Two questions about solar panels.

[JohnW2]

I’m probably missing lots of things, but as I currently see it, there is only one thing which is guaranteed to happen at the moment of loss of the grid supply – namely that the (essentially ‘infinite’) ‘sink capacity' of what the inverters ‘see’ of the grid will drop dramatically. In nearly all situations, the inverter will therefore see a sudden change in the ‘load’ – either a sudden drop (if there were previously a net export of electricity from the affected segment of the network) or a sudden rise (if there were previously a net import of electricity). That sudden change in load (i.e. inverter output current) could be used to sense the loss of the grid supply, hence the need to shutdown.

However, as I’ve said, the one exception arises if, by chance, the total load in the affected network segment is almost exactly the same as the total inverter capacity in the segment at the moment of grid supply loss – in which case I would not expect the inverters to notice any change in loading. Whether it would be considered acceptable to ignore that very unlikely chance happening, I don’t know – but with DNO staff lives at risk, I wouldn’t have thought so – so I doubt that this, alone, could be the means of sensing the need for inverter shutdown.

Kind Regards, John

 

[DetlefSchmitz]

John, you should probably look at a completed G83 certificate, there are loads of them on the web. (Unfortunately G83 itself is too expensive to buy on a whim.)

This will indicate the tests required for compliance. As expected the two key tests are shutdown on over/under voltage and shutdown on over/under frequency.

I believe G83 also discusses the density of permissible installations, but as I said, I can't afford it.

 

[westie101]

G83

If I get a chance I'll have a look during the week as I have access to it

 

[JohnW2]

John, you should probably look at a completed G83 certificate, there are loads of them on the web. (Unfortunately G83 itself is too expensive to buy on a whim.) ... This will indicate the tests required for compliance. As expected the two key tests are shutdown on over/under voltage and shutdown on over/under frequency.

I'll have a look, but will that actually help me? I have no doubt that those (along with shutdown on supply failure,which is what we are discussing) are amongst the most important required features of the inverters, and I have no doubt that all inverters are required to (and presumably do) pass such tests. However, I wouldn't have thought a certificate would tell me about the conditions of the test or,more important (since that's what I'm asking about) is how the inverters actually achieve guaranteed shutdown in the case of grid supply failure, would it?

I believe G83 also discusses the density of permissible installations, but as I said, I can't afford it.

That makes sense. One would imagine that, almost regardless of the method used to sense grid supply failure, the task would get more difficult as the number (and power) of inverters attached to a 'segment' of the networked increased.

Kind Regards, John

 

[JohnW2]

G83 If I get a chance I'll have a look during the week as I have access to it

Many thanks. Needless to say, I'd be particularly intererested to hear about anything it has to say about how grid supply failure should/must/can be sensed and also possibly about the nature/conditions of the test to confirm shutdown in the event of grid supply failure. Thanks.

Kind Regards, John

 

[westie101]

I believe G83 also discusses the density of permissible installations,

Unlikely as the below 4kW installations can be connected and the DNOs informed afterwards (within 28 days)
The only control we have is if the installations are found to be non-compliant (unlikely) or cause unacceptable voltage rises at other customer's terminals

 

[Lectrician]

The 11kv turbine we are involve with as a separate G83 control panel which operates the remote electrical switching of the 11kv RMU (ring main unit). Who knows exactly how they work.....but they do!

 

[DetlefSchmitz]

I believe G83 also discusses the density of permissible installations,

Unlikely as the below 4kW installations can be connected and the DNOs informed afterwards (within 28 days)
The only control we have is if the installations are found to be non-compliant (unlikely) or cause unacceptable voltage rises at other customer's terminals

This was my source information (chatter on the internet about a year ago). As it relates to a draft I don't know what made it into the final recommendation.

A new draft version of the G83/2 regulations has been released for consultation including major changes to the rules about when advance permission is needed for connecting a solar PV system to the grid. These changes have massive implications for the industry as a whole, but particularly for local installation companies with a large local customer base.

The key changes are:-

Definition of 'Close Geographic Region'

Any installations of more than 1 solar PV system by the same installer that fall within either of these categories would require advance permission to be connected, using a revised Appendix 2 multiple installations application form.

1) The postcodes of any of the premises where a SSEG installation is planned by the same organisation are the same when the last two letters are ignored…ie AB1 2xx where xx could be any pair of letters or where x could be any letter.

2) The premises where a SSEG installation is planned by the same organisation are within 500m of each other.

 

[westie101]

This was my source information (chatter on the internet about a year ago). As it relates to a draft I don't know what made it into the final recommendation.

Yes it was, the situation as far as I can see is that there is no limitation on individual customers connecting systems within a specific area.
If, however, the installations are by the same organisation as part of a large scheme (roof rental from a housing authority perhaps?) that needs to be notified in advance for network checks

 

[JohnW2]

This was my source information (chatter on the internet about a year ago). As it relates to a draft I don't know what made it into the final recommendation.

Maybe someone had the sense to see that it shouldn't get into the final version?! Unless I'm missing something, that draft was suggesting that 50 different companies could install small PV installations in 50 premises in the same postcode area without prior authorisation, but if one company wanted to install just two PV installations in the same postcode area, then they would nened to get prior authorisation.

Is it just me, or is there something a bit daft about that proposal?

Kind Regards, John

 

[westie101]

Is it just me, or is there something a bit daft about that proposal?

Read the first three words of my post above yours, it is part of G83

Some useful information here: -

http://www.ukpowernetworks.co.uk/internet/en/connections/electricity-generation/

 

[JohnW2]

Is it just me, or is there something a bit daft about that proposal?

Read the first three words of my post above yours, it is part of G83

Yes, I now know that - but, as you will probably have guessed from the timestamps of our posts, you posted that whilst I was typing mine, so I hadn't then seen it!

However, even if that wording did get into G83, I think my comment/ question which you've quoted above still stands!

Kind Regards, John

 

[plugwash]

However, as I’ve said, the one exception arises if, by chance, the total load in the affected network segment is almost exactly the same as the total inverter capacity in the segment at the moment of grid supply loss – in which case I would not expect the inverters to notice any change in loading.

I could be wrong but I strongly suspect that assuming the inverters are set up as "constant power" sources that follow the voltage frequency of the grid (rather than trying to set them) the chances of both voltage and frequency remaining stable without the grid to force them to be stable would be negligible.

Whether it would be considered acceptable to ignore that very unlikely chance happening, I don’t know – but with DNO staff lives at risk

I believe that nowadays the DNO have procedures to mitigate the dangers from back feeding anyway.

 

[westie101]

Had a read of G83 basically it gives the parameters required to disconnect for over/under voltage & frequency and for loss of mains, how that is achieved is up to the designer/manufacturer of the inverter.
One this design is complete it is then type tested and approved.

The Loss of mains mainly seems to be suggesting vector shift & rate of change of frequency as the methods used

 

[JohnW2]

I could be wrong but I strongly suspect that assuming the inverters are set up as "constant power" sources that follow the voltage frequency of the grid (rather than trying to set them) the chances of both voltage and frequency remaining stable without the grid to force them to be stable would be negligible.

Maybe. However, as I've been saying, if there are a number of connected inverters, all frequency/phased synched to the grid (hence to each other) up to the moment of its failure, I would have thought that they would continue to synch to one another for a potentially appreciable period of time before they (jointly) started drifting in frequency. In practice, of course, unless the load on the affected segment is 'right', loss of grid power will probably result in a voltage change which will result in disconnection - but that's not totally guaranteed.

Kind Regards, John