Photo from Daniel Pascoa
With summer approaching, we’re all getting excited to see our systems generating to their full potential. So it’s a bit of a downer when the clouds come over and spoil the party. But how much difference do they really make? It’s a question we get asked often, so we’ve run the numbers to shed some light on the topic.
First, a bit of context. (If you can’t wait, scroll down for the numbers!)
How much can your system generate on a cloudy day? There’s no straight answer because every day is different and so is every cloud. It’s actually fairly intuitive though, because solar cells mostly work in the same light wavelength range as our eyes. So if the sun is behind a thin and wispy cloud where it’s still too bright to look at, the solar production will be only slightly affected. If it’s very overcast and but you can still find a somewhat bright spot where the sun is, then you’re probably down to about 50% output. If you’ve got no idea where the sun is, you’re probably down to 20-30% and if it’s dark, the-heavens-are-about-to-open type skies, you might be below 10% or even shut off completely.
If you’ve got your Solar Analytics My Solar App (iOS or Android) handy, then of course you can just check the live monitor any time you want. But you can impress your friends by guessing the output based on contrast of shadows. The images below correspond to the sun-spots on the live monitor chart.
The silver lining
Surprisingly, clouds can sometimes increase solar output. The best case is when you have no clouds in front of the sun, but clouds everywhere else. This is because clouds reflect some scattered light back down to the ground. This is what happens when it feels very “glary” even when you’re not looking toward the sun.
An extreme case of this is when clouds surround the sun and the reflection creates a sort of lensing effect so that some of the direct sunlight, which would otherwise be sent out sideways and not reach your roof, is reflected down onto your array. This usually only lasts a few seconds so it doesn’t have a big impact on your overall output (and in most cases, your inverter would limit the output anyway).
But it’s kinda cool as it literally creates a silver lining on the clouds. You definitely shouldn’t try to look at this as the sun intensity can double during these times and damage your eyes. Speaking of damage - this is also one of the reasons that people get horribly sunburned on cloudy days. So if it’s a partly cloudy day and you suddenly feel like you’re being fried, then point your camera up and hope for the best, like I did.
Now for the numbers.
Ok, so I looked at estimated output based on data known as Typical Meteorological Year (TMY). This is the most representative solar data, as it looks at the distribution across the last ~25 years, rather than any particular year.
If we look at December - January, we can break it down into quarters, so we have the sunniest 25% of days, the cloudiest 25% and the two quarters in between. I looked at four capitals where we have lots of Solar Analytics sites.
If we then take the average output in each of those groups, we can see that in Adelaide, the cloudiest days produce around 57% of the output on the sunniest days. Adelaide is a very sunny place though and we can see that the 2nd-cloudiest quarter of days still produces 90% peak output compared to the sunniest quarter.
Sydney-siders will be all-too familiar with a storm ruining the family BBQ or washing out the Sydney Test and we can see that in the data. The cloudiest quarter of days produces around 37% of the output of the sunniest days, while the 2nd-cloudiest quarter produces around 69%.
Winter of course gives us less solar output since the sun is lower (and therefore gets absorbed by more atmosphere on it’s way to us) and the days are shorter. The difference between summer and winter also depends on our latitude. The closer you are to the equator, the smaller the difference. So in Sydney, the best Winter days produce around 60% of the best summer day output, while in Brisbane, it’s 70% and in Melbourne, about 50%.
In Adelaide, Winters appear to be cloudier than Summers with the cloudiest quartile of days producing 36% of the sunniest quarter output. In Sydney, however, the ratios are quite similar to summer time.
And yes, the solar panels have spoken - and they say that summer in Melbourne is indeed sunnier than it is in Sydney!
 Silicon solar cells work from near-infrared upwards to UV and beyond, so they respond to a wider spectrum than our eyes, but because the strongest sunlight intensities occur in the visible range, these are the wavelengths that have the biggest impact. Most solar cells also look dark blue. That’s because their anti-reflection layers are optimised to absorb the most-plentiful green-red light, which means they reflect more of the dark-blue-violet light.
 My only friends are other solar nerds.
 Also because UV light is worse for sunburn and this light can penetrate some clouds more than visible light.
 I didn’t analyse everywhere. I’ve got limited time and have to get back to finding system faults! But if you’re super keen to know your region, then add a comment and I’ll take a look.
Dr Jonathon Dore swapped International-level trampolining for data science. Lucky for us! Jono is now our Head of Data Strategy and is responsible for the sophisticated algorithms that we use to make our monitoring system the most intelligent available.