AC / DC isolators

[GaryM]

I've seen mentioned on this forum AC isolators modified for use as DC isolators.
I reckon that's exactly what I have, even the lid complete with rotary isolator is interchangeable!
The only difference I can see, apart from the different coloured rotary switchers is the AC is a 3 pole unit and DC a 4 pole using 2 poles in series for both positive and negative.

Is this common practice, even down to the lids being interchangeable?

AC Isolator

DC Isolator

 

[JulianC]

Standard practice for Kraus & Naimer and, despite protestations from those with vested interests in other manufacturers' equipment, I am yet to see evidence of it being a problem, providing the isolators are suitabley rated for the jobs in hand.

 

[GaryM]

Well the 'DC' isolator is rated at 25A, no problem there and 600 VAC.
I have a single string of 16 modules, each with a Voc of 37.82 VDC, totalling 605.12 VDC at STC.

I guess the only way to know would be to ask the manufacturer of the isolators voltage rating with DC.

 

[JulianC]

http://www.krausnaimer.co.uk/images/stories/pv_g_2012_27_2_lowres.pdf

Page 13 of the above guide gives the DC-21B rating for DC operation. Looks like you have a KG20 T104 so good for 18A @ 800v

 

[Glen Powell]

It is still a modified AC isolator and although they say it is up for the job the internal contacts will have to be switched very fast to stop failure, the question is would you or have you installed one in your house with your wife and kids, this may sound dramatic and over the top but I will never promote or condone the use of a modified AC switch.

It is more than the little daisy chain wire, the contact mechanism also needs changing in its entirity, It goes to show that it is more costly to produce a true DC switch than to modify an existing switch, surely if all the components were right for the job this would not make economic sense. Take a switch and manually add more parts to it ?? two production processes ? Keeps people in work I suppose.

 

[JulianC]

Not being an expert in isolator design, I tend to rely on what a manufacturer tells me. If they have tested these isolators to a suitable standard then I have no qualms in fitting them. Like I say, show me evidence that it's a problem and I will reconsider.

 

[Glen Powell]

I will take a picture of a true DC isolator on wednesday when I am back in work you will see the difference straight away, unless someone else here beats me to it, A true DC isolator is wire in diagnal

 

[Glen Powell]

This is an answer to post #6

Install one on a test bed and switch it about 100 times and see if A) it is still working b) there is lots of residue / corrosion on the contacts, 100 times probably overkill

 

[JulianC]

I know, I've fitted IMO too.

 

[Gavin A]

This is an answer to post #6

Install one on a test bed and switch it about 100 times and see if A) it is still working b) there is lots of residue / corrosion on the contacts, 100 times probably overkill

at what ampage and voltage?

we'd only install the 16amp version for a single string, so max ampage would be around 8-8.5amps, usually at well below 450V. I'd strongly suspect that if you ran that on a test bed it'd last several hundred uses, particularly if you then made allowances for it mostly not being used at full capacity, which in my book makes it fit for purpose given that it's average usage will be around 15-20 times in it's lifetime, and maximum probably around 50.

I'd agree that it's probably best to give them a fair degree of leeway and not push them to their limits, but beyond that I personally think there's some scaremongering going on about this issue, though I do think it's worth everyone just reviewing their specification processes and installer training in light of the recent fire to be sure the DC isolators are being specced and installed to allow for the potential voltages and ampages they could be exposed to, with good margins of error.

 

[Glen Powell]

Gavin

we seem to be butting heads on every topic, if you are happy with it then so be it, but since a true DC isolator is a small cost to the overall isn't it best to be safe than sorry, I have not mentioned the fire deliberately (ok I have now), I just do not like the word modified nor would I invest in something that is modified, not even a car, obviously a car has massive insurance implications.

I am not trying to force people down a route but I do really want to know how many of the electricians in this forum have a modified AC in their home ?

 

[SolarCity]

I have a modified AC in my home.

Like others have said, I don't really see a problem with it. However, I'm no expert on isolators and I go with what a manufacturer tells me.

I very much doubt that the cause of the fire will have been the use of a modified AC isolator - especially considering that the isolator is very unlikely to have been operated since it's installation.

 

[LightGEN]

GaryM, looking at the picture, which you've provided of the wired DC isolator, I'm afraid it may to be wired incorrectly.

If you refer to the K&N wiring instructions, for DC use, you should use two pairs of links to bridge the contacts - the K&N instructions aren't as clear as they might be in this area and K&N only provide one pair with the isolator. As follows:

Link one between L1 and L2.
Link two between L3 and N.
Link three between T1 and T2.
Link four between T3 and N.
+ supply from PV panels goes to L1.
- supply from panels goes to T3.
+ output to inverter goes to T1.
- output to inverter goes to T3.

This arrangement doubles the contact surface area in use and, in doing so, halves the current passing through the contacts, therefore, reducing the effects of arcing.

Hope that helps.
Roger

....Hmm, actually, this has now caused me to relook at my own interperitation of the drawing....and in doing so, it's not very clear which approach K&N is recommending. In my own mind, paralleling the contacts to double the contact surface area, seems to make more sense than putting two in series, however, the drawing is open to interpretation!

 

[SibertSolar]

We always add 15% to Voc values, and 25% to Isc values of a string when calculating the required ratings of a DC disconector for our clients, if such support is needed. This subject has been batted back and forth many times, the fact is that simply achieving the DC21B ratings is not enough to provide a product that reliably operates time and again with zero deterioration of the contact surfaces. Adequate current/voltage ratings is not only what makes a switch a DC disconnector. Think about why an AC isolator is typically wired straight-through (terminal arrangement) and DC switches aren't. Think about why arc-cooling chambers are built-in to IMO/Santon switches but don't existing on most ABB and K&N modified AC units. Think about why contact surfaces are self-cleaning, knife-edge arrangements in a DC disconnector, rather than separating in a 180degree plane as per most AC switches.

Personally, I believe the reason for the historical and habitual use of inadequately designed "DC switches" (which more often than not are modified AC isolators) is that there has been a significant lack of understanding in the marketplace when it comes to switching DC and some manufacturers have simply pulled the wool over some people's eyes. Some brands have had a significant share of the market because of this, NOT necesarily because of the commercial difference between an AC isolator and a DC disconnector. When looking at the overall cost of a system, £10 or £20 more spent for a true DC disconnector should not compromise the competitiveness of the end-customer's pricing. In fact, using "built for purpose" equipment should be a selling point from the designers/installers point of view! I always advise people not to compare "apples with pears" when purely looking at the commercial differences.

There was a chap on here recently who worked for K&N Tech dept who, when asked about the small print "DC21B when switched with rapid handle operation" on their spec sheets said that it was perhaps removed from their documentation now and that it was not really relevent anyway. In my opinion, it is extremely relevent and was somewhat glossed over for a very specific reason.

There's plenty of info on this subject available, both on here in previous threads and on various websites including our own. Yes, I am biased as we sell IMO and Santon switches. However, the reason we only promote those brands is due to the considerable time I spent a couple of years back investigating the nature of the design of "other" units and their short-falls in performance and reliability. So, biased yes, but only from a technical point of view!

 

[Gavin A]

GaryM, looking at the picture, which you've provided of the wired DC isolator, I'm afraid it may to be wired incorrectly.

it's not.

well, we generally wire the panels into T1 & T3 rather than T2 & T4, but am pretty sure that's just our convention rather than serving any practical purpose.

 

[GaryM]

it's not.

well, we generally wire the panels into T1 & T3 rather than T2 & T4, but am pretty sure that's just our convention rather than serving any practical purpose.

Panels are wired into T1 & T3, the outgoing side to the inverter is T2 & N (T4).

Looking at the PDF created by the manufacturer that Julian linked to on Page 1 suggests the termination order isn't to manufacturers instructions.

They suggest:

+ in from panels to T1
+out to inverter to T2
-out to inverter to T3
-in from panels to N (T4)

 

[Gavin A]

Thought I'd check this with K&N...

[FONT=&quot]I can confirm that it is okay to connected the + and – leads from the pv string as stated in your email below.[/FONT]

so as I thought, it's fine either way round.

 

[Ramjam]

I use those isolators and I think they're great, as has been pointed out by breaking the dc circuit in two places rather than just one the arcing is halfed. DC isolators on PV systems are for maintenance purposes not general use so will see little action in reality.

But I've seen Bosch dishwashers and IBM monitors burst into flames, electiricty is dangerous and there is little point in stressing over what is beyond our control, if a reputable manufacturer sells an isolator which is rated for DC use then why lose sleep over whether or not it is a modified AC design?

My only criticism is that I run each DC cable through it's own stuffing gland to maintain the IP rating, 2 cables through 1 stuffing gland does not seal!

 

[Gavin A]

surely that 2 into 1 problem is only an issue for isolators situated externally, or possibly those with the cables running into the top of the isolator if you're worried about condensation forming on the cable and running down it.

no other household electric switches that are inside are fully water proofed as a rule other than special locations.

 

[Ramjam]

I agree, I just like to maintain the IP rating, it satisfies my OCD!

 

[SibertSolar]

Sorry, how does series linking multiple poles together reduce arcing? All it does is increase the voltage ratings, does nothing for the current rating. In fact, increasing the number of contacts in series actually increases the number of contact surfaces potentially affected by arcing. In the case of a properly designed DC disconnector this isn't an issue (when series-linking only to achieve 1000vdc ratings, by the way) as they tend to employ sliding, knife-edge contacts that self clean and result in the final contact point not being the normal (closed) contact faces. An AC isolator tends to separate in a 180degree plane so arcing has a direct impact on the actual contact surfaces....

 

[mr murdock]

Have to agree with Sibertsolar - K&N do not have a true snap action 1000V DC switch and are only fooling people into taking a AC switch

 

[LightGEN]

Sorry, how does series linking multiple poles together reduce arcing? All it does is increase the voltage ratings, does nothing for the current rating. In fact, increasing the number of contacts in series actually increases the number of contact surfaces potentially affected by arcing. In the case of a properly designed DC disconnector this isn't an issue (when series-linking only to achieve 1000vdc ratings, by the way) as they tend to employ sliding, knife-edge contacts that self clean and result in the final contact point not being the normal (closed) contact faces. An AC isolator tends to separate in a 180degree plane so arcing has a direct impact on the actual contact surfaces....

This is exactly why I thought K&N were suggesting that two poles should be operated in parallel, thus halving the current - seems I missread their instructions....however, I still feel my approach is better than two contacts in serial.

 

[Gavin A]

Sorry, how does series linking multiple poles together reduce arcing? All it does is increase the voltage ratings, does nothing for the current rating.

K&E disagree.

They give the voltage and current ratings for their isolators when doubled up, or in single breaker mode, as the single and 2 string installation options (page 13 and 14 of this guide).

KG32-T104 is rated at
dual string (single breaker per side) = 15A at 460V
single string (dual breakers per side) = 28A at 450V

the permitted voltage also increases with the dual breakers option, so essentially it's limiting factor would seem to be related to the watts per breaker.