Reddy
Posts: 1414
Joined: Fri Feb 11, 2011 3:09 pm
Delivery Date: 18 Aug 2011
Leaf Number: 006828
Location: Pasco, WA

Re: Enphase field MTBF: M190: ~38 Years M215: ~460 Years M250: >300 Years

Thu Jun 08, 2017 11:47 am

I don't have solar or follow this thread, but thought this article might be applicable:

https://www.homepower.com/articles/home-efficiency/project-profiles/heliospiti-5-years-lessons-learned?v=print
Contrary to initial expectations, we have not been as lucky with the reliability of our PV system as we have with the SWH system. Our largest trouble spot was with our Enphase M190 microinverters. As of this writing, seven of 20 microinverters have failed, and others are showing intermittent output. While the Enphase warranty provides replacement microinverters, it does not cover the labor cost of hiring a solar contractor to replace the equipment, nor the expense of the boom lift necessary to service modules on a second-floor, 38° metal roof. Replacement of the failed units is averaging $180 per unit. The graph illustrates the decline of our PV production due to the periodic inverter failures, which began in 2012.
Reddy
2011 SL; 9 bar, 46.44 AHr; 40,067 mi; rcv'd Aug 18, 2011
Long: http://www.mynissanleaf.com/viewtopic.p ... al#p226115"
Cold: http://www.mynissanleaf.com/viewtopic.p ... 60#p243033"

User avatar
RegGuheert
Posts: 5522
Joined: Mon Mar 19, 2012 4:12 am
Delivery Date: 16 Mar 2012
Leaf Number: 5926
Location: Northern VA

Re: Enphase field MTBF: M190: ~38 Years M215: ~460 Years M250: >300 Years

Thu Jun 08, 2017 12:59 pm

GetOffYourGas wrote:Line 182 - no further failures. Still waiting for the weather to get hot (it was 40F yesterday morning - at least that's good for production!)
Thanks! Updated. The MTBF of the M215 microinverters in your system is approaching 30 years.

As you know, you have reported the majority of the M215 failures of which I am aware, so I am particularly interested to see if this trend continues or not.

Along the same lines, the one failure of an M215IG that I experienced in my system in January is ALL of the M215IG failures that I know about. The difference is that the 29 M215IGs and 13 M250s in my system come from six different batches (12 + 12 + 3 + 2 + 9 + 4 units) and have five different hardware versions, so I may be more immune to the possibility of having an entire batch of inverters which have a common manufacturing problem. Of course, if there is a common design problem, I'm not immune. In your case, the MTBF of the M215s is established by so many units for enough years that I seriously doubt that a design flaw is to blame. Of course we will learn more as time moves on.
RegGuheert
2011 Leaf SL Demo vehicle
2011 miles at purchase. 10K miles on Apr 14, 2013. 20K miles (55.7Ah) on Aug 7, 2014, 30K miles (52.0Ah) on Dec 30, 2015, 40K miles (49.8Ah) on Feb 8, 2017.
Enphase Inverter Measured MTBF: M190, M215, M250, S280

User avatar
RegGuheert
Posts: 5522
Joined: Mon Mar 19, 2012 4:12 am
Delivery Date: 16 Mar 2012
Leaf Number: 5926
Location: Northern VA

Re: Enphase field MTBF: M190: ~38 Years M215: ~460 Years M250: >300 Years

Thu Jun 08, 2017 2:00 pm

Reddy wrote:I don't have solar or follow this thread, but thought this article might be applicable:

https://www.homepower.com/articles/home-efficiency/project-profiles/heliospiti-5-years-lessons-learned?v=print
Contrary to initial expectations, we have not been as lucky with the reliability of our PV system as we have with the SWH system. Our largest trouble spot was with our Enphase M190 microinverters. As of this writing, seven of 20 microinverters have failed, and others are showing intermittent output. While the Enphase warranty provides replacement microinverters, it does not cover the labor cost of hiring a solar contractor to replace the equipment, nor the expense of the boom lift necessary to service modules on a second-floor, 38° metal roof. Replacement of the failed units is averaging $180 per unit. The graph illustrates the decline of our PV production due to the periodic inverter failures, which began in 2012.
Thanks for the link! Great article and a very interesting home design! It makes me wish I had chosen a drainback system for a solar thermal design which I did a few years ago. I'm not achieving the kind of thermal performance that I see reported there, even in a milder climate. Impressive!

The conclusions the author makes about what he should have done about the inverter failures warrant further discussion:
Jim Riggins in Home Power Magazine wrote:Our takeaway? Scrutinize published reliability claims, and always plan maintenance and accessibility into the house design.
That's very good advice. But back in 2011 (or even 2012) when many of us were installing our M190-based systems, there was no indication of any reliability issues. Enphase had done fairly detailed analysis to try to predict the MTBF of these inverters and summarized their efforts in a couple of white papers on the subject. The 15-year warranty sealed the deal for many of us. Unfortunately, while the measured MTBF of the M190s achieved the promised 331 years that had been promised on the datasheet (at about two years' into my analysis), the failures quickly accumulated after that point. The overall M190 MTBF has settled in at around 40 years and it appears that my system is right about at that point. Unfortunately, some systems have experienced failures rates nearly 10X that average value (over 50X the datasheet value).

Enphase has since decided that printing an MTBF number on their datasheets was a dumb idea. But the good news is that their M215 microinverter currently has a measured MTBF of nearly 500 years in actual operation and the M250 may have a similar result given that no failures have been reported, yet.

But the simple fact is that there is STILL not much information available for new solar purchasers to go on regarding inverter reliability. Enphase certainly has this information, but they do not publish it. And neither do the other manufacturers. My data is the only data which I know of an it is only for the products from a single manufacturer. But even if you want to purchase an Enphase-based system, my spreadsheet does not contain a sufficient number of inverters nor time for the current-generation product to be of any use to anyone. (Although, it might help someone who is interested in purchasing new M215s from a warehouse which is selling them for $59 each right now!)

Ultimately, there is still a trade-off between heavily-shadowed performance (like when there is snow cover) and whether you have any electronics up on the roof.

Side Note: Unless he has experienced a failure and not reported it here (which I doubt), it appears that QueenBee has now operated his system containing 73 M215s for over 320 device-years without a single inverter failure. If his system can maintain that level of reliability, or higher, that should be quite acceptable for long-term operation.
RegGuheert
2011 Leaf SL Demo vehicle
2011 miles at purchase. 10K miles on Apr 14, 2013. 20K miles (55.7Ah) on Aug 7, 2014, 30K miles (52.0Ah) on Dec 30, 2015, 40K miles (49.8Ah) on Feb 8, 2017.
Enphase Inverter Measured MTBF: M190, M215, M250, S280

pclifton
Forum Supporter
Posts: 59
Joined: Thu Jan 26, 2012 4:51 pm
Delivery Date: 28 Jan 2012
Leaf Number: 17731
Location: Salisbury, NC
Contact: Website

Re: Enphase field MTBF: M190: ~38 Years M215: ~460 Years M250: >300 Years

Sun Jun 11, 2017 2:24 pm

Thanks again to RegGuheert for helping test out the Python for my Envoy (Pill-Style).

About two months ago I upgraded one of my Envoys to the Envoy-S . The Envoy-S provides more data in JSON format. That allowed me to create several Python scripts for the newer Envoy-S at: https://tinyurl.com/y83ma46a . I created repositories on GitHub for the code to display production, inventory and archive the eventlog. My favourite script displays per-inverter production locally with no Internet required.

Of course the Envoy-S can have it's internal AP enabled to get to things with the toolkit, but these scripts are really platform-independent. The inventory and per-panel production even work on my iPad and iPhone using the Pythonista app, and they are brutally fast.

Here is a screenshot of my per-panel production, each html table cell has the rgb value set to reflect the power production of that particular inverter:
Image

Ken Clifton
2 PHEVs Powered by 10.2 kW Solar
Topaz Blue 2013 Chevy Volt and Silver 2014 Chevy Volt
Traded My Ocean Blue 2012 LEAF SL on 2-2013
Wife Traded Her Cayenne Red 2012 LEAF SL on 11-2013
Web site: http://www.kenclifton.com

ltbighorn
Posts: 83
Joined: Thu Nov 05, 2015 5:24 am
Delivery Date: 08 Nov 2015
Location: SF Bay Area

Re: Enphase field MTBF: M190: ~38 Years M215: ~460 Years M250: >300 Years

Wed Jun 14, 2017 12:46 am

Today was a bit odd -- in the late afternoon I checked my system on pvoutput and saw that the power output had apparently dropped to zero at 3:30pm.

Checking with my Envoy-S's CTs, the system was actually still producing power the entire time, but from 3:30pm-6:30pm it thought all the microinverters were failing to report. Eventually they all started reporting again, one by one, over a 20 minute window.

Enlighten has since updated to show some flat-line power graph than approximates the average output during the outage time, presumably based on the increased energy total, and this is the data that gets shared with pvoutput via the API. Viewing the full system specifically under "graphs" though and a real power curve is visible -- it must be getting those values from the CTs.

Have others run into this before? I was wondering if maybe the devices were non-responsive while getting a firmware update, but an inventory.json pulled during the period when a couple had started reporting again and the rest were still non-responsive showed no difference in version between the two sets. I hadn't been previously archiving inventory.json (I am now), so I'm not 100% sure. Either way, the system

pvoutput tends to pull from the Enphase API in the morning one final time for the previous day, so tomorrow I'll correct the pvoutput data with my measured values from the Envoy-S CTs. (I have some cron scripts which save and archive the json output every few minutes, just in case.)

GetOffYourGas
Posts: 1609
Joined: Tue Dec 20, 2011 6:56 pm
Delivery Date: 09 Mar 2012
Location: Syracuse, NY

Re: Enphase field MTBF: M190: ~38 Years M215: ~460 Years M250: >300 Years

Wed Jun 14, 2017 7:11 am

ltbighorn wrote:Have others run into this before?


I have seen this many times over the past 5.5 years. It seems to me that the communications between the microinverters and the Envoy get cut off, but the microinverters continue to monitor themselves. When they do finally report, it is the right number of kWh (or it seems to be, AFAICT), but spread over a longer interval. Therefore the Envoy reports just an average value. The integrated power (area under the curve) is correct.
~Brian

EV Fleet:
2011 Torqeedo Travel 1003 electric outboard on a 22' sailboat
2012 Leaf SV
2015 C-Max Energi (302A package)

ltbighorn
Posts: 83
Joined: Thu Nov 05, 2015 5:24 am
Delivery Date: 08 Nov 2015
Location: SF Bay Area

Re: Enphase field MTBF: M190: ~38 Years M215: ~460 Years M250: >300 Years

Wed Jun 14, 2017 4:54 pm

I've had one or two micros fail to report for a few minutes over the last year, but only 1-3 times max. This is also the first time that all of them have gone offline. Sounds like from your experience that's a normal occurrence though and doesn't indicate anything special is going on. Good to know. :) Thanks!

FairwoodRed
Posts: 455
Joined: Wed May 18, 2011 10:25 am
Location: Fairwood, WA

Re: Enphase field MTBF: M190: ~38 Years M215: ~460 Years M250: >300 Years

Fri Jun 16, 2017 8:45 am

This probably isn’t your case, but I routinely lose communication between my Envoy and my microinverters. After a while, I determined that it was because I had a treadmill on the same branch circuit as the Envoy. Everytime I’d turn it on (even not in active use), I’d lose comms. Which just goes to say that powerline communication is subject to noise.

On the reliability side, I just had another failure of the D380’s That is now 10 out of 60 inverters that have failed in 6 years :( I had failures in June 2017, Aug 2016, Jan 2016, Aug 2015, April 2015. I am on line 28.

User avatar
RegGuheert
Posts: 5522
Joined: Mon Mar 19, 2012 4:12 am
Delivery Date: 16 Mar 2012
Leaf Number: 5926
Location: Northern VA

Re: Enphase field MTBF: M190: ~38 Years M215: ~460 Years M250: >300 Years

Sun Jun 18, 2017 6:19 am

FairwoodRed wrote:This probably isn’t your case, but I routinely lose communication between my Envoy and my microinverters. After a while, I determined that it was because I had a treadmill on the same branch circuit as the Envoy. Everytime I’d turn it on (even not in active use), I’d lose comms. Which just goes to say that powerline communication is subject to noise.
Absolutely! So far, I've had two loads in my house which have majorly disrupted commonications between my Envoy and the microinverters:

1) The LEAF's OBC: Our LEAF disrupts our PV system while charging at L2
2) I had a wall-wart for our SlingBox Pro HD fail in a way which seriously disrupted communications with the Envoy
FairwoodRed wrote:On the reliability side, I just had another failure of the D380’s That is now 10 out of 60 inverters that have failed in 6 years :( I had failures in June 2017, Aug 2016, Jan 2016, Aug 2015, April 2015. I am on line 28.
Thanks for the update! I have updated the spreadsheet to reflect these failures. Since each failure of a D380 counts as two, your MTBF now stands at 35 years. That is close to the average I see for the overall population of M190s and D380s.

It's interesting that you have had one of every six D380s fail in six years while I have had one of every seven M190s fail in about six-and-a-half years. (And I likely would have experienced more failures if I had not placed most of my M190s into storage as cold spares about a year ago.)

BTW, what does Enphase send you as replacements for the failed D380s? If they are M190IGs, I'm interested in knowing about them since I now also track their failure rate in my spreadsheet. (Unfortunately, those appear to have the highest failure rate of all Enphase microinverters.)
RegGuheert
2011 Leaf SL Demo vehicle
2011 miles at purchase. 10K miles on Apr 14, 2013. 20K miles (55.7Ah) on Aug 7, 2014, 30K miles (52.0Ah) on Dec 30, 2015, 40K miles (49.8Ah) on Feb 8, 2017.
Enphase Inverter Measured MTBF: M190, M215, M250, S280

iPlug
Posts: 15
Joined: Thu Apr 28, 2016 6:47 pm
Delivery Date: 25 Apr 2016
Location: Rocklin, CA

Re: Enphase field MTBF: M190: ~38 Years M215: ~460 Years M250: >300 Years

Sun Jun 18, 2017 8:25 pm

Reg,

Line 183 update.

Prior reported existing system with 28 Suntech 245 Pluto with 28 M215 Enphase microinverters: still no inverter failures.

System upgraded with additional installation 4/5/17 with 15 LG305 Neon X with 15 M250 Enphase microinverters. No inverter failures. Using same Envoy.
2016 Leaf SV + 2012 Plug-in Prius, 11.43 kW Solar PV, Two Clipper Creek Level-2 7.7 kW charging stations, PG&E E-6 TOU/Tiered rate plan

Return to “Solar”