Solar Inverter impedance matching

[MikeSharp01]

Just been thinking through our PV installation - too tired to visit site so thought I would exercise the grey cells instead.

 

The context: 50% of our PV array is on the garden room and connected to its distribution board, as I have mentioned elsewhere, this is a 50m run of 16mm cable from the incomer and that is itself 15m of 25mm cable from the house distribution board. The issue: I think I get the basic physics but I cannot seem to find out much about what range of matching the average inverter (String or otherwise) can manage as the impedance in these two cables and the loads on them must have an effect on the challenges for the inverter in supplying power (current etc) across this network. Anybody got any insights - including, don't be daft it just works!

[ProDave]

Impedance only becomes an issue in regards to voltage drop over it's length, which in the case of an inverter actually means voltage RISE

 

The impedance at the supply point will be so low that in practical purposes very little volt drop or rise will occur, but add a significant length of cable and you will vet volt drop (rise)

 

The issue for an inverter is when the voltage that it sees approaches it's maximum limit. It will either throttle back the generation or shut down completely.

 

This is exacerbated if your supply voltage is high to start with. Don't for one moment be led to believe we work on a 230V supply in the UK. Most places it is still 240V and 245V is pretty normal here. The limit for the inverter is 253V so here we only have 8V of "volt drop" to play with.

 

This is the volt drop on the ac side I am talking about. Long runs of dc cable from the panels to the inverter won't result in tripping or power limiting, just a bit of wasted power in the cables.

[MikeSharp01]

Thanks Dave about what I thought I need to calculate the voltage drop along that cable and measure the base voltage at the incomer then I will get an impression for the max voltage I will need. Looking at it Enphase recommend +/- 1% but it seems they allow +12/-10% voltage rise limits in their micro inverters. I guess max current flowing will be 4Kw if nothing is running in the Garden room. Looking at it I guess the loss will be about 1% at 20A / 240V (more than will be available from the array).

Edited June 9, 2017 by MikeSharp01
Added voltage basis

[Jeremy Harris]

It doesn't hurt to just oversize the cable if you can.  Our inverter could have connected with a bit of 4mm², as the worst case maximum current is only around 27A, but I had a spare length of 6mm² so used that.  As we have a pretty high local grid voltage, never less than 245V, even in mid winter, every tiny saving in voltage drop helps.

[MikeSharp01]

Thanks @JSHarris & @ProDave I have already purchased the cable and going to 25mm was going to cost me double the price so 16mm it will have to be it, and I think I can get into the 1% voltage drop @4Kw zone. The house supply, much shorter has 25mm and its own 4Kw array (assuming we have the budget) so the whole system will be an interesting 'installation' from a which way the electrons will flow point of view!

 

[SteamyTea]

I seem to remember that the permissible voltage drop was 1% for PV, not the normal 5%.  More to do with metering that anything else.

 

Inverters sense the impedance of the incoming supply and if it goes outside some set limit (no idea what), it shuts down for safety reasons.  That is nothing to do with generation or metering, purely a safety issue.  I have no idea how it is sensed of where it happens, other than inside the inverter.  All black magic to me.

[Jeremy Harris]

Inverters do a loop impedance check, as a part of the anti-islanding shutdown that's required to comply with the regs.  The check is usually done by one of two methods, the most common being to periodically force the output current waveform to increase and use that to measure the effective impedance of the loop, the other way is to periodically switch a capacitor across the supply loop and measure changes in source impedance that way (not that common, I believe).

 

Ours displays the loop impedance as a part of the cyclic display of data on the front.  My guess is that the loop impedance has to be quite high for that alone to cause the inverter to shutdown, and it's not the primary anti-islanding protection method, voltage, phase and harmonic detection are usually the front line defence mechanisms.