Effect of Thermal Derating on PV Output

11 10 2013

I’ve mentioned Thermal Derating on PV Output caused by hot weather several times on this blog.  It affects us more and more as the temperature refuses to be normal this year.  Then out of the blue, we did have a normal Spring day last Wednesday (it didn’t go over 26ºC), so I made the effort to chart our PV output all day, taking readings every half hour from the moment the inverters woke up (around 5:30AM these days…) until they fall asleep again some time after 5:30PM.  All the data went into a spreadsheet which is how I generated the chart…….

Two days later, on the 11th of October, we had another stinker, it was 36ºC for most of the afternoon with strong Norwesterly winds hitting us from the inland desert……  well, you get the picture!  The day was entirely cloud free too.  So I went through the whole exercise one more time.

The chart below has two lots of data.  Data taken on the 11th of October have red bars (for HOT!), while all the others are for data from the cool day.

The reason those cool day data bars are different colours I will now explain……  most of the bars are blue (for COOL)……..

Around 10AM, dark clouds rolled in causing havoc with the readings.  That’s the grey (for CLOUD) bar, the only one.

The yellow bars are ‘false readings’ caused by a phenomenon I call ‘lensing’Lensing is what happens when the Sun has

All four LEDS on

All four LEDS on

been behind clouds for a while, and then makes a sudden reappearance.  Two things happen.  First, as the Sun starts shining through the edge of the cloud, extra sunlight seems to hit the panels as the cloud edge acts like a lens and ‘concentrates’ the light on the panels.  Secondly, while the panels are shaded by the clouds, they cool down; and when the Sun reappears, the cooler panels momentarily overproduce causing ‘spikes’ in the data.  The 9th was a particularly difficult day to decide exactly when I would use a reading, as those readings went up and down like yoyos….  but I decided to use the high ones to demonstrate what occurs when this happens.  To differentiate those from the blue bars, I decided to use yellow……..  If anything it makes for a pretty chart!

The 9th was the first day all four lights on the new PV Edge inverter (indicating it was operating at between 87.5% and 100% capacity) came on, and stayed lit.  They’d come on before, but for no more than quick blinks, probably caused by the lensing effect described above.


The interesting things to take home from this chart (which you will have to click on to get a sharper view)

As you can see, our 3.5kW solar system doesn’t even come close to producing its rated output……

Had there been no lensing or clouds on the cool day, a clean bell shaped curve (like the red one) would have been created with blue bars.  It is obvious though, that even ignoring the highest yellow bars, more power is generated on the cool day than the hot day.  To further prove my point, you may notice that some of the early red bars are in fact taller than the blue ones (by almost 2%).  I think the reason this happened is because while it was very windy all day long, that wind was actually quite cool in the morning (until about 8AM), and that this wind aircooled the panels to an even lower temperature than occurred on Wednesday.

The hot day produced a nice bell shaped curve from the red bars thanks to no interference from clouds coming and going… but even though we had more sun, we actually produced less energy.

I worked for a crowd called Future Sustainability (yeah, I know……) in Brisbane for about six or seven weeks.  They refused to acknowledge this occurs, lying through their reps’ teeth that panels bought from them would produce their rated output all the time…..  obviously, I didn’t last, can’t stand unethical behaviour!  But then again, very few solar companies will tell you the truth.

Memories of doing the Renewable Energy course at TAFE in the 90’s………..  I just love doing stuff like this.

I did a few other calculations, and worked out that on the cool day, our panels produced between 8% and 11.5% more power than on the hot day.

I also worked out that our old thin film US64 panels operated at 79% of their rated output, while the newer monocrystalline ones only reached 73.8% of capacity.  This is NOT efficiency I hasten to add.  What I mean here is that the 64W panels produced 50.5W (on average), whilst the 185W panels only produced 136.5W.  As I also keep going on about, thin film panels, while less efficient, work better in hot weather.  Pity they are so hard to get these days.



7 responses

12 10 2013

Interesting stuff, Mike, thanks. I’ve just put solar on, so anything and everything I can learn is of benefit.

(Bev from ROEOZ)

14 10 2013

Interesting analysis Mike. Note that your roof tilt angle and latitude also has a large effect on when you are likely to achieve peak solar output. ie, to achieve maximimum solar output, then you want your solar panels pointing directly at the sun. So if your roof angle has a tilt angle of around 20 degrees, then unless you live up in Nth Australia, the sun is probably still too low in the sky even at noon to achieve maximum output. You may find that on cooler days next month or in Dec you will get closer to rated output. Having said that, you are unlikely to reach much beyond ~90% of the rated output of your panels.

Most mono or polycrystalline pv panels have a temperature coefficient of around 0.5%, meaning they will produce about 0.5% less electricity for each degree above 25 degC.

If you want to play with a lot more data, from thousands of systems around Australia & the world, see

14 10 2013

Hi and welcome to the blog……. having a diploma in Renewable Energy Technology, I am well aware of all the variables! Our Northern roof is on a 12.5º tilt actually, so we may not reach optimum insolation angle for a few more weeks. The Summer Solstice angle here is 9º from the vertical, so I expect 12.5º will occur maybe four weeks before then. We are only two and a bit months from then right now…. I doubt we will notice much difference, especially as it’s bound to get hotter.

I have measured our panel temperature using an infrared non contact digital thermometer a friend lent me, and in the middle of the day they can reach 75ºC, or 50ºC higher than what they are rated at. Going on 0.5% derating per ºC makes it 25% lower than what you are led to believe!

I’ve looked at, and registered with that http://www.pvoutput.org/ website, but I don’t have the means of collecting data automatically, and I’m sure not spending the rest of my life glued to my energy monitor!

14 10 2013

cheers Mike, I suspect you are right, that there’s unlikely to be much difference in the max output between now and late November. But worth looking back on again in a month’s time just to see.

Luckily I can connect to my inverter with Bluetooth, which I normally do about once a week, and then download the data and upload it to PV output. It’s nice to be able to compare my system’s output to those of nearby systems.

I’m down here in Canberra, and have never tried measuring surface temperature using an infrared thermometer. But I’m pretty sure I haven’t seen anything like a temperature 25% penalty. Will have a look through my data to see if I can make an estimate.

29 10 2013
Dave Kimble

It would be interesting to plot the (Actual / Expected) of the panels over a range of temperatures, ideally under the same conditions of sun’s angle to the perpendicular to the panel.

You know the direction and tilt of your panels, and the corresponding angles for the sun can be found from http://www.davekimble.org.au/ephemeris/solar.poisition.htm . The loss of sunlight due to big angles from the perpendicular is given by (1/Cosine(angle)). When you measure two angles, you need to do this twice and multiply the two results:
(1/Cos(North – Sun’s Azimuth)) * (1/Cos((90-Tilt) – Sun’s Altitude))

Then there is the strength of the sun to consider. The panels’ output are rated at 1,000 W per square metre (and 25 C), but is the sun that strong where you are? The angle of the sun makes a big difference to the amount of air it has to pass through, and the variable effect of water vapour in the air makes a big difference too – that’s why temperatures in the Wet Tropics are less variable than where you are.

29 10 2013

It’s actually been VERY dry here…..

29 10 2013

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s