Solar Panel output: Rating versus Reality

Introductions, general chit chat and off-topic banter.
T1 Terry
Senior Member
Posts: 771
Joined: Thu, 30 Sep 2010, 20:11
Real Name: Terry Covill
Location: Mannum SA

Re: Solar Panel output: Rating versus Reality

Post by T1 Terry » Sat, 12 Jan 2019, 12:17

We had a lot of the semi flexible panels fail with wide cracks showing on at least one module in each panel. At first we put it down to the uneven expansion pulling the modules apart, but then after another lot failed that weren't glued down to the Themoclear we stumbled upon the reverse current flow problem actually heating the weakest modules until they failed. We fitted blocking Schottky diodes in the positive panel cables and that appears to have solved the problem. Well, we sure hope it has anyway, it costs us a lot of money and time replacing all the failed panels so it was a very expensive learning curve.
It will be interesting to see if the glued down panels with blocking diodes fitted survive for more than 12 mths, that was they will have been through all the seasonal temperature variation that could cause the expansion cracking the modules or if that theory is now debunked.

T1 Terry
Green but want to learn

rhills
Site Admin
Posts: 365
Joined: Fri, 25 Jul 2008, 01:57
Real Name: Rob Hills
Location: Waikiki, WA

Re: Solar Panel output: Rating versus Reality

Post by rhills » Sun, 13 Jan 2019, 18:12

Hi Terry,
I'd actually used some "JCB Joint Connector" bolts and matching Barrel Nuts from Anzor Fasteners to connect my panel to its square-section aluminium frame so I will repurpose these..

Interestingly, my experience with the Armourflex has not been that wonderful. The one panel I've glued to the Thermoclear has started to separate from the Thermoclear already. The Armourflex bead has remained stuck to the back of the panel but seems to be separating from the Thermoclear quite easily. I've picked at one of the beads on the back of the panel with my fingernail and it seems it won't be too hard to pick off.

I'm not sure what went wrong with the Armourflex. I cleaned both the back of the panel and the thermoclear quite thoroughly with metho, removing all visible marks. I sanded both surfaces lightly with 180 grit sandpaper and then cleaned again with metho. I applied a bead of Armourflex around the edge of the panel and then applied quite a lot more in the middle, zig-zagging down both halves, so quite a bit more than the large cross you mentioned. FWIW, the beads didn't spread anywhere near as much as I expected, so maybe I didn't put enough weight on it? I also wondered if the Armourflex might be a bit "stale", but I couldn't see any use-by date or manufacture date on the tube anywhere but Selleys' web page on the product indicates a 12 month shelf life so stale product could be an issue. Then again, maybe I didn't leave it long enough to cure, I'd removed the weights after about 12 hours and started moving the panel around (gently) after about 24 hours. Re-reading your earlier post, I note you said that it can take up to a week to cure "if it is thick". Not sure how thick "thick" is though :^)

In any case, today I tried using the panel, feeding its output through a Morningstar Sunsaver Duo PWM regulator into two AGM batteries. I saw up to 20.9V from the panel and current between panel and controller started at 5A and quickly dropped to about 3.5A. Unfortunately, this was all around 3:30pm so while the panel was at right-angles to the sun, unshaded and with clear sky, the sun was obviously fading as time passed. Also, I only had the one volt meter so I couldn't measure voltage and current at the same time, I had to re-jig the wiring in between. However, at one point, I did measure the panel surface temperature at 65°C (using a Benetech GM300 Infrared Thermometer), this was around the time the current had dropped to 3.5A. Unfortunately I'd not measured it when the current was at 4.5-5A. However, I think it is possible that the Thermoclear is not cooling the panel as much as it might. I note that the glue bead is causing a 1-2mm separation between Thermoclear and panel wherever there isn't glue so maybe that's preventing effective heat transfer away from the panel?

As I mentioned earlier, I think that for several reasons, I should split this panel from the Thermoclear if I can do it as easily as seems likely, remove the glue and try bolting it together instead.

Cheers,
Rob Hills
AEVA Webmaster
2014 Mitsubishi Outlander Aspire PHEV
Jul 2014 - Dec 2018
Total Petrol: 622.0L
ODO: 44964
Av Consumption: 1.38 L/100km

T1 Terry
Senior Member
Posts: 771
Joined: Thu, 30 Sep 2010, 20:11
Real Name: Terry Covill
Location: Mannum SA

Re: Solar Panel output: Rating versus Reality

Post by T1 Terry » Mon, 14 Jan 2019, 11:03

Lay the panel and thermoclear in the sun, it will pass 70*C where the Amourflex meets the thermoclear and let go if the glue is more than about 2mm thick. Lift the corner, the foot placed on the Thermoclear and carefully lift the panel off the Thermoclear, just make sure you do create any tight radius curves. When the modules are warm/hot they are quite flexible so very little risk of cracking/breaking them, but careful as you go and they will separate easily.
Use a sharpened scraper to carefully remove the Armourflex from the back of the panel, again, much easier when it is hot. Clean the remainder off with turps and revert to the free contact method for heat transfer to the Thermoclear.

T1 Terry
Green but want to learn

rhills
Site Admin
Posts: 365
Joined: Fri, 25 Jul 2008, 01:57
Real Name: Rob Hills
Location: Waikiki, WA

Re: Solar Panel output: Rating versus Reality

Post by rhills » Mon, 14 Jan 2019, 13:24

Thanks Terry, I managed to separate the two last night, carefully working my hands between the Thermoclear and panel without any curves. I'd avoided any more sun on the panel fearing it might cure the Armourflex further but I'll definitely use the sun to help get the glue off the back of the panel now. I had been using a sharp, broad chisel which seems to work quite well: some of the Armourflex comes off quite easily, some is very well stuck so hopefully heat +/- turps will help with that.
Rob Hills
AEVA Webmaster
2014 Mitsubishi Outlander Aspire PHEV
Jul 2014 - Dec 2018
Total Petrol: 622.0L
ODO: 44964
Av Consumption: 1.38 L/100km

T1 Terry
Senior Member
Posts: 771
Joined: Thu, 30 Sep 2010, 20:11
Real Name: Terry Covill
Location: Mannum SA

Re: Solar Panel output: Rating versus Reality

Post by T1 Terry » Mon, 14 Jan 2019, 13:50

Just don't leave it too long, once it fully cures you will have a real battle on your hands. It is moisture and time that cures the stuff, not heat, the greener it is the more it is affected by heat. just be thankful you didn't use Simson 7003, that doesn't let go until you pass 125*C

T1 Terry
Green but want to learn

rhills
Site Admin
Posts: 365
Joined: Fri, 25 Jul 2008, 01:57
Real Name: Rob Hills
Location: Waikiki, WA

Re: Solar Panel output: Rating versus Reality

Post by rhills » Mon, 14 Jan 2019, 14:05

OK, thanks, I'll get onto it tonight (at work ATM)
Rob Hills
AEVA Webmaster
2014 Mitsubishi Outlander Aspire PHEV
Jul 2014 - Dec 2018
Total Petrol: 622.0L
ODO: 44964
Av Consumption: 1.38 L/100km

User avatar
Richo
Senior Member
Posts: 3373
Joined: Mon, 16 Jun 2008, 00:19
Real Name: Richard
Location: Perth, WA

Re: Solar Panel output: Rating versus Reality

Post by Richo » Fri, 18 Jan 2019, 12:38

T1 Terry wrote:
Thu, 03 Jan 2019, 13:08
I know the percentage loss is for total power, not just the Vmp, so the math is vague at best, but you get the general idea.
The graph for the 150W panel indicates the peak power output drops ~10% every 25DegC above 25DegC.
On a clear sunny summer day in Australia if you can't take a panel from inside at 25DegC then outside facing at the sun and not get with 5-10% of the rating to me this sounds like a consumer law issue.
Doesn't meet specification.
I'd be P'd off if I only got 60%.

Are you sure that you are actually loading the panel up after the MPPT to get the full peak rating?
Help prevent road rage - get outta my way! Blasphemy is a swear word. Magnetic North is a south Pole.

T1 Terry
Senior Member
Posts: 771
Joined: Thu, 30 Sep 2010, 20:11
Real Name: Terry Covill
Location: Mannum SA

Re: Solar Panel output: Rating versus Reality

Post by T1 Terry » Yesterday, 09:29

:lol: It does meet specification, the specification is under the STC test conditions and if you duplicate those test conditions you would most likely get a similar figure give or take a few %. The sly trick the solar industry has used for ever is to not give true output figures in a real world application. Even the NOCT (Normal Cell Operating Temperature) specs are not real world with a 20*C ambient temp and the manufacturer can cherry pick the cell temperature as long as the specify just what that temperature is.
The real laugh is the MPPT controller mob ran with the bogus STC figures and came up with claims of a 30% harvest increase without actually testing that they could achieve such figures in the real world. They simply duplicated the STC figures with a power supply and designed a circuit that would improve the harvest over and above that of a PWM controller using the same lab base equipment.
In reality in the harsh Australian conditions that solar operates in, the MPPT controller's only advantage is the higher solar voltage gained by a series string of solar panels on the supply side means the cabling can be smaller. Side by side testing using adequate cabling, parallel connection and PWM control of solid state relays against an MPPT controller with exactly the same number of panels connected in series shows the PWM set up outperforms on an instantaneous basis as well as over the full days harvest. This test rig is on an off grid set up in Crystal Brook SA and has been running continuously for 18 mths now and the PWM system still outperforms the MPPT with both systems charging a 15.5kWh lithium battery pack.

T1 Terry
Green but want to learn

Post Reply