PCE) requires use of multiple disconnection devices.
para 7: which conductors the disconnection devices must operate in
(also defines "live conductor"; which should be in the definitions)
para 8: that a required load-breaking function must be in all poles of the device
para 9: any switches that are not load-break type to be labelled accordingly

We would normally skip; as it's not applicable in NZ.

Clause says a switch disconnector is not required at the array location provided certain conditions are met.
Note that nothing in the preceding sub-clauses indicated any such requirement; all the rest of 4.4.1 is about isolation at / near the PCE
The relevance is by comparison to the [Australian]
(Basically is written as an Exception to; but it has been separated out as a standalone [NZ] clause. No wonder people get confused!)
by reading right through; we find that in Oz an array switch disconnector is required, located at the array. If the array & PCE are either more than 3 m apart, or not in sight of each other; this means 2 switch disconnectors - and the one at the PCE end must be mechanically interlocked with the PCE as per option (b) of
(this is actually a contradiction of &; which say one is always required at the PCE end)

So by reading the [A] subclause, suddenly the meaning of the [NZ] subclause becomes clear:
In Oz, an array isolator is always required at the array end; and in most cases a 2nd one is required interlocked with the PCE.
But in NZ, provided we comply with the conditions (a) to (d); we don't need to install one at the array end, but we always need one at the PCE.

Now for the conditions. (a) , (b) & (d) are straightforward.
But item (c), which requires array cabling to be exterior to building isn't.
(In fact it's very poorly written; but the meaning isn't too hard to extract.)
The underlying requirement is that the array cabling must be outside the building.
First exception is we can have up to 300 mm inside - enough to come through a wall and into the PCE.
2nd Exception is we can have more than 300 mm inside the building if the cable is in a "metallic wiring enclosure" that is earthed.
A Note states that PVC conduit is not permitted.

However the citation in schedule 2 of ESRs changes this item; so that the requirement is not just "metallic" but "metallic high density PVC"
Taken literally, there is no such product; so this is clearly an error.
Most people assume the intent was to provide a choice between "metallic" or "high density PVC".
Two problems result:
- the citation doesn't remove the obligation to earth this wiring enclosure, and earthing HDPVC isn't possible
- The Note also continues to prohibit using any kind of PVC (officially a Note to text can't set a requirement; though this one is just reinforcing that the actual requirement is for "metallic").

Basically there's NO way of complying with the condition as modified by the citation change.
So the ONLY way I can see of being certain of compliance is with all array cabling outside the building (except 300 mm).
If wanting to run inside for longer than 300 mm, earthed metal wiring enclosure would comply with the Standard as published, but technically not as changed by citation.

The re-write in 2014 edition doesn't help much for interpreting ; and can't be relied on anyway 'cos we have to comply with the cited version. clarifies that in NZ an array-end switch disconnector is not required. No conditions; just simply "not required".
There must still be a means of isolation at the array end; but it doesn't have to be a switch disconnector.

As far as isolation goes; it remains 1 switch disconnector per array / sub-array / string that arrive at the PCE. covers the requirements where there;s more than one disconnection device for the d.c. input(s) to the PCE.

The requirement to keep the array cabling external is removed; but a new requirement comes in for "no joins in the wiring enclosure".
This requirement should not be in a clause about isolation; as it's completely unrelated. It should be in 4.4.4 ,that covers wiring systems for arrays.

My understanding is that many inverters being installed have the d.c cabling terminated to them directly, with additional terminals for the switch disconnector(s) to be connected to. If the inverter requires servicing, the main part unplugs from the base, without any need to disconnect any cabling.
That would comply with option (a) - adjacent but physically separate switch-disconnector.
But not with option (b) - ; because there's no mechanical interlock between the switch-disconnector and the removable module.
These users thanked the author AlecK for the post:
Slovett (Mon May 17, 2021 1:44 pm)
Rating: 16.67%
Post Reply