PowerWall 3 with optimizers or microinverters

[OppositeLock]

Hi All,

I've got a house in a place with frequent outages, so I had 3 PowerWall 3's go in today, it'll give me a few days of power, as I don't have solar. I need a new roof in a couple of years, so I'm delaying any kind of solar until then, but the complication is that I have 3 different roof planes where solar panels would go, and giant redwoods would be covering some part of the roof at all times. This is where you need microinverters or optimizers, since your whole string produces as much power as the lowest producing panel in the string. The installer told me that Tesla isn't supporting the PW3 with either optimizers or microinverters. Does anyone know anything about this? It's kinda lame if true, because I'd have argued for a PW2 install had I know this, ah well. I'm hoping each PW3 can drive a separate panel string on each portion of the roof.

I have a PW2 as well, in a different house, and that thing is super versatile. It doesn't care what your solar array looks like, and when there is an outage, it simply absorbs any excess energy into the battery. It doesn't need to know where this even comes from, so during long outages, I drive a bunch of solar inverters with a generator and the powerwall charges off those just fine. I'm hoping the PW3 can behave this flexibly too. It doesn't need to know solar is there as long as it can store excess energy from battery inverter side.

 

[arnolddeleon]

As I understand it, Powerwall 3 Specifications (tesla.com), there 6 MPPTs (Maximum Power Point Tracking) inputs on PW3. So, it could easily handle 6 different roof planes for MPPT purposes. Bypass diodes would for the most part take care of shading of individual panels as long as the total string voltage doesn't drop below the minimum string input.

The Powerwall 3 is supposed to support AC coupling but @Vines indicated that it didn't ship with it, but it should come someday soon. This will allow you "hack" charging the Powerwalls.

For absolute maximum output, a microinverter or optimizer system will probably produce slightly more energy in a partial shading situation. However, the figure of merit that you should care about is kWh produced/dollar spent. Basically, you can usually make any difference by adding one more panel which likely still cost less the microinverter/optimizer solution. If @nwdiver is still reading the forums he can elaborate.

[edited to correctly say 6 planes instead of 3 to align w/ number MPPT inputs]

 

[nwdiver]

This is where you need microinverters or optimizers, since your whole string produces as much power as the lowest producing panel in the string.

That's 1000% baloney. Misinformation propagated in no small part by Enphase.

 

[CrazyRabbit]

Tesla LOVES to install two strings with one homerun, I believe that will affect production even more when shading occurs. I would like to see that test with different azimuths and pitch on a single string.
But in a nutshell, optimizers (including enphase) produces about 10-13% more energy.
System design is simpler with optimizers..

 

[Vines]

As I understand it, Powerwall 3 Specifications (tesla.com), there 6 MPPTs (Maximum Power Point Tracking) inputs on PW3. So, it could easily handle 6 different roof planes for MPPT purposes. Bypass diodes would for the most part take care of shading of individual panels as long as the total string voltage doesn't drop below the minimum string input.

The Powerwall 3 is supposed to support AC coupling but @Vines indicated that it didn't ship with it, but it should come someday soon. This will allow you "hack" charging the Powerwalls.

For absolute maximum output, a microinverter or optimizer system will probably produce slightly more energy in a partial shading situation. However, the figure of merit that you should care about is kWh produced/dollar spent. Basically, you can usually make any difference by adding one more panel which likely still cost less the microinverter/optimizer solution. If @nwdiver is still reading the forums he can elaborate.

[edited to correctly say 6 planes instead of 3 to align w/ number MPPT inputs]

AC coupling is only on hold for PW3 plus the Backup Switch. With the Gateway, AC coupling works now.

These days with inverters like the PW3 have a silly wide range of DC voltages they will take. A string of 3 qcell modules will work and so will 13 of them so you have massive flexibility.

As pointed out above a microinverter/optimizer solution has to pay for itself before it is a better solution regardless of the energy harvest. For instance lets take the simplest example where we connect a single Tesla Inverter to the same (25) qcell 400w panels, then compare with (25) IQ8A microinverters.

If we compare a Tesla PV inverter to an enphase IQ8A solution we get a delta in cost of about $3000 for the hardware. How much production is required to make up this difference?

But in a nutshell, optimizers (including enphase) produces about 10-13% more energy.
System design is simpler with optimizers..

I do not believe for a minute that the solar edge system would perform 10% better unless the string design was terrible. Lets assume the $3000 system works 10% better for a moment.

PV watts says a 10 kW system in mountain view produces about 16,555 kWh per year, so the theoretical system with 10% better production makes 18,210 kWh per year. This extra production cost about $3k in hardware.

If your system was under NEM 2.0 and you got $0.25 per kWh then the value of the extra power from Enphase is about $415 per year. It would take 7.2 years approximately for that 10% more efficient system to break even with the hardware costs. I still do not believe an Enphase system would produce 10% more energy except for an edge case so YMMV. If you had a Solar Edge system you probably went through 2 inverters in the same time, so I'm not counting that lost production lol.

Also most people don't know that the Solar Edge inverters will not produce power if the string is shaded so only about 4-5 modules are in the sun. The optimizers are playing with the voltage and amperage but they can only produce so much voltage and the inverter wants to see 350 vdc. Solar Edge recommended string length minimum is 8 modules for a residential installation. The string will stop producing power once 3-4 of those 8 modules are shaded.

 

[BGbreeder]

Tesla LOVES to install two strings with one homerun, I believe that will affect production even more when shading occurs. I would like to see that test with different azimuths and pitch on a single string.
But in a nutshell, optimizers (including enphase) produces about 10-13% more energy.
System design is simpler with optimizers..

Do you have a reference for that efficiency number?

@Vines makes some great points about the ROI for the added cost of microinverters that I think are completely valid.

Despite those points, I wanted and have microinverters, knowing the extra cost, for two reasons; 1) longer MTBF (mean time between failure/ and longer warranty), and 2) reducing a single point of failure for the majority of our solar production, neither of which is a straight economic ROI. The panel level reporting is reassuring, and would make troubleshooting easier for me. If replacing a blown string inverter was as easy as switching a light bulb, I might have gone that route, and stocked a spare inverter, but it isn't. Our roof is shaded a lot but I would not know how to do an apples to apples comparison of microinverters to string outputs under some levels of shading, except in a lab.

All the best,

BG

 

[nwdiver]

I do not believe for a minute that the solar edge system would perform 10% better unless the string design was terrible. Lets assume the $3000 system works 10% better for a moment.

PV watts says a 10 kW system in mountain view produces about 16,555 kWh per year, so the theoretical system with 10% better production makes 18,210 kWh per year. This extra production cost about $3k in hardware.

Payback is even longer when you have an oversized array, which is becoming more common without NEM and with cheaper panels. DC side efficiency is irrelevant when the inverter is maxed out. It's almost always going to make more sense to add more panels than to add optimizers.

 

[CrazyRabbit]

Do you have a reference for that efficiency number?
…

That was out of the video, with some shading.
Vines said Tesla inverters have a wide range on each mppt, and they do. But just know you will lose efficiency as the power will have to be converted by a buck/boost converter before the inverter and you will lose efficiency. I guarantee the efficiency numbers on the string inverters is as best as it gets.

 

[Vines]

That was out of the video, with some shading.
Vines said Tesla inverters have a wide range on each mppt, and they do. But just know you will lose efficiency as the power will have to be converted by a buck/boost converter before the inverter and you will lose efficiency. I guarantee the efficiency numbers on the string inverters is as best as it gets.

I am not 100% sure here but I think it's unlikely there is a dc/dc converter increasing the DC voltages inside the Tesla PV inverter. Maybe someone else with more hardware experience can speak to this.

The Solar Edge optimizer does have a buck/boost converter in it, and it is around 99% efficient at it's job.

The CEC weighted efficiency is not the best efficiency, it is weighted with different multipliers for various loading from 10% to 100%

Here is the formula: 0.04 x Eff10% + 0.05 x Eff20% + 0.12 x Eff30% + 0.21 x Eff50% + 0.53 x Eff75%. + 0.05 x Eff100%.

 

[OppositeLock]

That's 1000% baloney. Misinformation propagated in no small part by Enphase.

I do know that in one place I own, microinverters helped quite a bit. Is it possible there are older panel designs which handle shading less gracefully? Mine are made by Canadian Solar, and shading even a small piece of one panel really knocks the output down on that panel, and affects the array noticeably.

 

[OppositeLock]

Sometimes, you want to maximize your generation because you have very little roof that's angled properly to make use of the sun, and you're willing to maximize watts, not watts/dollar.

 

[nwdiver]

I do know that in one place I own, microinverters helped quite a bit. Is it possible there are older panel designs which handle shading less gracefully? Mine are made by Canadian Solar, and shading even a small piece of one panel really knocks the output down on that panel, and affects the array noticeably.

No. All panels that are UL listed and designed to be incorporated into a grid-tied array of >100v will have bypass diodes. The lack of these diodes would be a significant fire hazard. If a shaded cell cannot be bypassed it becomes a load and temperatures quickly rise to >200F.

Older inverters with only a single MPPT would have issues. If you have strings in parallel and one panel is in shade that string is now 'shorter'. You can't have strings of different length in parallel so the unshaded portion of that string will stop producing. That's why modern inverters usually have 3-4 MPPT trackers so each string can operate independently. As long as each string is able to operate independently shaded panels will not reduce the performance of unshaded panels.

 

[CrazyRabbit]

That's why modern inverters usually have 3-4 MPPT trackers so each string can operate independently. As long as each string is able to operate independently shaded panels will not reduce the performance of unshaded panels.

please educate Tesla solar designers about this. They keep wiring arrays in parallel on the roof.

 

[arnolddeleon]

Sometimes, you want to maximize your generation because you have very little roof that's angled properly to make use of the sun, and you're willing to maximize watts, not watts/dollar.

Yes, this is sometimes true. But not nearly as often as most people think. Non-optimal roof planes these days may not be as bad as people think. You could also go down the rabbit hole of calculating the value of a given kWh delivered. For example, there is potentially an increased value of a kWh produced by a western facing panel versus a southern facing one. So, while a south facing array might produce more the western facing one, they might be comparable in value. For simplicity, I generally recommend the basic $/kWh approach.

The point about watts/dollar also applies to choices beyond the inverter. It also applies to the model of the solar panels, wiring, etc. When I bought my first PV system 20 years ago, I was obsessed by panel efficiency, wire losses, matching panels, etc. All fun and good engineering challenges. Today, for most (but not all) people, I recommend figuring what gives the best price/performance for kWh. In general, there is a premium for the highest performance panel. It's not unusual to make up the difference with an additional panel or two for much less cost (again assuming no space constraints) by choosing the second-tier model.

There are definitively other factors, like being space limited. There could aesthetic considerations. There could be supplier/installer limitations/preferences. Just be aware of

 

[offandonagain]

2) reducing a single point of failure for the majority of our solar production, neither of which is a straight economic ROI. The panel level reporting is reassuring, and would make troubleshooting easier for me. If replacing a blown string inverter was as easy as switching a light bulb, I might have gone that route, and stocked a spare inverter, but it isn't.

Agreed, the single point of failure is a big drawback for me and would make me consider microinverters in the future. I've had two issues in the three years of operation where incorrect string wiring or a failed MCI took out most or all of my production, in one case for months. String inverters can also fail. Even if the lost production for those months doesn't make up the price difference, reducing the probability of a major loss in production would make it worth it for me. Not sure if it's loss aversion or just irrational resistance to pulling from the grid

 

[sunwarriors]

As a biased fan of micro-inverters in general, one thing I'd add that I don't think I saw in this thread is with the 30% tax credit, the additional "cost" is not as much. That and of course, panel monitoring to quickly see what's busted, the 25 year warranty, UL certification (I think some are missing this). I think some of those hybrid ones had lower warranty years also and we know about SE from the past.

I'm in the pay once, cry once group. I don't think a lot of will care longer term as long as the darn thing just works. 10 year warranty to me means it'll break in year 11. Of course, this isn't always the case, but there is some value in your own time, not having to be your own project manager or advocate to get stuff fixed.

I've called Enphase directly before on an issue and the first person on the phone fixed the problem. That's sorta nice being able to open a case by phone, in the app/monitoring screen, etc. I also have my installer to fall back on if I needed it.

For residential projects especially I feel in most cases, unlike commercial if you don't have a massive system, there just isn't as big of a cost in the scheme of a complete project/install, especially with batteries. Some installers didn't even charge much more vs. strings when I ran my quotes, but if you are a DIY, then yes, you can/will save $$.

 

[nwdiver]

If replacing a blown string inverter was as easy as switching a light bulb, I might have gone that route, and stocked a spare inverter, but it isn't.

Depends on the string inverter. Fronious has a 'swing lock' so you can swap an inverter in ~2 minutes. I replaced a SMA inverter last week. That took ~10 minutes.

I really dread the day that I have to replace a micro inverter in a system that wasn't mapped correctly or not mapped. THAT would be finding a needle in a haystack! And neither of the Enphase systems I've had to troubleshoot were mapped or even had monitoring. Most installers are terrible. At least with most string systems you have a display you can look at. Worst case you have DC lines you can check.

That's probably the worst feature of micro-inverters. They serve as a crutch so installers don't have to educate themselves on how solar works.

 

[BGbreeder]

Depends on the string inverter. Fronious has a 'swing lock' so you can swap an inverter in ~2 minutes. I replaced a SMA inverter last week. That took ~10 minutes.

I really dread the day that I have to replace a micro inverter in a system that wasn't mapped correctly or not mapped. THAT would be finding a needle in a haystack! And neither of the Enphase systems I've had to troubleshoot were mapped or even had monitoring. Most installers are terrible. At least with most string systems you have a display you can look at. Worst case you have DC lines you can check.

That's probably the worst feature of micro-inverters. They serve as a crutch so installers don't have to educate themselves on how solar works.

Always happy to learn. Do you have a link to that Fronius feature? I'm not finding it.

I guess that we had a good installer. Our microinverters are well mapped.

Like @sunwarriors there was essentially no price difference for microinverters when we had solar installed, so to me, trading up from a seven year to 25 year warranty was a no brainer. However, I do think that the differences are not black and white, and it is very much a "horses for courses" subject.

BG

 

[nwdiver]

Always happy to learn. Do you have a link to that Fronius feature? I'm not finding it.

The inverter just lifts off the backplate. All the wire termination are on the back plate. There's ~2 screws and you lift it off.

 

[jjrandorin]

The moral of the story for me is, especially as you get into this type of stuff, with higher dollar figures etc, for a lot of people there is a LOT more going into it than simple "Return on Investment". Some people just want "the best stuff", even if it would be better to step down a notch or two and get more panels or something.

Each persons goal is individual for them, and as long as someone is making the choice(s) "eyes wide open" knowing they might be paying extra etc, its all good in my book.