Partial shading of installation

Talk Electrician Forum

Help Support Talk Electrician Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

binky

retired and loving it!
Supporting Member
Joined
May 7, 2008
Messages
14,579
Reaction score
1,739
Location
Sunny Plymouth
If one part of installation is a little more prone to shading than the rest of it, does this 'drag the system down' if connected to single inverter? My friend has a typical 30's type property with a bay window and small apex roof bit above the bay that protrudes forward from the main roof. He has been told that shadow from this will effect panels mounted beside it and would therefore be better off with 2 inverters to seperate panels from main array further up roof to optimise performance. Given cost of inverters I would prefer to avoid buying 2 units, or was considering possible fitment of diodes to prevent back-flow to panels not pushing out voltage. Your opinions please.

 
Yes, if panels are in 1 string, shading will affect the whole string. Much better to avoid the area altogether, if possible.......

Blocking diodes used to be standard fitment; not anymore due to the high failure rate. Splitting the string could cause more issues than the worth; dependant on the exact location, degree of shading and panel spec / invertor spec.

HTH

KME

 
^^

what he said,

needs a better planning design to avoid shading if possible,

or get panels with blocking diodes and hope they do what they are meant to do.

 
Guys,

I'm amazed the blocking diodes are unreliable!

There should be no issues IF they are spec'd correctly and have been desinged & made satisfactorily!

Or, is this another case of value engineering?

 
It appears to be, mate.

Modules used to be fitted with blocking diodes as a matter of course; then they started fitting `em per string......

Now they aren`t used, as a rule, because the reverse bias from a string seems to routinely wipe `em out..........

( this is after a german study of 200,000 properties, and problems associated with the installs - invertors - as you would expect, other than stuffed diodes ( which could render the whole string as a load on the D.C. system; which may not be noticed / realised for many months!)

I have some "in-depth" info out in the van, but it just started raining here, and it was a pretty intensive day, so I`m not going to go find it now............

 
Guys,I'm amazed the blocking diodes are unreliable!

There should be no issues IF they are spec'd correctly and have been desinged & made satisfactorily!

Or, is this another case of value engineering?
its maybe got something to do with how do you rate them,

they could be the subject of 50v to 250v through them,

and power would have an even greater variable depending on how many panels you have in an array and if they are in parallel or series I suppose.

I dunno, just a thought.

 
Steps, I get your point & I think you get mine.

They should be rated in excess of the maximum voltage they can "see" whether this is fwd or reverse biased.

Also they need to take the max fwd or rev current they can "see".

This may be where the value engineering comes in?

Blocking diodes are successfully applied in many industries.

 
Yes, snake hips, I agree with you.

If it IS down to cost-effectiveness, and not enough headroom in selection, then rewtro-fitting your own will work; I wonder if there is anything in the nature of the load presented by the PV that causes an issue; or even local heating effects by the modules themselves.........? I don`t know, maybe the reference materials in the van go to that extent; maybe not.....I will have a look, just not right now.

KME

 
Sidewinder & KME

I think its got to be down to economics really, as you say.

that was why they started to limit the amount actually fitted inside each panel and now none at all in general.

 
Right, after much consulatation of various books I have a cunning plan (or two). If I wire each panel individually ie in parallel back to single point before inverter, this should overcome any issues with shading on any one panel, however, I have also read that the inverter works better if shown higher voltages ie panels wired in series. I am planning on using a sunny boy inverter, and Kyocera panels with built in diodes. So the questions is,

1/ will lower voltage affect peformance of inverter??? Blurb gives no indictation of this so I am asking for real world experience.

2/ is wiring in this fashion a bit of a mistake - again experinece is what i am interested in

3/ would I be better off using mini inverters connected to say 2 panels at a time, and what make of mini - inverter would people recommend?

 
Inverter comes with min / max input voltage spec, which you won`t get anywhere near with all panels parallel - plus you`re creating a huge Isc(tot), as the Isc from each panel will be added. You NEED the string in series to generate the necessary voltage for the invertor. (i.e. if you have 10 panels, with say Voc of 21.3V, and Isc of 3.2A, you`d have a 21.3V supply, with 32A behind it.

In answer to all questions - don`t do it that way; good size string with decent voltage, and matched invertor

KME

 
I have spoken to my uni electronics lecturer about this. It's more to do with the temperature variation. Semiconductors don't like high temps, it changes the electrical make-up of the material.

Diodes are normally located on the back of the panel and are therefore subjected to serious temps in summer. Think of any hi-tech electronic gizmo, for example the pc you are sitting at, all have fans and heat sinks sitting on top of the imporant semiconductor part...

At the price we're paying for panels you'd hope the diodes were correctly spec'd.

 
The Kyocera panels don`t have diodes in anymore - certain of that.

It could be worth speaking to the panel mfrs about it; as you may find that it is classed as a modification of the mfrs spec. , and MCS might not be happy about it......don`t know this for sure, but it might be worth checking.......

KME

 
Something just occurred to me, not having seem back of panel, what are the options for fitting doides??? I am assuming they are all sealed units etc etc and diode would have to be external to connections box thingy on back of panel????

 
I never had a really close look at the Kyocera panel, but I think that you can get into the termination box; so may be able to fit the diode there?

If it`s a biggish install, with a few strings, it may be better to fit the diodes internal to the property, on the string rather than individuated panels? finding & repairing a blown diode on the array would be a `mare!

KME

 
Top