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

My Nissan Leaf Forum

Help Support My Nissan Leaf Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
Just a quick note to say that the 54 fourth-generation inverters (M215IGs and M250s) in my system have attained an MTBF over twice that of the 42 original third-generation inverters (M190s) which were finally fully retired in October 2017:

Fourth-generation inverter MTBF: 93 years (with 1 failure)
Third-generation inverter MTBF: 42 years (with 6 failures)

With more device-years on the overall population, the numbers are even more stark:

Fourth-generation inverter MTBF (INCLUDING original M215s): 332 years (with 9 failure)
Fourth-generation inverter MTBF (EXCLUDING original M215s): 471 years (with 1 failures)
Third-generation inverter MTBF: 36 years (with 246 failures)

So far it looks like the fourth-generation Enphase microinverters are achieving an MTBF which is about 10X that of the third-generation inverters. Time will tell if this improvement holds up or not.

Please update the status of the systems you are tracking, including row number from my spreadsheet.
 
iPlug said:
Row 200 and row 215, (Rocklin, CA): no system changes, all original inverters as noted, still no failures.
Thanks! That brings the MTBF of the inverters in your system over 161 years.

BTW, with no failures an actual failure rate cannot be established, just a suspected lower bound. I would feel best to have at least five failures to establish some sort of failure rate for a given population. Even then, I find the failure rate of the original M215s in my spreadsheet to be a bit suspect since the failures are nearly all clustered around two systems near North Syracuse, NY. Outside of those two systems, the original M215s have a very impressive record.

Simply put, calculating MTBF is a bit of a messy business. But as time goes on the data here should prove to be more and more useful, particularly if the fourth-generation inverters prove to have a long operating life.
 
philip said:
Row 213 - No failures.
Thanks! Updated. Your measured MTBF is over 50 years now.

The MTBF of the 283 M250s we are tracking is about to pass 400 device-years without a single failure. I'd say that's a pretty impressive feat!
 
RegGuheert said:
Paging Tony Williams...

I looked at your Enlighten site today and noticed a couple of things:

1) Your system stopped reporting on October 20. Did your Envoy die?
2) Enlighten lists 37 inverters even though you have 35 in your array. Have you had 2 M190s die?
Even without any response from Tony, I have decided to record two M190 failures and one Envoy failure for his system. The signature of having additional inverters introduced to the system seems like a clear indication of failures. What it DOESN'T tell me is what replaced the M190s. This updated brings the MTBF for the M190s in Tony's system to 118 years. That is much better than most M190 systems out there.

I will use this approach with other systems in the spreadsheet, but I will NOT update the monitored date if the number of inverters shown matches how many are in the array. In that case, there are really three possibilities: 1) No failures, 2) There were failures but no inverters have been replaced, or 3) There were failures, the inverters were replaced, and Enphase properly removed the old inverters from Enlighten.

In looking around the spreadsheet for other signatures like I see with Tony's system, I ran across a large D380-based system in row 26 where the owners had the ENTIRE system replaced and upgraded in August 2016. I have no idea what possessed Enphase to think that putting two M190s into a single package would be a good idea. Understanding the impact of failures dictates that the D380 had to have a failure rate significantly higher than the M190s. Perhaps the M190 failures had not yet kicked in at the time.
 
ltbighorn said:
Row 212, no failures or other issues.
Thanks! Updated. The M250s will cross over 400 device-years within the next few months and we still have yet to record a single failure.

This week marks a year since I have experienced a failure of any of the inverters in my system: an M215IG which failed on January 25, 2017. Outside of the four failures of M190IGs (which are not properly grounded through their adapter cables) that is the only failure of an "integrated ground" fourth-generation Enphase inverter connected (properly connected using Engage) that I know about. That is after 500 total device-years of operation of M215IGs and M250s.

This week also marks 100 device-years of operation of fourth-generation Enphase inverters in my system: M190IG: 2.5 device-years, M215IG: 69.6 device-years, M250-60: 24.7 device-years, and M250-72: 3.2 device-years. Given the one failure of an M215IG, that puts the MTBF for fourth-generation inverters in my system at 100 years. The oldest fourth-generation inverter has been on my roof for a bit over 3.5 years now. Hopefully it will be many years before a few more units fail, thus establishing an accurate MTBF number for the fourth-generation inverters in this system.

By comparison, the original M190s (third generation) racked up about 249 devices-years before being replaced and experienced 6 failures over 6.8 calendar years, giving an MTBF of about 41 years. It will be about three more years before I have that much calendar- and device-time on the fourth-generation units.

Would anyone else like to provide an update on your Enphase system? TIA!
 
RegGuheert on November 27 said:
These new inverters went into service on October 23, 2017. Over one month later, I am still seeing the "Cycles Skipped" messages associated with large power dropouts. It is not always the same inverters showing this issue.
The new 72-cell M250 inverters have now been producing for over three months and these "Cycles Skipped" messages are still happening. Yesterday there were two different instances involving a total of four of the twelve new inverters.

It's no big deal in terms of production because the inverters recover fairly quickly. But it is odd to see these malfunctions so regularly. My only real concern is that there could be some additional stress induced during these incidents that might eventually damage these inverters.

BTW, on Thursday I reached a milestone of one year since the last microinverter failure (an M215IG) in my system. Let's hope this stretch continues for many more years!
 
SurfHawk said:
Row 221 still going strong with no failures on my 12x S280 inverters.
Thanks, SurfHawk! Updated.

That brings the total number of device-years for all S280s up to 86 with no known failures. So far so good!

The M250s are the current MTBF leaders. They are approaching 400 years with no known failures. The oldest units are nearly four years old. By that point, many M190s were already starting to fail.

pclifton has told me separately that he has a friend who owns many M215s as well as many M250s. He reports that quite a few M215s have failed, but only one M250. Unfortunately, I am unable to add that data into my database since I do not have any visibility of it. It's fine, since we are accumulating more and more data as time goes on.
 
Now that the snow is melting, I have begun watching my panels again (it's really boring to see 0Wh produced every day). I noticed that another M215 inverter has bitten the dust. Looking through its history, it stopped reporting on December 19, 2017 at 10:20am, and has not produced anything since. I contacted my installer, and he says that they are in high demand because of the large number of failures (which of course doesn't match Reg's history). He expects my replacement to arrive in about 60 days. That means I'll miss out on some of the best production months of the year.

I also noticed another inverter which operates normally until it is in direct sunlight. On cloudy days, it will produce the same 50-60 W as the rest of the panels. On sunny days, it will be producing 2-5 W while all other panels are producing 180-200 W. I have never seen this failure mode. It clearly isn't meeting the spec, so I've reported it to my installer as well. I assume enphase will replace it.
 
GetOffYourGas said:
Now that the snow is melting, I have begun watching my panels again (it's really boring to see 0Wh produced every day).
We've been in the "Nor'easter Hole" this year and only one of the four dumped on us. Our array had snow on it only for about two days in March and we beat our previous March production record by 61 kWh.
GetOffYourGas said:
I noticed that another M215 inverter has bitten the dust. Looking through its history, it stopped reporting on December 19, 2017 at 10:20am, and has not produced anything since.
I recorded that as a failure of one of your original inverters on the date you gave.
GetOffYourGas said:
I contacted my installer, and he says that they are in high demand because of the large number of failures (which of course doesn't match Reg's history). He expects my replacement to arrive in about 60 days. That means I'll miss out on some of the best production months of the year.
I think the real issue is that the M215s are no longer a standard product so they probably only produce a batch now and then.
GetOffYourGas said:
I also noticed another inverter which operates normally until it is in direct sunlight. On cloudy days, it will produce the same 50-60 W as the rest of the panels. On sunny days, it will be producing 2-5 W while all other panels are producing 180-200 W. I have never seen this failure mode. It clearly isn't meeting the spec, so I've reported it to my installer as well. I assume enphase will replace it.
I'm sure they will replace it. The key is that it needs to produce less than 90% of its neighbors over a 7-day period. You didn't mention which type of inverter this was so I recorded it as a failure of one of the original M215s, just like the other one.

I will add two new M215IGs once those are installed.

Assuming these were both failures of the original M215s you had installed, you now have had 7 out of 16 fail after about 6.3 years of service. The spreadsheet is not overly friendly for calculating the MTBF of your system since each replacement is with a different type, but I did the calcs and I come up with 92.33 total device years and an MTBF of only 13.2 years after seven failures for the original M215s. You only have about 8.2 device-years on your M215IGs and no failures, so MTBF is really not established there.

By way of comparison, I have seen the following MTBFs in my system to date:

- M190s: 249 device-years - 6 failures - MTBF = 35.5 years (These are out of service now, but I intend to run 12 this summer.)
- M190IGs: 2.7 device-years - 0 failures - MTBF = ??
- M215IGs: 75 device-years - 1 failure - MTBF = 75 years (not well-established)
- M250s: 32 device-years - 0 failures - MTBF = ??
 
RegGuheert said:
We've been in the "Nor'easter Hole" this year and only one of the four dumped on us. Our array had snow on it only for about two days in March and we beat our previous March production record by 61 kWh.

Yeah, the back-to-back Nor'easters have been crazy. Although most of them haven't reached inland as far as Syracuse. It's mostly lake effect this year. We've had very few thaws through the winter and most were immediately followed by more lake effect snow.

RegGuheert said:
I'm sure they will replace it. The key is that it needs to produce less than 90% of its neighbors over a 7-day period. You didn't mention which type of inverter this was so I recorded it as a failure of one of the original M215s, just like the other one.

In that case, it is definitely covered. The question is whether my contractor is paying enough attention to the details to realize that there are two in need of replacement.

Over the past 7 days, it has produced 1.47kWh. Its next lowest neighbor has produced 2.50kWh. That's 59%. It is an M215. I haven't had any issues with M215IGs yet.
 
RegGuheert on October 3 said:
But I have now upgraded the wiring on the field array to allow for the transition from M190s to M250-72s. Once I started working on the wiring, I realized that I can easily accommodate BOTH M190-style wiring AND Enphase Engage wiring. This arrangement allows me to install any arbitrary combination of third- or fourth-generation Enphase inverters on that array.

Given this newfound flexibility, I am now in a bit of a quandary. The M250-72s have a peak production capability which is 612 W higher than that of the M190s (3000 W versus 2388 W). Since the array is ideally pointed during the coldest months, this will mean some additional production when it is lowest. OTOH, the M190s do not limit their output power during the other half of the year. Since my preference would be to keep the hours OFF the M250-72s as much as possible AND I would also prefer to put as many hours on the M190s as possible, I am thinking it may be best to operate the M190s during the summertime between the Spring and Autumn equinoxes when I adjust the array pointing and operate the M250-72s during the other half of the year. The goal is to try to encourage as many M190 failures as possible (to get fourth-generation replacements as spares) while limiting M250-72 failures as much as possible.

I'll probably give this a try during 2018 and see how much of a PITA it turns out to be.
Well, I made the swap today before the sun made it above the horizon. It only took about 15 minutes to do: I unconnected the DC from the 12 M250-72s, moved the weatherproof plugs from the M4 connectors on the M190s to the M4 connectors on the M250s and then connected the DC wires from the PV modules to the M190s. Done! Finally, I plugged in the original "pill" Envoy to talk to the old inverters and told the new one to ignore the 12 M250-72s for now.

It will be harder to swap them back in the Fall only because it is difficult to unmate the MC M4 connectors on the PV cables from the Amphenol H4 connectors which came on the M190s.

Let's see if I can get any M190s to fail during the summertime this year.
 
RegGuheert said:
The M250-72s have a peak production capability which is 612 W higher than that of the M190s (3000 W versus 2388 W). Since the array is ideally pointed during the coldest months, this will mean some additional production when it is lowest.
I take it you have actually observed clipping with the M190s? What is the economic value of the foregone energy if you left them in all year round?

Cheers, Wayne
 
wwhitney said:
I take it you have actually observed clipping with the M190s? What is the economic value of the foregone energy if you left them in all year round?
It's quite small. I've done the calcs once-upon-a-time for the 235 Wp PV modules on the roof which have an elevation angle of about 30 degrees. I believe I calculated that the lost production was less than about 300 kWh/year (out of about 14.5 MWh/year that array produces), which is around 2%.

But the field array gets an elevation angle of 60 degrees in wintertime, so it is well-pointed in the colder months. I have even seen all of the inverters in the field array producing 250 W each. I wouldn't be surprised if the field array might produce 5% or more additional electricity in the wintertime with the higher-power inverters. It's still not a huge amount of electricity.

Just to be clear, I did not purchase the M250s to harvest more electricity. I purchased them to extend the warranty by 17 years to cover the period from January 2026 to December 2042. But since I bought them, I realized that I can easily set them up to swap back and forth to take advantage of the higher output power capability. The main drawbacks are the effort to swap them back and forth and the risk to the not-active inverters by being connected to A/C even when they are not in use.
 
RegGuheert said:
Let's see if I can get any M190s to fail during the summertime this year.
Wow! That was fast! One of the M190s is now reporting a "Hardware Failure".

My evil plan is working! :twisted:
 
Back
Top