DC string testing at the end of the day

Gavin A · Nov 30, 2012

We were installing a system on an east facing roof today, so by the time I got to doing the DC string tests the sun was well and truely off the roof and close to setting to the west. The string voltage was so low at 408V as opposed to the nominal 444Voc that it could well have been an entire panel missing, particularly given how cold the panels were, so I had to go and individually inspect each of the panels to be certain we'd not got one missing.

We're back on Wednesday next week anyway so will redo the string test then, but this really is a problem at this time of year, and makes a bit of a mockery of the idea of the string tests really, but yet we can't be going back to every job the next day just to get the readings when the light's a bit better (though in practice we do end up doing this a lot in winter).

Not sure quite what the point of this thread is other than to see what other installers do about this issue at this time of year - are you all going back to get a better reading the next day, or just banging the numbers down, checking the panels are all connected and noting the low light levels down?

SolarCity · Nov 30, 2012

We had it happen a few times last year. We went back to check - one of them was in North Lincolnshire.....

This year we have been installing over two days anyway and it's not been an issue.

dansk · Dec 1, 2012

If its too dark to test, we go back, like SolarCity we been taking two days to install when the jobs come in.

Gavin A · Dec 2, 2012

We've usually worked on the basis of if we can get the inverter to fire up and check the G83 settings then we'd consider that light enough, but if we can't then we'd go back and test it, which we're doing here anyway.

I guess my query really is around what the cut off point is that people use.

t4sltd · Dec 5, 2012

MCS auditor said 300W/m2 which effectively means wait till spring! not sure if customer wants scaff up that long. Tend to go back when sunny if less than a few hundred W/m2

SolarCity · Dec 6, 2012

300W/m2?! Ridiculous. Where did he get that from?

Worcester · Dec 6, 2012

We'd previously heard 200W/m2 though that is also ridiculous.

Had an intersting discussion over Light Meter and DC meter calibration, - he came to the conclusion (after a sensible discussion) that seeing as readings can vary / fluctuate so rapidly as the light / sun changes that calibration wasn't nescessary as the results were 'relative to each other'. - Bear in mind though that our jobs typically never have less than 6 stings on them. So it's easy to compare one string with another.
Different when you've only got a single string on one inverter..

Gavin A · Dec 6, 2012

it's a farce.

No useful data can be had without temperature data anyway, and after the thread on here related to this where I ended up taking a long call from the head guy at NIC who's also a lead author of the new DTI guide, there's no intention to collect panel temp data, so it's really down to the experience of the installer to spot a voltage that appears suspect and double check that all the panels are definitely in series.