Testing the Inverter

[DomB]

Hi,
An old customer has asked me to run a health check on their Solar PV system which we installed some 5 years ago. Not something I've done before, beyond the standard installation testing, so I was going to:

• Test each panel for Vout against irradiance
• Check all electrical connections
• Test the inverter.. somehow!

They want to move to micro inverters but I'm saying that if the existing string inverter is still working they should run with that until it pops. Any thoughts on how I can test their inverter, show its still being efficient etc? I was hoping to maybe test DC in v's AC out and run a quick % calc. What do you think?

Thanks in advance for any advice!

 

[Gaz92552]

I used to do a bit of maintenance/service calls on solar systems and all we did, was same tests as when you install it. Waste of time testing individualy imho. Have a quick look through the history on the inverter to check it hasn’t been dropping out. Sounds like these people have money to burn if they want to start changing things after 5years?!

 

[DomB]

I'll see if i can get any logs off the inverter but was looking to test efficiency, if I can.

Any thoughts on testing efficiency? If I get DC total in and AC out what could I derive from those numbers?

 

[Gaz92552]

Well if it’s a 3kw system you should be getting 3kw under standard test conditions, just calculate according to high bright it is outside

 

[Worcester]

When we do 'health checks' for people buying a house with PV, we do a full standard installation commissioning check, Voc, Isc, Irradiance, Resistance to earth, cable routing, condition, component rating, labelling etc.
And have come a cross a few nightmares, like 6mm AC T&E used for the DC from inverter to loft, AC isolators instead of DC etc, inverters mounted on non fire resistant boards etc.
Then look at their generation meter returns to their FiT supplier - they'll have quarterly logs, (and what the inverter logs say) then knowing the installation postcode, elevation, orientation, and a shading analysis, compare that to the MCS predictions (they should over time be within 10%) compare that to the original predictions given to the client at quotation stage and in the handover pack.
Also go to PVoutput.org and see if anyone locally is logging daily outputs and compare those to the installation.