More sun in January?

I sat in on a sales pitch for a solar photo-voltaic system a few days ago.  I don’t know what the decision of the owner will be yet, so I will not comment on the economics of the proposal until later.  But one little comment did get me do some research.

As we were walking around the yard with the Solar Pathfinder ™ instrument looking for a good site (we had decided that a ground-mount installation would be preferable to roof mounting) the sales guy mentioned that we got more sun in January than June because “we were closer to the sun” at that time.

OK – so the earth doesn’t have an exactly circular orbit around the sun, but I always thought the difference in distance was so small as to be insignificant.  Turns out I’m wrong – which is not an unusual occurrence.

According to Wikipedia the difference in insolation between aphelion (the furthest out) and perihelion (closest in) is about 6.9%.  I was surprised at that figure (I expected some fraction of a percent) and spent a few minutes scribbling on scrap paper trying to disprove it, but it is correct.  There are lots more detail at Uni of Colorado, further Wikipedia articles on Milankovitch cycles and, in a simplified form, at

By coincidence – it will change in a few thousand years – current perihelion is the 3rd of January, so the earth (in total!) does indeed get almost 7% more sun energy in the northern winter.  But most of that extra energy goes to the southern hemisphere, where it is summer and they don’t really need it.  The effect is completely overshadowed by the tilt of the earth relative to the sun, so it remains winter in the north – due to the angle of incidence and the absorption by the light traveling through more atmosphere.

One thought is that this effect should cause comparatively big differences between summer and winter conditions in the southern hemisphere and relatively smaller differences between summer and winter here in the north.  I have never noticed or read about any such difference.  The difference up here (near New York) seems brutal (we’re sweating it out in record heat just now, and I know we’ll be back in jackets and boots in October), while Australia seems almost boring by comparison (I know, I know – we need to compare like latitudes etc. etc. but the observation does hold)

So what’s the answer?  It’s the huge amount of open water in the south (as compared to the northern hemisphere) which absorbs, distributes and otherwise masks the effect almost completely.

OK – so the whole issue is settled – sales rep is only part right, total insolation is greater in January, but we do get most of the solar energy in summer.

But life is never that easy.  There can still be debate on the angle of the solar panels, and then we get the added complication of snow.  Again I always thought of snow as a bad thing for solar PV – it covers the panels and needs to be cleared off.  Then, on the Solar Pathfinder site, I read: “The average increase of solar radiation used for passive solar heating due to snow cover is only around five percent, but PV panels on a powder-snow-covered, clear, cold winter day often [my italics] produce more watts than on the much longer sunny summer days”.  What does that mean, and why?  First, we have reflection from the snow – new, dry powder is best – and then we have dry, clear air with little absorption by water vapor.

It’s comparable to getting more sunburn on a mid-day hour on the ski slope, than on a comparable mid-day hour on the beach (even allowing for altitude…).  But – you will get a lot more burnt if you lie on the beach all day in summer, against spending a day on the slopes in winter.

So at any instant the panels could indeed produce more power than in a comparable summer instant – but there will still be more hours of sun in summer than winter.

Unless of course you forgot about the big deciduous tree, almost invisible in winter, that spreads a huge shadow in summer over half your panels.  But nobody would make a mistake like that…

Which brings us to considering partial shading of a PV array.  I really need to get a more complete section on that topic, but here is a start.

Leave a Reply