Cu & Pv Configuration

[Sidewinder]

There are 4 issues here.

1. RCD may not trip as inverter output is not true pure sine wave.

2. Inverter will continue back feeding other circuits on the RCD after it has tripped for up to 5 seconds.

3. The inverter is a generator. It is earthed, it has to be. Now, is there an N-E link at the inverter as would be expected at a normal generator. If so then, there is a danger. The earth is continuous from the inverter frame to true earth, it has to be. Whilst the N of the DNO supply is disconnected, the inverter N is still connected to the fault side of the RCD. Remember, this is a generator, not a consumer, so an internal N-E link is NOT prohibited. Is this why most inverters are not allowed on TT?...  Remember, this link could switch on loss of synchronisation feed, so would not be seen by an RCD protecting the DNO supply as it would not be in place until the DNO feed was off.

4. IT systems are not allowed in consumer installs in the UK.

The issue is that there is not enough information on the internal configuration of the inverters for us to make a definitive judgement, we can do all the tests and make all the measurements we like, but, this will simply be empirical data.  The definite specification of the inverter will give the actual danger, and no one is telling.

So, we have to err on the side of caution IMHO.

 

[binky]

Binky,,, I never said that you can't get a L-N shock,, but you'll never get a L-E shock after an RCD has tripped

Noz, in domestic CU from a shared RCD it links to other circuit MCB via nice copper bus bar, so you could be mowing grass, cut cable L-E short, RCD trips, but Inverter is still pushing leccy down L through MCB for its circuit, through busbar to MCB for socket circuit you plugged lawnmower into rendering RCD useless for around 5 seconds - enough time to kill!

 

[NozSpark]

Sidey,,

Wouldn't any N-E link within an inverter effectively make the circuit have a CNE conductor which isn't allowed?

I was assuming that an inverter wouldn't have a N-E link BTW....

..,..I was thinking what Slips said,,, but he said it a lot better than I

 

[steptoe]

Noz, in domestic CU from a shared RCD it links to other circuit MCB via nice copper bus bar, so you could be mowing grass, cut cable L-E short, RCD trips, but Inverter is still pushing leccy down L through MCB for its circuit, through busbar to MCB for socket circuit you plugged lawnmower into rendering RCD useless for around 5 seconds - enough time to kill!

this is pot luck ime

Ive threw 500mA down a 30mA RCD and still not got it to trip if the PV is connected.

 

[SPECIAL LOCATION]

Sidey,,

Wouldn't any N-E link within an inverter effectively make the circuit have a CNE conductor which isn't allowed?

I was assuming that an inverter wouldn't have a N-E link BTW....

..,..I was thinking what Slips said,,, but he said it a lot better than I

Whether its a Solar PV inverter or any other bog standard accessory or load...

I thought N & E were NOT permitted to be rejoined back together at any point in an installation once they have been seperated..

Reg 543.4.3 pg 164...

So how could Solar be allowed to re-join them???

:C

 

[steptoe]

Whether its a Solar PV inverter or any other bog standard accessory or load...

I thought N & E were NOT permitted to be rejoined back together at any point in an installation once they have been seperated..

Reg 543.4.3 pg 164...

So how could Solar be allowed to re-join them???

:C

its a weird thing, although a solar inverter is classed as a load for installation purposes,

we know it is not a load, it is in fact a generator, so what happens inside it is a mystery AFAIAC

:C

 

[binky]

Not sure sidey is right about N-E link in inverter, but they dissipate energy when shut down somehow

 

[Sidewinder]

Not sure sidey is right about N-E link in inverter, but they dissipate energy when shut down somehow

I haven't said there is, an N-E link, I have merely asked if, there is, now, there could be.

As I also said, it could switch in upon loss of synchronisation feed, to act as a bus discharge

As you say Binky, they discharge energy somehow.

That either has to be to earth, or back into the panels.

Which is it?

Anyone got some internal schematics & design details?

 

[steptoe]

If they do discharge to earth is this perhaps why only certain RCDs are permitted,? 

 

[binky]

Thinking about it, to shut inverter down, we isolate AC and DC, then it dissipates energy internally - instructions have comments about leaving well alone for a couple of minutes before opening covers. In the case of AC power fail, relay to isolate AC side can take 5 seconds, but DC keeps the whole inverter awake so that it continues to monitor AC side with a view to restarting ASAP - there aren't N-E links in inverter, earthing is as per any other electrical appliance.

 

[Sidewinder]

Binky are you, sure, about the N-E link?

I would expect this to be buried in the internals, rather than obvious.

 

[binky]

If there is, it's along the lines of most appliances ie 'smoothing circuitry' and no doubt leaks a few milliAmps. As per most appliances earthing is there for 'extraneous conductive parts' and functions in that way. The internal circuitry I've never really studied, it's mostly enclosed or consists of PCBs.

What makes you think there is a N-E link sidey?

 

[Sidewinder]

Generation reference, and the way the power electronics work.

 

[binky]

  This any use?

http://www.edn.com/design/power-management/4368876/Teardown-The-power-inverter--from-sunlight-to-power-grid

All I see is lots of things to keep my fingers out of :^O

 

[Andy™]

Binky are you, sure, about the N-E link?

I would expect this to be buried in the internals, rather than obvious.

if there was any N-E links in the inverter, then it would trip RCD's, same as it would if you had an N-E fault

 

[steptoe]

if there was any N-E links in the inverter, then it would trip RCD's, same as it would if you had an N-E fault

Not quite, an inverter could bleed to earth any excess it has and still keep the L and N balanced.

Remember, it is generating, as long as the L N is balanced it could shove 100A down the earth if it wanted to. 

 

[Phoenix]

4. IT systems are not allowed in consumer installs in the UK.

Are you are its not TN-C you are thinking of SW?. Prohibited by supply regs unless granted permission by the secretary of state... however an old teacher from college told me that back in the 60s/70s, there was one particular department store in the UK which used TNC systems in its stores, wired in single core MICC...

Back to IT systems, I see quite a few of them in schools, extra point to anyone who can suggest which part of the school they might be in....not sure whether to let M107 answer or not... . Aside from that I understand they crop up in certain areas of hospitals as well, but that's just what I've been told

 

[Sidewinder]

Phoenix,

Not quite, but I did not explain myself correctly.

You are right, no TN-C anywhere without special permission.

IT are not allowed in domestic, only in areas under "control", yes OT in healthcare, and certain school labs.

When I wrote that, consumer meant to me domestic consumer, but, obviously not what I writ ( ) sorry!

if there was any N-E links in the inverter, then it would trip RCD's, same as it would if you had an N-E fault

No it would not Andy, not if you read my full comments, the N-E link could only be in place when there is no mains to sync to.

IF, it was in place at all times, then it could trip an RCD, remember, you can do clever things with electronics.

You can do an earth fault loop test without tripping an RCD remember.

  This any use?

http://www.edn.com/design/power-management/4368876/Teardown-The-power-inverter--from-sunlight-to-power-grid

All I see is lots of things to keep my fingers out of :^O

Don't know Binky, not sure as there is a lot of supposition.

How do the components behave under loss of synchronisation, I'm not convinced that in the event of loss of mains the generator immediately becomes islanded, yet.

Until I am, then I will err on the side of caution.

The output can't float until it "sees" a sync supply, else how will it have a 0V N reference immediately that the ac kicks in, I believe that there must be some sort of reference inside the inverter for the electronics when the device is islanded.

 

[binky]

wat the inverter is does on loss of mains is use DC to keep itself working, monitors the AC side, so that when mains is re-connected it automatically restarts after 180seconds

 

[Sidewinder]

binky,

You're thinking on a systems level, I'm thinking on a component level of the actual implementation of what they achieve and how they achieve it.

That is the bit that is unclear, do any of the gates drop down to ground?