Enphase field MTBF: M190: ~36 Years M215: ~316 Years M250: >357 Years

[pclifton]

Hey Ken did you have a power outage yesterday? Just curious. Looks like both systems were off until 1pm? Also looks like you had production 6-8pm on Thursday, around 400 watts continuous?!

The 4 year old Envoy I mentioned in my original post died Thursday at about 4PM. I had ordered a replacement on eBay for much less than the $500 price most places are asking. The replacement came the same day that the Envoy died.

I installed the new one at 7 AM Friday, and left it scanning for devices. I shut the other small system down to eliminate any chance of those inverters being picked up by the new Envoy until it's new device scan completed. The new Envoy updated its 2013 firmware about 4 times. Looks like a sequential update process that rolls through each new version released.

I had a break at 2 PM and spoke with Enphase support to have the new Envoy added to my system. Thankfully that kept me from having to pay the additional $249 for per-panel monitoring. I put everything back online about 4 PM. That is the long version of what went on.

Unfortunately, my four year old TED5000 that I had programmed to auto-post every 5 minutes to PVOutput also died Thursday. The new "wall plug" is supposed to be here Monday. My PHP plugin that I wrote for Wordpress pulls from the TED5000 so my PVOutput and web site will not be displaying correctly until Monday evening.

Interesting that two unrelated systems decided to fail on the same day. I did have advanced notice that the Envoy was on its way out... Maybe in the summer I will do a tear down on the old Envoy. I would like to see what is inside it.

Ken Clifton

 

[RegGuheert]

The 4 year old Envoy I mentioned in my original post died Thursday at about 4PM.

Unfortunately, my four year old TED5000 that I had programmed to auto-post every 5 minutes to PVOutput also died Thursday.

We had a thunderstorm either Wednesday or Thursday night here in VA. I'm wondering if lightning might have been the culprit in your case.

 

[pclifton]

We had a thunderstorm either Wednesday or Thursday night here in VA. I'm wondering if lightning might have been the culprit in your case.

No, all the storms missed us. I was actually home when the failures occurred. No nearby lightening for several days prior. That was something I considered as well.

We have a layered approach to surge and power spikes. A delta-T on the main meter. A Square-D surge-breaker panel protector in the main panel. Each sub panel for the solar arrays has a large panel guard protector. Each disconnect on the pole has a square-D protector on it as well. Of course nothing is going to help with a direct hit except the homeowners policy.

Ken Clifton

 

[RegGuheert]

We have a layered approach to surge and power spikes. A delta-T on the main meter. A Square-D surge-breaker panel protector in the main panel. Each sub panel for the solar arrays has a large panel guard protector. Each disconnect on the pole has a square-D protector on it as well.

Wow! You're all protected-up! I've got nothing here except whatever protection is built into the inverters and the Envoy and various other equipment around the house here.

Of course nothing is going to help with a direct hit except the homeowners policy.

Sure, but you certainly don't need a direct hit to damage something connected to the power lines, especially PLC devices.

 

[billg]

My original Envoy bricked at just short of 5 years and I couldn't believe the replacement cost.

It does appear to me that the very earliest Envoy units are the ones most likely to fail, but I will reserve final judgment on that until a few more years have passed.

Just to confirm your supposition, my Envoy that failed was an old "square corner" unit, shaped much like the brick it now resembles.

 

[RegGuheert]

My original Envoy bricked at just short of 5 years and I couldn't believe the replacement cost.

It does appear to me that the very earliest Envoy units are the ones most likely to fail, but I will reserve final judgment on that until a few more years have passed.

Just to confirm your supposition, my Envoy that failed was an old "square corner" unit, shaped much like the brick it now resembles.

O.K. Thanks!

For reference, my Envoy, which was made in 2010, but which hasn't yet failed, is in the shape of an oval.

Has anyone had an oval Envoy fail, yet?

 

[DaveEV]

Mine was an oval purchased in 2010 that was proactively replaced by Enphase shortly before the two year warranty was up. I did not notice any issues so not sure how they detected issues with it.

 

[pclifton]

For reference, my Envoy, which was made in 2010, but which hasn't yet failed, is in the shape of an oval.
Has anyone had an oval Envoy fail, yet?

Count my Envoy failure last week as one of the "oval" shaped units. It was purchased in 2010. According to Enphase, the latest firmware that it would run was R3.7.26. It's replacement updated all the way to R3.11.30. In Enlighten the new Envoy reports as 800-00069-r05 (Envoy-R-NA), whereas the old unit reported as 800-00024-r02 (Envoy)

Ken Clifton

 

[RegGuheert]

Count my Envoy failure last week as one of the "oval" shaped units. It was purchased in 2010. According to Enphase, the latest firmware that it would run was R3.7.26. It's replacement updated all the way to R3.11.30.

My Envoy shows Current Software Version as R3.7.26 (7888b3). But when I went to Enlighten, it said "A software update is available for your Envoy," so I clicked "Update Software." It now says "Upgrade software task for Envoy (serial number #blah) has been scheduled." We'll see what the result of that is.

In Enlighten the new Envoy reports as 800-00069-r05 (Envoy-R-NA), whereas the old unit reported as 800-00024-r02 (Envoy)

My Envoy reports as a 800-00024-r02 (Envoy), same as your old unit.

It seems my Envoy, also purchased in 2010, is the same as your old unit. That's not overly comforting!

 

[dsinned]

Once a firmware update is "scheduled", do you have to power cycle the Envoy in order for the update to actually occur?

 

[billg]

Count my Envoy failure last week as one of the "oval" shaped units. It was purchased in 2010. According to Enphase, the latest firmware that it would run was R3.7.26. It's replacement updated all the way to R3.11.30.

My Envoy shows Current Software Version as R3.7.26 (7888b3). But when I went to Enlighten, it said "A software update is available for your Envoy," so I clicked "Update Software." It now says "Upgrade software task for Envoy (serial number #blah) has been scheduled." We'll see what the result of that is.

In Enlighten the new Envoy reports as 800-00069-r05 (Envoy-R-NA), whereas the old unit reported as 800-00024-r02 (Envoy)

My Envoy reports as a 800-00024-r02 (Envoy), same as your old unit.

It seems my Envoy, also purchased in 2010, is the same as your old unit. That's not overly comforting!

My new one is also identified as 800-00069-r05 (Envoy-R-NA), while my first one was identified as 860-00003-r03 (EMU).

 

[RegGuheert]

My Envoy shows Current Software Version as R3.7.26 (7888b3). But when I went to Enlighten, it said "A software update is available for your Envoy," so I clicked "Update Software." It now says "Upgrade software task for Envoy (serial number #blah) has been scheduled." We'll see what the result of that is.

Just to close the loop on this: Today I checked and it says "Your Envoy software is up to date." This is true even though none of the numbers have changed from what I reported previously.

 

[RegGuheert]

All other inverters are reporting normally except for the two which are currently on Enphase's "limp-along" firmware. They are producing relatively normally, although they produce "Grid Gone" events daily or multiple times each day:

Limp-Along Firmware:
- 121045414867 - Garage, Middle, Second from the left
- 121045415714 - House, Middle, Third from the left

As such, I currently have 10% of the original M190s on the Primary Watch List, 10% on the Secondary Watch List and 5% on Enphase's Limp-Along Firmware. We'll see if the next failures come from these groups or from elsewhere in the array.

It looks like the next failure is the "Limp-Along" unit on the garage. It stopped reporting yesterday morning. Now only one inverter is spitting out daily "Grid Gone" events. I notified Enphase this morning, but I did not receive even the normal auto-reply. We'll see how they handle this one.

This will be the most difficult replacement to date. If I can get up there from the side of the house safely, I will remove both the one to the left of this unit and the unit itself and move the other new-style inverter up with it (to have a more reliable unit in this middle row). That will allow me to later upgrade that run to an Engage cable at a later date and replace the other five inverters in that string. The other alternative is to simply remove only the PV module directly *below* that unit (in the bottom row) and replace the dead unit that way. That may be safer and easier. But Enphase needs to send me one first...

 

[Weatherman]

I, finally, got Enphase to agree to replace three of my worst grid-goners. That would, officially, make 16 failures in less then three years on my 30-panel array.

 

[RegGuheert]

I, finally, got Enphase to agree to replace three of my worst grid-goners. That would, officially, make 16 failures in less then three years on my 30-panel array.

Good news! OTOH, that brings your MTBF down to 5 years. My recent failure brings the MTBF down to 45 years for my system.

Speaking of MTBF, I have updated the thread title. The MTBF of the M215s has steadily increased and is currently above 1800 years (!) while the MTBF of the M190s has steadily decreased and now stands at 50 years. I suppose the M215 MTBF would be 900 years if I included dsinned's failure, but I won't add it to the spreadsheet without a link. Still 900 years is more than adequate for this product. For a 10 kW system, anything above about 400 years limits the failures to about one every 10 years.

Weatherman, since you have 30 microinverters, it is possible that your installer connected them simply as two strings of 15 inverters each. You can verify whether this is true or not by counting how many 2-pole 15A breakers are assigned to the solar array in your electrical panel. If there are only two, then this has to be how they are arranged. Then looking at your array layout on page 18 of this thread, it is possible (likely?) that they connected the entire 15 in series. If all that conjecture is correct, then the voltage rise of the last inverters in each of the strings will be 4V(!) on top of any other wiring losses, meaning you would be losing about 2% of the production from some of the units in your array just to copper losses, or a bit less than 1% overall. It's not a huge number, but I wonder if the higher voltage might also be contributing to failures. Anyway, the reason I bring this up is that (if all my conjecture above is correct) you will be able to reduce that maximum voltage rise from 4V to about 1.15V if you replace each of your M190 strings with a length of portrait Engage cable. You would need to have at least 15 of the black M190IGs before you can do that with one of your strings, however.

Just a thought...

 

[RegGuheert]

You'll be the first to know if I have a failure have friends with about 80 M215s/M250s with no failures as well.

Do I have these microinverters in my spreadsheet? If not, can you provide the links? That would be great since I would have a way to find out about any failures. TIA!

 

[Weatherman]

Weatherman, since you have 30 microinverters, it is possible that your installer connected them simply as two strings of 15 inverters each. You can verify whether this is true or not by counting how many 2-pole 15A breakers are assigned to the solar array in your electrical panel.

Yes, there are two branches, 15 micros, each. If I remember correctly how my installer set things up, the top row and the two on the bottom left are one branch, and the middle row and the three on the lower right are the second. The last panel in each branch is on the left side. The failures have been scattered all over the place, but I suppose you could say there's a focus in the top row and left side.


Here's the latest pic of my array, with the 12 replacements obvious as lower lifetime energy values. The three marked with X's are slated to be replaced.

 

[dsinned]

Speaking of MTBF, I have updated the thread title. The MTBF of the M215s has steadily increased and is currently above 1800 years (!) while the MTBF of the M190s has steadily decreased and now stands at 50 years. I suppose the M215 MTBF would be 900 years if I included dsinned's failure, but I won't add it to the spreadsheet without a link. Still 900 years is more than adequate for this product. For a 10 kW system, anything above about 400 years limits the failures to about one every 10 years.

Just a thought...

RegGuheert, here's the (public) link to my Enphase system. Is this what you need?

https://enlighten.enphaseenergy.com/public/systems/TRVh161550

 

[RegGuheert]

RegGuheert, here's the (public) link to my Enphase system. Is this what you need?

https://enlighten.enphaseenergy.com/public/systems/TRVh161550

That's it! Thanks! I've added your system on lines 141 and 142 of my spreadsheet. The MTBF of your system is currently 37 years. Please have a look and ensure that all of my entries are correct.

BTW, thanks for putting a detailed description and pictures on your system page! It's nice to learn more about other systems.

I have modified the thread title to reflect an MTBF of 925 years for the M215s.

 

[dsinned]

Yes, the information you added to the spreadsheet is correct. Note, my array consists of 17 M215s. The initial installation (February 12, 2013) included the first 14, of which one failed, followed by a subsequent installation (May 3, 2013) of the final 3 PV modules. The one M215 that failed was initially located in the upper right corner.

The most unusual aspect of my system is that it also incorporates an Enphase "RGM" (Residential Grade Meter). However, that wasn't installed until May 17, 2013.

With the RGM, Enlighten has two ways of presenting production measurements. When there is a "M" shown next to the measurements, this denotes measurements based on the RGM. When the "M" is not shown, then the measurements are based on the Microinverters themselves.

Surprisingly, the accuracy of the latter has been very good compared to the RGM, although there should be an obvious small measureable difference caused by copper losses in the wiring. The RGM is located right next to the main power (utility) meter, so its measurement data is AFTER those losses between the individual micros on the roof and the RGM's connection to the AC Mains just before the distribution power buss on the load side.

Another interesting revelation is that the two meters read out almost exactly in perfect correlation. How do I know this? Easy, I just turn OFF all the main power distribution breakers to all the loads in the house except the set of breakers to the PV system and compare the readings on the two meters. They both read almost exactly the same number of kilowatts at peak sun load, which means the utility company's meter as is Enphase's RGM must be extremely accurate.