Shading of PV arrays

[johnxgoodall]

I have had two companies quoting for PV installations. One has told me that he would fit them on four of my roof surfaces with differing orientations from south to west but the other has told me this is impractical since as soon as one panel on one array becomes shades all the other panels suffer a similar effect. For this reason the latter company has said that it is only practical to install arrays on roofs with identical orientations.  Different parts of my roof are shaded at different times of the day by the house next door. Can anyone enlighten me?

[billi]

hi there are Grid tie inverters available with more then one Internal MPPT Tracker , so each string should then work independently from the other

Billi

[StBarnabas]

I agree with Billi. You would need to wire the different roof sections as separate strings and have a grid tie inverter (GTI) with multiple Maximum Power Point trackers - one for each string. Alternatively you could  use a number of smaller GTIs

[johnxgoodall]

Thanks for those replies - I assume that the company who would fit on differently orientated rules would use the methods you mentioned and that if the other guy did not know about these then he shouldn't be in the business.

John

[djh]

I assume

Assuming is dangerous! Check

[Paulh_Boats]

Those 4 strings could be in parallel with blocking diodes on each string - then each string (roof section) would not interfere with another. Not as good as 4 mini-inverters though.

Steca have a good range of mini-inverters:

http://www.stecasolar.com/index.php?StecaGrid_300_500_Control_en

You would have 4 of these in parallel, one for each roof section.

cheers
Paul

[billi]

or Kostal Pico 5.5  has 3 inbuilt tracker  and the Mastervolt Sunmaster QS 6400  has 4 trackers

But donot know if allowed in the UK

Billi

[Outtasight]

The strings in parallel with diodes idea will be the cheaper option and will work with a cheaper GTI with only one MPP tracker if the strings are nominally the same voltage (i.e. same number of PV modules) per string.  If one is much bigger than another then it may not work very well as the high voltage string will block the low voltage strings from working.  In any case, the strings would have to all be within the working voltage range of the GTI.

Assuming they are reasonably balanced strings then it may give you some advantage in extending the generating day if you have a east, south and west facing slope roof.  The peak output will be lower (meaning for a given power of PV you can use a smaller and cheaper GTI) and you'll get a flatter power output profile across the morning, mid-day and afternoon. 

Have the shadows surveyed properly though.  Even small shadows (TV aerial or satellite dishes) can cause the power output to fall off quite a bit.

My father-in-laws system is on two roof slopes (S and W) and has 15 modules on one and 13 on the other.  This required a combiner / voltage converter before the 5.5kW single MPP GTI.  The PVC-SNA7C2 combiner he has is explained on this page (well you can see the diagram at least...)

http://jp.sanyo.com/solar/system/connection/index.html

[johnrae]

Think there's a spelling error in the second line/column of the link

Seriously though, I would have thought that blocking diodes from parallel power sources act as OR gates and should only permit the panel with the highest voltage to power your system.  However I'm sure others will enlighten me as to why this isn't the case - unless it is.

jack

[Paulh_Boats]

Jack,

The inverter load will drag down the voltages of each parallel string to a common voltage point where the diodes join. Kirchoff's law says that all the currents flowing into that point from the strings will combine into a single current going into the inverter.

Each string could have different currents due to slight variations of manufacture and temperature....but they would all be producing the same voltage at the common diode point. In other words their currents would adjust to match the voltage at the common point.

The diodes prevent current flowing from one string to another and the panel voltage drop under load keeps things balanced. Then if one string shuts down due to shading the other 3 carry on regardless.

I hope that explains it!

cheers
Paul

[johnrae]

Having a degree in a relevant subject i fully understand the point you're making but only if the panels are outputing similar voltages so that the droop factor permits current flow from each - even if it is grossly unbalanced.  If the shaded panel output voltage is substantially less that those in the sun their output will be zero and therefore totally ineffective. 
jack

[Outtasight]

That's why it would be best if the parallel strings are matched voltages or you go with two trackers.  Say the first string is 5 modules (175V at 5.7A or 1000W) and the second is 7 modules (245V at 5.7A or 1400W).  Parallel them together with both in full sun and two things could happen...

Say the inverter locks on to the 245V peak as its maximum power operating point.  The 7 module string will deliver 245 x 5.7A = 1400W but the 5 module string will try to get to 245V by going to Voc (where it produces 220V at ZERO power).

Say the inverter scans the IV curve and spots the peak at 175V and another at 245V and decides to settle at 175V so that both strings can contribute current...  Now you get 175 x 5.7A = 1000W from the 5 module string and 175 x 5.7A = 1000W (not 1400W) from the 7 module string for a total of 2000W.  So it would be cheaper just to only have 5 module strings in that situation (as you never get to use the last 400W of a 7 module string).

If you used a dual tracker inverter then you'd get the full 2400W of both strings as they could operate at their independent maximum power voltages (but at the cost of a more expensive inverter).  This is what I have in my system.  There are two arrays - one works at Vmp=35 and the other works at Vmp=46 into two MPPT controllers.

If one string points in a different direction then it's less of an issue as they'll produce peak power at different times of the day and so you'll not "waste" power and the diodes will stop power from the active string leaking through the inactive one.

[Paulh_Boats]

Jack, Outtasight,

Sorry I forgot to mention my assumption that all 4 strings have identical panels.

I do wonder though if the money for east/west facing panels is better spent on 1 solar thermal panel on the south side.....solar thermal can be almost 80% efficient.

cheers
Paul

[billi]

f you used a dual tracker inverter then you'd get the full 2400W of both strings as they could operate at their independent maximum power voltages (but at the cost of a more expensive inverter).

Outtasight   i cannot say anything about blocking diodes (cause  haven't got the electrical background  you have   )

But the 300- 500  Pounds more for a dual or  a "quadruplets" GTI  for a 4-5 kwp PV, does not sound too shocking to me


But how are these Blocking Diodes behave ( are they called bypass diodes as well ?) Do you need them on each panel or only on each string ?  And when do they kick in ? And if they then kick in do the close off a whole string ?

cheers Billi

[EccentricAnomaly]

But how are these Blocking Diodes behave ( are they called bypass diodes as well ?) Do you need them on each panel or only on each string ?  And when do they kick in ? And if they then kick in do the close off a whole string ?

Nobody else has bitten so I'll have a go.  Hopefully somebody will correct or refine this:

Bypass and blocking diodes are different things.  Imagine you have four panels in two strings.  Each string has two panels connected in series.  The two strings are connected in parallel.  Inside each panel is a bunch of cells, I've just shown three in series in the diagram at the bottom of this post.

The bypass diodes are the ones in the panel I've shown for each cell.  Their purpose is to allow current generated by any of the other cells in the panel (or the other panel(s) in the series string) to flow in the case that the individual cell is shaded.  Of course, though the same current may flow the overall voltage (and therefore the power) of the string will be reduced - that's part of what MPPTs are supposed to track.

The blocking diodes are the ones at the top of each of the strings.  Their purpose is to prevent current flowing in the "wrong" direction through the string when another string in the parallel array is generating more voltage.  Remember that the equivalent circuit of a solar cell



contains a diode which is forward biased to short out the current the cell generates and for reversed voltage pushed in this way.  (Note, this diode is intrinsic to the construction of the cell and is separate again from the bypass and blocking diodes.)  If it was an ideal diode then PV wouldn't work, we rely on the fact that it needs a bias of about 0.6 or 0.7 volts to make it conduct significantly.

Without the blocking diodes the output voltage of the array would be limited to the Voc of the most shaded string.  That wouldn't be a complete disaster as even this is well above zero volts but still it's likely to be better to put up with the permanent voltage drop of the blocking diode than to suffer this loss.

Something you have to be careful of in selecting the blocking diode is to make sure that its reverse breakdown voltage is safely above the maximum array voltage.  AIUI, you'd ideally like a diode with a small forward bias voltage drop (Schottky?) but these tend to also have low reverse breakdown voltages.

What I'm not sure about is which panels have bypass diodes and whether they're really put in one per cell as I've shown in the diagram or if a bunch of cells are bypassed together.