Solar Edge Vs Enphase Power optimizers

[Naptown]

Some installers cover labor. My installer warranty covers enphase micros for 20 years labor in addition to the 25 year manufacturer warranty on parts.

I would be happy to give you a 30 year warranty as an installer.
Costs me nothing. Incorporated so when solar falls off the cliff at the end of 2016 I will dissolve the corporation
Oh and that 20 year warranty just turned into a year and a half warranty.
I doubt your installer has even been around more than 6 years.

[makarowski]

very nice...

I would think that since the SDGE meter data from Wattvision could be "read/timed" to coincide with the SE data, the timing portion could be solved without a revenue grade meter at PVOutput... then it would be just the accuracy of the SE data (1-5%)? not sure there is the control there down to the "second" on data polling..

looks like you are enjoying the same dark weather here in san diego ... solar generation went to about 0 at 3pm today...blah, its supposed to be summer!

Cheers
BMak

[sensij]

I know a few people have been digging into the data to find where the errors are occurring... likely just the inaccuracy of SE data (rounded up and/or poor accuracy)... there are some wiered data points where my homes usage goes to 0 (which I know is not occurring)... "timestamps" of the data could also be a factor.

I'll eventually dig into it, but too much other stuff going on right now.

Yes, it is timestamps and synchonization error, with a little bit of calibration error (maybe 1%) thrown in. Every time you get a "0" for consumption you will increase the error with respect to SDG&E.

To prove this, I wrote my own PVOutput uploader with a revenue grade data source so I have more control over synchronization. I also left a system on PVOutput that uses the SE data untouched. Yesterday was a decent example... You can see the good consumption data here, and the spikey consumption data here. My agreement to SDG&E for the first full month I ran the uploader was within 1 kWh.

[makarowski]

I had looked at his system (on PVOutput) at the time the optimizer wasn't reporting, and have been following it since it is a bit of an outlier for production efficiency. In my opinion, the optimizer was properly producing power, even when it wasn't communicating. I suspect there are failure modes that would result in no power out, and there might even be modes that take down the string, but as far as I know, none of those have been reported here. There was another forum member recently with the same problem as makarowski, but I'll have to check my notes to see who it was.

As for the outlier in efficiency... it looks like better thermal performance to me, but I'm not sure. Some ideas:

1) The panels manufactured at the very top end of the + tolerance for power (someone's got to get the good tail of the distribution, right?).
2) The panels are installed high off the roof, so there is better air flow.
3) The location is near the coast, perhaps there is more wind, or more moisture in the air helping the heat transfer rate.

All of the zip code 9200X systems have generally been performing well, so it might be a combination of things that make his system stand out. Calibration is always on the short list too, but the calculated consumption data in PVOutput doesn't look skewed when the array is producing, which suggests that at least some of the outperformance is real.

I'd have to agree the old optimizer was generating power...

As for the efficiency, its nice to end up on the high side... although not sure how real it is. I think you have something on point about being on the high part of the roof. I have two strings where the string on the lower roof with same orientation/elevation, (and gets wind protection from the next house) is not as efficient. no real shading on either strings... maybe I should have put more panels on my upper roof

PVOutput is using data based on SE monitoring data... I have tried to correlate with my SDGE bill... with some limited success

first month: SDGE measured a net of 32 kwh of usage, while the net from pvoutput was 109 (966 usage, 857 generation)
last month: sdge had a net of -194, while PVOutput had -203 (a bit closer, 1031 usage, 1234 generation)

I know a few people have been digging into the data to find where the errors are occurring... likely just the inaccuracy of SE data (rounded up and/or poor accuracy)... there are some wiered data points where my homes usage goes to 0 (which I know is not occurring)... "timestamps" of the data could also be a factor.

I'll eventually dig into it, but too much other stuff going on right now.

Cheers
BMak

[Pomonabill220]

My system installed 3/2015

I have been lurking for a while, but decided to signup and make my first post about my system.
I had researched about 5 different installers/systems and decided to go with a SolarEdge SE5000, 20-250watt Qcell QPRo BFR-G3 250 panels with power optimizers.

My installer that I choose was great! VERY good contact, answered all my questions, did all the legwork.
I had to get my electrical service upgraded which included 2 new panels. I stuck by my guns and got 2 SquareD QO 250amp panels!!! Had to fight a little because they wanted to install Homeline, but I knew that QO was top of the line, and was familiar with them, and I got them.

I was on the fence with SolarEdge and Enphase, but when I found that panel level monitoring was going to go away, or cost, and my cousin had Enphase and has had problems with his install (of only 5 years), SolarEdge was my final choice, and I am glad I did.

Having the inverter on the roof in the HEAT, I knew this was a disaster waiting to happen. Being an electronics eng., I knew that heat is the number ONE enemy of electronics, and the inverter generates heat by itself.

I am a little disappointed that the most I have seen to date (excluding days with clouds) is only about 3.5Kw at high noon.
My layout is 11 pnls facing east, 4 facing south, 2 horiz (almost flat) at 7 deg., 3 facing west. Roof tilt is 27deg.

Also, my installer said that IF I wanted to get better monitoring rights on the web portal, I would have to become the installer and tell them when something failed.... so I plugged a USB cable into the inverter and am monitoring the inverter's data, can't log though....

I would like to know how to post a public link to my monitoring portal for all to watch, and signup to PVoutput.

Any ideas?

Thanks!

[sensij]

Thanks for the explanation. Can you look back at the data and guess if the failed optimizer was producing power? I realize you had little baseline to make a judgement at the time it broke.

I had looked at his system (on PVOutput) at the time the optimizer wasn't reporting, and have been following it since it is a bit of an outlier for production efficiency. In my opinion, the optimizer was properly producing power, even when it wasn't communicating. I suspect there are failure modes that would result in no power out, and there might even be modes that take down the string, but as far as I know, none of those have been reported here. There was another forum member recently with the same problem as makarowski, but I'll have to check my notes to see who it was.

As for the outlier in efficiency... it looks like better thermal performance to me, but I'm not sure. Some ideas:

1) The panels manufactured at the very top end of the + tolerance for power (someone's got to get the good tail of the distribution, right?).
2) The panels are installed high off the roof, so there is better air flow.
3) The location is near the coast, perhaps there is more wind, or more moisture in the air helping the heat transfer rate.

All of the zip code 9200X systems have generally been performing well, so it might be a combination of things that make his system stand out. Calibration is always on the short list too, but the calculated consumption data in PVOutput doesn't look skewed when the array is producing, which suggests that at least some of the outperformance is real.

[donald]

Hi Everyone,

I had one optimizer fail on about day 3 ... infant mortality. my luck.

The optimizer stopped communicating and looked to be not generating power ... at least in a string of 13 it was a guess. Since I was doing some hyper monitoring in the early days, I noticed it right away. The LCD panel on the inverter also showed "23/24 optimizers" so at a min the comms were down. My installer responded quickly over email, acknowledged the problem and came out about a week later -> had to coordinate with a SE technician on the phone to troubleshoot on-site. "yes you have a bad optimizer, new one on order".

Seems like they needed to do a formal site visit just to confirm bad inverter and get a new one under warranty. And also cost SE two "truck rolls"..

About a week after that, since the replacement still didn't arrive from SE, my installer pulled one from their stock and came to do the repair. I was not at home due to travel, but when I got back I saw "24/24" optimizers on the inverter LCD, but online there was still the "bad" optimizer in the system. I think they forgot to update the portal with the new optimizer ID. I think it took them about 3 weeks (only in the last week or so) to finally get around to updating the on-line system. I was not pushing them since the system seemed to be working well (I have some amazing efficiency numbers as reported by PVOutput in team sandiego).

Let me know if you have any more questions...

BMak

Thanks for the explanation. Can you look back at the data and guess if the failed optimizer was producing power? I realize you had little baseline to make a judgement at the time it broke.

What a remarkably inefficient way to do warranty service.

[thejq]

Thanks, BMak, for the update. Now we know.

[makarowski]

I have only seen one confirmed failure (a recent system by @makarowski), but it wasn't clear to me if it's because of communication or bad optimizer. But in his case, even the web portal reports zero output, the overall output suggested that the panel+optimizer was still working. I just checked his public site, it seems the problem has now been fixed. Maybe you can PM him directly and report back or ask him to comment here for everyone's benefit.

Hi Everyone,

I had one optimizer fail on about day 3 ... infant mortality. my luck.

The optimizer stopped communicating and looked to be not generating power ... at least in a string of 13 it was a guess. Since I was doing some hyper monitoring in the early days, I noticed it right away. The LCD panel on the inverter also showed "23/24 optimizers" so at a min the comms were down. My installer responded quickly over email, acknowledged the problem and came out about a week later -> had to coordinate with a SE technician on the phone to troubleshoot on-site. "yes you have a bad optimizer, new one on order".

Seems like they needed to do a formal site visit just to confirm bad inverter and get a new one under warranty. And also cost SE two "truck rolls"..

About a week after that, since the replacement still didn't arrive from SE, my installer pulled one from their stock and came to do the repair. I was not at home due to travel, but when I got back I saw "24/24" optimizers on the inverter LCD, but online there was still the "bad" optimizer in the system. I think they forgot to update the portal with the new optimizer ID. I think it took them about 3 weeks (only in the last week or so) to finally get around to updating the on-line system. I was not pushing them since the system seemed to be working well (I have some amazing efficiency numbers as reported by PVOutput in team sandiego).

Let me know if you have any more questions...

BMak

[kevcor620]

Some installers cover labor. My installer warranty covers enphase micros for 20 years labor in addition to the 25 year manufacturer warranty on parts.

[thejq]

Does a failed Solaredge optimizer stop producing power, or does it just stop optimizing?

I have only seen one confirmed failure (a recent system by @makarowski), but it wasn't clear to me if it's because of communication or bad optimizer. But in his case, even the web portal reports zero output, the overall output suggested that the panel+optimizer was still working. I just checked his public site, it seems the problem has now been fixed. Maybe you can PM him directly and report back or ask him to comment here for everyone's benefit.

[sdold]

Yikes! So no monitoring for self installs? Really? If that's true I won't bother with Enphase.....

No, for self installs, you get the full monitoring (Enlighten Manager) at no cost. You list yourself as the installer when setting up your account.

[donald]

I guess Solaredge would have 43 then. I suppose that the OP would need to weigh Solaredge's better warranty against possibly not being able to replace a failed optimizer or inverter if they are no longer in business.

Does a failed Solaredge optimizer stop producing power, or does it just stop optimizing?

[GoingSolar]

Enphase also stopped the panel level monitoring except for installers.

Yikes! So no monitoring for self installs? Really? If that's true I won't bother with Enphase.....

[sdold]

Oh and no single point of failure?
How about taking the number of panels and adding the number of microinverters to it. 20 panels string inverter 21 points of failure
21 panels with microinverters 42 points of failure.

I guess Solaredge would have 43 then. I suppose that the OP would need to weigh Solaredge's better warranty against possibly not being able to replace a failed optimizer or inverter if they are no longer in business.