Why does shading over one solar panel reduce the output of the whole string?

[Malvineous]

Hi all,

I've been reading up about solar, and one thing confuses me (well lots of things do, but you have to start somewhere!)

From what I have read, if you have a string of identical panels in series and one of the panels becomes shaded, the power output of the whole string drops to match the shaded panel.  In other words, if the panels are producing 8 A each and one becomes shaded and drops back to producing only 4 A, all the other panels in the string also drop back to supplying only 4 A each as well.  It's like each panel also acts as a current limiter, only passing up to whatever current it happens to be producing at that moment, and no more.

This confuses me because I was under the impression that the panels worked like unregulated power supplies, producing as high a voltage as possible (up to the Voc / open circuit rating), and then as power is drawn from them the voltage drops until the Isc / short circuit current is reached, by which time they are presumably at an extremely low voltage but at their full current rating.

If that's correct, then when one panel becomes shaded and it limits the current through the string of panels, shouldn't that make the voltage of the other unshaded panels increase accordingly from the reduction in current?  Shouldn't this result in the inverter seeing an increase in total voltage across the string, since the panels are all in series?  The inverter would presumably draw more current, dragging the voltage down low again but getting the extra power from the unshaded cells.

Apparently it doesn't work like this, but can anyone point out where my misunderstanding is?  If the unshaded panels suddenly can't deliver half the power they are generating, where does it go?

[tszaboo]

One image is worth a thousand words.
mod: they remove the linked picture of course.

[capt bullshot]

The drop in power output depends on how the load (inverter) handles the MPP algorithm.

Usual PV panels have one (or multiple) reverse diodes installed. Either one diode for the hole panel or e.g. 3 diodes for each third of the number of cells in the panel. The purpose of these diodes is: if one or more cells is shaded, the generated current drops for this cell. If other cells / panels are unshaded, they still can supply the full current.
The inverter can now try to still draw the maximum current from the non-shaded panels, this will drive the shaded panel into reversed voltage. The reverse diodes now start to conduct and reverse voltage drop across the shaded panel is limited to the diode drop. If the diodes wouldn't be there, the solar cells would see a quite high reverse voltage and may overheat due to this. So the diodes protect the shaded panel from overheating / damage.

Now, if one panel (or part of the panel) is shaded, these cells deliver less current. The inverter can now either draw this reduced current and keep the voltage across the shaded panel - this results in a gross reduction of delivered power, or just continue to draw the full current from the non-shaded panels. Due to the reverse diodes, the voltage across the shaded panel drops to zero (or a bit negative), reducing the total output of the string by just the amount of one panel.

So it's up to the MPP algorithm in the inverter to try lower voltages and see if there's more power available to maximize to strings power output. Some inverters do, others don't ...

[capt bullshot]

Apparently it doesn't work like this, but can anyone point out where my misunderstanding is?  If the unshaded panels suddenly can't deliver half the power they are generating, where does it go?

It just heats up the panel, since the power isn't taken away from the panel.

If that's correct, then when one panel becomes shaded and it limits the current through the string of panels, shouldn't that make the voltage of the other unshaded panels increase accordingly from the reduction in current? 

Yes, voltage rises, but just a little bit (see the MPP diagrams above)

[Seekonk]

It actually doesn't take much to turn a panels output to crap.  MPPT of a big string is just about as bad an idea as a PWM controller.  The world is going to micro controllers for each panel.

[IanMacdonald]

The reason is that a solar cell acts as a forward biased diode when illuminated. That is why they never give more than 0.7v or so. It follows that a dark cell can be driven into reverse bias by the others in series with it. Since the PIV of the reverse biased junction is more than the panel o/c voltage, no current will flow.

Exactly the same mechanism as the 'bully boy effect' in which reverse voltage will destroy a weaker NiMh or lithium cell in a series chain if you don't have a charge balancing circuit. Except that with the solar cell no harm is normally done.

[f4eru]

An example :


There are two power maximas, and often the mppt will select the wrong one, depending on it's software.

MPPT of a big string is just about as bad an idea as a PWM controller.  The world is going to micro controllers for each panel.

it's a bad idea only if:
- the panel string can be partially shaded by an obstacle
- the MPPT is not smart enough
- the installation of micro-inverters or local MPPT's is financially viable, which is often not hte case.

[f4eru]

just about as bad an idea as a PWM controller.

What's the problem with a PWM controller

[Malvineous]

Many thanks for the detailed replies!  That makes so much more sense now.

I hadn't considered that when the current drops, the voltage of the panel can only rise so far before it hits the Voc and even before then the available current at any given voltage starts to rapidly drop, just as a result of the way solar panels work.

Per-panel solutions (e.g. microinverters or DC-DC solar optimisers) make a lot more sense now.  On this note, is it naive of me to think that a cheap constant-current DC-DC converter will do a similar job to a much more expensive solar optimiser?  Ignoring for a moment that these cheap converters don't do MPPT tracking, they take a range of input voltages and produce a constant output current.  If the output of these were all connected in series, then presumably you'd have a constant current at varying voltage, even as panels became shaded, with the voltage range no longer being a problem?

The reason I ask is that I'm thinking about trying out cheap second-hand panels as a bit of an experiment, so they will all be different specs making a direct series string of them impractical.  Because it's "for fun" so to speak (and unlikely to be grid connected), microinverters won't be any good (too expensive and they all must be grid-tied) and solar optimisers aren't so practical either (they either need a special inverter to control them or they are too expensive for this kind of project.)  But if I could somehow combine a cheap MPPT device with a DC-DC converter...

[f4eru]

It really depends on your configuration!

If you have no shading ( tree, obstacles), then you can skip the microinverters.
If you have two roofs at different angles, you have to put them on separate MPPTs
etcetc...

[Seekonk]

If you have any shading, everything turns to crap fast. MPPT ain't going to save you except in a really long string and parallel strings are just lost. A lot of opinions from people that have never used solar. Being off grid makes you look at every watt.

[mcbota]

Hello. I have a problem which I cannot solve it and I ask you your help , please. On one string from my  two strings PV system, consisting of 14 panels of 400 W, 37 Voc, after the system starts, about one hour later, the voltage decrease to about  220 V, the current on string increases much higher than normal (like something trying to adjust the drop of voltage), stays so for about three hours and then come back to the normal voltage. Looking at the position of the panel I observed that it is a little bit shaded. Is this the effect of the shadow that the inverter tries to adjust it by increasing the power? It is for any help to try to put that shaded panel somehow to the end of the string?. Thanks!

[Berni]

This happens because solar panels are simply current sources with a zenner diode in parallel.

When one gets in shade it stops pushing current trough it, since they are in series this drops the total current.

The fix for this are so called "Optimizer Modules":
https://www.solaredge.com/sites/default/files/se-p-series-commercial-add-on-power-optimizer-datasheet.pdf

These modules sit between the solar panel and the series chain. They monitor each individual panel and bypass the current past it when it gets shaded. Additionally they also monitor the performance of individual panels and send that information over powerline communication back to the solar inverter. These are typically only seen in modern large solar arrays that operate with a series open circuit voltage in the 200 to 1000V

EDIT: Additionally these modules enhance the safety of the system since the inverter can command the whole array to shut down by disconnecting all panels. Leaving the formerly 1000V wires perfectly safe to touch even during mid day sun. Making maintenance safer.

[trobbins]

One would hope that the MPP used in the inverter can move to the actual MPP of the string.  The MPP of a string of panels is a series combination of voltage and current curves from each panel (as the inverter only sees the resulting series sum of panel voltages, and the sole current level through all panels). 

If the string comprised just the shaded panel, then the MPP would be at a lower current and voltage than for an unshaded panel.  If the string comprised 2 panels, one shaded and one unshaded, then the resulting MPP is likely to be at a V and I that is between the two individual panel MPP's (if the inverter can resolve the best V-I MPP, and not lock on to a local MPP related to one or other of the 2 panels).  If the string comprises more and more unshaded panels, then the MPP moves more to a V,I operating point that is closer to just if it had unshaded panels, but modified such that the total series string voltage is lower by an amount related to the section of shaded cells that are diode bypassed (eg. typically circa 12V or multiples).

[fourfathom]

It seems to me that discussing MPPT is likely confusing the OP.  The question was about shaded panels in a series string, and to understand what is going on I suggest that the OP just look at the panel characteristics (the basic V-I curve) and then consider the V-I curve of a string of panels with and without partial shading.  At this point one can calculate the maximum power point of the full string to see what a MPPT controller might do with it.

Note that the panel is not a constant-power device, but is much closer to a voltage-limited constant-current source. The V-I curves will show this.

[NiHaoMike]

I wonder if anyone has tried to implement a sub string MPPT using MOSFET switching, with a single series inductor for the whole string. Doing it per panel should be relatively easy, it could also be done with multiple MPPT zones per panel or at an extreme, per cell using MOSFETs etched right onto the back side of the silicon.

An open hardware implementation of the per panel type would be interesting.

[ledtester]

This happens because solar panels are simply current sources with a zenner diode in parallel.

FesZ Electronics just did a video on this last week:

Modeling Photovoltaic Cells - Theory 1/2 - FesZ Electronics
https://youtu.be/uV_z1ptufa4

From what I have read, if you have a string of identical panels in series and one of the panels becomes shaded, the power output of the whole string drops to match the shaded panel. ...

At about 10:15 he talks about this phenomena.

[fourfathom]

From what I have read, if you have a string of identical panels in series and one of the panels becomes shaded, the power output of the whole string drops to match the shaded panel. ...

The *current* in the string drops to whatever the shaded panel is delivering.  Since in practice the panel voltage stays fairly constant, shaded or not, the available power also drops proportionally.

[mcbota]

Thanks for all your answer. This is the situation in this morning, after the drop of the voltage. Normally it has to be above 400 V and around/under 1A, at this hour, on that string, being 14 panels. And this voltage and current stays for about two-three hours.
Later edit: the second picture is made after that period  of time passed and the voltage became normal on that string.

[Siwastaja]

Hi all,

I've been reading up about solar, and one thing confuses me (well lots of things do, but you have to start somewhere!)

From what I have read, if you have a string of identical panels in series and one of the panels becomes shaded, the power output of the whole string drops to match the shaded panel.

Obviously this is confusing, because it is a total and utter made up lie. It does not work that way. For example, I have a string of 10 panels where I get 1 to 2 completely shaded panels in late afternoon. No such effect.

Sometimes it is just people who have no idea what they are talking about trying to be "helpful" but actually doing harm. Sometimes it's a microinverter or "optimizer" FUD BS market guy.

I'm surprised such bullcrap is seen in this thread, too:

This happens

No it won't, so the rest of the post is BS too.

How can people be so oblivious to such commonplace urban myth so that they keep writing responses happily assuming it was true, instead of shooting it down as the very first thing?

In reality, shaded panel cannot produce the same current, that much is true. This causes the voltage to change, look at the U-I curve set of the panel. The voltage tries to go negative, and current starts flowing through the bypass diodes installed in the junction box.

As a result, shaded panel production goes to zero, but rest of the panels keep producing optimally, assuming non-broken MPPT algorithm.

Actual panels usually consist of 3 "sub-panels" (with 3 bypass diodes, too), so that if you partially shade 1/3rd of the panel, only that part goes to zero, and rest of the panel keeps optimal production.

This is also why microinverters and "optimizers" rarely make economic sense: if you have say 10 * 300W panels, each producing 300W for 3000W total, and you partially shade one of them to production potential of 30W, in a microinverter system you could have 9*300W + 1*30W, and with string inverter you could have 9*300W + 1*0W. It is simple as that. However, string inverters tend to have a few %-point better efficiency than micros, this evens out the difference in typical installation where there is only very minor shading. Microinverters are recommendable when there is a lot of shading, sometimes half of the panels, or the panels are not in the same plane but point in whatever directions.

[Marco]

I wonder if anyone has tried to implement a sub string MPPT using MOSFET switching, with a single series inductor for the whole string.

Can only work if you give each panel a time slot. Seems a lot of trouble to save some inductors.

[NiHaoMike]

Can only work if you give each panel a time slot. Seems a lot of trouble to save some inductors.

Conceptually, take a bunch of buck converters (without output capacitors and with independent inputs) with the outputs in series and a single large inductor would be the same as many smaller inductors. At the extreme of per cell MPPT, having one inductor per panel (or one inductor per string) would be a substantial savings over each of the 70 or so cells in a modern panel having its own inductor.

Where things get interesting would be developing an algorithm for each of the MPPT switchers to sense the ripple and try to minimize it. Probably the easiest would be to have the charge controller provide a timing reference and each switcher to slowly vary its phase shift at a pseudorandom rate until it settles down to a minimum ripple level.

Come to think of it, a per cell MPPT wouldn't even a proper MPPT algorithm, as it could detect the cell temperature and light level to figure out what the Vmp should be.

[Berni]

Can only work if you give each panel a time slot. Seems a lot of trouble to save some inductors.

Conceptually, take a bunch of buck converters (without output capacitors and with independent inputs) with the outputs in series and a single large inductor would be the same as many smaller inductors. At the extreme of per cell MPPT, having one inductor per panel (or one inductor per string) would be a substantial savings over each of the 70 or so cells in a modern panel having its own inductor.

Where things get interesting would be developing an algorithm for each of the MPPT switchers to sense the ripple and try to minimize it. Probably the easiest would be to have the charge controller provide a timing reference and each switcher to slowly vary its phase shift at a pseudorandom rate until it settles down to a minimum ripple level.

Come to think of it, a per cell MPPT wouldn't even a proper MPPT algorithm, as it could detect the cell temperature and light level to figure out what the Vmp should be.

You would need an inductor at each panel if you wanted to MPPT each panel individualy.

What you are trying to regulate is not the voltage at the end of the string. It is the voltage across the actual solar panel. If you PWM the panel onto the series string you get either full current or 0 current trough the panel. This would give you an average of the two extremes in the panels IV characteristic. What you actually want to do is hold the operating point at the maximum power level. So for this you would need an inductor at the panel so that on one cycle the panel is charging the inductor with energy, then once switched over the inductor and panel are sending energy back out to the string. Changing the PWM duty cycle moves the inductors operating point up and down and so also rises and lowers the current trough the particular panel. Do this for every panel individually and each string will be operating at ideal current and voltage, the series string has the current flowing trough all of it, just that shaded panels are adding less voltage to the series string.

This is what these solar optimizer modules are. Here is the insides of a SolarEdge brand one from a presentation:


Here is the full presentation. It explains how the optimization works and a bit how the communication to the modules works:
https://www.glavaenergycenter.se/wp-content/uploads/2018/11/solaredge.pdf

[Siwastaja]

The problem with both "optimizers" and microinverters is that they still increase complexity and cost too much.

Most installations that do use them, only use them because of serious misconceptions and false data they are marketed with. Of course, there are some legit cases, too.

I do believe that microinverters will eventually win, but that will be after more rigorous integration, cost reduction, and reliability improvements. What I really mean, inverter becomes part of the panel, so you just basically have a sleek panel with AC output.

It's worth noting how important the suitable, weather-proof DC power plug (MC4 and similar) was to the proliferation of solar installations. We also need the same to happen with AC power. These things need to happen:

* Microinverter is integrated with the panel,
* Output is with weather-proof, safe AC plug anyone can plug in (no electrician needed legally),
* Pre-made wiring harnesses are made in different shapes, so that electrician only does one connection in house fuse box.

I hope this happens soon. In the meantime, good old string inverter seems the best option, or maybe microinverters in special cases such as panels all over the places in random directions, severe shading, etc. Or very small installations where you only have a few panels. Modern cheap string inverters have evolved and support two strings (2 x MPPT) which reduces the gap to microinverters in partially shaded installations.

In any case, this is a completely different discussion, and more like micro-optimization. For the OP, it is most important to realize they have been fed completely wrong information. In actual reality, string inverters, microinverters and power optimizers usually all work well, and differences in production is small, except in some pathological cases. The problem with pathological cases is, if you are having so much shading, total output will be low in any case, and hence spending more money for the system is even worse. Best course of action would be to find a better location. The problem is psychological: you like PV a lot but don't own a suitable spot. In this case, you could mentally or even financially support others to do the installs in optimum locations.

[fourfathom]

From what I have read, if you have a string of identical panels in series and one of the panels becomes shaded, the power output of the whole string drops to match the shaded panel. ...

The *current* in the string drops to whatever the shaded panel is delivering.  Since in practice the panel voltage stays fairly constant, shaded or not, the available power also drops proportionally.

As Siwastaja states, bypass diodes can "bypass" a shaded panel, or a shaded section of a panel (this depends no how the bypass diodes are configured).  I should have mentioned that.  How well this works depends on the number of panels in your string, and I suppose your controller. 

I am used to boat installations where we typically have 2/3/4 panels and the voltage reduction when a panel section is bypassed is a significant fraction of the full string voltage.  On my sailboat I have three "12V" panels, with per-panel MPPT controllers.  The shadowing problem on a sailboat is so bad that connecting the panels in series would not be a good idea, and the wire lengths are relatively short so there's one less compelling reason to go serial.  Home rooftop installations often have many panels in series, and shadowing typically isn't as severe as on a sailboat.

But I still think that one should first understand the behavior of a solar cell, then a panel (without diodes), then a string of panels.  *Then* see what adding bypass diodes will do.