Extreme Shading!

[kenilworthian]

If anyone doubts whether shading might affect the output capability of their prospective PV system the photo montage below will serve as an example that has cost much time and effort to prove and address.

My problem is my neighbour's gable end and chimney which did not present a problem in summer when I was deciding to proceed with solar PV. I was told that "it would be better to avoid as much shading as possible" but I wrongly assumed that the consequential loss of output would be proportional to the area shaded.  How wrong I was!

On a clear blue cloudless day in November I took regular photographs and captured the operational data from my inverter (Diehl Ako with Solar Control 1.1). The vertical scale on the chart below is percentage of maximum rating x 10.

Thus it can be seen that at 10:29 my system gives around 4% output in a hard shading environment.  At 12:15 with around 10% shading output has climbed to 13%. In the more defuse light at 12:54 due to haze softening the shadows, my systems gets to around 50%.

The practical way to address the problem as others have found is to split the PV array into two and use two smaller inverters.
(Taking a bazooka shot at my neighbour's roof never really was an option.)
In summer the problem will be less due to higher sun angles.

So - if you haven't already got your PV system and are doing some homework Caveat Emptor when it comes to shading! When in doubt - don't.

[guydewdney]

wow - thanks.

[mespilus]

Reading on here, that poor December production
can be matched by just a good single day in peak sunny weather,
what annual production do you expect?

[billi]

Great info  Great Photo montage

I guess shading  it is a much bigger Problem in the UK than i am used to it ,from Gridtied installs in Germany ,
cause of  differing roof architecture and space to the neighbour

so the individual  system design  has to be  looked at very carefully , while the price race is on  ( who supplies the cheapest  )

How is it now 2 Month later with higher Sun

Billi

[Outtasight]

Good work.  The shading is pretty severe... more than just the roof and chimney, though.  A significant culprit is the TV aerial and mast.

It only casts a small line shadow (an inch wide, plus the grey penumbra, as it's a small object) but this can knock out a whole module from the string.

Each module may have only two banks of cells with bypass diodes.  If ANY cell in a bank is severely shaded, the whole bank will be bypassed.  So a line shadow going vertically up the length of a module may cause one bank to bypass.  The same line crossing diagonally can cause both banks to bypass.

I counted the modules that had definitely both halves shaded and discounted them.  I also counted any module with a line shadow or a corner or edge clipped by a shadow as being half bypassed.  The counts are:

09:21 = 3.5 active modules (8.5 bypassed)
10:29 = 4 active modules (8 bypassed)
12:15 = 7.5 active modules (4.5 bypassed)
12:54 = 8 active modules (4 bypassed)  

With only 4 active modules, the inverter may struggle to start or run.

The TV mast will cause shading events for a large part of the winter and even spring, as it seems to be a very tall one (much higher than the chimney).

Some new modules coming on to the market will have integrated per cell bypass diodes that will make the situation better, eliminating the problem of one or two shaded cells causing a whole bank or module to drop out.  Unisolar triple junction thin film modules already have per cell bypass diodes.  Trouble is, even with triple junctions, thin film amorphous panels are much bigger than crystalline ones.  But with their better tolerance of shading and better diffuse light performance, it may generate more total energy overall.

Micro inverters (one per module) can also help a lot.  It doesn't stop the bank shading issues, but as each inverter can work independently and at a low DC Voltage, you would get more like 3-4 modules worth of power in the situations where a central inverter would fail to start or bump along at <5% output because the array is out of the working Voltage range of the inverters MPP tracker.

On my ground array, I struggle with the line shadows from my wifes pot plants, that she insists on decorating the front of the array with.  They sometimes get re-arranged to move the tall plants elsewhere (or sometimes they just get a secret pruning when she's not looking ).  

[wyleu]

Lovely piece of work!

Now heres a question... Given that the energy that isn't being taken out of the array presumably ends up as heat, how much and for how long will that heat gain reduce the efficiency of the panels when the shading is removed?

[rondurrans]

On a positive note your array, like mine, should produce 70% of it's forecast energy between March and October and therefore should still be worthwhile from both a 'green energy' and 'FiT' point of view..................

[martin W]

well done, excellent information that very clearly shows how little shadow can have a big effect..... a very clear explaination that would show anyone thinking of pv how important assessing shadows is...

[HalcyonRichard]

Hi Kenilworthian,
                       Thats an excellent post. Good information - and a picture is worth a thousand words. I have a polytunnel which is shaded by the house. I would not take a midwinter shading as typical. Over 5 years I have become accustomed to the suns trajectory changing over the year. In summer the sun rises "faster" and is higher in the sky. I wish I could podt a picture of this. It will be interesting to compare the summer output and shading.

Regards Richard

[Nicedad]

Excellent post, as said by others a picture paints a thousand words. Thanks to taking the time to do it.   

[MR GUS]

What the others said! great example well illustrated, please pass a smite on to the aerial (or a lightning bolt) 

I hope you post some more in he coming months, could be good to do a solid years worth, very handy info.

cheers
Gus

[renewablejohn]

Would it not be worth having a word with the neighbour to remove his now obsolete tv aerial

[JohnS]

I see that you are proposing to change the inverter to two smaller ones doing half the panels each.  This will definitely help.  Bit difficult to see what combination will be best - perhaps 4 from the top row and two from the bottom row into one and 2/4 into the other.

Who is picking up the cost for this?

[wyleu]

Could you route the DC panel lines into differing inverters to avoid the shadowed panels efficiently?
How would you switch and mix those sort of currents and voltages?

[StBarnabas]

Thanks
a very nice illustrstion. Shading is bad! Even with a very optimistic assumption that the cell will avalanche at -5V then it only takes 2 cells to knock out 1/2 of a panel (as Outta says most panels have only 2  bypass diodes). For my Yingli Solar panels with 36 cells the bypass diodes will be connected across two sets of 18 cells. Attached is a sliver of MATLAB code (need to pay) but should run in OCTAVE (freeware). Can post if anyone is interested

-------------------------------
function err=IV2(I,v,T,I_ph,I_0,n,R_s,R_sh)
% PV array model current source with a Schottky diode,
% Series Resistor R_s and Shunt resistor R_sh
% Avalanche breakdown at V_b from S. Silvestre et al. / Applied Energy 86
% (2009) 1632–1640
q=1.6e-19;k=1.38e-23;V_b=-5;beta=3;a=1e-3;
err=-I+I_ph-I_0*(exp(q*(v+I*R_s)/(n*k*T)-1))-((v+I*R_s)/R_sh).*(1+a*(1-(v+R_s*I)/V_b).^-beta);
err=abs(err);
---------------------------------------