Degradation of domestic solar panels

JKenH

I have today edited the discussion title as there may be a more widespread problem with degradation than I am aware of and would like to see the experience of others good or bad, preferably based on statistics adjusted for annual variation in insolaation.

2nd_time_buyer

I wonder how they measure the degradation of the panels. I assume it is averaged across a number of panels. 

If so and it is a mean average and there is scatter, then the system will default to the worst performing panel.

If it is a median average then an outlier would not affect the median value but in practice would significantly bring the whole system down.

Either way, they clearly can't test them for 25 years. So presumably they extrapolate from 1 or 2 years. But that assumes the degradation is linear which is no certainty.

Martyn1981

I've relied on the (old) NREL report to conclude that for the UK it's largely minimal/negligible.

The report finds 0.8% annual degradation for all types of PV and ages, but for post 2000 multi silicon it's about 0.36% (pre 2000 0.47%).

Possibly fair to assume that degradation rates are still improving since that report 10yrs ago, but there again there's the thought that mass production can sometimes lead to cheaper products with slightly less endurance?

Personally, after 10yrs with PV, comparing annual generation to PVGIS targets, and also cross referencing to annual sunshine anomaly maps from the Met Office, I can't (yet) see any drop off.

Now, I appreciate that doesn't sound reasonable, and I actually find it strange. Even assuming just 0.2% to 0.3% degradation, I'd have thought I might be able to see a drop off of 2-3% in the figures, but perhaps it is there, just too low to stand out yet for me.

I'm confident that I wouldn't miss a loss of 5%+ in the figures, so less than 0.5%pa is an unscientific position for me, for now.

pete-20-11

The solcast website suggests a 2% drop off for each year they’ve been in operation. May vary, but a good estimate perhaps. 

ispookie666

JKenH said:

I have today edited the discussion title as there may be a more widespread problem with degradation than I am aware of and would like to see the experience of others good or bad, preferably based on statistics adjusted for annual variation in insolaation.

Is there an easy way to calculate the insolation? I can only see daily insolation on PVoutput and I am not sure if it takes into account the prevailing weather condition.   

ispookie666

Looks like another panel bites the dust!  That's 2 out of 14 in nearly 7 years

JKenH

Thispookie666 said:

Looks like another panel bites the dust!  That's 2 out of 14 in nearly 7 years

That’s quite sudden then. It’s a good job you have micro inverters. 

One of my strings on one roof is now down to 76% of the other. On the other roof the poor string is 87% of good one. It could be both the other strings are down but I have no reference level of output to compare against. 

I was pleasantly surprised to make PVGIS target last month. It just shows how sunny March was - second sunniest on record according to ITV weather.

QrizB

ispookie666 said:

Looks like another panel bites the dust!  That's 2 out of 14 in nearly 7 years

That's a shame

How accessible are they, and can you still buy equivalent panels? (From your signature I think they're 250 watt panels which seem to be available now and again as new-old-stock.)

Assuming you're on a FIT it could be worth replacing them.

ispookie666

@QrizB
Will need to put scaffolding up for any change and that would be significant cost (likely a years FIT payment).   Might wait till a few more degrade or die completely to consider changing them.    I'll have to  check the M215 micro inverter spec sheet to find out if newer panels would be ok.  

JKenH

Update: I have been exchanging emails with the panel manufacturers in Italy. Initially they requested serial numbers of the panels and lots of info about the individual panels performance that could only be obtained by getting on the roof and behind the panels. Given my installer has gone out of business and other installers are very busy, coupled with the need for scaffold they have scaled that back (for now) to requesting open circuit voltages for the strings which after a bit of help from YouTube I have managed to measure today.

On the east roof, which has 6 panels on each string, DC1 measured 218v and DC2 176v. I swapped the two strings round when I reconnected the panels to the inverter to satisfy myself the problem was with the panels rather than the inverters so DC1 will now appear as the weaker string. 

On the west roof which has 7 panels on each string, DC1 was 216v and DC2 251v. 

The ratios for both roofs are pretty much in line with the difference in wattage being generated, although the west roof OCVs are a little wider apart.

The readings were taken at around 06.20 so the East roof was facing the sun whereas the west roof was in full shade. There was intermittent cloud and as I was in the loft I cannot be sure whether the sun was shining on the panels at the time I took the readings. There would only be seconds between reading each string on the same roof so I am hoping the amount of sun didn’t interfere.

The above is partly an update but any comments would be most welcome.

Thanks in anticipation.