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

pclifton said:

Often I have wanted to look back further in time, but I don't know how to do that.

Click to expand...

I shared my Python 3 script to archive events with RegGuheert a few days ago. We tweaked the script to add some delays for timeouts with slower Envoys. The script works with the first and second generation Envoys (apparently not the Envoy-S).

Click to expand...

Hi Ken,
I took a look at the script and saw the first call is to /event. On the Envoy-S, this is restricted (401 Unauthorized) and requires the envoy/last-6-serial-num to pass the request. However even then a request never resolves and simply times out.

That all said, /datatab/event_dt.rb seems to update OK on my Envoy-S without any extra calls to /event. It appears to keep the last 500 records (based on some of the json values), but I only have one on both the Envoy-S and Enlighten Manager, so I can't personally verify that.

I took a look at the script and saw the first call is to /event. On the Envoy-S, this is restricted (401 Unauthorized) and requires the envoy/last-6-serial-num to pass the request. However even then a request never resolves and simply times out.

Click to expand...

Thanks ltbighorn.

If I am understanding you correctly, then folks with the Envoy-S could just comment out the first call to the EventLog page and the script should work.

It was difficult to find out that the EventLog page had to be visited first on the older Envoys. If you think about it, many of us are hitting the EventLog page daily, and that was refreshing the table. I discovered the issue after not hitting the EventLog page for a few days and I noticed that I was getting three day old information in the archive table.

Thanks for the information on the Envoy-S

Ken Clifton

I may have just experienced the first failure of a four-generation inverter including an integrated ground. (To my knowledge, all failures of M215s that I have recorded to date were the early ones without the "IG" suffix.) I currently have four of the latest-style fourth-generation M215IGs on my roof that include the metal case. They were built in week 45 of 2015. One of those four inverters malfunctioned yesterday. It had a dropout at about 12:30 PM with no event recorded in the log. Then at about 2:00 PM it stopped producing power. About an hour later it reported "DC Power Too Low" in the event log and did not produce any power for the rest of the day. This was true even though yesterday was the only fully-sunny day we have had so far this month. Very strange behavior.

This inverter did wake up this morning and start producing the same power as its neighbors. If this unit is failed, that will be a rather disappointing development. If so, hopefully it is just a random failure quite early in life (this unit was installed almost exactly six months ago). Six months is about how long it was before my first M190 failure.

I'll update on this microinverter if there are any further malfunctions.

RegGuheert said:

This was true even though yesterday was the only fully-sunny day we have had so far this month. Very strange behavior.

Click to expand...

Maybe, maybe not. One hypothesis could be that the panels are producing a surge of power, which is stressing the inverters.

I noticed that all of my failures seemed to happen on cold sunny days. Electronics are more efficient in the cold, and the panels could be producing more power than usual.

Like I said, just a hypothesis. I don't know if it makes sense, but it seems to fit my small number of data points.

GetOffYourGas said:

RegGuheert said:

This was true even though yesterday was the only fully-sunny day we have had so far this month. Very strange behavior.

Click to expand...

Maybe, maybe not. One hypothesis could be that the panels are producing a surge of power, which is stressing the inverters.

I noticed that all of my failures seemed to happen on cold sunny days.

Click to expand...

Interesting. This event excepted, I have never had a failure in the wintertime. The six M190 failures I have experienced happened between April and September.

GetOffYourGas said:

Electronics are more efficient in the cold, and the panels could be producing more power than usual.

Click to expand...

True enough. In particular with photovoltaics, the voltage goes up when it gets cold. But it wasn't cold here yesterday. The temperature at the time of the malfunctions was above 55F. Production of the malfunctioning microinverter and its neighbors at the time the power dropped was about 190W. I wouldn't think such a low power level would put undue stress on an M215IG, which can put out 225W.

GetOffYourGas said:

Like I said, just a hypothesis. I don't know if it makes sense, but it seems to fit my small number of data points.

Click to expand...

It's interesting that our experiences are so different in this regard.

Well, there goes that! Thought I'd throw it out there and see if it sticks.

I'm a bit concerned that the M215IG's are failing so soon. You probably don't have enough data to come up with a meaningful MTBF for them, is that correct?

Please excuse the tangent...
I'm trying to settle on which micro-inverters to buy for my 340 watt panels.
I realize you Enphase folks have experience with what you own but if you could it again and wanted module level MPPT, what manufacturer would you choose ?

GetOffYourGas said:

Well, there goes that! Thought I'd throw it out there and see if it sticks.

Click to expand...

After looking at the notes in my spreadsheet, I went back and looked at your posts about your failures:

GetOffYourGas on August 29 said:

Your note says that I report a failure in January 2015. I think the failure was actually August 2014. I tried looking it up, but my contractor removed the old microinverter from the list and I'm having a hard time figuring out exactly when it happened.

Click to expand...

GetOffYourGas on June 20 said:

BTW, unrelated, but I had a second M215 fail on me last week. It started to sputter out in direct sun on Tuesday and Wednesday (still worked in the morning/evening). Then Thursday it finally gave up the ghost and quit for good. Time to call the installer back.

Click to expand...

GetOffYourGas on August 1 said:

I have officially lost the third M215 inverter. It was a very hot July, at least for Syracuse (day time highs breaking 90F for about two weeks out of the month). The inverter would start out the day just fine. As the day got hotter and the sun got more direct, the inverter would flicker for about an hour and then cut out completely. Sometimes it would come back to life in the evening, other times not until the following day. It finally gave up the ghost and hasn't been on at all for about a week.

Click to expand...

Those three posts indicate that your failures all occurred in hot weather, not cold weather. That matches my experience with M190s fairly well.

GetOffYourGas said:

I'm a bit concerned that the M215IG's are failing so soon. You probably don't have enough data to come up with a meaningful MTBF for them, is that correct?

Click to expand...

That's right. Unfortunately, Enphase cut off the spigot for data on all systems about the same time the M215IGs and M250s came out in 2013. As such, most of the systems in my spreadsheet are the older M215s like the ones you have. The only M215IGs which I am sure about are my 28 and QueenBee's 15. OTOH, the M250s should have nearly-identical architecture and currently indicate no failures after nearly 300 device-years.

I don't consider MTBF numbers to be overly meaningful until there are at least five failures. Early in the life of the M190s when there were very few failures, the calculated MTBF was over about 200 years. This is likely because the failure rate may not be constant over the life of the products. The other confounding factor with Enphase microinverters is that some installations have a much higher failure rate than others. We've theorized about why that might be, but the fact is that we are just guessing. Unfortunately, you own the system with the highest failure rate of M215s that I know about.

BTW, the M215IG that malfunctioned yesterday is producing power today, but it is producing about 8% less than the adjacent units (which were from the same manufacturing batch).

So much for going by memory

Looks more like it's my memory that's going :lol:

SageBrush said:

I'm trying to settle on which micro-inverters to buy for my 340 watt panels.
I realize you Enphase folks have experience with what you own but if you could it again and wanted module level MPPT, what manufacturer would you choose ?

Click to expand...

For me anyway, that is a real tough call. The things that worry me right now don't just relate to the reliability of the product. The warranty is only as good as the longevity of the company and it's financial condition. You might want to do some due diligence on Enphase's current financial situation.

I have been continuously looking at all the micro-inverter options. There are APS, Sunny, Chilicon Power and now Magnium Energy has a new storage-ready micro-inverter (Micro GT-500) too.

Like many folks here I went through the battery degradation with two 2012 LEAFs then the Enphase M190 failures. As RegGuheert will tell you the later series Enphase micros (so far) have been much better. There are more Enphase micros out in the wild than any of the rest... But, I can't say "yes" right now with the current financials... Hopefully a year from now things will be different. I really like the new Enphase design coming down the pike.

Ken Clifton

pclifton said:

SageBrush said:

I'm trying to settle on which micro-inverters to buy for my 340 watt panels.
I realize you Enphase folks have experience with what you own but if you could it again and wanted module level MPPT, what manufacturer would you choose ?

Click to expand...

For me anyway, that is a real tough call. The things that worry me right now don't just relate to the reliability of the product. The warranty is only as good as the longevity of the company and it's financial condition. You might want to do some due diligence on Enphase's current financial situation.

I have been continuously looking at all the micro-inverter options. There are APS, Sunny, Chilicon Power and now Magnium Energy has a new storage-ready micro-inverter (Micro GT-500) too.

Like many folks here I went through the battery degradation with two 2012 LEAFs then the Enphase M190 failures. As RegGuheert will tell you the later series Enphase micros (so far) have been much better. There are more Enphase micros out in the wild than any of the rest... But, I can't say "yes" right now with the current financials... Hopefully a year from now things will be different. I really like the new Enphase design coming down the pike.

Ken Clifton

Click to expand...

Thanks Ken.

Do you have an opinion about SolarEdge ?

SageBrush said:

Please excuse the tangent...
I'm trying to settle on which micro-inverters to buy for my 340 watt panels.
I realize you Enphase folks have experience with what you own but if you could it again and wanted module level MPPT, what manufacturer would you choose ?

Click to expand...

I have to agree with Ken, the main concern really comes down to their financial situation and their ability to support and warranty devices going forward. They have great products, but it's a significant downside if they're not in business anymore, and with their current financials it's hard to say what the outcome is going to be.

Anyone seeing strange data from Enlighten lately?

Two days in the last week there's been periods in the afternoon where pvoutput, pulling from the Enphase API, shows 0W for my array. On the 24rd from 3:35pm PST to 4:35pm PST, and on the 25th from 3:30pm-3:35pm PST. Readings outside these windows are normal.

Power produced as show in Enlighten Manager and MyEnlighten both show no such gaps. The pull to pvoutput isn't just one time either -- if I correct the values manually after the sun goes down, the next morning pvoutput pulls one more time and the gaps are repopulated.

Even more oddly...
In investigating, I find the "Site Energy Production" report on Enlighten Manager also shows very strange output. It purports to show me the values for the dates I've requested, off into future (2018!), but in reality it's really giving me the values from system install (July 2016) through yesterday. Anyone else seeing similar?

Site Energy Production
Generated on 1/26/2017
This report summarizes the daily energy production of the system from 1/20/2017 to 1/2/25/2017
Date/Time Energy Delivered (Wh)
01/20/2017 9,765
01/21/2017 10,219
01/22/2017 16,286
01/23/2017 15,897
01/24/2017 24,237
01/25/2017 23,913
01/26/2017 21,456
01/27/2017 23,644
01/28/2017 18,479
01/29/2017 14,637
01/30/2017 17,269
01/31/2017 9,814
02/01/2017 16,801
02/02/2017 19,216
02/03/2017 19,274
02/04/2017 20,782
02/05/2017 22,759
02/06/2017 24,085
02/07/2017 21,374
02/08/2017 19,059
....
08/11/2017 11,718
Total 2,706,326

SageBrush said:

pclifton said:

<snip>There are APS, Sunny, Chilicon Power and now Magnium Energy has a new storage-ready micro-inverter (Micro GT-500) too.<Snip>

Click to expand...

Do you have an opinion about SolarEdge ?

Click to expand...

I think SolarEdge is a very good option. We have a few threads on SolarEdge here that you can review:

Solar Edge Optimizers?
SolarEdge HD-Wave Inverter Technology
Failed SolarEdge optimizer causes entire string to fail

That last link recently had a drive-by posting which read:

12345678t said:

I have a small commercial system that is about 4 months old. and have had 6 optmizers fail to date. System was working fine and just today tripped offline.

Click to expand...

The bottom line is that we know that all of the options for module-based optimization will experience some failure rate. The question is this: Which ones are best? My spreadsheet attempts to quantify Enphase microinverter real-life MTBF numbers. It's not perfect, but I believe it is in the ballpark for the M190s. It should be closing in on the M215s and M250s. Additionally, it gives you an idea of the RANGE of experiences different users have. How much of that is related to the installer rather than the manufacturer is anyone's guess. (Don't discount the ability of a sufficiently-bad installer to ruin the reliability of your equipment! GetOffYourGas suspects that might account for his issues, at least partially.)

But there is NOTHING to compare against the Enphase results on my spreadsheet. While it's true that the M190 failure rates are way too high, it currently looks like the failure rates for the M215s and M250s are quite good. Unfortunately, we do not have nearly enough data for the S280s to form any opinion. We know that the architecture is completely different from the M215s and M250s, but we also can expect that Enphase has applied all of their experience to try to reduce field failures.

I suspect the same type of analysis I have done with Enphase could be done with SolarEdge, but I have not pursued it.

QueenBee has this little bit of information on APS.

Since you have not included your location in your signature, it is difficult to for us to say whether or not the S280 would be sufficient for 340W PV modules. My original installation employed PV modules rated for 20% more power than the microinverters were capable of producing and the production throughout the year was only reduced by a nominal amount (likely less than 1%). Your power ratio is only slightly higher, so your result should be similar or better if you live at a similar or lower latitude where temperatures are higher.

The bottom line is that if I were installing 340Wp PV modules, I might be inclined to go with 300W or 400W Optimizers from SolarEdge and purchase a central inverter that could take a battery from either Tesla or one of their competitors. (Just beware that adding the battery will both increase the initial cost AND the recurring cost, since it will likely be the least durable component.)

ltbighorn said:

Anyone seeing strange data from Enlighten lately?

Click to expand...

Yes. That problem is still unresolved.

ltbighorn said:

Two days in the last week there's been periods in the afternoon where pvoutput, pulling from the Enphase API, shows 0W for my array. On the 24rd from 3:35pm PST to 4:35pm PST, and on the 25th from 3:30pm-3:35pm PST. Readings outside these windows are normal.

Click to expand...

I see those types of anomalies from time to time on Enlighten, not just for my data, but sometimes for others' data, as well. They typically get cleared up after a while.

ltbighorn said:

Even more oddly...
In investigating, I find the "Site Energy Production" report on Enlighten Manager also shows very strange output. It purports to show me the values for the dates I've requested, off into future (2018!), but in reality it's really giving me the values from system install (July 2016) through yesterday. Anyone else seeing similar?

Click to expand...

No, but I like the sound of that feature! What do I need to type into my browser to get the Envoy to tell me my future solar production? If I had that information, I'm pretty sure I could get rich very quickly!

RegGuheert said:

No, but I like the sound of that feature! What do I need to type into my browser to get the Envoy to tell me my future solar production? If I had that information, I'm pretty sure I could get rich very quickly!

Click to expand...

It's very user-friendly. Just follow these three steps: Click "Site Energy Report", click "Submit!", and you're done. There's is no step 3! The best part is that it also enables July-level production in January. :mrgreen:

Thanks very much everybody for the input and links.

One more question, if I may: The SE inverters vary little in price from 3.4 to 6.0 kW.
Any downsides to be aware of in oversizing the inverter for the expected production ?

If not I'd like to buy the bigger inverter for future proofing, a larger resale market if needed, and perhaps longer longevity ?

SageBrush said:

Thanks very much everybody for the input and links.

One more question, if I may: The SE inverters vary little in price from 3.4 to 6.0 kW.
Any downsides to be aware of in oversizing the inverter for the expected production ?

If not I'd like to buy the bigger inverter for future proofing, a larger resale market if needed, and perhaps longer longevity ?

Click to expand...

There may be other factors, but at least one is that an inverter is less efficient below certain output levels. For SolarEdge you can find some inverter efficiency charts here:

http://www.solaredge.com/sites/default/files/application_note_solaredge_inverters_efficiency.pdf

The difference may not matter much in practice, especially when you consider that the worst efficiency comes when your power output levels are low. i.e. you're losing an extra percent or two of small amounts, rather than large amounts.

There could be other concerns like minimum startup levels but someone more familiar with SolarEdge would have to tell you. That may not be an issue with the optimizers.

SageBrush said:

One more question, if I may: The SE inverters vary little in price from 3.4 to 6.0 kW.
Any downsides to be aware of in oversizing the inverter for the expected production ?

Click to expand...

Since SolarEdge Optimizers allow for constant-string voltage with a wide range of the number of modules in each string, I think that's a good idea.

SageBrush said:

If not I'd like to buy the bigger inverter for future proofing, a larger resale market if needed, and perhaps longer longevity ?

Click to expand...

Bigger power MOSFETs will run cooler because of their lower on resistance (assuming losses are not dominated by switching losses). However, the driver for the larger power MOSFETs will dissipate MORE power. If the different units employ the same gate drivers, then those components will run hotter in the higher-power units regardless of the power being processed. So I would expect the MOSFETs to be slightly more reliable and the gate drivers to be slightly less reliable. Neither of those two types of components should dominate the reliability calculations, however. Rather, it is likely to be dominated by the electrolytic capacitors (unless you get the HD Wave inverter, which has none). Those should operate at a lower RMS current and therefore should run cooler and experience less-severe thermal cycles every day. That could make a real difference in device reliability.

I'd be inclined to oversize the inverter, myself.

ltbighorn said:

SageBrush said:

There may be other factors, but at least one is that an inverter is less efficient below certain output levels. For SolarEdge you can find some inverter efficiency charts here:

http://www.solaredge.com/sites/default/files/application_note_solaredge_inverters_efficiency.pdf

Click to expand...

Click to expand...

Wow! That's the first time I've seen the HD Wave efficiency curves. If those things are available today, note that you'll be throwing away a couple of percent of your energy at ALL power levels if you go with any of the other inverters. Since the HD Wave inverters are not yet available in the larger sizes, you'll have to trade that off against possible future expansion.

(Note that those efficiencies are for European & Asia Pacific units. U.S. units should be similar.)