Solar PV fault

[Daf Thomas]

About a year ago we installed a 3.99kW PV system to a farm shed roof. We used a sunnyboy SB4000TL-20 inverter and used 2 strings, 1 of 11 and 1 of 10. Up until a month ago it had worked fine since we installed it, but now during midday when the sun is strong the inverter stops working. Ive been back to see whats happening and it appears that each string makes the AC voltage rise by 4 volts. The supply voltage at the property is allready quite high at 254 volts so when both strings are on during sunny periods the voltage coming back through the AC side of the inverter goes up to 262 volts which knocks it off. The supply from the house to the shed is 10mm run over 70M.
Any suggestions on what is making the voltage go so high, or how to fix it? Or is it something like the inverter has gone faulty?

Thanks

 

[JulianC]

Off the top of my head, without doing any sums, I can almost guarantee that 10mm cable is far too small over 70m. With the inverter running at or close to full power, the volt drop is enough to send the output high enough to trip

 

[Daf Thomas]

Why is it that normally the further away from the AC supply you get voltage drop but with a PV system the voltage rises? Are there calculations for this? For a normal AC installation the voltage drop for the 10mm cable over 70M with a 16A load reference method C is 4.928 volts. Does this have anything to do with the voltage increase?

 

[Earthstore]

Julian is correct, and this is very common.

I will expand if I may.

Electricians tend to work on a 3% volt drop for cables, however if you incoming supply is on the high side and your inverter has to match this on the other end of a long cable, it has too raise its voltage to match the incoming supply, sending it over voltage.
It is just basic maths, oh and an expensive cable to rectify the problem.

 

[SolarCity]

Also, it is a good idea to work to 1% VD when installing PV systems.

 

[SolarCity]

Having said all this, if the grid voltage is 254v then it goes beyond the voltage that the DNO is obliged to produce at (10%+ 5%-). If you can monitor the voltage with the PV system off then you will have a better idea if this is causing it. Have you measured this voltage at the incomer?

 

[Earthstore]

[h=2]Does this help?
Voltage Drop and Cable Size Calculator for ELV Solar Power Systems[/h][TABLE="class: data, width: 992"]
[TR]
[TD="bgcolor: #DCDCDC"]Allowable Voltage Drop[/TD]
[TD="bgcolor: #DCDCDC"]1% %[/TD]
[/TR]
[TR]
[TD="bgcolor: #DCDCDC"]Run Length[/TD]
[TD="bgcolor: #DCDCDC"]70 m[/TD]
[/TR]
[TR]
[TD="bgcolor: #DCDCDC"]Current Demand[/TD]
[TD="bgcolor: #DCDCDC"]16 A[/TD]
[/TR]
[TR]
[TD="bgcolor: #DCDCDC"]System Voltage[/TD]
[TD="bgcolor: #DCDCDC"]254 V[/TD]
[/TR]
[TR]
[TD="bgcolor: #DCDCDC"]Voltage Drop[/TD]
[TD="bgcolor: #DCDCDC"]2.27 mV/A.m[/TD]
[/TR]
[TR]
[TD="bgcolor: #DCDCDC"]Recommended Cable Size[/TD]
[TD="bgcolor: #DCDCDC"]25 mm[SUP]2







[/SUP][/TD]
[/TR]
[/TABLE]

 

[JulianC]

Volt drop calculations are the same as you'd normally use. The voltage from the inverter rises to overcome the volt drop across the cable. Look at it this way - if your inverter is kicking out 16A down 70m of 10mm cable then, as you've calculated, the volt drop across the cable is 5v (did you software limit the SB4000TL? If not it will kick out over 20A flat out, further adding to your problem). If your mains is 254v then the inverter, when generating, is seeing 261v at its terminals. It therefore has to kick out a higher voltage than this in order to feed in to the grid. Hello trip.

This is a prime example of why the AC volt drop design limit is set at 1%

 

[JulianC]

You will have probably also already worked out by now that I don't type fast enough. :goofy:

 

[SibertSolar]

Does this help?
Voltage Drop and Cable Size Calculator for ELV Solar Power Systems

[TABLE="class: data, width: 992"]
[TR]
[TD="bgcolor: #DCDCDC"]Allowable Voltage Drop
[/TD]
[TD="bgcolor: #DCDCDC"]1% %
[/TD]
[/TR]
[TR]
[TD="bgcolor: #DCDCDC"]Run Length
[/TD]
[TD="bgcolor: #DCDCDC"]70 m
[/TD]
[/TR]
[TR]
[TD="bgcolor: #DCDCDC"]Current Demand
[/TD]
[TD="bgcolor: #DCDCDC"]16 A
[/TD]
[/TR]
[TR]
[TD="bgcolor: #DCDCDC"]System Voltage
[/TD]
[TD="bgcolor: #DCDCDC"]254 V
[/TD]
[/TR]
[TR]
[TD="bgcolor: #DCDCDC"]Voltage Drop
[/TD]
[TD="bgcolor: #DCDCDC"]2.27 mV/A.m
[/TD]
[/TR]
[TR]
[TD="bgcolor: #DCDCDC"]Recommended Cable Size
[/TD]
[TD="bgcolor: #DCDCDC"]25 mm[SUP]2







[/SUP]
[/TD]
[/TR]
[/TABLE]

That's a handy volt drop calculator Earthstore, is it from the Kilowatts website? Does anyone have a quick/handy one for DC volt drop?

 

[moggy1968]

the sunny design software, available online, does it all for you, very handy!

 

[Eco Deal]

Had this before did you change the country setting to G83 one year ago????/ thats the first port of call before you start swapping cables!!!!

 

[Eco Deal]

or change the parameters on inverter.

 

[tiddywiddy]

Your starting voltage is too high, as said before, contact the DNO and ask them to monitor the voltage and then turn down the transformer or swap the property to a different phase. Also something else to consider is that sunnyboy state in thier manual that the resistance of the cable must not be more than 1ohm over total length so if it's TT then you have got a problem. I have contacted them with exactly the same problem, first thing they asked was the size of the cable and length of run. Also said before, your cable size is too small and 25mm is absolute minimum working to 1% volt drop as stated in the DTI guide to installation of solar photovoltaics. 25mm cable (if its armoured buried in the ground, correction factor 0.9) is exactly a 1% voltdrop with the design current being 22A (from sunnyboy manual) and this is not including the voltdrop you will get from the .5M of tails and flex which takes it over. My two penny worth would be to contact DNO and get voltage turned down and increase cable size as an absolute minimum and look at the resistance of the cable. Also, you will need a grid guard from sunnyboy if you want to change the settings if its not set to G83.

 

[tiddywiddy]

forgot to mention, also with the DC side we can work to a 3% voltdrop so not as bad!!!