pv system connected to sub-panel

[J.P.M.]

I don't think your statements track logically.
If the modules fail at about the same rate they will fail about the same rate whether it's a string inverter or solaredge.
From my install I have *fewer* panels because I have solaredge than I would if I had gone with a string inverter. Because I can get more kwh out of the same panels because partial shading is handled better.
But even if we ignore that, I don't see a reason you would have MORE modules than a string inverter. Only reasoning I can see is if it were less efficient. And I haven't seen evidence that a Solaredge system has a lower efficiency than a string inverter system. (If there were, I'd think that'd be a marketting point that would be played up by SMA and others.)

The other thing is you seem to think that it'd be more noticeable on a string inverter. I would think the opposite.
If I have a set of 12 modules and one of them is providing 10% less power IMO you will *only* see that when looking at per-module monitoring data. If I can only look at string-level data I won't be able to tell the difference between 10% lower on one module and all modules being less than 1% lower.

If it's an inverter failure for string vs. micro-inverter, then I would agree that the string is more likely to be noticed quickly.
But we are talking about module replacements under the module's warranty.

I was referring to Micros as modules as I perhaps incorrectly assumed Butch was doing I was actually referring to micro inverter failures. I'll admit to bad and confusing wording with apologies. My thinking is that 250 Watt failure on a 6 kW array may go unnoticed but a string inverter failure is likely to be noticed much more quickly.

I'm also not including discussions about partial shading and the partial advantage that micros and other systems such as SolarEdge can bring to the table, although I think that advantage may be a bit overstated in the hype. A 50% shaded array will not have 50 % of it's unshaded instantaneous output restored by use of micros or a SolarEdge system.

I also never wrote that a sting inverter will allow a smaller array than a micro inverter equipped system.

One of my points might be better explained this way: With a micro inverter system equipped array of 16, ea. 300 Watt modules, one micro inverter failure may well not be noticed as quickly, or at all maybe as if the inverter dies on a similar paneled but equipped with a string inverter.

See further posting. I bet there's a difference of opinion, but maybe some terms need to be cleared up before things get more out of hand and confused.,

[ButchDeal]

Naw, only if its already there. That's why I offered peak current on a sunny day. Software, its supposed to
be available for the job requirements, though that may be far from reality. How frequent are your present
checks? Bruce Roe

The default alletrs check daily for issues and send emails. I usually change some of them though. Like connectivity to two days, as homeowners sometimes have internet outages etc that can set it off. Two days rarely gives alerts that way.

[bcroe]

Really you think we should install two inverters at every site across the US and compare them constantly. Got any software to do all that?

Of course we would have to balance them all with equal parts on each face so they are identical and compensate for shade etc. yeah that will work and not cost too much either.

Naw, only if its already there. That's why I offered peak current on a sunny day. Software, its supposed to
be available for the job requirements, though that may be far from reality. How frequent are your present
checks? Bruce Roe

[ButchDeal]

I appreciate and respect the difference of opinion.

But - there's always a but I suppose - you post raises some questions in my mind:

1.) If you've been told that you have "the most module replacements by far...", how would you answer those who suggest (and accepting for the time being the idea that all such modules are ostensibly created approx. equal) that rather than blaming the product, the higher replacement rate is due to the quality of the install ? Seems by your own admission, you have more replacements. Are your modules of inferior quality to those from the same source, or inferior to those used by other vendors ?

We sell many different brands as stated, multiple distributors, and many different installers (national remember). We drop ship the equipment from the distributors to the installers or sites.
Why are you assuming that we are having a higher failure rate than others who do not monitor. Most would assume the failures are fairly evenly distributed, but only a few are found (by those looking). I am not talking about workmanship issues, but PV module failures. They are really easy to see on SolarEdge. Optimizers we have had very few failures on, and most were with two installers who we have stopped using after sending different installers out to investigate.

2.) How "small" (a failure/service call rate) as compared to string inverters ? Small seems a relative thing.

not sure what you mean here.

3.) If the modules are indeed created about equal, which seems at least plausible, and your failure rate is indeed due to better monitoring, could that mean the other such modules may be failing at something like an approx. equal but unreported or unknown rate? If so, that would seem to give at least some credence to my opinion and conjecture that small(er) failures of, say, one panel, would have a higher probability of going unnoticed by users. If so, why would I or any potential customer want to use a product or rely on a vendor who would sell such a product that might fail at a higher rate per install than a string inverter (if for no other reason than their being more of them per install), and have such failures slip under the radar, or, as an alternative, require a customer to perhaps be more diligent in their own monitoring. Seems to me a string inverter failure, being something like an order of magnitude larger and thus more prominent and more quickly noticeable, might be easier for an owner to catch rather than waiting/hoping/relying on monitoring from some company not as diligent as yours.

The failure rate is not due to the increased monitoring. Knowing about the failure rate is due to the increased monitoring.
Again we have had very very few optimizer failures. We have actually had more string inverters fail than optimizers. Which is incredibly surprising since we have installed very few string inverters.

Most installers do not like to give module level access to homeowners either. We set up all our monitoring (because they do such a poor job of it) and always give module level.
We also set them all up in PVOutput and give them a managed account. Even with all this very few of our homeowners notice the issues, when they come up. We call them and let them know we are sending out a tech to look at it and they are surprised. We let them know that they are only loosing a small amount (in the case of just a module failure).
I am not trying to toot our horn here just saying what can be done.
A diligent homeowner that is familiar with electronics could maintain a string inverter well as you have and that is clearly the best choice for such a person (I have a string bimodal system myself), but for the homeowner that doesn't want to play with the thing all the time or really wants an appliance, SolarEdge makes a strong argument.
Still waiting to see if SMA can get as integrated with their optimizer solution after a few generations.

[ButchDeal]

Sounds like your company is doing a very conscientious job of maintaining somewhat failure prone equipment. You could
probably detect a failure as well by monitoring peak current on a sunny day, or matching outputs of 2 string inverters, like me.
Service might be more complex in finding the exact fault, but it wouldn't happen as often.

A clamp on ammeter and a voltmeter could locate 24 bypassed cells in a string of 720 here; a non contact thermal gauge
also might locate the fault. I suppose the micro readouts look more attractive when working on a roof. The ammeter does
check for best performance here several times a year. But I've run 5 million panel hours and 100,000 inverter hours with
zero failures so far. Bruce Roe

Really you think we should install two inverters at every site across the US and compare them constantly. Got any software to do all that?

Of course we would have to balance them all with equal parts on each face so they are identical and compensate for shade etc. yeah that will work and not cost too much either.

[ButchDeal]

Out of curiosity, when you use the word "module", is it safe to assume that you are referring to the PV panel, not the optimizer or inverter? If so, I think it is bad logic to infer that the panels you are installing are of inferior quality, when your ability to detect bad modules is superior to those who are using the typical monitoring system provided with string inverters. I respect Bruce's point that string inverters *could* be effectively monitored, but in practice, they aren't. Most monitoring systems are looking only at output power, and not individual string voltage and current. This is a benefit of SolarEdge (and microinverters) that is probably not going to be appreciated by most owners (since the module failure rate is low), but in aggregate, and the volumes you work with, it seems like a actual source of value.

A module is made up of solar cells, what many call a panel, but....
Panel is a group of modules, an array is a group of arrays ( for example two faces on a roof would be an array of two panels)

this is my assessment as well. Modules fail, squarely chew threw wires, **** generally happens, good monitoring helps find these issues AND diagnose them. On top of that we use revenue grade meters built into the inverters which simplifies install and SREC monitoring.

[Jest Waitn]

FWIW, it seems to me that the chief point of failure in these micro-inverters is due to roof mounting; they receive far less cooling. i live in the country and whenever i visit an urban area with few trees and much asphalt (roads and roofs), the difference in temperature is dramatic! it takes all evening for a hot asphalt roof to cool down. imagine how much heat radiation hits one of these inverters.

[sensij]

JPM and others have pointed out that to a homeowner that monitors system level closely, a bad or partially failed module would often show up. Possibly but harder to automate an alert on that.

Out of curiosity, when you use the word "module", is it safe to assume that you are referring to the PV panel, not the optimizer or inverter? If so, I think it is bad logic to infer that the panels you are installing are of inferior quality, when your ability to detect bad modules is superior to those who are using the typical monitoring system provided with string inverters. I respect Bruce's point that string inverters *could* be effectively monitored, but in practice, they aren't. Most monitoring systems are looking only at output power, and not individual string voltage and current. This is a benefit of SolarEdge (and microinverters) that is probably not going to be appreciated by most owners (since the module failure rate is low), but in aggregate, and the volumes you work with, it seems like a actual source of value.

[foo1bar]

If so, why would I or any potential customer want to use a product or rely on a vendor who would sell such a product that might fail at a higher rate per install than a string inverter (if for no other reason than their being more of them per install), and have such failures slip under the radar, or, as an alternative, require a customer to perhaps be more diligent in their own monitoring. Seems to me a string inverter failure ...

I don't think your statements track logically.
If the modules fail at about the same rate they will fail about the same rate whether it's a string inverter or solaredge.
From my install I have *fewer* panels because I have solaredge than I would if I had gone with a string inverter. Because I can get more kwh out of the same panels because partial shading is handled better.
But even if we ignore that, I don't see a reason you would have MORE modules than a string inverter. Only reasoning I can see is if it were less efficient. And I haven't seen evidence that a Solaredge system has a lower efficiency than a string inverter system. (If there were, I'd think that'd be a marketting point that would be played up by SMA and others.)

The other thing is you seem to think that it'd be more noticeable on a string inverter. I would think the opposite.
If I have a set of 12 modules and one of them is providing 10% less power IMO you will *only* see that when looking at per-module monitoring data. If I can only look at string-level data I won't be able to tell the difference between 10% lower on one module and all modules being less than 1% lower.

If it's an inverter failure for string vs. micro-inverter, then I would agree that the string is more likely to be noticed quickly.
But we are talking about module replacements under the module's warranty.

[bcroe]

You set up alerts such that individual module data issues trip an alert. We have this and we monitor all our installs. We have been told that we have had the most module replacements by far over any other company, and all of them warranty issues. This is because of the module level monitoring. That said it is a small percentage but I have to imagine that since we are the only ones monitoring actively like this and we have the highest number of modules warranty replacements that other installers and even the big guys like SolarCity, SunEd, etc would benefit from monitoring at the module level.

Sounds like your company is doing a very conscientious job of maintaining somewhat failure prone equipment. You could
probably detect a failure as well by monitoring peak current on a sunny day, or matching outputs of 2 string inverters, like me.
Service might be more complex in finding the exact fault, but it wouldn't happen as often.

A clamp on ammeter and a voltmeter could locate 24 bypassed cells in a string of 720 here; a non contact thermal gauge
also might locate the fault. I suppose the micro readouts look more attractive when working on a roof. The ammeter does
check for best performance here several times a year. But I've run 5 million panel hours and 100,000 inverter hours with
zero failures so far. Bruce Roe

[J.P.M.]

I will just speak to this one. I personally disagree with J.P.M. on this one detail. though I agree looking at individual module data gets old. You set up alerts such that individual module data issues trip an alert. We have this and we monitor all our installs. We have been told that we have had the most module replacements by far over any other company, and all of them warranty issues. This is because of the module level monitoring. That said it is a small percentage but I have to imagine that since we are the only ones monitoring actively like this and we have the highest number of modules warranty replacements that other installers and even the big guys like SolarCity, SunEd, etc would benefit from monitoring at the module level.
SunEd specifically only monitors at the system level even though they install SolarEdge and Enphase systems, and I have been told that SolarCity does similarly.

JPM and others have pointed out that to a homeowner that monitors system level closely, a bad or partially failed module would often show up. Possibly but harder to automate an alert on that.

I appreciate and respect the difference of opinion.

But - there's always a but I suppose - you post raises some questions in my mind:

1.) If you've been told that you have "the most module replacements by far...", how would you answer those who suggest (and accepting for the time being the idea that all such modules are ostensibly created approx. equal) that rather than blaming the product, the higher replacement rate is due to the quality of the install ? Seems by your own admission, you have more replacements. Are your modules of inferior quality to those from the same source, or inferior to those used by other vendors ?

2.) How "small" (a failure/service call rate) as compared to string inverters ? Small seems a relative thing.

3.) If the modules are indeed created about equal, which seems at least plausible, and your failure rate is indeed due to better monitoring, could that mean the other such modules may be failing at something like an approx. equal but unreported or unknown rate? If so, that would seem to give at least some credence to my opinion and conjecture that small(er) failures of, say, one panel, would have a higher probability of going unnoticed by users. If so, why would I or any potential customer want to use a product or rely on a vendor who would sell such a product that might fail at a higher rate per install than a string inverter (if for no other reason than their being more of them per install), and have such failures slip under the radar, or, as an alternative, require a customer to perhaps be more diligent in their own monitoring. Seems to me a string inverter failure, being something like an order of magnitude larger and thus more prominent and more quickly noticeable, might be easier for an owner to catch rather than waiting/hoping/relying on monitoring from some company not as diligent as yours.

See my further posts for acknowledgement and correction of an error in referring to micro inverters or SolarEdge roof equipment as "modules". Patch : Replace the word "module" with the words "micro inverter or optimizer".

[ButchDeal]

Individual panel monitoring novelty usually wears off quickly.

I will just speak to this one. I personally disagree with J.P.M. on this one detail. though I agree looking at individual module data gets old. You set up alerts such that individual module data issues trip an alert. We have this and we monitor all our installs. We have been told that we have had the most module replacements by far over any other company, and all of them warranty issues. This is because of the module level monitoring. That said it is a small percentage but I have to imagine that since we are the only ones monitoring actively like this and we have the highest number of modules warranty replacements that other installers and even the big guys like SolarCity, SunEd, etc would benefit from monitoring at the module level.
SunEd specifically only monitors at the system level even though they install SolarEdge and Enphase systems, and I have been told that SolarCity does similarly.

JPM and others have pointed out that to a homeowner that monitors system level closely, a bad or partially failed module would often show up. Possibly but harder to automate an alert on that.

[J.P.M.]

no one wants to commit to an answer on this?

I'll add to the chorus: More reliable, less up front cost. Individual panel monitoring novelty usually wears off quickly. What more do you want ?

[ButchDeal]

no one wants to commit to an answer on this?

string inverters are cheaper, single point of failure (instead of many), more efficient, more reliable.

Further there are other options as well, like SolarEdge optimized system which is kind of a mixing of micros and string inverter.
This is my preference as it is most reliable, low cost, meets all safety requirements easily, easy to monitor, and most efficient.

[bcroe]

no one wants to commit to an answer on this?

There are plenty of advantages to high voltage strings; some may not apply to your situation. Assuming you
have no compelling need for micros, the cost will be less in inverters and wire. There are fewer parts to fail,
and generally there will be a little bit higher efficiency. A lot of things can be done with panels to better match
up to your inverter plant and your weather, even after the initial build. See my thread Sun Hours for more
on that. Bruce Roe