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.
ltbighorn said:
I'm probably going to pull the trigger in the next week or two on the solar install. Probably going to go with 12x 280W LGs (LG280N1C-G3) with Enphase M250s, as recommended by my installer. It sounds like they intend to pair it with the regular 2-year warranty Envoy.
Awesome! I'm sure you'll be happy with it! I don't have enough data on the M250s to establish the MTBF of those units yet, but I also do not know of any failures. If their MTBF is anything like the M215s (>700 years) then those should be great! (I just picked up another new M250 on ebay this week for $99.99 shipped. I find it hard to pass them up at that price!)
ltbighorn said:
I'm hoping I won't have to preemptively go to quite such lengths re: spares.
I don't see why you (or most people, for that matter) would need to do that. It's just that I have these 12 72-cell panels which I would like to keep running for a long time to come. For 60-cell panels, there will be *something* available for a long time to come, as they are still the standard.
ltbighorn said:
Are there any good reasons to push for and/or pay the additional cost for an Envoy-S (besides the 5-year warranty)?
Since you are not getting the S-series inverters, the only reason I can think to go with the Envoy-S would be for utility-grade data accuracy, and you only get that with the higher-end unit.
ltbighorn said:
Not sure how much the cost spread is for new Envoy vs new Envoy-S. MTBF looks OK for the Envoy on the spreadsheet, but anecdotally it seems like a common failure point.
Frankly, it seems most of the failures are with the older units. I suspect the newer Envoys have improvements that help them to last longer. My Envoy is older than the length of the warranty period on the new Envoy-S, so I suspect all you really get for your money is a longer warranty and support for the S-series inverters.
ltbighorn said:
I expect to go ahead and pay the extra $1k or so it'll cost to have the roofer seal ~20 solar penetrations and include in my roof guarantee. Still on the fence whether to pony $250 for Enlighten Manager and individual panel data. With the M190 it seems like a necessity to be informed of the fairly frequent issues, but with the M215 and M250, perhaps it's unnecessary? It just makes me nervous not being able to tell if a panel is underperforming, given that without per-panel stats, I have no ability to tell whether variations over time are due to panel issues or just variations in solar radiation.
If it's a one-time payment of $250, then I'd recommend you spring for it. If your system lasts 25 years, that only comes to $10/year.

Good luck with your installation!
 
RegGuheert said:
Frankly, it seems most of the failures are with the older units. I suspect the newer Envoys have improvements that help them to last longer. My Envoy is older than the length of the warranty period on the new Envoy-S, so I suspect all you really get for your money is a longer warranty and support for the S-series inverters.
That's reassuring to hear. I'll certainly pay attention more attention to eBay if I hear about Enphase stopping production of anything (without a suitable replacement).

One thing I noticed when looking at PG&E's NEM Program Tracking page today ( http://www.pge.com/en/mybusiness/save/solar/nemtracking/index.page ) is a section with the following info, which I hadn't noticed before:

Total NEM Requests Pending as of 4/24/16: 16,193 Quantity, 432.15 MW
Remaining MWs to NEM Cap Assuming All Pending Requests on the Line Immediately Above Complete Their Projects and Provide PG&E With the FInal Building Permits Required To Be Approved for NEM: 0.00 MW

I'm a bit confused about why people would be submitting to PG&E unless they had their final permits/inspections approved etc. And if they're just waiting for a permit, wouldn't most of that happen within a month? I can't believe 432.15 MW is going to get installed next month, when the last 6+ months seem to be about 40-50 MW/month, so they must be early in the projects? But why inform PG&E if it's so early?

I had previously figured that at 50 MW/month and 396 MW remaining, there'd be about 8 months of breathing room. Even if installs rose to 75 MW/month it'd be a little over 5 months (September '16) remaining.

It looks like I may not be able to get the roofer out to redo the roof until June 22nd. Assuming I can't get the roof done any earlier than that (8 weeks from now), how big is the risk I'll miss NEM 1.0? My solar installer seems pretty quick, so I imagine they'd finish their work within a week of that.
 
RegGuheert said:
ltbighorn said:
Still on the fence whether to pony $250 for Enlighten Manager and individual panel data. With the M190 it seems like a necessity to be informed of the fairly frequent issues, but with the M215 and M250, perhaps it's unnecessary? It just makes me nervous not being able to tell if a panel is underperforming, given that without per-panel stats, I have no ability to tell whether variations over time are due to panel issues or just variations in solar radiation.
If it's a one-time payment of $250, then I'd recommend you spring for it. If your system lasts 25 years, that only comes to $10/year.
Realized on the upside that the installer has to pay this fee for me, so I should be able to roll it into the overall cost for the system and benefit from the 30% ITC tax credit, lowering the effective cost to $175. Will ask for it.
 
Envoy-S and consumption metering
So looks like the cost difference between the Envoy and the Envoy-S are effectively nil - my installer will provide an Envoy-S at no additional cost. Cool. It looks like the Envoy-S Metered is about $100 more, and adds revenue-grade metering plus the capability of optional consumption monitoring ($50 for consumption sensor coils).

While it'd be nice to see my own consumption data, I'm not a fan of depending Enphase to feed it to me on terms I'd like. Getting a Rainforest EAGLE ZigBee gateway that pulls data from the Smart Meter seems to make more sense in terms of wiring (none) and data independence -- can this data be imported into Enphase's site, or something similar to get a whole-house view?

It seems like the Metered model would mainly be of benefit for someone who planned to take advantage of Enphase AC batteries, as it would allow the system to understand the load side of things and thus optimize battery charge/discharge patterns.

S280 microinverters
I was also looking at the new S280 inverters (vs the M250s). At wholesale prices, they seem similarly priced or perhaps "only" $10-20/unit more, (though I don't know if my installer is getting better pricing on the M250s), with the benefits of greater efficiency, higher max input power, and some sort of smart-grid capabilities that seem like they primarily benefit the utilities.

On the downside, the S280 is probably more than I need for 280W panels, and because the AC output rating is 280 VA, I could only put 14 of these these on a 20A branch circuit. The S230 doesn't seem quite yet available and has too low a max input rating (275W). And of course these have an untested reliability history. Meanwhile the M250 would allow up to 16 inverters in a 20A branch circuit. While today I'll probably only install 12-13 panels, under NEM 1.0 I could add up to 1 kW, or ~3 panel panels. With the S280 I'd be limited to 1-2 extra, with the M250s there'd be no issue. My panel is pretty full, and a smaller input feed makes it easier to deal with the panel backfeed limits/requirements.

So I'll probably skip inquiring to the solar installer about the S230/S280s.


The other thing to figure out is if I'll be able to squeeze a useful number of extra slots for circuits on a sub-panel, for potential future use. Not sure how the rules on sub-panel sizes/feeds work when you don't plan to connect anything to them soon (no change in demand load calculations), but I guess that's something to research.
 
Nice post, ltbighorn!
ltbighorn said:
Envoy-S and consumption metering
So looks like the cost difference between the Envoy and the Envoy-S are effectively nil - my installer will provide an Envoy-S at no additional cost. Cool. It looks like the Envoy-S Metered is about $100 more, and adds revenue-grade metering plus the capability of optional consumption monitoring ($50 for consumption sensor coils).
The cost difference is nil, but the old Envoy has a display while the new one does not. A small thing, but something. OTOH, the new one has a better warranty.
ltbighorn said:
While it'd be nice to see my own consumption data, I'm not a fan of depending Enphase to feed it to me on terms I'd like. Getting a Rainforest EAGLE ZigBee gateway that pulls data from the Smart Meter seems to make more sense in terms of wiring (none) and data independence -- can this data be imported into Enphase's site, or something similar to get a whole-house view?
Thanks! I didn't know about this option. Here is the link to the RainForest Automation website. So you mean that the radio on my meter is completely open and can be read by anyone? Interesting... I wonder how the power company reads it. Through PLC, I suppose.

To answer your question, I seriously doubt that you can import it into Enlighten. That is part of Enphase' business model (owning the data). I don't know about third-party options, but I'm interested to learn more about this from you or others.
ltbighorn said:
It seems like the Metered model would mainly be of benefit for someone who planned to take advantage of Enphase AC batteries, as it would allow the system to understand the load side of things and thus optimize battery charge/discharge patterns.
Agreed, but as someone who lives in CA, you should take that into consideration. Net metering will certainly come under fire from your utility sooner rather than later. Definitely within the 25-year life of your system. As such, the AC batteries are something you may have need of before you would like. If the Envoy-S with metering is a requirement for the AC Battery, then you really might want to consider springing for that.
ltbighorn said:
S280 microinverters
I was also looking at the new S280 inverters (vs the M250s). At wholesale prices, they seem similarly priced or perhaps "only" $10-20/unit more, (though I don't know if my installer is getting better pricing on the M250s), with the benefits of greater efficiency, higher max input power, and some sort of smart-grid capabilities that seem like they primarily benefit the utilities.
Agreed this is primarily for the (future) benefit of the utilities. But putting that another way, power factor correction *might* have *real* value to them in the future. I don't know how much, but you never know. Still, since this is an unknown possible benefit, it is best to focus on YOUR requirements first.
ltbighorn said:
On the downside, the S280 is probably more than I need for 280W panels, and because the AC output rating is 280 VA, I could only put 14 of these these on a 20A branch circuit.
You are correct: You do NOT need a 280W inverter for a 280W PV module. You can find proof in this thread. (And it is MORE true in your climate than in mine.) You are also correct that it is better to allow for more inverters on a single string, particularly since you are tight on panel space.
ltbighorn said:
The S230 doesn't seem quite yet available and has too low a max input rating (275W).
Yeah, I'm not sure why they chose 230W. It's a bit small for the current and future PV generations. Why not an S250? My guess it has to do with the ratings on the available MOSFET transistors, but who knows?
ltbighorn said:
And of course these have an untested reliability history.
Yep. Completely unknown. And the next generation sounds like it will be a complete departure from where they are today, which increases the risk for both Enphase and their customers, IMO. (More here.)
ltbighorn said:
The other thing to figure out is if I'll be able to squeeze a useful number of extra slots for circuits on a sub-panel, for potential future use. Not sure how the rules on sub-panel sizes/feeds work when you don't plan to connect anything to them soon (no change in demand load calculations), but I guess that's something to research.
If you'll tell us what brand and model of panel you have, someone might be able to give you some tips.

Frankly, I'm impressed that you are able to digest and comprehend all the implications of these details BEFORE your purchase. Good on you! (Clearly you've gotten WAY beyond "Ooh! Shiny!" since you are posting in a thread about the failure rate of the inverters you are considering. :) )
 
Appreciate your feedback and insights as usual RegGuheert!

Envoy's and metering
RegGuheert said:
The cost difference is nil, but the old Envoy has a display while the new one does not. A small thing, but something. OTOH, the new one has a better warranty.
Good point. In my case I don't anticipate visiting the display in the garage regularly (though it'd certainly be nicer to have it than not for quick sanity checks). As long as both have the local API I can query for basic production status, then I'll be polling and graphing that info myself.

RegGuheert said:
ltbighorn said:
While it'd be nice to see my own consumption data, I'm not a fan of depending Enphase to feed it to me on terms I'd like. Getting a Rainforest EAGLE ZigBee gateway that pulls data from the Smart Meter seems to make more sense in terms of wiring (none) and data independence -- can this data be imported into Enphase's site, or something similar to get a whole-house view?
Thanks! I didn't know about this option. Here is the link to the RainForest Automation website. So you mean that the radio on my meter is completely open and can be read by anyone? Interesting... I wonder how the power company reads it. Through PLC, I suppose.
Yes, that'd be the one. The radio is secured, important as it's considered to have personal behavior revealing data. You need the utility to approve and enable access from your device. With PG&E, as long as it's an approved (certified?) device, that's no problem.

They already have a revenue grade meter installed, so why not, eh?

RegGuheert said:
To answer your question, I seriously doubt that you can import it into Enlighten. That is part of Enphase' business model (owning the data). I don't know about third-party options, but I'm interested to learn more about this from you or others.
I think PVOutput ( http://pvoutput.org/ ) may be the third-party option I was thinking of. With an open API and a lot of auto-uploaders, I believe it can automatically pull the production data from the Enphase Enlighten API, as well as an Eagle Rainforest account (or Wattvision, if you get that version), to generate a net consumption view.

Here's an example from someone at solarpaneltalk, showing consumption and production, both real-time and cumulative for the day. I don't think temperature and voltage are actually being graphed -- not sure if that can be pulled from Enlighten if you have Manager access, or if it has to come from another input source.
fetch


And live version: http://pvoutput.org/intraday.jsp?id=34829&sid=33867&t=0&gs=3&s=1

AC batteries/future proofing
RegGuheert said:
ltbighorn said:
It seems like the Metered model would mainly be of benefit for someone who planned to take advantage of Enphase AC batteries, as it would allow the system to understand the load side of things and thus optimize battery charge/discharge patterns.
Agreed, but as someone who lives in CA, you should take that into consideration. Net metering will certainly come under fire from your utility sooner rather than later. Definitely within the 25-year life of your system. As such, the AC batteries are something you may have need of before you would like. If the Envoy-S with metering is a requirement for the AC Battery, then you really might want to consider springing for that.
Hmmmmmmmm. I do imagine it will come under fire, though with 20-year grandfathering on NEM 1.0 it'd probably be primarily in the form of TOU rate shifting. On NEM 1.0, as long as a non-TOU plan was offered, I would have that option as a fallback. With my system I'd always be low in the lowest tier on a non-TOU plan.

Right now the battery prices look to be pretty extreme, aside from the space costs & safety risks. Obviously battery prices will likely come down dramatically over the next 5-15 years, but I wonder whether I might not end up having to replace the Envoy-S over that time period, or prices come down on them a lot anyway by the time AC batteries are cost effective. The skeptic in me suspects 10 years before it becomes cost effective would be optimistic.

I hope to be in my current home in 10 years, but it's a lot time to try and predict, and longer than my panel payback period (was ~7-8 years, closer to 8-9 with the additional costs of having the roofer do the flashing, Enlighten Manager view, etc, not factoring in PG&E rate increases or severe negative impacts on returns).

Basically I could imagine springing for the Metered unit, and then having it die or become obsolete before I take advantage of it. i.e. a newer Envoy ends up becoming required, or how it's integrated changes and the Envoy installed becomes irrelevant. Still, $150 isn't a terrible premium to risk, needs more thought.. All the little costs do add up though. :)

S280 micros
RegGuheert said:
ltbighorn said:
S280 microinverters
I was also looking at the new S280 inverters (vs the M250s). At wholesale prices, they seem similarly priced or perhaps "only" $10-20/unit more, (though I don't know if my installer is getting better pricing on the M250s), with the benefits of greater efficiency, higher max input power, and some sort of smart-grid capabilities that seem like they primarily benefit the utilities.
Agreed this is primarily for the (future) benefit of the utilities. But putting that another way, power factor correction *might* have *real* value to them in the future. I don't know how much, but you never know. Still, since this is an unknown possible benefit, it is best to focus on YOUR requirements first.
Yeah, I definitely imagine there are benefits providing real value to the grid, but given these are just coming to market, the presence of lack of mine in the system isn't going to break the camel's back. I have trouble imagining PG&E effectively compensating me for the value it provides, though. It seems like it's challenging enough for them to compensate people for requested demand-response behaviors, let alone something subtle like power factor tweaks.

Long detail on my panel situation, can skip here for those not interested
RegGuheert said:
ltbighorn said:
The other thing to figure out is if I'll be able to squeeze a useful number of extra slots for circuits on a sub-panel, for potential future use. Not sure how the rules on sub-panel sizes/feeds work when you don't plan to connect anything to them soon (no change in demand load calculations), but I guess that's something to research.
If you'll tell us what brand and model of panel you have, someone might be able to give you some tips.
Westinghouse 125A end-fed panel with a 100A input feed/breaker. Underground feed from the street (only about 10 ft to street). I believe with the 120% backfeed rule, this means I could have up to a 50A backfeed (125*1.2 - 100 = 50), though I should only need 20A for 12-16 M250s. Because my panel is within 3-ft of the gas meter, upgrading it would not only require getting the utility feed upgraded, but also completely relocating the panel, probably a major project.

16 slots, but 4 are consumed by the 240V input-breaker and electric oven breakers. Another 2 are consumed by 240V hot-tub breaker (currently off, came with the house, haven't gotten around to trying it out). That leaves 10 slots for house-hold circuits, 3 of which already have tandem breakers installed. At least 2, maybe 3-4 additional could be switched to tandems, it's a bit tricky because phase needs to be maintained when shifting circuits around, as shared neutrals are used on many circuits. This is assuming I'm not forced into AFCI+GFCI combo requirements, as the availability of those in tandems for my breaker type is currently very limited (unavailable?)

The three existing tandems look to have been added when the kitchen was remodeled by the previous owner, with the various dedicated circuits being needed for things like disposal, microwave, fridge, etc.

We have enough circuits now that we're not really over-loaded on anything, though I had to do some juggling to get things I'd rather not share a circuit on separate ones. Primarily keeping motors (like the gas washer/dryer) off circuits with sensitive electronics, or moving the L1 EVSE to the essentially idle 20A garage door opener circuit (plenty of room for both). Getting an available 240V circuit if I ever wanted an L2 EVSE (in no hurry, but probably eventually) would currently be difficult. And my home file-server shares a circuit with a currently unused kitchen island outlet, etc.

If everything was staying as it was today I'd leave it as. However, there's unfinished space on the ground floor in most of the homes in the area (including mine), many of which have been finished to add some subset/combination of den/bed/bath. Good chance that's something I'll tackle 3-8 years down the line, and I'd definitely need more circuits for that job. More for sanity reasons (not having weird combinations of rooms sharing circuits) and probably code requirements, vs overloading issues. If new rooms were added, there'd probably be effective 0A increase in peak demand. If there were future residents (let's say my parents moved in someday), then it's probably still at most 25A increase, on what's probably currently 40A peak usage (in theory electric oven + multiple space heaters + EVSE -- has never has actually happened).

There are some small benefits to a sub-panel even I didn't need it for the slots -- I've heard that Enphase's microinverter powerline communications with the Envoy can interfere with other powerline networking sometimes (which I use), and being able to ferrite-choke isolate the micros+Envoy can solve this issue. There's additional cost to a sub-panel obviously, but I would be able to take advantage of the tax credit for the work.

If I can get a meaningful number of additional slots for future circuits, then that feels like a pretty strong incentive to go ahead with it.

My research on this point is just starting, but it seems like the size of the sub-feed to the sub-panel isn't restricted by the backfeed rules, as long the solar feed has its own appropriately sized breaker (e..g 20A). Thus in theory I could send a 80A 240V feed to the sub-panel and give myself room for 8x 120V 20A circuits (2 consumed by the solar feed), which would be very nice.

But based on my early readings around sub-panels, it sounds like inspectors generally limit sub-panel sizes based on the estimated total load at the main panel. I'm not sure how this would be handled since I don't plan to add any circuits or load in the near-term. Maybe the L2 EVSE circuit at most.

So I really don't know yet how this would be treated. Some questions that come to mind are: Would they give me grief for a sub-panel with a bunch of empty slots and an 80A feed? What if I wire it to handle 80A, but only have 40A feed breakers installed now? If I do 40A now, and try to increase later when doing the additions, will it be more difficult? Then they'll be more work being done, possibly new codes, and actual anticipated load -- seems more likely they'd try to make me replace the main panel then if the electrical had to be touched. vs if I add one room at a time over the years, with 1-2 additional circuits used each time, to an existing panel. :) But frankly I don't know if at some point they might force a major overhaul down the road anyway, because I've pushed past some sort of threshold in their mind.

RegGuheert said:
Frankly, I'm impressed that you are able to digest and comprehend all the implications of these details BEFORE your purchase. Good on you! (Clearly you've gotten WAY beyond "Ooh! Shiny!" since you are posting in a thread about the failure rate of the inverters you are considering. :) )
Hah, you flatter me. :) I try to do the most I can before the big checks are written, then stop stressing about those choices once they're made. Hopefully I haven't hijacked this thread too terribly. This thread is what got me comfortable with going with Enphase M250 microinverters, based on the failure data and discussion of the benefits/cons. Definitely useful. Thanks again to all who have shared their knowledge and experience.
 
ltbighorn said:
My research on this point is just starting, but it seems like the size of the sub-feed to the sub-panel isn't restricted by the backfeed rules, as long the solar feed has its own appropriately sized breaker (e..g 20A). Thus in theory I could send a 80A 240V feed to the sub-panel and give myself room for 8x 120V 20A circuits (2 consumed by the solar feed), which would be very nice.
Your 50A calculation was correct, earlier. So whatever solar you install can be protected by a maximum of a 50A breaker.

As an example, I have a 100A service panel which feeds into a 100A sub-panel dedicated for two Enphase 15A branches. But the sub-panel feeds a 20A breaker in my main service panel. I have 18 M190 inverters with 9 on each string which have a rating of 14.4A maximum continuous, which is protected by the 20A breaker despite having a 15A breaker for each branch.

I suppose technically, I could put two 120V 20A load circuits on the solar sub-panel if I wanted to, but there is a utility solar disconnect switch between the sub-panel and backfeed breaker, so probably best to avoid it.

In your situation, it's similar, though if you also want to use your sub-panel for loads then your sub-panel also has to meet the 120% rule. So if you use a 100A sub-panel protected by a 80A breaker, then you can only feed 40A of solar into that sub-panel.
 
ltbighorn said:
Westinghouse 125A end-fed panel with a 100A input feed/breaker. Underground feed from the street (only about 10 ft to street). I believe with the 120% backfeed rule, this means I could have up to a 50A backfeed (125*1.2 - 100 = 50), though I should only need 20A for 12-16 M250s.
That's correct, although you have to ensure that the two supplies to the panel bus are on opposite ends. That is, the main breaker and the solar breaker.

ltbighorn said:
My research on this point is just starting, but it seems like the size of the sub-feed to the sub-panel isn't restricted by the backfeed rules, as long the solar feed has its own appropriately sized breaker (e..g 20A).
The rules when there are more than two breakers between the inverter and the utility are a bit complicated and seem to change each code cycle. California is currently on the 2011 NEC, I've not studied that version.

But the 120% rule definitely applies to the sub-panel. So if you use a panel with a 100A rated bus, then its feed and the solar breaker have to be on opposite ends. With a 20A solar breaker in the sub-panel, you could feed the sub-panel with up to 100A.

ltbighorn said:
Thus in theory I could send a 80A 240V feed to the sub-panel and give myself room for 8x 120V 20A circuits (2 consumed by the solar feed), which would be very nice.
You could give yourself room for 40x 120V 20A circuits. There's no direct relationship between the size of the feeder to the subpanel and the sum of the breakers in the subpanel. The relationship is that the calculated load on the subpanel has to be less than the size of the feeder. That calculated load is almost always much less than the sum of the breakers.

Since your main panel is pretty full, going with a subpanel is a good move. And you can make your feeder as big as you want, subject to the maximum size breaker that fits in your main panel. If it will take 100A breakers, you could have a 100A subpanel off a 100A main panel. You may be asked to provide a load calculation, though, for both panels.

Cheers, Wayne
 
drees said:
In your situation, it's similar, though if you also want to use your sub-panel for loads then your sub-panel also has to meet the 120% rule. So if you use a 100A sub-panel protected by a 80A breaker, then you can only feed 40A of solar into that sub-panel.
Interesting, thanks for pointing that out. Since I don't expect to need more than 20A for solar, that should be easy to manage.

wwhitney said:
ltbighorn said:
Thus in theory I could send a 80A 240V feed to the sub-panel and give myself room for 8x 120V 20A circuits (2 consumed by the solar feed), which would be very nice.
You could give yourself room for 40x 120V 20A circuits. There's no direct relationship between the size of the feeder to the subpanel and the sum of the breakers in the subpanel. The relationship is that the calculated load on the subpanel has to be less than the size of the feeder. That calculated load is almost always much less than the sum of the breakers.

Since your main panel is pretty full, going with a subpanel is a good move. And you can make your feeder as big as you want, subject to the maximum size breaker that fits in your main panel. If it will take 100A breakers, you could have a 100A subpanel off a 100A main panel. You may be asked to provide a load calculation, though, for both panels.
Thanks for explaining Wayne. In my readings I get the impression that inspectors sometimes won't approve installs that they believe are tempting to convert into dangerous situations. For example, too much load for the utility feed. Or de-rating the main panel utility feed breaker in order to allow more solar feed-in (too easy to put the original breaker back in, I suppose?)

So that's where I wondered if there'd be difficulty getting, say, a 100A, 20- or 40-slot sub-panel off the 125A (100A utility feed) main panel. While today there'd be negligible additional load (basically just solar input, maybe shifting my already connected L1 EVSE to it), would they be hesitant to approve a panel that could invite much larger loads, which may then exceed what's approved for my main panel's 100A input?

However, if I understand correctly, it sounds like in practice I'll probably get no grief over this, with the natural caveat that probably only a local electrician would know exactly what hoops inspectors here make people jump through. If they made me do a new load calculation for both panels, it seems that the 100A service would still be fine, unless the load calculation rules have changed so much as to make what was OK during the install in the 80s be considered insufficient today.

Then again, I don't know if the remodeled kitchen or hot tub were ever included in a load calc, but I can't imagine it'd be too bad -- we have gas furnace, gas water heater, gas dryer and gas range. Only big electricals I can think of are the electric oven, standard appliances like fridge/disposal/microwave/dishwasher, motors for the washer/dryer, furnace blower, the disconnected hot tub and the L1 EVSE. I think the most we've ever used in an hour according to PG&E is 2.5 kWh (L1 EVSE + space heater).

I'll probably go for the largest panel that is physically reasonable for the space - maybe something in the area of 100-125A and 12-space (24-circuit) . Can't see a reason to go for less?
 
Somewhat unexpectedly, PG&E updated their NEM tracking page twice in the same month. Once on 4/24 and again 6 days later on 4/30. They approved about 14 MW in that time, or 2.365 MW/day. At that rate (70.95 MW/month), the remaining 381.85 MW would last until October 11th. Hopefully that remains true, as think I should be able to get turned on my mid- or late July. (Install end of June, inspectors etc.). It's 72 days until July 15th. At 100 MW/month it'd fill around Aug 24th. The rate would have to increase to 159+ MW/month, starting from May, to fill by July 15th. Hopefully it doesn't end up being near that stressful.

NEM 2.0 isn't too terrible as proposed (the increase in non-bypassable charges is eh), but removes the safety net of being able to back-off to a non-TOU rate schedule (E-1) if PG&E turns TOU schedules upside-down (sell solar kwh low, buy high). I think the biggest risk may be that the utilities are challenging the finalized NEM 2.0 as it stands, which could end up being reviewed by the Court of Appeals if they're not happy with the re-review. One thing they're asking for is the elimination of 20-year grandfathering under NEM 2.0.

I think the only thing I can do right now is make sure there are some incentives to deliver on time in the solar contract (i.e. discounts for delayed installs). If I knew that a truly final decision on what NEM 2.0 will be would come in May, it might be worth trying to hold off on signing a solar contract until then, but it seems like that could drag on for some time, and at some point the lack of a contract will put me at risk for delays.
 
A small update and resolving a couple of the questions I asked earlier:

Sub-panel
As whitney and drees advised, there shouldn't be any amp/slot-count issues on the subpanel, as confirmed by local installer/electrician. Will slap in 100-125A panel w/24 slots. Hoping to squeeze it next to the existing panel (it'll be tight), otherwise it has to go about 6 feet the other direction to stay >3 ft from the gas. A load calculation won't be necessary now, though could come up some day when if I add more finished space. Permits should be over the counter, no plans required for a <4 kW install. They're going to have to install an additional grounding rod to bring it up to current code.

NEM tracking
PG&E appears to be updating the NEM tracking page weekly now. If the rest of May proceeds like the previous two weeks (last of April, first of May), then ~61 MW will be consumed. At that rate, the cap would be exhausted Nov 5th. Still seems like October is a reasonable expectation, aiming to get my PTO hopefully in the week or two of July.

Envoy-S vs Envoy-S metered
I'm leaning towards skipping on the Metered version of the Envoy-S -- I feel like by the time the utility rate/rules and battery costs converge to provide a significant advantage, it'll be a minimum of 5 years and more likely 10+. By then, it may be a whole new field of tech and the right solution may not involve an Envoy-S Metered. If it does, I can probably get one used for a good price then. Buying tech today, to essentially leave idle for 10 years on speculation that I might eventually use it, even at a modest incremental cost, just doesn't make a lot of sense to me.

Future expansion
Under NEM 1.0, I'm limited to max 1 kW increase in capacity. I'm guessing that's AC rating, so possibly 1.12 kW (4x250W) on 4xM250's. Otherwise 3 panels. Since I'm having the solar mounts installed with the new roof, and sealed/flashed by the roofers, it made me wonder if I should consider having additional mounts installed now, for potential future expansion?

Under today's usage (~4800 kWh/year), 3.3 kW will likely offset 100% of my kWh usage, >100% of my bill thanks to the TOU multiplier. However in 18 months my No Charge to Charge program ends, and I'll probably add about 1200 kWh/year in home charging, at which point I'd be producing about 80% of my kWh. With today's TOU situation, I'm WAG'ing that that would still be close to 100% bill offset, and with minimum charges of $10/m, if it's a little under that's no big deal.

However, with the potential for the TOU multiplier to be reduced, or even flipped upside-down as rate periods change (particularly after 2020 when the E-6 rate plan I'm on starts to change), perhaps there may reach a point where putting an additional couple panels could be called for.

Today, for 12 panels, they're planning to do 2 rows of 6 panels. This would leave roof space for a potential 3rd row of up to 6 panels.

I'm guessing installing mounts for another row (of up to 4) would probably add at least $500 to the cost (assuming at least 8 mounts at $50/sealed penetration by my roofer). It'd be nice to be able to expand in the future without a 3rd party putting new holes in the roof (simplifies warranty/roof guarantee issues), but it's up-front money for a maybe-used-down-the-line item. It's also simply more roof penetrations. Roof-wise, would it be lower risk to have penetrations done later, when they're needed, rather than as part of the new roof install, but sitting idle for years? Perhaps the extra penetrations now would be more of a liability than prevention.

All these estimates could be crushed if I decide to try and get that hot tub going though. ;)
 
ltbighorn said:
Under today's usage (~4800 kWh/year), 3.3 kW will likely offset 100% of my kWh usage, >100% of my bill thanks to the TOU multiplier. However in 18 months my No Charge to Charge program ends, and I'll probably add about 1200 kWh/year in home charging, at which point I'd be producing about 80% of my kWh. With today's TOU situation, I'm WAG'ing that that would still be close to 100% bill offset, and with minimum charges of $10/m, if it's a little under that's no big deal.
A few things to consider regarding future expansion:

1) Likely you will not be able to purchase the same panels in the future. In other words, the expansion panels likely will not match. My wife made this point when we built our system and it convinced me to max out our space. (But we have plenty of usage to match our production.) Those panels are no longer manufactured today.
2) Adding a third row of panels now will protect the new roof beneath while adding ONLY the mounts will add penetrations with no overlying protection.
3) Overproducing electricity does cost extra money (until you add BEV loads and a hot tub), but it's for a good cause! :)

It sounds like you have a good plan!
 
RegGuheert said:
ltbighorn said:
Under today's usage (~4800 kWh/year), 3.3 kW will likely offset 100% of my kWh usage, >100% of my bill thanks to the TOU multiplier. However in 18 months my No Charge to Charge program ends, and I'll probably add about 1200 kWh/year in home charging, at which point I'd be producing about 80% of my kWh. With today's TOU situation, I'm WAG'ing that that would still be close to 100% bill offset, and with minimum charges of $10/m, if it's a little under that's no big deal.
A few things to consider regarding future expansion:

1) Likely you will not be able to purchase the same panels in the future. In other words, the expansion panels likely will not match. My wife made this point when we built our system and it convinced me to max out our space. (But we have plenty of usage to match our production.) Those panels are no longer manufactured today.
2) Adding a third row of panels now will protect the new roof beneath while adding ONLY the mounts will add penetrations with no overlying protection.
3) Overproducing electricity does cost extra money (until you add BEV loads and a hot tub), but it's for a good cause! :)
#1. Given microinverter usage, is this primarily a concern for aesthetics? My solar panels can't be seen from the ground, so in my case I think of a reason that would be a downside. Future panels might be better/cheaper (of course, they'll have to make up for the eventual expiration of the ITC)

#2. My friend joked that I should get the mounts installed, and then just mount sacrificial plywood "panels" in the meantime to shield the roof.

#3. Adding just the BEV probably still leaves me sized right. From what I read hot tubs can drive people crazy with just the maintenance. :D Though they just a lot of extra panels (200-300 kWh/month worth!). The main risk would be BEV + TOU rates changing enough to bring the value of my production credits below 6000 - $120/year minimum charge equivalent kWh.

At $0.22/kWh (maybe a little high -- need to re-comprehend how TOU + tiers + NEM interact), $120/yr min charge would be equiv to 550 kWh, meaning I'm slightly under-generating once the TOU multiplier drops below 1.14.
 
It has now been over 13 months since I have experienced a failure of an M190 in my system! The MTBF has clawed its way up to 57 years. Amazingly, Enphase has been able to come up with new firmware which greatly reduces the failure rate, at least in my application.

But I am still adjusting my sparing strategy. Here is a bit of background:
RegGuheert said:
I have decided to make some changes to my strategy for the original M190s since I have 12 72-cell panels which would like to keep running for as long as possible. After over five years of operation, I have had four M190 failures and currently about seven more units are misbehaving. As I see it, ALL of my M190s will be out of warranty in a little under 10 years regardless of what I do. If I keep them all in service, it is likely that most or all will die and be replaced with newer-style inverters during that period (assuming that Enphase remains viable). That is certainly preferable for the 60-cell PV modules on the roof. But those newer-style modules will likely fail in cold weather if used with the 72-cell modules.

So I have a dilemma: Can I keep 12 old-style M190s alive for significantly longer than 15 years? I certainly don't know the answer, but I have decided to do the following to try to get 30 or more years out of this system (perhaps a bit less from the 12 72-cell panels):

- I currently have nine used (no-warranty) M250s. I will put those into service on my roof to replace nine old-style M190s. Hopefully these M250s will last a long time since they will have low electrical and thermal stresses on them due to being paired with 235 Wp PV modules. (It seems to work for drees!)
<snip>
RegGuheert said:
I picked up four NEW M215IGs on eBay for $91.00/ea. shipped, so I will put a few more M190s up as cold spares. I simply couldn't pass up that price considering the warranty on those lasts until 2041! The remaining warranty on the original M190s is now less than 10 years.
I recently purchased 12 NEW M215IGs on eBay for $93.75 each. This brings the total number of fourth-generation (IG) Enphase inverters in my inventory to 42, which is exactly the number of 60-cell PV modules I have on my roof! I have also already purchased sufficient Engage sections to complete the transition. I now have in inventory 79 microinverters with 54 in operation and 25 cold spares. Currently, all the spares are fourth-generation PV modules, but soon I will remove all of the M190s from the roof and make them the cold spares. I also have added a cold spare Envoy.

So, why did I decide to purchase so many spares when all of the operating inverters are still under warranty? Here are my reasons:
- Purchasing new fourth-generation inverters for about $110/each (with Engage cable) provides a warranty extension of 15 years (about $7/inverter/year). Perhaps that's a bit pricey, since each module produces a bit under $40 of electricity each year, but the price of electricity in 20 years will likely be twice that.
- The measured failure rate of the fourth-generation Enphase inverters in the field is significantly lower than the M190s. They have an MTBF of over 700 years versus below 50 years for the M190s.
- The M190s which are on the roof ALL now slightly underproduce due to behavior of the new firmware. I expect to produce as much as an additional 0.5 kWh some days by having fourth-generation inverters in place instead. There should also be a *very* slight improvement in overall production due to the elimination of clipping on the roof on very cold, sunny winter days.
- Currently, every failure of an M190 requires a trip up onto the roof, no matter whether the failure occurs in the field or on the roof. This is because the replacements MUST go on the roof since the field array contains 72-cell PV modules. With all PV modules on the roof upgraded to fourth-generation, hopefully all failures in the foreseeable future will occur in the field where replacements are extremely simple.
- Any failures of M190s will be replaced by M190s in stock and the replacements from Enphase will become spares for the roof arrays. Put another way, I expect to have a growing number of fourth-generation spares and a shrinking number of M190 spares as time goes on.
- This spares strategy should take this system well beyond the warranty period for both the M190s and the fourth-generation inverters.
- I should be less affected in the case where Enphase goes bankrupt. (If that does happen, I hope they can hold off until I can get a few of the new-style spares in hand!)
- Failed M190s can be replaced immediately following diagnosis by Enphase rather than having to wait for a replacement to be shipped. This will be true for the fourth-generation types once I have at least one spare in hand through replacement of a failed M190 by an M190IG. This is not a big deal, but it is nice not to have those dropouts.

Anyway, I'll be looking for a few cloudy days to get up there and complete the conversion of the roof to all fourth-generation inverters.
 
Entered contract today for my system. Stanchions will be installed during the reroofing in 1 week. Panels and electrical (everything besides stanchions) will go in after the roofers have finished. Luckily permits here for this size system are over the counter.

PG&E still has ~310 MW left to NEM cap, filling at about 50MW/month for the last couple months, so not stressing about making NEM 1.0. Should be done by mid-July, and cap seems like it won't be hit until at least September, if not October+.

I was originally quoted for LG 280W Neon's (the original, not Neon2's), but they've since been discontinued. My installer agreed to substitute LG285S1C-G4's, which are not Neons but they are a generation newer than the first gen Neons, and have a lot of the same benefits as the Neon2's that surpass the original Neon. Namely 2% max initial year degradation instead of 3%, 0.6%/year instead of 0.7%/year, and a 12 year limited warranty instead of 10. (25 year performance warranty). Between that and the extra 5W/panel, I think it makes up for any potential low-light gains I might have gotten out of the Neon's double-sided cells.

Final config as planned:
12x LG285S1C-G4 (3.42 kW, slight bump from 3.36)
12x Enphase M250 microinverters
Envoy-S Metered with consumption metering
Enphase Manager access
100A subpanel added
15-year workmanship guarantee
Exterior conduits will be painted to match
50%+ of installer's payroll is in the city, so slightly bigger city rebate :)

Penetrations will all be sealed by the roofing company which has been in business here for 30 years and is the primary roofing partner for one of the top commercial solar installers here. Definitely costs me a bit extra but worth it for the peace of mind of having it done as best can be (rather than punching a bunch of holes in a brand new roof), and one party responsible for water-tightness.

Next few weeks will be interesting! Thanks for all the help. Tremendously valuable, especially in weighing the Enphase specific pros & cons. I'll soon have M250 inverter reliability data to contribute to this thread.
 
RegGuheert said:
Anyway, I'll be looking for a few cloudy days to get up there and complete the conversion of the roof to all fourth-generation inverters.
I have just completed the conversion of the Garage Array to fourth-generation inverters. Here's a breakdown of the costs for this upgrade (all from eBay):

- 4 - New (2014) M250 @ $129 ea. + $27.96 shipping = $543.96 (Though I sold and installed four old M190s for $500 to offset some of this cost.)
- 4 - New (2015) M215IG @ $91 ea. = $364.00
- 12 - New (2015) M215IG @ $93.75 ea. = $1125.00
- 1 - New (2014) M190IG @ Free (Replacement from Enphase for failed M190.)
- 25 - New Engage Portrait Positions @ $12 ea. + $20 shipping = $320.00 (I have to cut off some positions to get full-length pigtails to make the wiring work for my system.)
- 50 - Stainless-steel clips = $32.50

Total cost for Garage Array inverter upgrade: $2385.46

The entire array produces about 18 MWh/year and this portion is 21/54 of the total, so it produces about 7 MWh/year. At $0.12/kWh, that portion of the array produces about $840 worth of electricity each year, so this upgrade is costing about three years' worth of electricity for that portion of the array. So what did I get by paying for three years' worth of electricity now? Here's a breakdown:

- 14 additional years of warranty coverage for 20 of the 21 inverters. (The M190IG still has the original warranty.)
- 20 spare original M190s to be used for the Field Array once the warranty expires.
- Slightly better energy harvest. This is a very nominal benefit of perhaps 100 Wh/day.
- Eliminate of trips onto the garage to replace inverters for the foreseeable future (I hope).

That just leaves the House Array of 21 inverters. I account for the costs for the upgrade to that array quite differently:

- 12 - New (2014) M215IG @ Free (I bought these twelve inverters (thanks again, QueenBee!)and their Engage cables and clips to allow the M190s to be used to energize the existing field array, so I count that as a separate project. They weren't free, but that project pays back its cost in only four years' time.)
- 9 - Used (2013) M250 @ $1000, including Engage cables. (No warranty.)

Total cost for House Array inverter upgrade: about $1000.00

So the House Array will be paid for with about 1.5 years of electricity. What did I get for that expense?

- The M215IGs came with 14 additional years of warranty.
- The M250s come with 10 FEWER years of warranty. (But I expect they will be more reliable and hopefully they will last a longer time.)
- Slightly better energy harvest. This is a very nominal benefit of perhaps 100 Wh/day.
- Elimination of trips onto the house to replace inverters for the foreseeable future (I hope).

Now I just need to get up on the house and install the remaining 10 fourth-generation microinverters which are not up there, yet. It may be a few weeks before I get around to that.

I know, it's pretty wonky accounting for the two arrays differently, but different inverters ended up in different locations on the roofs. You can average the total together if that makes more sense to you.

One final note on M190 reliability: While there still have been no failures of the old M190s in about 14 months, I noticed that some of the nuts (which you tighten with your fingers) on the M190 cables made a cracking sound when I turned them to disconnect them from the roof. It sounded as if the plastic had gotten hard and was cracking or breaking. None of the nuts broke to the point where they fell into pieces, but there are still 10 years remaining on the 15-year warranty period for these inverters, so I wonder if this may become a failure mode for the old M190s in the not-so-distant future. Just be forewarned and be gentle with these nuts if you have to do any repairs or replacements of the old M190s. I guess it's nothing that a bit of duct tape won't fix, however.
 
ltbighorn said:
Entered contract today for my system.
Congratulations! That's exciting!
ltbighorn said:
I'll soon have M250 inverter reliability data to contribute to this thread.
Great! You'll accumulate MTBF information at a rate of 12 years/year.

Good luck with your install!
 
RegGuheert said:
I have just completed the conversion of the Garage Array to fourth-generation inverters.
...
That just leaves the House Array of 21 inverters. I account for the costs for the upgrade to that array quite differently:
I recall thinking on your last update that we might need a color coded chart showing the migration pattern of your inverters. I must admit that I've become a bit unclear on what is going where. It does sound like good fun though. Any recent pictures?
 
ltbighorn said:
I recall thinking on your last update that we might need a color coded chart showing the migration pattern of your inverters. I must admit that I've become a bit unclear on what is going where. It does sound like good fun though. Any recent pictures?
Good idea!

Here is what the final arrangement will be once all the upgrades to the inverters on the house are complete:

The Guheert Microinverter Menagerie

Inverter_Types_Map.png


Brief description:
- Each of the seven different colors shown represents a different hardware part number from Enphase.
- 12 of the old M190s (dark blue) were built in late 2010 and are in the field array. 10 of these inverters are currently indicating some form of malfunction (but still harvest nearly as much energy as a properly-functioning unit).
- 4 of the old M190s (dark blue) have been sold and are in an array nearby. One was built in late 2010, one in late 2011 and two in mid 2014.
- 25 of the old M190s were built in late 2010 and are now cold spares.
- 4 M250s (pink) were built in early 2015 and are new with warranty.
- 9 M250s (green) were built in late 2013 and are used with NO warranty. They are NOT like most M250s. The metal plate is shaped like the original M215s that required a ground and they include the Amphenol H4 connectors like were found on the old M190s (which I do not like since they don't mate well with the Multimate M4s found on the PV modules.).
- 4 M215IGs (goldenrod) were built in late 2015 and are new with warranty. These are the newest inverters in the system and they are the only fourth-generation inverters I have with metal cases. (I'm hoping Enphase' switch to metal cases does not mean I should expect the plastic ones to fail in the future!)
- 12 M215s (yellow) were built in mid 2014 and are new with warranty.
- 12 M215s (light blue) were build in early 2014 and are new with warranty.
- 1 M190IG (purple) was built in mid 2014 and carries the original warranty from the old M190s. It is a replacement sent by Enphase for a failed M190.

If and when old M190s fail, I expect to receive fourth-generation M190IGs as replacements. As a result, I expect that over time the number of spares for the Field Array will shrink and number of spares for the House and Garage Arrays will grow (from zero). The good news is that I won't have to go onto the roof unless an M250, M215IG or M190IG fails, which should not happen for quite some time.
 
Wow, those are quite some arrays! The graphic and overview helps a lot in understanding your changes. Thanks.

It seems like you're preferring to cycle out units that still have warranty left on them for new units (characterizing it as "extending the warranty by X years"), but I presume that essentially sacrifices the warranty on the existing ones. What's your thinking behind that? Is it primarily a side-effect of trying to get more failure-prone M190s off your house roof, or is it also for its own sake? If not the former, then why keep so many spares right now, vs. buying them later? Are you concerned that secondary supplies will dry up?
 
Back
Top