Comparing Solar Power In Summer And Winter ..........

Chasn

Full Member

Messages
44
Whilst there are big differences in solar output summer/winter and latitude/tilting my recent experience with an off-grid fixed system shows that the differences between sunny and overcast can be equally significant. I have 36 panels (325w each equating to11.7kw) set at an angle of 30 degrees at 51.5 degrees north. At midday on sunny days at this time of year they will still put out 10.5kw. However this can drop by 30-40% by a passing cloud and can drop to 1kw in overcast conditions. To show the real panel output capability the batteries being charged need to be significantly less than full (certainly below 90%) to enable a big enough voltage difference between panels and batteries to create a meaningful current flow.

As well as providing electrical power to a remote building I use this system to charge an EV. By connecting up with a 7kw charger every 4-5 days I can add about 150 miles range on each occasion thus avoiding paying for grid electricity. If the conditions are sunny any initial loss in battery bank power is recovered by the end of the day. However with the current heavy overcast conditions this can take several days to recover. As we progress towards winter I will have to rely more on my home based charger. In the last month about 800 of 1000 miles driven has been provided by the sun so that is very satisfying.
 

AvalonAdventurers

Full Member

Messages
19
.........A Theoretical Approach.

Everyone in the UK who has solar panels finds that significantly more solar energy is harvested in the summer than in the winter.
And those who live at more northerly latitudes expect to harvest less solar energy than those living in the south.

To put this another way, it's generally accepted that we can harvest more solar energy in the summer and more solar energy at more southerly latitudes.
As a rule of thumb the guidance often given for the UK is something like 'In the winter expect to get about 10% of the solar energy you get in the summer' and 'If you're lucky enough to get to Spain in the winter you'll do better than in the UK'.

Previously I've had a wander around the internet searching for a theoretical approach that confirms these anecdotal comments, but haven't managed to find the sort of thing I was looking for. At the time I decided that if I couldn't find what I wanted, then I would have a go at theoretically quantifying the effect of season and latitude on solar harvesting. The Autumn Equinox is fast approaching and I have some time available while Moho-ing in the West Country - so the time is nigh!
Please bear in mind when judging the following that I'm in the Moho and only have a phone (occasionally) connected to the internet, paper, pencil, pens and the sort of basic calculator that an 'A' level maths student might have had as of about 10 years ago.

I've deliberately kept this post a maths-free zone but I should add right here and now that some people will find some of the technical content that follows somewhat tedious and/or of little interest.
If that's the case for you, then please feel free to scroll down to the graph in the last photo, which summarises the outcome of this earlier stuff.

As always with these things, one has to make a few assumptions. Here are mine:

1. The solar panel is used horizontally.
2. The weather in the winter is generally the same as in summer. That is to say, there is roughly the same proportion of cloud and clear sky in summer and winter.
3. The sun crosses the sky following all or part of something called a sine curve (the sort of pattern that non-breaking waves follow on the surface of the sea or that ripples follow on a pond).
4. Light falling on the solar panel isn't obstructed by trees, hills etc. at any time of the day. That is, light from the sun reaches the solar panel as the sun rises above the true horizon at sunrise and similarly leaves the solar panel as the sun sinks below the true horizon at sunset.
5. The amount of solar energy arriving at a solar panel is proportional to the angle of the sun above the horizon and how long it's at that position.
6. A solar panel and controller starts charging a battery at sunrise and stops at sunset. This assumption agrees extremely well with what my own solar panel and controller does throughout the year.

There are other assumptions such as what happens to light falling on a solar surface at low angles, but I feel that the six assumptions above are those that can be recognised intuitively and are those probably having the greatest impact on the approach I'm taking on solar harvesting.

If you've been lucky enough to be on or above the Artic Circle at midsummer, you may have seen the midnight sun. It's a wonderful sight.
This is a time lapse photo (not produced by me) of the passage of the sun as it crosses the sky in a 24 hour period (fig. 1):

View attachment 56547

It can clearly be seen that the path of the sun follows a sine curve as described in assumption 3.
Considering fig.1 and assumption 5 together leads us to the conclusion that under these conditions the solar energy arriving at a solar panel in a two hour period (there are 12 strips representing 24 hours) will be proportional to the area between the sun and the horizon of one of the strips on the photo. And the total energy arriving at the solar panel in a 24 hour period would be proportional to the total area of all 12 strips.

That's all very interesting, but what happens at lower latitudes south of the Artic Circle where most of us use our mohos?
(From now on the images are hand drawn so please bear with me!).
If we were to travel further south to the latitude of the Artic Circle (about latitude 67°) and take similar photos on midsummer day, we'd find that the path of the sun just clips the horizon and the path of the sun would look thus (fig.2):

View attachment 56551

Again, the total energy arriving at the solar panel will be proportional to the area under the curve (the red hatched area is a precise area under the sine curve, whereas the 12 strips described above are an approximation).

If we now travel south of the Artic Circle at midsummer, the horizon blocks the sunlight for part of the day and the red hatched area becomes smaller; the solar energy arriving at the solar panel is diminished (fig.3):

View attachment 56552

Finally, if we were to revisit this latitude at midwinter the horizon would be blocking even more of our view of the sun's path and the red hatched area would be smaller still (fig.4):

View attachment 56550

Fortunately there's a mathematical method that allows us to calculate the area below curves (such as the sine curve) extremely precisely. It's one of those things where the outcome is far more interesting than the method so I'll miss the method out here. Having said that, should anyone want to discuss the maths, I'm happy to do so at a meet.

So how can we use this information?
I've selected a latitude (48°) and established the times of sunrise and sunset at midsummer for that latitude.
Using the mathematical method hinted at above, I've then calculated the red hatched area (it looks similar to that shown in fig.3) for midsummer.
For the same latitude (48°), I've then repeated the above for the times of sunrise and sunset at midwinter (this time the hatched area looks similar to that shown in fig 4) to establish the area under the sine curve at midwinter.

In order to compare the solar power available at midwinter with that at midsummer, we can divide the red hatched area for midwinter by the red hatched area for midsummer. In this case we get the ratio 0.21. In other words, the output at midwinter at latitude 48° is 0.21 (or 21%) of the output at midsummer.

I've then repeated the calculations described above at midsummer and midwinter for latitudes 30°, 58° and the Artic Circle at approximately latitude 67°.
At latitude 30° the proportion is 0.46 or 46%.
At latitude 58° the proportion is 0.081 or 8.1%.
At the Artic Circle (about latitude 67°) the proportion is 0.0 or 0% - which is to be expected since the sun doesn't rise at the Artic Circle at midwinter.

And finally, to hopefully make at least some of this information useful, I've plotted proportion of solar output (midwinter divided by midsummer) versus latitude for the 4 latitudes I've mentioned above (fig.5):

View attachment 56553

For information I've marked the latitude of a few cities on this graph. If the location of interest to you isn't included, then simply establish your latitude and from the graph read off the theoretical solar proportion figure for that location.

I'm somewhat pleased to find that the mean figure for the UK (at a mean latitude of about 55°) is 0.11 or 11%, which agrees very well with the anecdotal comments one hears and is mentioned in the third paragraph at the beginning of this post.

Other observations I thought interesting:
1. The north of Scotland has a winter/summer proportion half that of southern England.
2. Faro has a winter/summer proportion over twice that of southern England.
3. Marrakech has a winter/summer proportion we should all be jealous of!

Do bear in mind that a solar panel rarely delivers it's rated output, simply because the rated output is measured under a particular set of test conditions that are rarely experienced outdoors in the UK.
If you have a 300 watt panel and the information presented here gives you a winter/summer proportion of 0.15 (15%) then in winter you can assume that your solar panel should be considered the equivalent of a 300 X 0.15 = 45 watt panel under the conditions that you generally use it.

If you've found this post interesting and/or useful I'd appreciate a happy smiley or similar. If the opposite is true then please post a sad smiley or other obviously negative emoji. Your feedback will help me decide whether it's worth expanding on this work to help members estimate, for instance, their winter solar output in a more southerly location compared with their summer solar output in the UK etc, etc.

© Please note that this post is copyrighted and may only be reproduced or distributed in total or part by any means including electronic by permission of the copyright holder who may be contacted via the owner of Motorhomer Ltd.

Colin🙂🙂🙂
😁🤵👰😁
 

harrytherid

Full Member

Messages
28
Unless I have missed something, no one has mentioned the facts that at low incidences more of the available light is reflected away and the angle of incidence dictates that the light has further to travel through the glass to get to the cells so the efficiency is going to be even less. The answer to that is the possibility of angling the panel to face the sun, difficult but possible and is it worth the effort.
Harry
 

Full Member

Full Member

Messages
4,158
My original post made some assumptions, one of which was that a horizontal solar panel and it's controller would be capable of charging a battery from sunrise to sunset.
At midsummer day here at approximately latitude 52° sunrise is at 04:42 and sunset at 21:24, resulting in a day length of 16 hours 42 minutes. Unfortunately I don't have a screenshot of the solar charger output on midsummer day to allow comparison of daylength and charging time, but do for the Autumn Equinox:

Screenshot_20200914-081005~2.jpg

As can be seen, the system was charging the leisure battery for 12 hours 48 minutes at the Autumn Equinox.
On the equinox one would expect 12 hours of daylight at pretty much any latitude, but there are several different ways (surprisingly!) of defining these things and using the definition I've used, sunrise was at 06:46 and sunset at 19:01, giving a day length of 12 hours 15 minutes.
There's reasonable agreement between daylength and charging time at the Autumn Equinox.

On midwinter day here at approximately latitude 52° sunrise is at 08:06 and sunset at 15:52, resulting in a day length of 7 hours 46 minutes. The solar system data on midwinter day was:

Screenshot_20201221-200549~2.jpg

As can be seen, the solar system was charging for a total of 7 hours 3 minutes.
Again, there wasn't bad agreement between daylength and charging time.
On midwinter day the weather here was appalling with dark skies which means that the only solar energy reaching the solar panel was light scattered by thick cloud - so the light intensity was much attenuated. This would affect the start and stop times of the solar charger, resulting in reduced overall charging time.

In summary the assumption I made in the original post was reasonable to a first approximation.

Colin 🙂🙂🙂
 

Full Member

Full Member

Messages
4,158
The solar panels that we currently use on our houses and Mohos have a conversion efficiency of about 20% (this is a figure for the better panels currently available). They convert mainly solar energy towards the red end of the Sun's spectrum to electricity leaving the energy towards the blue end of the spectrum largely wasted. A new class of panels is being developed which harvest power from the red and blue end of the solar spectrum and promise to increase efficiency to somewhere between 30% and 40%. These panels are being developed by, amongst others, Oxford PV and could be brought to market by next year.
As part of the series 'Why I'm Feeling Hopeful About The Environment In 2021', BBC Radio 4 has a piece on this development which can be found here.

Colin :):):)
 

SquirrellCook

Full Member

Messages
360
My original post made some assumptions, one of which was that a horizontal solar panel and it's controller would be capable of charging a battery from sunrise to sunset.
At midsummer day here at approximately latitude 52° sunrise is at 04:42 and sunset at 21:24, resulting in a day length of 16 hours 42 minutes. Unfortunately I don't have a screenshot of the solar charger output on midsummer day to allow comparison of daylength and charging time, but do for the Autumn Equinox:

View attachment 57923

As can be seen, the system was charging the leisure battery for 12 hours 48 minutes at the Autumn Equinox.
On the equinox one would expect 12 hours of daylight at pretty much any latitude, but there are several different ways (surprisingly!) of defining these things and using the definition I've used, sunrise was at 06:46 and sunset at 19:01, giving a day length of 12 hours 15 minutes.
There's reasonable agreement between daylength and charging time at the Autumn Equinox.

On midwinter day here at approximately latitude 52° sunrise is at 08:06 and sunset at 15:52, resulting in a day length of 7 hours 46 minutes. The solar system data on midwinter day was:

View attachment 57921

As can be seen, the solar system was charging for a total of 7 hours 3 minutes.
Again, there wasn't bad agreement between daylength and charging time.
On midwinter day the weather here was appalling with dark skies which means that the only solar energy reaching the solar panel was light scattered by thick cloud - so the light intensity was much attenuated. This would affect the start and stop times of the solar charger, resulting in reduced overall charging time.

In summary the assumption I made in the original post was reasonable to a first approximation.

Colin 🙂🙂🙂
Am I missing something, it looks like your claiming to charge already charged batteries?
 

Full Member

Full Member

Messages
4,158
Am I missing something, it looks like your claiming to charge already charged batteries?

The solar panel and controller are constantly monitoring the condition of the batteries and making good any discharge, whether that discharge is intentional or not.
I posted something on this topic a while ago:


I hope this helps.

Colin :):):)
 

Obanboy666

Full Member

Messages
419
Reading through this thread with a great deal of interest and it amazes me the detail monitoring some members log and use to improve their solar setups.
In layman's terms what I have found using my system which is 200 amps of batteries, 210 watts of solar and a Victron mppt controller is no problems with maintaining battery charge spring and summer when I don't need much charge as not running heating and lights and not much tv but crap in the autumn and winter when I need it lol.
So always carry my ever reliable Honda generator 😁😁😁😁😁😁.
 
Last edited:

Molly 3

Full Member

Messages
968
lol Baz, i just posted on a thread on WC to say where I am today it has been overcast and wet all day and best I have seen coming in is 4W off 300W of panels. Thats North Yorkshire
Sound right BUT i know you have a plan B I have rigged spare panel to charge a spare battery some days not getting enough to power the mppt controller , I have rigged up a monitor so no guess work , in perfect sunshine panel facing the sun I got 4.4 amps for a few minutes, Bazz
 

Lee

Rally Organiser

Messages
1,403
lol Baz, i just posted on a thread on WC to say where I am today it has been overcast and wet all day and best I have seen coming in is 4W off 300W of panels. Thats North Yorkshire

Charlie summed it up years ago quote solar in the winter is a much use as a chocolate fire guard.
 

Trotter

Full Member

Messages
2,263
My new toy tells me my batteries are at 13.08v at the moment, 0800. By midday they will be back to 13.8. Same every day.
This would change, IF I COULD USE IT. Sorry for shouting. But I’m, Bee Oh Are EE Dee. In the extreme.
 
Top