Connecting modules on different roofs

[DanPhillips]

Hello all

I am looking for a bit of advice. If I wanted to connect a system of 12 modules, 10 of which are on a pitched roof and 2 of which are on a frame on a flat roof, can all the modules be connected together into 1 or 2 strings as normal. All of the modules will be at the same incline and facing the same direction.

Many thanks

Dan Phillips

 

[jerriais]

Basically yes, as you say the incline and orientation are the same, the only other issue to consider is if there is different shading on the two sets of modules. If that's fine too then you can consider them as part of the same string.

 

[SpitfireW'sale]

You can do that providing that all the panels have the same inclination and orientation along with shading issues. If you split your system into two strings, you should have an equal number of panels per string. Unbalanced design results in losses at your system.

SpitfireWholesale

 

[DanPhillips]

Thanks for your replies, that's what I was thinking.

Out of curiousity, if one part of a string did fall into shade whilst another part was not shaded at all would this cause a drop in output of all panels in the string, or just those in the shade?

Many thanks

 

[SpitfireW'sale]

Shading of all or some of the modules in a string reduces the output of the whole series string and any other parts of the array that are electrically connected to that string. The percentage reduction in output can be much more than the percentage of the area that is shaded, since the current through the string is reduced to the level of the poorest performing module. It is not always easy to quantify exactly what the shading losses will be for any specific case, especially as the effect varies with sun position.

It's hoped that this answers your question.

 

[Worcester]

since the current through the string is reduced to the level of the poorest performing module.

I'm confused now ....

If a part of a panel, or even a whole panel is shaded how does that affect the current through the string?

I can see how it could affect the string voltage, how is the current affected?

And how would bypass diodes affect that - for example most panels have 3 bypass diodes (Yingly up to 6 and Day 4 9 diodes) surely the purpose of those is just to bypass that part of the panel that is affected by the shading, and so the aim would therefors by to reduce the effect of the shading, by only bypassing / cutting out a proportion of the panel?

 

[SpitfireW'sale]

For example, we assume a system with 20 panels. I am going to use e.g ET Solar Panel 190W (black or silver frame). I-V characteristic is presented in the data sheet of this panel.

2 Strings of 10 modules. Vmp=36.68V & Imp=5.18A


Maximum power per string at maximum irradiation (1kW/m2) is: Pm=10 * 36.68V * 5.18A=1900W


Let's assume that there is 40% partial shading of one panel (not low enough to activate the bypass diodes of the panel):
The string without shading keeps having Pm=1900W
The string with the shaded panel is forced to have voltage 36.68*10=366.8V (voltage is not affected by the irradiation) but based on the data sheet and particularly the V-I, the current of the string is reduced at ca 3.1A. Pm'=366.8V * 3.1A=1137.08W.
The losses due to shade are: Losses(%)=100 - 100*[1900+1137.08/(1900+1900)]=20%


Assuming that we have got 100% shading of the panel or enough shade to activate the bypass diodes gives the following losses:
The string without shading effect: Pm=1900W
The second string is lossing one panel, forcing to operate at the same voltage V=10 * 36.68V=366.8V, which means 366.8V/9 ET panels = 40.75V per panel.
Based on V-I, for maximum irradiation and voltage at 40.75V, the current reduces to 4A. So, Pm''=40.75 * 9 * 4=1467W
Thus, the losses are = 100 - 100*[1467+1900/(1900+1900)] = 12% Less impact than partial shading!!!

 

[PvNewbie]

Spitfire ...that almost made sense to me, but where did you get the 3.1A for the partial shaded string example? I can't quite deduce that from the V-I given the irradiance of 1,000W/m2

 

[Kannonchiku]

Hello everybody, this is my first post.
I am about to have a system installed . It consists of 16 250w Sanyo panels. 10 on the main south facing roof and 6 on adjacent SE facing roof. The shadow of a chimney cuts across the corner of the main roof early in the morning. The installer is fitting power boxes on each panel for 'independent operation' and no power loss over the whole aray.
I would be very interested to hear any comments on this arrangement

 

[MrBod]

Hello everybody, this is my first post.
I am about to have a system installed . It consists of 16 250w Sanyo panels. 10 on the main south facing roof and 6 on adjacent SE facing roof. The shadow of a chimney cuts across the corner of the main roof early in the morning. The installer is fitting power boxes on each panel for 'independent operation' and no power loss over the whole aray.
I would be very interested to hear any comments on this arrangement

I was interested in using DC optimisers (I assume this is the same thing as the power boxes you mention) for my planned 16 x 250W Sanyo install. However my (potential) installer could not find any that were MCS approved. Are you able to tell me the model of the the power boxes you are having installed?

 

[BruceB]

I was interested in using DC optimisers (I assume this is the same thing as the power boxes you mention) for my planned 16 x 250W Sanyo install. However my (potential) installer could not find any that were MCS approved. Are you able to tell me the model of the the power boxes you are having installed?

Why would DC optimisers need to be MCS approved? They are not as far as I understand it one of the categories that needs MCS approval. For example inverters do not get MCS approval.
Regards
Bruce

 

[BruceB]

Spitfire, your assumptions are not quite right in the partially shaded example. For the shaded string, consider the non-shaded panels. If for whatever reason the current through the string is reduced whilst full illumination continues, then the voltage across each of the unshaded panels in the shaded string will increase along the path in the curve on the data sheet you cited. You need a more complex model than you are using. I believe it very unlikely that more shading in the way you describe will increase the output.
Regards
Bruce

 

[Kannonchiku]

I was interested in using DC optimisers (I assume this is the same thing as the power boxes you mention) for my planned 16 x 250W Sanyo install. However my (potential) installer could not find any that were MCS approved. Are you able to tell me the model of the the power boxes you are having installed?

As a complete novice I know nothing about 'power boxes' or 'powerpack' I'll try and find out from my installer. Sorry I am able to help. Maybe others on this forum will be able to enlighten us. Thanks in advance if you can.

All the best - Keith

 

[Worcester]

Spitfire ... Also almost made sense to me, however, along with Bruce's comments if your going to go to that level you also need to take into account the temperature effects, at 1kW/m2, the panels will be operating as Sanyo call it under NOC (Normal Operating Conditions) as opposed to STC (Standard Test Conditions) so you also need to take into account the losses due to the panel operating at 45°C (NOC) as opposed to STC, similarly NOC would be 800W/m2 I appreciate you might have simplified it for the maths sake, however that skews the real results.

 

[Werneroeder]

Hi Dan,

you can try using solar power voltage optimisers. They cost around £60 each and can be connected to each module or a few, depending on system design. It copes well with different roof aspects and shading issues and claim to increase the yield up to 25%.

 

[SpitfireW'sale]

Spitfire, your assumptions are not quite right in the partially shaded example. For the shaded string, consider the non-shaded panels. If for whatever reason the current through the string is reduced whilst full illumination continues, then the voltage across each of the unshaded panels in the shaded string will increase along the path in the curve on the data sheet you cited. You need a more complex model than you are using. I believe it very unlikely that more shading in the way you describe will increase the output.
Regards
Bruce

Bruce, I know that it's not such an easy calculation because I have not taken into account the coefficients of the panel and in addition everything is calculated based on STC conditions. Actually, using that simple maths, I tried to show the difference between the partial and fully shading of a panel in two strings connected to one MPPT.

I believe it very unlikely that more shading in the way you describe will increase the output.Bruce

I have not mentioned something like that. I just said that two uneven strings may have less losses than two even strings with a shaded panel.

I hope that this makes sense to you.

 

[BruceB]

Then I have misunderstood you, but your words say that completely shading one panel " = 12% Less impact than partial shading!!!"

 

[SpitfireW'sale]

Partial shading results in ca. 20% losses of power.
Completely shading of one panel results in ca. 12% losses of power.

So, it's more efficient to loss one panel instead of partial shading of that.

 

[BruceB]

So you are saying that increasing the shading from partial to full on a panel will increase the overall output of the array.

My point is that is almost certainly wrong, because of the wrong assumptions you are making about voltages in your example.

Regards
Bruce

 

[solarcat]

If both sets of panels have the same direction and tilt, then yes, no problem.
But beware of long DC cable runs. The longer the DC cable, tho more power you loose.

 

[SpitfireW'sale]

The thing is if we take into account all the factors (coefficients, etc), they impact both cases equally. the fluctuation of the irradiance affects only the current, not the voltage. So, the MPPT tries to maximize the power based on the current of each string, not the voltage. The full shading does not increase the overall output of the array, it does only result in less losses than partial shading.

What's your opinion in terms of full and partial shading? In which case you have got less losses?

 

[BruceB]

You are saying that partially shade a panel and you have 20% losses, then fully shade it and you have 12% losses. OK that is less losses as you say, but in the conventional use of English you have increased the amount of shading from partial to full and the output of the array has increased from 80% of max to 88% of max. As I have said, I do not believe it.

If your example assumptions had been accurate I might have believed it, but they are not. For instance, you are assuming the voltage generated by the unshaded string determines the voltage across both strings. It does not. They interact, but settle at a single value of course. You assume the voltage across each of the panels in the partially shaded string stays the same when the current is reduced. It will not. The voltages on the fully illuminated panels will increase along the line of the appropriate curve in the data sheet. That is after all how mppt tracking works. That means that the power reduction is not as large as you calculate.

If your example were true, that less energy in (ie more shading) gave more energy out then you have the basis of a perpetual motion machine that could make you very rich.

Regards
Bruce