Flat roof installation in Netherlands - considering glass-glass, bifacial, manufacturer, dimensions

[NeoX]

So, I'm now at a stage where I'm considering SolarEdge string inverter again over Enphase microinverters. I also read this very interesting topic by @FilterGuy who did an analysis on MPPT controller in series considering bypass diodes and shading. While the topic in the end wasn't conclusive, especially as most panels on offer today have half-cut cells which means the panel is typically split into two, it did make me wonder if I should opt for a string inverter system.

To add some context to my situation:

• An Enphase system roughly adds about 3.3% to the price of the system (comparing top SolarEdge Home Hub to top Enphase microinverter), but can also be as much as 18% if comparing to a more basic SolarEdge system).
• To get as many panels on my roof as possible, I'm going to go for an east-west system, which I understand optimisers will resolve - no need for a dual MPPT (SolarEdge doesn't have this).
• There is a chimney on my roof which will cause shade on one or two panels.
• There is a small "wall" about 15cm high around the edge of my roof.

My concern with Enphase is that it seems to be a bit of a fad with a lot of marketing hype born more out of the shading issues with string inverters before MPPT and MLPE software/hardware caught up. I like my system to be as simple as possible (less points of failure).

My aim is to generate the maximum power output possible with minimal future maintenance (replacing/repairing microinverters under warranty whack-a-mole style is not my idea of efficiency). I understand string inverters are more efficient on the AC/DC conversion.

My other concern is that a string inverter must be attuned to the size of the PV array, but there is some scope for oversizing upfront as long as the DC/AC sizing ratio is no more than 73% (3-phase) according to here. In my system, it's unlikely I'll add more than two additional panels in future, and I may upgrade panels to higher Wp panels if any need replacing later, but this is where I really need clarification.

If I'm running a PV array producing 9460 Wp, my supplier recommends an SolarEdge SE10K inverter. This allows for a maximum of 15,000W input, so a lot more than my array produces, and an AC output of 10,000 W total (3333 W per phase). The DC/AC sizing ratio in this case would be 9460/10000 = 94.6%. However, if I go for an SE5K inverter, which has an AC output of 5000 W total, then I would be oversizing the system with a DC/AC sizing ratio of 9460/5000 = 189%.

I had a chat with ChatGPT to get my head around this (search for "I am considering a string inverter" at this link for where this part of the conversation starts), and it said that oversizing allows for greater potential energy capture when my panels are not operating at their maximum capacity (e.g., cloudy days), but that I could lose out on energy capture due to curtailment during peak conditions. So, this complicates the choice further. The SE5K is a lot cheaper than the SE10K, but there's an in-between option with the SE8K where the DC/AC sizing ratio would be 9460/8000 = 118%.

The Netherlands is often cloudy, but it has lots of bright sunny days in the summer in the southern parts where I live. So, I just don't know what would be the most optimal choice. And this is leading me back towards microinverters again so I don't have to think about it.

Any thoughts or insight on the above would be greatly appreciated.

 

[Hedges]

Research Solar Edge. I started a thread about an installer's blog; he and others have reported higher failure rates than other tier-1 brands. But some people seem to have no problem. Could be different models and different environments. They push optimizers, some of us question the value.

SMA? Is Eindhoven enough in the Netherlands for you?
Check their website to see if retailers or installers are available, vs. just this one large project.

Eindhoven, Netherlands | SMA Solar

www.sma.de

 

[houseofancients]

I've recently been getting sold SolarEdge string inverters a lot more by some suppliers which enticed me to have another look at them based on the MPPT and MLPE concepts which mitigate shading and allow per-panel monitoring.

---

I've updated the link to this in my earlier message as the original one was dead. The diagram at the link relates to microinverters, but seems to a "combiner" to allow all the power from the microinverters to be channelled into one or more batteries. I think the reference to "IQ8 series microinverters" just means they're from the IQ8 series, not that anything is in series as opposed to parallel.

SMA doesn't seem to be a sold brand in the Netherlands (never seen it offered), but the SolarEdge Home Hub is offered, and this is a hybrid inverter. In fact, one supplier I'm speaking to spoke to SolarEdge on my behalf today, and this is what they recommend to futureproof my system.

---

Tbf, a few suppliers are saying similar - I'm working with 18 suppliers at the moment (a lot, I know, but I'm trying to get a good overall picture). The supplier in question would make more money by selling me a battery, but he assured me he'd rather sell me an economical system right now than one that maximises his revenue. Similarly, I find many suppliers in the Netherlands are resistant when I ask for more panels to squeeze on every bit of space on the roof. They'd rather sell me less panels to minimise shading from the roof edging and the chimney. I guess they're not as hard-selling as in other countries (the Dutch are known for spending as little money as possible, but that also seems to be the case with companies selling to customers in a lot of cases ? - it seems to be an ingrained cultural thing, which means it's annoying for me when I'm trying to spend more money!).

Noted on the energy companies. I'm aware of Vandebron doing it, but I'll see how it goes. I'd rather just get a battery the moment it becomes economical for reasons such as this, but note what @Hedges said here:

---

Not different. Completely agree. Doesn't seem to be a downside in deferring if I get a futureproof system such as the SolarEdge Home Hub mentioned above. There is certainly no issue with power cuts in the Netherlands (the infrastructure is second-to-none), but being independent of the grid given the constantly changing government policies around energy and taxation might be a useful peace of mind to have. My house is powered by the battery, and nothing goes through the smart meter unless the battery is depleted.

You do know I live in the Netherlands right ?

Had many many supplier telling me the same...
They are banking on their margins, and hoping the "eerste Kamer" won't scrap the "saldering"..

Just mind you, you're not buying a system for the next few years, but this is the kind of investment that should pay itself back within 7 years, and I guarantee you that "salderen" as we know it now will be gone long by then..
It is simply putting too big of a dent in the powerco's bottom line

 

[zanydroid]

Did you price out say Hoymiles instead of Enphase? 4 port 2 MPPT like HM-1500. You might find that these destroy Enphase and SolarEdge on price.

There’s two flavors of HMS2000, one with with 2 MPPT instead of 4.
Also note Enphase batteries are probably technologically jankier than a high end hybrid like Victron or Deye for comparable price.

Can’t remember whether you have shading.

TBH I don’t think there’s so many solutions out there that you need to talk to 20 people. Sounds like a ton of work to talk to probably less sophisticated sales people than forum. Unless your install is so huge they assign their best salesperson to you. Or you work better talking through a person than online, enough to compensate for their bias/lower knowledge level than the forum. TBH sounds like they’re blowing smoke at you and you don’t have enough skepticism antibodies yet

Or you don’t have the capacity to do the engineering/design that forum people can do, which is somewhat sort of needed for string in some countries because installers would rather make their lives easier with MLPE. Even with that you could still go with a cheaper microinverter setup.

 

[zanydroid]

ChatGPT is not good for researching something as fast changing as solar

Overpaneling and clipping is pretty well understood in some common cases.

If you over panel with diversity in orientation (EG parallel a W string and a E string, it is more cost effective to do it with string vs microinverters because the capacity is shared across full string in the string inverter case.

If you overpanel with everything in symmetric conditions then they are similar.

 

[zanydroid]

If I was you I would flip the approach.

Figure out what kind of systems are viable for you to either (1) DIY install, (2) find an installer. This would then generate the set of viable options. If there are options you'd like to install and are viable from regulatory angle but not from viable from a able to find an installer angle, then find out whether it's worth trying to hustle harder to find an installer, or just give up and take your second top preference.

Seize the initiative from the distributors rather than sort of drifting around with them in the driver's seat.

 

[SeaGal]

I had a chat with ChatGPT to get my head around this (search for "I am considering a string inverter" at this link for where this part of the conversation starts)

Any thoughts or insight on the above would be greatly appreciated.

With all due respect... you cannot be serious? What a load of b**lox that ChatGPT regurgitates " .. it's full of "depends on the specific configuration", "can vary widely depending on several factors,", "can vary based on various factors", "The specific outcome can depend on various factors"...

FFS (excuse my French) but just listen to the RI (Real Intelligence) of the likes of @houseofancients, @Hedges, @OffGridForGood, @zanydroid and others.

What has the world come to? ?‍

 

[zanydroid]

With all due respect... you cannot be serious? What a load of b**lox that ChatGPT regurgitates " .. it's full of "depends on the specific configuration", "can vary widely depending on several factors,", "can vary based on various factors", "The specific outcome can depend on various factors"...

FFS (excuse my French) but just listen to the RI (Real Intelligence) of the likes of @houseofancients, @Hedges, @OffGridForGood, @zanydroid and others.

What has the world come to? ?‍

TBH for my day job I do use ChatGPT to do quick reading of stuff I should be familiar with to get a starting point for research, and I also need to understand what it's good at for targeted applications. As a research tool when you already know something I think it's promising since it can bulk-read more stuff than you can bulk-read.

But I have no idea how one would use ChatGPT to productively learn when you know nothing.

Google does use a LLM now to try to summarize results, but presumably they do more prompt engineering/post-processing on the results (not that I really read that section of the results, so I don't know what the quality is).

 

[Hedges]

With all due respect... you cannot be serious? What a load of b**lox that ChatGPT regurgitates " .. it's full of "depends on the specific configuration", "can vary widely depending on several factors,", "can vary based on various factors", "The specific outcome can depend on various factors"...

FFS (excuse my French) but just listen to the RI (Real Intelligence) of the likes of @houseofancients, @Hedges, @OffGridForGood, @zanydroid and others.

What has the world come to? ?‍

It is probably smarter than the average salesman.

Chat GPT already behaves just like a typical person, getting angry and blithely lying, making things up. Plus has a few deeply entrenched beliefs, accepted as irrefutable fact.

 

[zanydroid]

It is probably smarter than the average salesman.

Chat GPT already behaves just like a typical person, getting angry and blithely lying, making things up. Plus has a few deeply entrenched beliefs, accepted as irrefutable fact.

ChatGPT is already better than asking people (IE students in gen ed classes) with low motivation to write essays and do research. ChatGPT may be dumb, but it has no motivation problems and it's willing to read a ton of crap (well to be precise it's already pre-read it during training) to repackage into an essay, and the essay is just tantalizingly on the useful side of useful vs garbage spectrum.

 

[NeoX]

With all due respect... you cannot be serious? What a load of b**lox that ChatGPT regurgitates " .. it's full of "depends on the specific configuration", "can vary widely depending on several factors,", "can vary based on various factors", "The specific outcome can depend on various factors"...

FFS (excuse my French) but just listen to the RI (Real Intelligence) of the likes of @houseofancients, @Hedges, @OffGridForGood, @zanydroid and others.

What has the world come to? ?‍

How rude.

Not sure if my reference to ChatGPT should result in an entire discussion about its value. I use an array of sources including ChatGPT, salespeople and installers from multiple suppliers, Google (leading to various blogs, articles, etc.), manufacturer websites and datasheets, simulation software (PVsyst), and forums. I know how to distil the information, so hating on ChatGPT because it uses certain phrases in between a mixture of accurate and inaccurate information is short-sighted to say the least. Did you even notice that I'm conversing with some of the names you mentioned in this very topic?

Stop getting so mad and emotional over this! You've contributed nothing to this topic so far, then you opt to bang on about ChatGPT because certain phrases butthurt you. The phrases you referenced are not even within the section of the chat I referred to, therefore completely irrelevant to this topic. If you can't be bothered to read my posts property, and you don't understand modern technology ("What has the world come to?"), and need to vent your frustration on forums, then please do that elsewhere instead of polluting my topic where I'm trying to have an intelligent discussion with other members (not you). I'm an AI software developer, so I probably know how to navigate and distil the information ChatGPT gives me infinitely more than you, hence how I questioned it at every turn until I got it to question itself and give me more accurate answers (which you clearly didn't even bother to notice because you were too eager to start keyboard jockeying). ?‍

I'm out of this thread now. If you disagree with SeaGal's behaviour, please also get out of the thread. I see no point in this topic carrying on.

 

[zanydroid]

All right I’m out of this thread.

The most cringe/dangerous use of LLM output I’ve seen on forums is electrical related where you don’t want guesswork on safety, so I and others are extra sensitive to that.

 

[NeoX]

All right I’m out of this thread.

The most cringe/dangerous use of LLM output I’ve seen on forums is electrical related where you don’t want guesswork on safety, so I and others are extra sensitive to that.

Pretty sure professional installers won't take my instructions on electrics that I got from ChatGPT or in any case, so calm yourself. But I also have electrical qualifications higher than yours, so don't worry. There's no guesswork here.

Btw, the landing page of ChatGPT says at the bottom, "ChatGPT can make mistakes. Verify important information." Pretty sure forums with emotionally-charged opinionated humans should have an even stronger disclaimer!

 

[SeaGal]

How rude.

Hands-up, I probably used stronger words than justified, but believe everyone should be wary of using AI for such situations. ?‍ Personally I find the trend to over-hype AI worrying, especially when used for technical purposes. My point, obviously put too strongly, was to listen to the real intelligence of people on here and not chatGPT. In your 'chat' chatGPT has given you incorrect and potentially dangerous (or at least financially wasteful) data - see my more detailed summary below.

Did you even notice that I'm conversing with some of the names you mentioned in this very topic?

Yes, which is why I referred to them.

You've contributed nothing to this topic so far, then you opt to bang on about ChatGPT because certain phrases butthurt you. The phrases you referenced are not even within the section of the chat I referred to, therefore completely irrelevant to this topic.

Excuse me? I was the first to respond to you in post #2, which you acknowledged in post #3.
The phrases I referenced were all extracted from the chat you linked to.
I don't have an issue with the phrases themselves, more the fact that chatGPT or any self-learning AI system has no inherent understanding or concept of the technical components of a solar system. Sure it can be used to quickly summarize wads of information from the internet as @zanydroid pointed out. But that is it. A summary of pre-written stuff, with no intelligence of what it is generating.

If you can't be bothered to read my posts property, and you don't understand modern technology...
I'm an AI software developer, so I probably know how to navigate and distil the information ChatGPT gives me infinitely more than you...

But I also have electrical qualifications higher than yours, so don't worry.

I'm sure you do have higher electrical qualifications as I only have a degree in electronic engineering... but I also have spent over 35 years in highly technical software development (yes, I am that old), so I doubt the adverb 'infinitely' is correct.

Anyway, to give some useful input to your overpanelling question...

Generally it makes sense for us with a lot of cloudy weather to overpanel as the gains (more production on cloudy days, or different orientations). I do the same with 6kWp of panels on a 3.6kW inverter.

There are 2 things to consider though when looking at your inverter and panel specifications;

a) what are your house needs? There is no point having only a 5kW inverter if you regularly have a house load of (say) 7kW.

b) you can overpanel with most inverters, but there is usually a maximum that the inverter will safely or correctly handle. My Solis, for example is a 3.6kW inverter (= AC output max), but it can accept 7kW of PV input. Any excess above the 3.6kW (if that is being used by house load) from the panels will be used to charge the batteries, up to a maximum charge rate of 3kW. Hence, the 'clipping' in peak day in summer, if sunny, is only a few hundred watts, until the batteries are charged up, of course.

Looking at the specification of the SE5K inverter, the datasheet states recommended max DC power of 7750W and specifically states "Limited to 155% DC oversizing"

I have highlighted in bold the statements chatGPT has stated which are blatantly wrong.

chatGPT has told you the following...
"Your interpretation of the sizing ratio for the SE5K inverter also aligns with the concept of oversizing. When you have an inverter with a lower AC output capacity (e.g., 5,000 W), you need to ensure that the total DC capacity of the panels is significantly higher (in your case, 10,000 W) to meet the minimum DC/AC sizing ratio guideline. This oversizing is considered beneficial because it allows for better energy capture during peak conditions."

then, when you questioned further.. chat GPT says...
"In your case:

• The SE5K inverter has a higher DC/AC sizing ratio of 189.2%. This means your solar panel array's DC capacity (9,460 Wp) significantly exceeds the inverter's AC output capacity (5,000 W), allowing for potential energy capture during peak moments.
• The SE10K inverter has a lower DC/AC sizing ratio of 94.6%. The solar panel array's DC capacity (9,460 Wp) is closer to the inverter's AC output capacity (10,000 W).

Both options are valid, and the choice between them depends on your specific goals and local conditions."

then when you specifically questioned...
YOU.... "But the inverter's capabilities are fixed. So, if there's a maximum amount of DC power that the inverter can handle before curtailment, then the capturing of more energy during peak conditions would never actually happen, right?"

chatGPT responded..
"You are correct. I apologize for any confusion. The inverter's capabilities are fixed, and once the maximum DC power input capacity is reached, any excess DC power will be curtailed, regardless of the DC/AC sizing ratio. In this context, having a higher DC/AC sizing ratio does not directly enable the capture of more energy during peak conditions if the inverter's DC input limit is already reached."

In all this to-ing and fro-ing, at no point has chatGPT referred to either the max DC power input of the inverter being 7750W, above that could cause mis-operation of the inverter or the fact that excess PV power over and above your AC output can be used to charge batteries.

IMHO this is dangerous. I rest my case.

 

[NeoX]

@SeaGal Apology accepted, but just some clarifications...

I'm sure you do have higher electrical qualifications as I only have a degree in electronic engineering... but I also have spent over 35 years in highly technical software development (yes, I am that old), so I doubt the adverb 'infinitely' is correct.

The University of Nottingham (the uni I went to) offers three similar courses: Electrical Engineering, Electronic Engineering, Electrical and Electronic Engineering. Each course includes modules in both electrical and electronic engineering, but the first two are more biased towards the title of the course. I did Electronic Engineering, but per above did modules in electrical engineering. I have also spent my entire career as a highly technical software developer working for electronics firms (e.g., Nokia) for the first part of my career followed by a series of top-tier investment banks (and still do). I have therefore attained charterships in both engineering and IT (CEng and CITP) which makes me more academically and professionally qualified than almost everyone I come across, PhDs aside.

But, I should also point out that my comment wasn't directed at you - it was directed at @zanydroid.

The phrases I referenced were all extracted from the chat you linked to.

I'll reiterate. I said, 'search for "I am considering a string inverter" at this link for where this part of the conversation starts.' All the phrases you mentioned are prior to the section I referenced. If I refer you to book X, page Y, you're not going to start telling me how you disagree with phrases in page Y - Z. Or maybe you are, to try to discredit page Y if page Y - Z is by the same author. But I don't think that was what happened here. You clicked the link, and didn't search for the text I mentioned, and proceeded to read top-down. But even if it was what happened, I'm fully aware of ChatGPT's sketchy behaviour under GPT-3.5, and really don't need to be told by you. But painting LLM or AI as having "no inherent understanding or concept of the technical components of a solar system" is short-sighted. It literally has all the information on this topic available online in 2021, and presents this when asked in natural language. Any inaccurate information it presents is because there was inaccurate information online in 2021. Prior to ChatGPT, I would've found the inaccurate information via Google. Now I save time, and it gets presented to me quicker so I can spend more time verifying and cross-referencing the information and less time searching for it. This is how you use an LLM - you don't rely on it as gospel. You verify everything that isn't obvious. That's what we did pre-LLM, that's what we do now. The only difference is you can get to the information quicker through natural conversation instead of sifting through pages of search results and reading countless articles.

On top of that, there is GPT-4, which Bing Chat and the paid version of ChatGPT use. These provide more accurate and dynamic answers. So, this is another reason why you shouldn't just paint LLM as rubbish. If you do, then you really don't understand what the purpose of an LLM is. In short, they aggregate masses of information (big data), and present it through natural language. The underlying data is where the accuracy issues are, not the LLM itself. The LLM will present the data and take different paths through the data when questioned, which is why you will see it apologise and correct itself. This is not that different to the old days where you read one article, then manually find a different search result that says something else. The difference now is that you will spend 1 day searching (via ChatGPT) and 3 days verifying instead of 3 days searching (manually via Google) and 3 days verifying. A saving of 2 days. For example.

---

Anyway, I wrote more than I expected to there, and explaining myself to a stranger is a poor use of my time, but you started this, and I really only want this topic to be about my solar panels situation, not any side topics. So let's go on to your analysis of my situation...

Looking at the specification of the SE5K inverter, the datasheet states recommended max DC power of 7750W and specifically states "Limited to 155% DC oversizing"

Where did you get this information from as both of your bolded assertions seem wrong? Do you have a link?

I have highlighted in bold the statements chatGPT has stated which are blatantly wrong.

chatGPT has told you the following...
"Your interpretation of the sizing ratio for the SE5K inverter also aligns with the concept of oversizing. When you have an inverter with a lower AC output capacity (e.g., 5,000 W), you need to ensure that the total DC capacity of the panels is significantly higher (in your case, 10,000 W) to meet the minimum DC/AC sizing ratio guideline. This oversizing is considered beneficial because it allows for better energy capture during peak conditions."

then, when you questioned further.. chat GPT says...
"In your case:

• The SE5K inverter has a higher DC/AC sizing ratio of 189.2%. This means your solar panel array's DC capacity (9,460 Wp) significantly exceeds the inverter's AC output capacity (5,000 W), allowing for potential energy capture during peak moments.
• The SE10K inverter has a lower DC/AC sizing ratio of 94.6%. The solar panel array's DC capacity (9,460 Wp) is closer to the inverter's AC output capacity (10,000 W).

Both options are valid, and the choice between them depends on your specific goals and local conditions."

then when you specifically questioned...
YOU.... "But the inverter's capabilities are fixed. So, if there's a maximum amount of DC power that the inverter can handle before curtailment, then the capturing of more energy during peak conditions would never actually happen, right?"

chatGPT responded..
"You are correct. I apologize for any confusion. The inverter's capabilities are fixed, and once the maximum DC power input capacity is reached, any excess DC power will be curtailed, regardless of the DC/AC sizing ratio. In this context, having a higher DC/AC sizing ratio does not directly enable the capture of more energy during peak conditions if the inverter's DC input limit is already reached."

In all this to-ing and fro-ing, at no point has chatGPT referred to either the max DC power input of the inverter being 7750W, above that could cause mis-operation of the inverter or the fact that excess PV power over and above your AC output can be used to charge batteries.

IMHO this is dangerous. I rest my case.

I still don't know why you're so bent on picking apart ChatGPT's responses. The bottom of the front page as I already stated says "ChatGPT can make mistakes. Verify important information." Do you have a problem with immature/free modern technology? Everyone knows ChatGPT's limitations. It provides a starting point, but nonetheless the disclaimer is right there on the very front page. I certainly know more than I did before I had that conversation, and I'm now here verifying things with real people, and unbeknown to you cross-referencing some things using other resources.

I'm also unsure why you're telling me what I already know about the chat with ChatGPT (I was the one questioning its mistakes). You questioning it secondarily to me adds no value here. If you want to add value to help me and others who might read this topic, then maybe just focus on the outcome that I'm trying to verify, and not copy-pasting what I already shared and questioned. In fact, the last part you said is all you needed to say:

at no point has chatGPT referred to either the max DC power input of the inverter being 7750W, above that could cause mis-operation of the inverter or the fact that excess PV power over and above your AC output can be used to charge batteries.

But even this (that the max DC power input is 7750W) seems wrong, and if that's the case, then I don't think we can start calling out an LLM in favour of humans while we ourselves are subject to making human errors.

But let's start with ^this. Where are you getting this information from?

 

[Supervstech]

Pretty sure professional installers won't take my instructions on electrics that I got from ChatGPT or in any case, so calm yourself. But I also have electrical qualifications higher than yours, so don't worry. There's no guesswork here.

Btw, the landing page of ChatGPT says at the bottom, "ChatGPT can make mistakes. Verify important information." Pretty sure forums with emotionally-charged opinionated humans should have an even stronger disclaimer!

What a professional installer would do is irrelevant in this DIY forum.
Please try and look at a post before you make it.

Think about someone with no understanding reading the response and following the advice.

 

[Hedges]

Speaking of over-paneling, consider this inverter:

Sunny Boy Smart Energy | SMA Benelux

www.sma-benelux.com

https://files.sma.de/downloads/SBSExx-50-DS-en-04.pdf

Spec of the 6kW inverter allows up to 12kW of PV panels. It optionally can be connected to a battery for up to 10kW charging, so could charge at 6kW while simultaneously inverting to 6kW AC.

Using PV panels of multiple orientations so peak PV power delivered is less than sum of panel ratings, I might connect 15kW STC of panels.

The battery would let it deliver full AC power to the grid into the evening.

It provides a backup-lite "Secure Power System" you can manually switch to.

"Optional battery backup function available with later hardware release", probably a transfer switch, would provide full backup.



If you have 3-phase, this one is up to 10kW, accepts 15kW PV, has 10kW battery charging.
It has internal transfer switch and backup function included.

Sunny Tripower Smart Energy hybride omvormer | SMA Benelux

De Sunny Tripower Smart Energy hybride omvormer is de 2-in-1 oplossing voor voorziening van zonne-energie in de woning van uw klant.

www.sma-benelux.com

https://files.sma.de/downloads/STPxx-3SE-40-DS-nl-20.pdf

Both of these have particular supported batteries. I think about twice the price of "server rack" batteries you will find.

 

[SeaGal]

@SeaGal Apology accepted,

I have therefore attained charterships in both engineering and IT (CEng and CITP) which makes me more academically and professionally qualified than almost everyone I come across, PhDs aside.

Good point, that must be a nice feeling to have... for the record I also have a BSc(Hons) in electronic engineering, member of the IET (MIET) but only have a CEng

I am not going to be drawn into any more discussion about AI or other off-topic matters. The only important matter was regarding the max PV input.

Where did you get this information from as both of your bolded assertions seem wrong? Do you have a link?

The docs I noticed that detail are here...

https://www.solaredge.com/sites/default/files/se-single-phase-HD-wave-inverter-datasheet.pdf

But if you are referring to the 3 phase version only has a recommended max PV input of 6750W.

https://www.solaredge.com/sites/default/files/se-three-phase-indoor-inverter-datasheet.pdf

 

[NeoX]

member of the IET (MIET)

Then we're fellow IET members. We should be more respectful of each other.

The docs I noticed that detail are here...
https://www.solaredge.com/sites/default/files/se-single-phase-HD-wave-inverter-datasheet.pdf
But if you are referring to the 3 phase version only has a recommended max PV input of 6750W.
https://www.solaredge.com/sites/default/files/se-three-phase-indoor-inverter-datasheet.pdf

OK, it seems we're singing from different datasheets. The HD-Wave has been superseded. On offer now are the Home Hub, Home Wave (HD-Wave rehashed), and Home Short String inverters. I mentioned at the start of the part of the conversation I had with ChatGPT that the SE10K has a max DC input of 15,000W, and the SE5K has a max DC input of 10,000W. While I didn't specifically mention this to ChatGPT (as it was unnecessary info), this would imply that I'm referring to only the Home Hub (datasheet here). I also said in ChatGPT (same message per above) that this is for a 3-phase system which is also clear from the OP given the number of panels and the Wp of those panels.

This is why I said your assertion "max DC power of 7750W" seemed wrong.

Now, I still don't know where you got the 155% oversizing limitation from as I can only find a reference to that here, which is based on single-phase inverters 8 years ago. The current datasheets don't state any limitation, but there is an official guidance which refers you to the specific model. IOW, SolarEdge are not really clear on this, but there is an indication on the Belgian guidance for the StorEdge SE5K (which Home Hub is a rehash of) that 200% oversizing is possible. This is corroborated on the landing page for Home Hub which clearly states "Up to 200% DC oversizing".

I'm not trying to prove a point, but what this proves is two things. 1) SolarEdge make the information confusing and difficult to find by using similar model numbers across their different lines, and not including all information in their datasheets. 2) ChatGPT wasn't so wrong after all; perhaps this topic will be a humble lesson that humans are not infallible, and computers are better than us at many things while also not being infallible (yet they only do what humans tell them to including their input data and programming).

Ultimately, I'd only like to get to the point where I know 1) whether to get SolarEdge Home Hub or Enphase microinverters given my minimal shading issues (the 15cm "wall" around my roof edge and the chimney, and the east-west configuration I want), and 2) if SolarEdge, given the climate of the Netherlands, would I be better going for the SE5K or SE8K rather than the SE10K that the supplier suggested?

 

[Hedges]

You keep referring to SolarEdge. Manufacturers control most of the narrative potential consumers get to see by creating it, and sometimes get heavy-handed trying to control what information is available from other sources (which should be a big red flag.)

I encountered and others reported experiencing or learning of various issues. Read through this thread, make sure you're confident in the quality of what you select.

Solar Edge - an installer's blog

Came across the following blog. It includes the company's installed system numbers and failure rates for SolarEdge, Fronius, Enphase. Also, information on limitations and best use of optimizers. https://www.mcelectrical.com.au/solaredge-inverter-optimiser-review/

diysolarforum.com