DC Isolator or no DC Isolator that is the question

T

Thegreenelec

Does anyone know if you are using a SMA Inverter with its built in pull off DC Isolator, if you need a seperate DC Isolator fitting next to the inverter? I know in Europe they don't fit them. If we don't need to fit one, where does it say for sure in black and white?

Thanks

Paul
 
Separate DC isolators are useful where you have more than one string coming into the inverter and to facilitate testing. But in my personal view, from a safety and operational perspective, they are not necessary where an inverter has a built in isolator. I think you can argue that the built in isolator satisfies 712.537.2.2.5. However I would acknowledge that it is a little bit grey. The only person likely to pick you up is a pedantic assessor.
Regards
Bruce
 
I'm not familiar with the SMA inverter, however, how would you safely isolate the d.c. side in order to replace the inverter if you have no external d.c. isolation?
 
Like Netblindpaul says, the inverter isn't isolated unless a DC isolator is fitted. The in-built DC isolator effectively kills the DC connection the workings of the inverter but there still exists a dangerous voltage AT the inverter.

Doesn't seem like a corner worth cutting in my opinion.
 
Sunny boys come with MC-T4 connectors that can only be reached once the front panel is removed. The panel con only be removed once the DC Isolator (Solar Switch) is removed. Was told on my course, technically compliant, but always better to fit additional DC Isolators.
 
I see no great issue either way and yes agree a large grey area however could you be condemned for for fitting a DC isolator regardless? It's not the big money item after all.
 
Nope, but when your've got 50 strings to wire up, the costs of the time and components do mount up, especially when the large inverters also have built in string fuses. Time for thought here.

I see that SMA have opened their UK office now, so time for a call.
 
With 50 strings, I agree that the cost does add up (although on a job with 50 strings I'd imagine it's a drop in the ocean) but we'd definitely still install the isolators.

What are your thoughts on testing? Without an isolator would this not make things tricky? Particularly Isc?
 
Worcester said:
Nope, but when your've got 50 strings to wire up, the costs of the time and components do mount up, especially when the large inverters also have built in string fuses. Time for thought here.

I see that SMA have opened their UK office now, so time for a call.
Click to expand...

I don't see any difference if it's one inverter, 50 inverters or 100 inverters. The actual cost of a DC isolator is minimal to the price of the job. Should such a small price item dictate weather it is used or not?

It's a grey area and a personal choice at the moment until the reg's catch up to the products. However there is a over ruling factor to any thing and thats good working practice.

I consider if in the future there's an inverter problem and the manufacture's appointed service agent is sent out. You know dame well that they will see an issue with the absence of a DC isolator regardless of if it's needed or not. How many times have we had this with boiler agents who don't under stand differences in wiring arrangements and just say it's wired wrong!

I personally will always fit. it's good practice. We use SMA and like the product, it's good stuff but regardless a DC isolator is fitted, every time.

Just my opinion.
 
Have always fitted separate DC isolators with SMA inverters and will continue to do so. As metioned above, the costs compared to the overall project cost are minimal and they facilitate much easier testing. Commissioned a 50kWp system this morning using 3 x SMA TriPowers all fitted with separate DC isolators to each string. Didn't add any significant time or money to the project and, as Markc says, no one will penalise you for doing a proper job.
 
@Markc, @JulianC

We always put DC isolators in, just thought I cause a little debate :)
 
I had a long chat with SMA today about this subject, and they are aware of a number of installations done this way (no DC isolators) - that's not to say it's right or they approve, just they know of some.

The problem is the fact that it is DC.

In THEORY, pulling the ESS switch on the Inverter will isolate the DC from the inverter, however, especially on their larger inverters this can only be done at night (little voltage / current) otherwise you could easily get massive arcing sparks and a bang! The inverter then goes into maintenance mode and you've probably lost 12 - 24 hours.

The above is why converted AC isolators don't work on DC, as the contacts break, (even in a DC isolator) you get arcing across the contacts, with AC this isn't a problem as the voltage goes to zero 100 times / second, with DC through an AC isolator the contacts can easily get welded, burnt, or even still be arcing, so although you think it's switched off, it isn't.

And live PV cables at up to 1000V and 15 amps isn't a good idea.
 
Worcester,
Is this not why you must use suitably d.c. rated isolators?
The voltage will sustain the arc and the current will fuel it, thus it could be a big bang in the wrong situation!!!
 
Very relevant and important messages being expressed by Worcester/Markc/JulianC/netblindpaul above. With AC isolators, even when series-linking multiple poles together to achieve a suitably-rated 2pole output, the fact remains that they are not typically designed to manage making/breaking DC currents/voltages.

Proper and correct DC isolation should not be a commercial decision - the important thing is to make sure that the product being used is designed for purpose.....

Andy
 
Last edited:
@netblindpaul

Yes!

Sorry guys been at Intersolar.. out at 4am Thursday Back at 3 am Friday .... Need sleep!

Andy at Sibert can supply DC isolators (the one's in my timelapse install video) they're made by IMO and his prices are good,

Here's the technical info from IMO:

IMO Precision Controls - DC Isolators

(No big bang then!)
 
Last edited:
KenHobbs said:
The SMA ESS DC Disconnect is fully compliant as a DC Isolator, so an additional isolator is not required.

Ken Hobbs
Click to expand...

Ken, I had a long chat with SMA technical about this and other matters the other day, especially regarding the situation when you have 50+ strings pumping out 800V (we have a few!), they were adamant that using the ESS as the primary DC isolator was not a good idea.

Yes it will do it, however the comment was that if you have a string (accidentally on hopes) connected with the wrong polarity, then if you disconnect it using the ESS you will get big sparks and a big bang, and the Inverter will go into maintenance mode and you wount be able to do anything with it until dark. You will also potentially invalidate the warranty.

Yes it will do it, however is not designed to cope with accidental incorrect wiring, so they recommended using seperate DC isolators.
 
@bhmike, the problem is in how the DC circuit is broken.

Some of the Isolators sold on the market as DC isolators are actually modifed AC isolators (read the earlier posts) and may not effectively isolate the DC after a period of use.

Switching DC is very different from switching AC.

There are also bigger issues also , you now have a pair of dangling cables that are 'live DC' (during daylight hours :) )

if you read the earlier posts you'll see the whole discussion.
 
Sorry to bring up an old thread, but this discussion is relevant to my situation.

We’ve just had a new PV system installed with 15 x 265W panels and a SB4000TL inverter. The installer has not fitted any additional DC isolators on the basis that the DC breaker is in the SMA ESS switch.

Reading this thread, it looks like opinions differ as to whether separate DC isolators are really necessary, but the general consensus appears to be that they are low cost parts so probably should be fitted.

The question is: is this something I should be concerned about?

Is it possible to retro-fit the DC isolators, and should I ask my installer to sort it out? (I’ve not been invoiced for the job yet).

I think if the truth be told, I would be happier if they were present, but I don’t want to cause too much hassle to my installer because for the most part they have done good a job.

 
for me it's about how you safely isolate the system at the point that the inverter needs replacing.

Though I suppose you could just connect both positive and negative connectors together, this assumes the replacement inverter uses the same connectors.

- - - Updated - - -

also I'm sure the ESS switch only breaks the positive side of the circuit.
 
sedgy34 said:
Could be a problem for who is doin maintanance, as no local isolation.
Click to expand...

Well, my first point of call for any maintenance will probably be the original installer. So if it's a problem for them, it's a problem of their own making!

The installer is bringing all the paperwork over later on this week, so I'll mention it to him then. But I know what he's going to say: SMA say it's ok. I discussed a few issues with him pre-install (such as cable spec, no of panels per string etc) he basically said that they work to the manufacturer's instructions. I actually phoned SMA myself this afternoon just to check with them and they basically said that you don't need a separate DC isolators if you have an ESS switch.

Personally, I would have thought it's sensible to fit them. But then what do I know?
 
if you prefer a dc isolator to be fitted too ask your installer to fit one when he turns up its not much drama as hes returning anyway. ;)
 
Sensible yes, good practice, yes, in a multi string environment almost essential for maintenance and testing. Mandatory ... No as SMA have said it is adequate :(

Version 3 of the guide says ..

A switch disconnector shall be provided on the dc side of the inverter.
* The switch must isolate all live conductors (typically double pole to isolate PV array positive and negative conductors)
* The switch must be rated for dc operation at the system voltage maxima as calculated in section 2.1.2.
* The switch must be rated for dc operation at the system current maxima as calculated in section 2.1.2.
* The switch must be labelled as ‘PV array d.c. isolator’, with the ON and OFF positions clearly marked. Switch enclosures must also be labelled with ‘Danger - contains live parts during daylight’. All labels must be clear, easily visible, constructed and affixed to last and remain legible for as long as the enclosure.
* The switch should be located adjacent to, or integrated into the inverter.
Click to expand...

So to the the letter it meets the final clause.

The guide goes on to say:
An additional d.c. switch may be specified for systems with long d.c. cable runs (typically at the point of cable entry into the building) – so as to provide a means of isolating the cable for safety reasons or maintenance works.
Click to expand...

We ALWAYS fit DC isolators, because on an industrial sized installation the thought of pulling the ESS to break 17+kW of DC is horrendous.. On a recent job we used 4 pole isolators to work on two strings, however we are unlikely to be doing that again because when you make one pair live to test the string you've now got the other pair live also. also some of our strings run close to 1000V Voc and you get a b*****y great spark if you try to seperate those after doing an Isc test without first opening the isolator!
 
does it actually break both positive and negative though?

I'm sure it's only a double break in the positive circuit, so not a dual pole isolator.
 
Worcester said:
Sensible yes, good practice, yes, in a multi string environment almost essential for maintenance and testing. Mandatory ... No as SMA have said it is adequate :(

Version 3 of the guide says ..



So to the the letter it meets the final clause.

The guide goes on to say:


We ALWAYS fit DC isolators, because on an industrial sized installation the thought of pulling the ESS to break 17+kW of DC is horrendous.. On a recent job we used 4 pole isolators to work on two strings, however we are unlikely to be doing that again because when you make one pair live to test the string you've now got the other pair live also. also some of our strings run close to 1000V Voc and you get a b*****y great spark if you try to seperate those after doing an Isc test without first opening the isolator!
Click to expand...


We have 4KWp panels (15 x 265w, 2 strings, one of 6 panels, and the other 9 panels). 15m DC cable run. The inverter is in the garage next to Consumer Unit.

Would you still be concerned an a 4kwp domestic install?
 
JH69 said:
We have 4KWp panels (15 x 265w, 2 strings, one of 6 panels, and the other 9 panels). 15m DC cable run. The inverter is in the garage next to Consumer Unit.

Would you still be concerned an a 4kwp domestic install?
Click to expand...

If your talking about the 15 mtr run of dc cable that's nothing to be concerned about
 
bought this up on my assessment, assessor would be happy on an SMA inverter to have no seperaste isolator. The SMA one breakes both -ve and +ve.
However, I always have and always will fit a seperate isolator for maintainance and testing purposes, and emergency isolation of the inverter (eg when it's on fire!) which is the whole point of isolators!
 
Hi any idea what make is this dc isolation switch?
 

Attachments

  • photo14.jpg
    photo14.jpg
    144.6 KB · Views: 84
nope, suggest you find an AUS forum to ask that question on. Also, is that not an inbuilt DC isolator in the inverter?
 
This is what the draft DTi (or what ever it is called) 3rd edition says:

To allow for maintenance and inspection tasks to be carried out safely, a means of isolation needs to be provided on the dc side of an inverter. The means of isolation shall:

  • Isolate all live conductors (PV array positive and negative)
  • Be readily accessible and immediately adjacent to or incorporated into the inverter

Note: For many smaller systems, inverter isolation and string isolation (see section 2.1.12) may be provided by the same devices (eg plug and socket connectors). As noted in section 2.1.7, plug and socket connectors must be labelled "Do not disconnect d.c. plugs and sockets under load".

A switch disconnector shall be provided on the dc side of the inverter.

  • The switch must isolate all live conductors (typically double pole to isolate PV array positive and negative conductors)
  • The switch must be rated for dc operation at the system voltage maxima as calculated in section 2.1.2.
  • The switch must be rated for dc operation at the system current maxima as calculated in section 2.1.2.
  • The switch must be labelled as "PV Array DC Isolator" with On and OFF positions clearly marked. Switch enclosures must also be labelled with "Danger - Contains live parts during daylight'. All labels must be clear, easily visible, constructed and affixed to last and remain legible for as long as the enclosure.
  • The switch should be located adjacent to, or integrated into the inverter.

A circuit breaker may be used as the d.c. switching device provided it meets all the above requirements.

Note: Switching a.c. is less demanding than switching d.c. (with an a.c. supply, the voltage passes through 0 V many times a second). A switch must be rated to break dc, an ac rated switch is not acceptable or safe.

An additional d.c. switch may be specified for systems with long d.c. cable runs (typically at the point of cable entry into the building) so as to provide a means of isolating the cable for safety reasons or maintenance works.

Systems using micro‐inverters may omit the d.c. switch disconnector where all of the following requirements are met:
  • The micro inverter is located immediately to the rear of the PV modules
  • The micro inverter is plugged directly into the flying leads provided by the module manufacturer (no extensions to the flying leads may be used)
  • The micro inverter and dc cables are generally inaccessible or only accessible to trained or authorised personnel
  • The d.c. conductors between the module and micro inverter are adequately protected against mechanical damage

NOTE: A roof mounted PV installation where the micro-inverters are located on the rear of the PV modules and where access is only possible by first removing some of the PV modules is a typical scenario where switch disconnectors can be omitted. Where the same inverters are brought into a loft void, it is expected that a switch disconnector would be required.


So integral isolators are OK if you meet the above.
Still awaiting this fabled document.
 

Similar threads

R
Isolator in the outdoor electrical box
Replies
15
Views
406
Risteard
Risteard
R
UK Consumer unit relocation
Replies
13
Views
1K
Bluetit
Bluetit

OFFICIAL SPONSORS

Electrical Goods - Electrical Tools - Brand Names Electrician Courses Green Electrical Goods PCB Way Green 2 Go Pushfit Wire Connectors Electric Underfloor Heating Electrician Courses Heating 2 Go
These Official Forum Sponsors May Provide Discounts to Regular Forum Members - If you would like to sponsor us then CLICK HERE and post a thread with who you are, and we'll send you some stats etc

Advert

YOUR Unread Posts

Daily, weekly or monthly email

Receive a daily, weekly, or monthly email containing a digest of new and recently updated threads.

Thread Information

Title
DC Isolator or no DC Isolator that is the question
Prefix
N/A
Forum
Solar PV Forum | Solar Panels Forum
Start date
Last reply date
Replies
40

Thread Tags

Tags Tags
electricians' talk isolator question

Advert

Thread statistics

Created
Thegreenelec,
Last reply from
Gavin A,
Replies
40
Views
18,148

Advert

TrueNAS JBOD Storage Server

Tilers | UK Tiling Forum | Tilers Forums | Find a Tiler
UK Tiling Forum by Tilers Forums UK
Plumbers Forums | Plumbing Advice | Gas Forum
Outside Play
Back
Top