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

[NiHaoMike]

Tho in practice this means that you are putting sharp high frequency square waves on the series string. In a typical installation this string has quite a large loop area as it snakes around panel to panel with the return then taking the shortest route (rather than retrace the string back along the same path). This means that we might not even need an inductor on the end since the string itself is likely going to have a self inductance that comes close to the miliHenry range. However making this pass EMI regulations is another thing. Just laying out a small low power DC/DC converter in a wrong enough way can make you fail EMI compliance tests pretty badly. So sensing square waves in such a long string might make it transmit more RF power than a WiFi router.

Some filtering would probably be necessary to limit the slew rate, but it would be much smaller than the smoothing inductor since it only needs to reduce the radiated EMI (30MHz and up) to acceptable levels.

[fourfathom]

Some filtering would probably be necessary to limit the slew rate, but it would be much smaller than the smoothing inductor since it only needs to reduce the radiated EMI (30MHz and up) to acceptable levels.

The FCC may not care about radiated EMI below 30 MHz, but you're going to have radio amateurs showing up with pitchforks and torches. 

[Marco]

Very strong "may happen". And this isn't relevant because global MPPT algorithm is needed in any case. All modern string inverters have it, but those optimizers need it too because they are also dealing with series strings like it or not

They might have some way to regulate the current down to make the string voltage more efficient for the inverter and to allow bypass of a subset of the optimizer DC/DC converters. In theory though simply for putting maximum power on the string DC link a nearly constant current output works just fine. Can output arbitrary power.

That's not really MPTT any more though, but minimum converter losses tracking.

[Berni]

Some filtering would probably be necessary to limit the slew rate, but it would be much smaller than the smoothing inductor since it only needs to reduce the radiated EMI (30MHz and up) to acceptable levels.

Not sure how it is with FCC. The European CE standard also includes conducted EMI that includes frequencies below 1MHz.

Very strong "may happen". And this isn't relevant because global MPPT algorithm is needed in any case. All modern string inverters have it, but those optimizers need it too because they are also dealing with series strings like it or not

They might have some way to regulate the current down to make the string voltage more efficient for the inverter and to allow bypass of a subset of the optimizer DC/DC converters. In theory though simply for putting maximum power on the string DC link a nearly constant current output works just fine. Can output arbitrary power.

That's not really MPTT any more though, but minimum converter losses tracking.

Well yeah each optimizer module becomes it's own little mini MPPT.

The only trick is that the MPPT can find the optimum operating point for each panel individually, so slightly shaded panels can still extract a good deal of power.

If the MPPT is smart enough to find a global or local maximum is another thing. But the peak power point is going to be more prominently seen on a single panel versus a large array in series. On a large array the panels are all going to have a slightly different optimum operating point so the resulting IV curve is going to be more fuzzy as a result. Sure the panel itself is also made up from multiple cells internally, but they do come from the same batch and are physically close together so they are more likely to act similarly. So all in all the IV curve of a single panel should be sharper and less wavey, so easier for the algorithm to find the right spot.

The big boy inverter on the end is basically a 3 channel D class audio amplifier that 'plays' 50/60Hz into the mains network. So it doesn't care about the precise voltage it is getting as long as it is within its operating range (Tho it likely is slightly more efficient around some ideal voltage)

[Marco]

I suspect if a lot of panels have near identical max power point, the optimizers bypass themselves to let the inverter determine max power point. Just to avoid double converter losses.

[tszaboo]

Yes, that attachment is an outright lie. It is not some kind of exaggeration or "optimal" case, it's total bogus and would belong to the "dodgy technology" section. There is no way I could buy any product from this kind of company. In fact, this is a scam that should be handled by the authorities.

It's not that bad.  The case they show is a simplistic version of what happens may happen if the string inverter doesn't have global MPPT, the global MPPT is not enabled or as in your example, the loss of one or more panels results in the string voltage going below the MPPT threshold voltage.  But in their simple example, even if the global MPPT was available and working, the output would be 0 + 100 +100 or 2/3 full capacity, whereas the per-panel optimizers would make that 50 + 100 + 100 or 5/6 full capacity.

Actually, this is the answer. If an inverter tracks MPPT, and it uses simplistic approach, it might drop the entire string to 50% if one panel is shaded. Normally new inverters are smarter than this, and they have global MPPT, scan the panel array for optimal and work on that range, but some early inverters didn't do this. And it might be marketing, but otherwise the technology works.
I have solaredge on the roof for the past year, and I'm very happy with it. It is more robust for panel failure, doesn't require global DC disconnector, and has a cheaper inverter than the more advanced 2 string inverters. So the total system cost is not that much higher. And unlike microinverters, the optimizers cost only like 40-50 EUR or so, and they can use the full power of the panels, unlike Enphase. Right now they make sense.
On the other hand, I'm seeing more and more microinverters with something like 4 MPPT inputs. They seem to be the price/performance champions now, or the near future.

[Marco]

BTW on micro inverters needing more cabling, there have been commercial series connected micro inverters, with the whole string adding up to mains voltage AC.

[Berni]

I suspect if a lot of panels have near identical max power point, the optimizers bypass themselves to let the inverter determine max power point. Just to avoid double converter losses.

Yep it does make sense to do it like that.

The inverter on the end has to constantly follow the mains voltage on its output, so it rarely operates in its ideal PWM duty anyway. But the optimizer can try to run there internal buckboost at near 100% (or even just stay on to bypass it) since that's where buck converters are the most efficient. It only has to do a lot of work on the shaded panels where it has to step down the input voltage in order to get the extra current for feeding the series chain at whatever current it wants to run according to the fully sunny panels. (And even then the input/output ratio is not huge, so it is easier to make a buck converter efficient at it)

The reason that these more complex ways make sense is because the panels are the most expensive part, so it might be beneficial to throw more electronics at the problem to make sure every drop of power is extracted from the panels. Maybe prices of solar panels will fall enough to a point where doing this does not make sense anymore.

[Siwastaja]

I don't understand why people try to somehow justify the wrong claims by trying to come up with some rare edge case in which their claims are true.

But you've provided a specific example of this very thing happening to you!

Wat? No! I have shown an example where shading 12 panels out of 16 (design # of panels for the inverter) causes string inverter to fall below minimum voltage. Or more exactly, in my case ignoring the misdimensioning of the inverter, shading 6 out of 10. I may add that once two panels are cleared, with 4/10 shaded, full production commences no problem.

They have shown two examples, where shading of (A) 1 panel out of 3, (B) 1 panel out of 6 fails the whole string.

12/16 = 75%
6/10 = 60%
1/3 = 33%
1/6 = 16.7%

Completely and utterly different.

Their optimizer will likely also fall below minimum voltage if you shade 45 of the 60 cells.

This does matter. Their central claim is: shading just one panel brings the whole string down. This is wrong, in reality shading most of the string brings the few remaining panels down. A huge difference because installations with one-two panels getting shaded for an hour-two makes total financial sense. But if you have most of your string shaded for a large part of the day, you should not be installing the damn thing to begin with. Especially not with their optimizers because it will only increase the cost and the system will still produce poorly.

[Siwastaja]

doesn't require global DC disconnector, and has a cheaper inverter than the more advanced 2 string inverters.

1) are you sure it doesn't require DC disconnector? I have not read the actual standards, but seeing the optimizers are non-isolated and suspect to failing short (or having configuration error, outputting a voltage), I would definitely want to have a DC disconnector to protect against arcing during maintenance work. In other words, I would not trust a software-controlled semiconductor switch in HVDC safety.

2) how much did the inverter cost? You might not realize how cheap string inverters are today (have been for years), they are manufactured in huge masses by companies like Huawei. 2-string feature seems to now be the feature in the lowest-cost string inverters. You can get a 3kW inverters around 700EUR or so, depending on where you live and taxes etc.

I am aware of the engineer reasoning that a PowerEdge inverter can be made simpler, and it theoretically can be of lower cost, but for that to become reality there has to be enough competition (so mix-matching different manufacturers must be possible, I don't know if it is) and large enough production quantities.

Also DC disconnection switch is often a feature of string inverter. At least mine has the switch built-in.

[tszaboo]

doesn't require global DC disconnector, and has a cheaper inverter than the more advanced 2 string inverters.

1) are you sure it doesn't require DC disconnector? I have not read the actual standards, but seeing the optimizers are non-isolated and suspect to failing short (or having configuration error, outputting a voltage), I would definitely want to have a DC disconnector to protect against arcing during maintenance work. In other words, I would not trust a software-controlled semiconductor switch in HVDC safety.

2) how much did the inverter cost? You might not realize how cheap string inverters are today (have been for years), they are manufactured in huge masses by companies like Huawei. 2-string feature seems to now be the feature in the lowest-cost string inverters. You can get a 3kW inverters around 700EUR or so, depending on where you live and taxes etc.

I am aware of the engineer reasoning that a PowerEdge inverter can be made simpler, and it theoretically can be of lower cost, but for that to become reality there has to be enough competition (so mix-matching different manufacturers must be possible, I don't know if it is) and large enough production quantities.

Also DC disconnection switch is often a feature of string inverter. At least mine has the switch built-in.

1) They had it up until a few years back, when they certified their system for safety. If the cables are disconnected, it drops the voltage to 1V/module.
2) I have the 3500W inverter, which was 750EUR ex BTW. I looked at some prices now, SMA sunnyboy 3000 costs 1100 EUR, Huawei 940 EUR. The interesting part is, the 25KW Solaredge is only 1700 EUR. Maybe its covid prices.
I don't think Solaredge will be the cheapest, but for me it was worth to spend extra ~5% on the system to get extra MPPT per panel, and the data. I get shading on the panels, plus the system is more flexible now.
Mixing panels is definitely possible, you have like 3-4 different optimizers that can work together, they have different wattage.
And I also get where you are coming from, at the beginning I was also doubting if it makes sense.

[Siwastaja]

Well, for +5% total cost it almost makes sense for just the monitoring alone.

Reliability is hit-or-miss. Microinverters have some reputation of failing, which is understandable when you mount electronics in worst conditions (specifically heat) and increase the number of modules by 10x compared to a string inverter, so also increase failures by 10x. This can only be solved by very good engineering, and it is of course doable.

You are looking at inflated prices now. 2.5-3kW inverters have ran as low as 500-600EUR at some point about 2 years ago. Panels have inflated a lot, too, for obvious reasons. The only comfort really is that price of energy has skyrocketed even more so PV is still more appealing than ever.

[NiHaoMike]

Not sure how it is with FCC. The European CE standard also includes conducted EMI that includes frequencies below 1MHz.

That doesn't apply for the DC connection since the panels only connect to a charge controller or inverter specifically designed for that arrangement, no different than a PWM motor controller. The conducted EMI applies to where the system connects to the public grid and/or to third party devices that might not be EMI hardened.

[Berni]

Not sure how it is with FCC. The European CE standard also includes conducted EMI that includes frequencies below 1MHz.

That doesn't apply for the DC connection since the panels only connect to a charge controller or inverter specifically designed for that arrangement, no different than a PWM motor controller. The conducted EMI applies to where the system connects to the public grid and/or to third party devices that might not be EMI hardened.

Even then if you wiggle one connection on a device it tends to end up on the other side as common mode noise instead. Both common and differential mode conducted noise is measured. The large array wiggling up and down has a lot of parasitic capacitance to its environment and so is capable of making a sizable common mode current. In the end it is not that hard to stick a few microhenries of inductance into the optimizer module, if it already has a MCU and beefy power electronics in it.

I don't know what standards exactly solar installations have to abide by, but they do seam to be more stringent that regular household devices like a TV.

Well, for +5% total cost it almost makes sense for just the monitoring alone.

Reliability is hit-or-miss. Microinverters have some reputation of failing, which is understandable when you mount electronics in worst conditions (specifically heat) and increase the number of modules by 10x compared to a string inverter, so also increase failures by 10x. This can only be solved by very good engineering, and it is of course doable.

You are looking at inflated prices now. 2.5-3kW inverters have ran as low as 500-600EUR at some point about 2 years ago. Panels have inflated a lot, too, for obvious reasons. The only comfort really is that price of energy has skyrocketed even more so PV is still more appealing than ever.

Yep the monitoring part is about ease of maintenance later on. There is no need to check up on the system at all, as long as all the panels are performing well according to monitoring that means the system is working. Even if a panel is knocked out by a fault the rest of the panels keep working.

Fact is that the inverter electronics are only a small part of the total system cost of a solar array. Most of the cost is in the panels themselves, so if you can use some extra electronics to squeeze more energy out of the existing panels you can actually save money since you can get the same amount of power from fewer panels.

We do have a 13kW SolarEdge inverter and optimizer system here. These SolarEdge boxes is simply what most solar installation companies offer here and they work fine. I am sure you can save some money on a cheaper inverter, but you are going to be shaving singles of a percent off the total system cost with it.

Microinverters are a perfectly valid solution. For small arrays they are likely the best solution even. But once you scale up to arrays with >30 panels they become a lot of unnecessarily duplicated electronics and mains wiring. Reliability also goes down indeed since you got 10s of little boxes doing pretty demanding power conversion in the hottest spot on a building while having to be cheap. Compared to a single big inverter that can put more money towards quality components (since you only need 1) while being placed in a nice cool dry shaded spot while having cooling fans to help keep it comfortable.

[Siwastaja]

For actual information about market shares, see the figure at
https://www.greentechmedia.com/articles/read/woodmac-last-year-saw-record-highs-for-the-inverter-market

Also: https://www.solarpowerworldonline.com/wp-content/uploads/2019/12/image004.png

Microinverters are totally negligible, but optimizers seem to have quite some market share, especially in US residential installations it seems. They are still small in the large scale of things, and even in households installations, in minority. The point I am making, if around you "everyone" is installing them, this is only a local exception to the general trend. For example, here no one has even heard about those things. Microinverters are at least a known thing.

[Berni]

Yep it seams like only SolarEdge is pushing heavily into these local optimizer modules.

This might be because they own a whole bunch of patents on the topic, so they might have a team of lawyers hunting down anyone who touches it:
https://patents.google.com/patent/US8587151?oq=9088178

Still regular oldschool series string inverters are perfectly fine. In the right cases with a well balanced array and limited shading they perform just as well.

Also not saying that SolarEdge is absolutely everywhere, just in this corner of Europe they seam to be pretty popular in residential installations for whatever reason. Maybe they have some magical EU seal of approval that some others don't. I have no idea

[NiHaoMike]

Even then if you wiggle one connection on a device it tends to end up on the other side as common mode noise instead. Both common and differential mode conducted noise is measured. The large array wiggling up and down has a lot of parasitic capacitance to its environment and so is capable of making a sizable common mode current. In the end it is not that hard to stick a few microhenries of inductance into the optimizer module, if it already has a MCU and beefy power electronics in it.

Tie the inverter ground to the panel ground (already required by code), then the common mode current just circulates in a loop between the inverter and panels, never reaching the mains connection.

[bdunham7]

Microinverters are totally negligible, but optimizers seem to have quite some market share, especially in US residential installations it seems. They are still small in the large scale of things, and even in households installations, in minority. The point I am making, if around you "everyone" is installing them, this is only a local exception to the general trend. For example, here no one has even heard about those things. Microinverters are at least a known thing.

I don't want to make this an interminable debate, but if you consider residential and small commercial (the only places any of this is realistically in play) then in the US, SolarEdge and Enphase have the majority of the market.  SolarEdge's share has increased dramatically in recent years and Enphase is second, but Enphase has pursued profits and shifted to a more expensive product with advanced features like battery backup, off-grid and so forth.  What is available and cost-effective in your market may vary, and cost is the biggest factor driving most of these decisions--as it should be.  But your market is small and not very sunny, so perhaps they aren't trying to compete there.  Enphase makes and ships more inverter capacity every quarter than the entire installed solar capacity of Finland, so I think they'd dispute your 'totally negligible' comment.  So your point about a 'local exception' seems to be correct. 

[Siwastaja]

I don't want to make this an interminable debate, but if you consider residential and small commercial (the only places any of this is realistically in play) then in the US, SolarEdge and Enphase have the majority of the market.

Still, USA is a huge market but does not dominate in world-wide installations; otherwise the numbers would look different. Minority share in global market can only be explained by largest Asian markets not buying SolarEdge and Enphase nearly at all.

'Murica however, this is where households have money at their disposal (big middle class), and where marketing works really well. For the similar reason cars must be large, PV systems must have some "premium" in them.

Compare this to Finland where everything is expensive, you pay premium for crap and premium^2 for premium, middle class is taxed to death and are only able to support their lifestyles by getting more in debt. It is going to be the cheapest panels with the cheapest inverter.

And there's nothing wrong in that, PV is most often money limited after all, so I think $/kWh (total cost of ownership per lifetime production) is the best metric. Also string inverters by the largest global players have the best track record for reliability, so there is no such compromise. It's like driving a Toyota instead of a Porsche, former is cheaper and better in getting reliably from A to B.

And this is the thing: if your local PV installer works with SolarEdge and this is all they sell, they will give you a competitive price. If you ask them for other choices, they are either hesitant to install them at all, or quote similar price for lower-tier system. But it also works the other way around, if you ask if you can get a SolarEdge system here, price will be significantly higher because not many sell them.

Ultimately, the problem with products like optimizers and microinverters is the economy of scales. It seems to me power optimizers have reached the point of being affordable enough that they are not necessarily a huge net loss. Have they reached this position by pushing the scale up by false marketing claims? Clearly yes. Is this morally right or wrong? I don't know, you decide.

But one thing is, I think, undisputable: false marketing claims have done definite harm to the knowledge about how PV systems work, this very thread being the proof. I know the "one panel brings whole string down" misconception is ages old and not invented by SolarEdge, but they sure have been feeding it.

And for me, truth and understanding really matters. Those who fight against it only for their own financial gains, will never see business with me.

[dirtcooker]

Panel level optimizers only marginally improve performance in most cases. However, in USA at least, NEC code requires module level shutdown for panels installed on a roof. Since you are required to purchase MLPE, you may as well buy modules that do MPPT as well. Hi voltage dc strings are also less safe due to fire and shock hazards. NEC code requires arc-fault interrupters for systems over 80vdc. Converting to 240 vac at each module avoids this requirement.
Longi and others have teamed with Enphase for their ac panels, which I think will be the dominant trend in the industry. I wonder how long that relationship will last. They don't really need Enphase for that. The technology is dead easy, so it is doubtful chinese panel makers will share profits for something they can easily do in house.

[NiHaoMike]

Panel level optimizers only marginally improve performance in most cases. However, in USA at least, NEC code requires module level shutdown for panels installed on a roof. Since you are required to purchase MLPE, you may as well buy modules that do MPPT as well.

Couldn't the same be achieved with mechanical relays? If an arc fault is detected, turn off all the relays.

[nowlan]

Micro and String Inverters in the Shade - Testing the True Impact of Shade on Solar Panels


Food for thought?
note: does appear to be Fronius marketing.

[Siwastaja]

Thanks for posting the video, the result is completely as expected - a small localized shadow like a chimney is totally meaningless for total production in either per-panel optimized system, or string inverter system, because it only affects the one panel in either case. Contrary to the marketing claims (which always have emphasis on small shadows, based on the lie about string inverter going to zero or half or whatever thanks to that small shadow), microinverters / optimizers would give the best benefit over string inverter when panels are aimed at random directions so that they run at significantly different power points all the time.

[Alti]

Contrary to the marketing claims (which always have emphasis on small shadows, based on the lie about string inverter going to zero or half or whatever thanks to that small shadow),

AFAIK the furthest going shading claim was 60% shade/40% Pstring by SolarEdge.
As for the attribution of the 100% shade/0% Pstring claim, this goes to OP and OP only, no reference for marketing involved provided so far:

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.(..)