diy solar

diy solar

battery grid feedback

Would love to see the math on that. I have not found any way to get a Solark-based system to break-even in 5 years unless your talking about a very large system in California (meaning offsetting a huge amount of consumption for at-conditioning).
It all depends on what your paying for electricity. As I have said before I would never have put up Solar when I was living in Florida as the utility rates were way to low.
You let me worry about the math on my system.
I was anti solar for the last 15 years based on the math. LIfePO4 batteries and the lower cost of panels per watt and a much higher utility rate changed that equation for me.
 
On average you can buy a 15 year old used car with nearly 100,000 miles on it for the price of a Sol-Ark 12K but then your going to have to spend thousands of dollars to keep it running for another 10 Years. Sure you can probably find cheaper cars but I bet you will probably be paying more and more money every year to keep them on the road.

On the other hand the Sol-Ark or any other PV system keeps money in your pocket year after year and peak shaving with batteries can make those savings a lot higher. In 5 years my system will be paid for and I still have 5 years of warranty life left that pretty much guarantees I will be making pure profit on my investment for another 5 years and possibly beyond that.
I am not trying to have you dislike your inverter. All I am saying is that it is (very) expensive, especially considering that it's a Deye clone, which apparently sold for 1/3 of that when was available. You need to make $3 a day for 5 years just to cover the cost of the SolArk. That's 20kWh every day at least at the average US rates.
 
I am not trying to have you dislike your inverter. All I am saying is that it is (very) expensive, especially considering that it's a Deye clone, which apparently sold for 1/3 of that when was available. You need to make $3 a day for 5 years just to cover the cost of the SolArk. That's 20kWh every day at least at the average US rates.
I use 50KWh per day. I am not saying the Sol-Ark is cheap but for me it made a lot of sense.
 
This is something that each person interested in solar power has to figure out for themselves.

The cost of electricity and the amount you use can make the difference between (don't bother) and (DO IT NOW!)

If you have cheap electric rates and only us 5 KWHs a day, then installing solar is probably a waste of time and money. Here in So Cal, we have high electric rates and long HOT summers. We also get a lot of sun. When you do the math, you quickly see that a basic grid tied solar PV system will pay off in about 10 years. It used to be 7 years, but they recently adjusted rates again, so the peak rate time as the sun goes down is getting more expensive. So for me, it was a bit of a shock when my calculated savings didn't show up. I was on a tiered plan, and I was forced to time of use to get the solar NEM agreement. My cheapest tier before was $0.15 per KWH. Now it is $0.17 and going up to $0.21 or $0.23 on the cheapest time of use. And the peak rate for 4 pm to 9 pm is going up to $0.43 per KWH and that is when the sun goes down, so a grid tie only system can't help there. So even though I was using less than half the power from the utility, my bill only dropped about 15%. That was a shocking eye opener. But now I know better, and you really need to look at what the rates will be after the solar is installed.

The only way around the time of use problem is to use storage. And this becomes a problem. Batteries are expensive. And here in So Cal, they threw in a few other wrenches. To be able to get a battery system permitted, they require extensive plans and all gear must be UL listed. And that includes the batteries. The one way around that is to have the batteries well separated from the living space. For my setup, it is not practical at all to get my batteries far enough from the house to make it fully legal. So for now, my battery system is just a giant UPS that runs a few appliances. I did use a UL inverter with the Schneider XW-Pro, and the wiring setup meets or exceeds all code requirements. The system is safe, and fail safe, but my battery cabinets are not technically legal. I now have two cabinets with 18 KWHs of battery in each. They are both on Anderson and wheels, near the overhead garage door. I have smoke and heat detectors above them, and they are in steel cabinets. I have never seen any part of my battery cabinet hit more than about 3 degrees above ambient. If UL listed LFP batteries become legal in the future as these star to lose capacity, I can change them out and make the system fully legal and get it inspected then. If I sell the house before then, I will just take the system out. The risk is that if the battery system does ever cause a fire, my insurance may not pay out. Because of that fact, I have it installed in a very safe manor and I am not stressing it at all. The batteries are under far less stress than if they were still in a Chevy Bolt. I am only charging and discharging at a maximum of a 0.15 C rate, and the cells stay between 40% and 85% at all times. Most days, they charge and discharge less than 30% of their capacity. The extra is overhead incase of a grid failure.

If your pre solar electric bill is under $100 a month average. Solar might not pay off, and a battery system certainly won't pay off.

Before I put in solar, my electric bills would top $300 in the summer. Now my top bill last year was $60 and I had 3 months go negative. With the battery bank time shifting the solar power to cover 4pm to 9pm I shaved off another $1.50 to $2.00 a day compared t the grid tied solar alone. Figure $1.75 x 360 days a year = over $600 per year saved. I spent under $10,000 total to add my battery system, including the surplus EV batteries. In theory, the whole battery system could pay off in 16 years, but I am realistic, and don't expect the cells to last that long. But it sure makes adding storage for backup power a lot cheaper when you can shave time of use charges. And if I roll in the battery cost with the solar, and take it as a single system, the pay off for the solar AND battery system comes down to 12 years, if the electric utility rates don't go up any more. Any increase in electric rates will bring the pay off time down even sooner.
 
It all depends on what your paying for electricity. As I have said before I would never have put up Solar when I was living in Florida as the utility rates were way to low.
You let me worry about the math on my system.
I was anti solar for the last 15 years based on the math. LIfePO4 batteries and the lower cost of panels per watt and a much higher utility rate changed that equation for me.
“robbby” said:
I use 50KWh per day. I am not saying the Sol-Ark is cheap but for me it made a lot of sense.

Well that goes a long way towards exposing how our math comes out to different conclusions. I only consume 14kWh per day and based on that consumption, there was no way I could concoct an off-grid system including the last expensive Solark Hybrid that would break-even in anywhere close to 5 years.

It really all boils down to the cost/value of your electrical bill, and my annual consumption boils down to ~$1000 for 5000kWh per year at base rate of $0.20/kWh plus ~$110 for the ~430kWh I consume during summer ‘peak hours’ when they charge a $0.25 premium ($0.45/kWh total).

So I’ve only got $1110 per year in annual savings to play with of $5550 over 5 years.

Adding in the 23% solar tax credit takes the budget up to $7210, but with the cheapest Solark 5kW consuming $4500 of that, the remaining $2710 wouldn’t even cover the cost of the battery I’d need, let alone the solar panels, racking and BOM as well.

With your daily consumption being more than 3.5X mine, your annual consumption probably totals to close to $4000 translating to a 5-year budget of $20,000 or $26,000 after tax credit. With the cheapest Solark consuming less than 20% of that budget, I can believe you’ve easily gotten the battery, PV array and BOM deigned for well under the remaining $20,000, probably even leaving you some budget for install.

Achieving the same at a much more constrained budget of $5-6K doesn’t leave any room for a Solark and I’ve determined the best I can do now is aim for a 3kW array to offset ~half of my total consumption plus all my peak consumption. That gives me a budget of $610 per year or $792 per year after soar tax credit. Meaning under $4000 over 5 years.

$1500 for 3kW of panels
$250 for racking
$1500 for a 5.1kWh LiFePO4 battery

That only leaves $750 for inverter/charger/hybrid.

And that’s not leaving anything for BOM, let alone install.

Today, the only way to achieve that is by cobbling together a system using cheap Chinese components, but between panel prices continuing to decline, battery prices continuing to decline, electricity rates continuing to increase and budget UL-listed offerings from companies like Growatt starting to penetrate the US market, I think a 5-year break even with a UL-listed system for consumption levels under 15kW/year will be possible within the next year or two.

But if Solark wants any part of that market, they are going to have to introduce much lower-cost offerings such as the $750 all-in-one 3kW hybrid Huayu is working on.
 
for many it’s only the cost of the electricity

not to wax poetic.. it’s also about the autonomy for me, that’s very valuable

i try to appraise the value of independence and somewhat factor that into the cost of the hardware.

paying for grid costs less, but costs more if autonomy is factored in so to speak.

all these systems are inspiring to me.

big thanks to @GXMnow for explaining the context of their build.

i don’t know how anyone can stand to have a solar system that shuts off when the grid goes down!
 
for many it’s only the cost of the electricity

not to wax poetic.. it’s also about the autonomy for me, that’s very valuable
I did not mean to suggest that breaking even in 5 years is the only motivation to go solar, merely that achieving a 5-year break-even is a big challenge and places severe constraints on equipment choice.

And in particular I was questioning the claim that a 5-year break-even could be achieved with a Solark-based system (which I now understand is easily possible for a large enough system serving large-enough consumption).
i try to appraise the value of independence and somewhat factor that into the cost of the hardware.

paying for grid costs less, but costs more if autonomy is factored in so to speak.

all these systems are inspiring to me.

big thanks to @GXMnow for explaining the context of their build.

i don’t know how anyone can stand to have a solar system that shuts off when the grid goes down!
I don’t know if you’ve already gotten your system in place yet or not, but I’ll add my experience in case it provides more insight.

I started with a 4kW Microinverter-based system under NEM in 2016. Overbuilt in anticipation of a hoped-for EV so electrical bills dropped from $700 annually to $0 and then back up to $120 when they changed the rules and decide to impose a minimum $10/month usage charge.

Between switching the oven from gas to electric, adding a kegerator and my son building a fancy new gaming PC he spends more hours than I want to admit gaming on, our consumption has now increased by ~25% consuming pretty much all the excess production I’d built to charge an EV.

I’m addition, our electrical rates have increased by ~25% since 2016 which on it’s face isn’t an issue since NEM means they credit at the same rate they charge, but then once you factor in the change in peak TOU window from noon to 6pm where it was in 2016 to the 4pm to 9pm peak window they are going to by 2023, and that translates to a ~17% reduction in annual generation credits coupled with an ~11% increase in consumption costs.

The net of all of this is that I went from paying $0 for electricity since 2016 (betting the $120/year in non-refundable charges) to having to pay $120 for electrical consumption at my last true-up this August.

Then you add in the new fire-prevention-motivated multi-day grid outages that result in good spoiling despite that 4kW of solar sitting idle, and the need to add a battery with capability for both backup during outages as well as time-shifting to cover increased consumption costs from 4-9pm was clear.

I ended up building a 14.3kWh LiFePO4 battery sufficiently big to power the full house for a full day, though my current 120V 3kW PSW is only large enough to power one leg of critical loads such as fridges during an outage (so no roast Turkey during a blackout for us ;).

To charge the battery, I added a new 1.1kW DC-coupled array on another face of my roof tied to a 60A MPPT charge controller (Epever AN).

During peak months of June to September, that array charges the battery with 6-8kWh of energy on Sunny days (more than enough to supply peak consumption from 4-9pm).

To offset that peak consumption, I added two 120V 1kW GTIL inverters, one per leg.

Those little inverters only cost me $275 each and they are powered from the LiFePO4 battery and can deliver up to 800W per leg to offset self-consumption. They use CT sensors on the L1 and L2 grid only wires to monitor grid consumption and generate offsetting power to maintain consumption if 5-10W per leg (as long as consumption is not greater than their 800W max output level).

I’m very happy with how they have performed so far (3 months in). They offset all of my peak consumption from 4-9pm over June-to-September peak months and consume all of the remaining day’s production by continuing to offset consumption into the night until the battery is depleted.

Efficiency of these little GTIL2 inverters is poor at 80%, but at a cost of $550 for 1.6kW output capacity, I can’t complain.

All-in, I’ve spent $500 on panels, $250 on SCC, $550 on inverters, and ~$500 on remaining BOM, or $1800 total for a system that should offset over $400 of consumption charges annually.

Another $300 for the PSW that only gets used once or twice a year during PSPS events and $1500 for the 14kW battery that is the heart of the system.

The battery is more three times the size I need right now but we’re still planning to get that EV, and once we do I’ll be able to add more panels to offset the increased consumption without making any other changes.

The $550 + $250 + $300 = $1100 I spent for GTILs + SCC + PSW may not last as long as I’d like and may eventually be more convenient being replaced with a hybrid inverter that would automatically switch to backup power when the grid goes down, but seeing the trend, I believe it’s likely we’ll see 3kW all-in-one hybrid inverters for under $800 within the next year or two.

Maxed out to a full 4kW DC-coupled array, I should be able to offset as much as 3750kWh of consumption with a system that cost me about $5000 in material cost.

That would translate to a best-case break-even of about 6 years or even 5 years if the solar tax credits remain in place when I eventually expand the system (ignoring installation costs since I did everything myself).
 
thank you for the context!

your systems sound awesome! very impressive that you are able to achieve one day autonomy from battery alone???

thanks for helping me understand about GTIL, and other

absolutely not being negative towards anyone doing cost amortization analysis, in fact totally important???

must feel good to be able to live with electricity and be able to generate a significant amount from just the sun rising ?

so far the biggest i’ve handled directly was a 1500W pure sine wave inverter (dumb type) running a hair dryer for a minute or two, off a 12V 100Ah DIY pack with 120A BMS. no permit type big things yet.

nowhere near close to covering my own daily electricity consumption. but a few hundred watts of panels and a handful of kWh LiFePO4 and still learning how to build.

one day i wish to generate enough photovoltaic electricity that battery export to grid would make sense for me ?
 
thank you for the context!

your systems sound awesome! very impressive that you are able to achieve one day autonomy from battery alone???

thanks for helping me understand about GTIL, and other

absolutely not being negative towards anyone doing cost amortization analysis, in fact totally important???

must feel good to be able to live with electricity and be able to generate a significant amount from just the sun rising ?

so far the biggest i’ve handled directly was a 1500W pure sine wave inverter (dumb type) running a hair dryer for a minute or two, off a 12V 100Ah DIY pack with 120A BMS. no permit type big things yet.
That’s the way to learn. My first effort was a DIY 90Ah LiFePO4 battery feeding my 3kW PSW powering a motorized grape crusher in the field for 2 hours…
nowhere near close to covering my own daily electricity consumption. but a few hundred watts of panels and a handful of kWh LiFePO4 and still learning how to build.

one day i wish to generate enough photovoltaic electricity that battery export to grid would make sense for me ?
Just to be clear, I never use battery energy to export to grid. I use battery energy to power self-consumption using my 2 GTIL inverters during peak hours of 4-9pm when electrical rates are jacked-up by +125% (from $0.20/kWh to $0.45/kWh).

Exporting to grid makes less and less sense (especially now that they are considering to change the terms governing export so that they credit at wholesale rates of $0.02/kWh rather than retail rates of $0.20/kWh).
 
TMy cheapest tier before was $0.15 per KWH. Now it is $0.17 and going up to $0.21 or $0.23 on the cheapest time of use. And the peak rate for 4 pm to 9 pm is going up to $0.43 per KWH and that is when the sun goes down, so a grid tie only system can't help there. So even though I was using less than half the power from the utility, my bill only dropped about 15%. That was a shocking eye opener. But now I know better, and you really need to look at what the rates will be after the solar is installed.

The only way around the time of use problem is to use storage.

Raising off-peak rates so span between peak and off-peak gets compressed isn't a problem for those PV users who produce 100% of their consumption, because the higher credits go further toward paying peak rates. It is for those with undersized GT PV systems. Having peak rates shifted to after the sun goes down does hurt.

If we can produce enough kWh off-peak to get credits sufficient to pay for peak rate consumption, that is another way around the problem, without storage. If off-peak is $0.15 and peak is $0.45, it just takes 3 kWh of production off peak to get 1 kWh of consumption on-peak in return.
Given how cheap PV is and how expensive batteries are, I think that is break in capital cost even compared to the LiFePO4 cells bought from China. Reliability, lifespan, and hassle factor of PV is much better than batteries.
 
Raising off-peak rates so span between peak and off-peak gets compressed isn't a problem for those PV users who produce 100% of their consumption, because the higher credits go further toward paying peak rates. It is for those with undersized GT PV systems. Having peak rates shifted to after the sun goes down does hurt.

If we can produce enough kWh off-peak to get credits sufficient to pay for peak rate consumption, that is another way around the problem, without storage. If off-peak is $0.15 and peak is $0.45, it just takes 3 kWh of production off peak to get 1 kWh of consumption on-peak in return.
Given how cheap PV is and how expensive batteries are, I think that is break in capital cost even compared to the LiFePO4 cells bought from China. Reliability, lifespan, and hassle factor of PV is much better than batteries.
I do agree with much of that. If it was only about the cost, that does make better economic sense. But I also wanted the backup power. Having only one full day power outage when I had solar but no battery, was enough for me to say, I want my system to work during a grid outage.

It becomes a balancing act to try and get by with the least amount of battery needed to fire up the solar. When I did the first battery, I truly felt that the 18 KWHs was going to be plenty. And honestly, it really is, if I hade a few more solar panels to keep it charged. The main reason I got the second bank of batteries was because the price was good, and by cycling the cells half as much, they may very well last much longer. I watched a bunch of videos about sizing a battery bank, and with Lead Acid cells, you can actually save money in the long run by buying double the batteries. Sure, it costs twice as much up front, but when the battery banks lasts more than twice as long, it is a savings. So I went from cycling almost 60% to cycling only 30%. How much does that extend the life? I don't know yet.

As many on here know, I have been trying to plan out a DC coupled addition to my solar array. Another 2 to 3 KW of solar panels will make a huge dent in my remaining 4pm to 9pm peak rate. As it is now, I have to rob a bit of morning power to cover the whole peak rate time. It was not a big issue in the spring and fall, but in winter, I am having to buy the power to run my home off peak, while the solar is charging the battery. But then I still get the "free" power from 4 to 9. I just started my charge cycle for today, and at 8 am in the winter, the solar is just barely making the 1,100 watts that I am feeding into the batteries. With the stuff running in the house, that means I am buying 1,000 watts to run my PC, Dish Network, TV, LED lighting, 3 iPhone chargers, etc., and now the furnace just kicked on, and 600 watts is going to the blower motor as well. Looking at the sky outside, even 3,000 watts of additional panel would not zero me out right now. The sun is just too low in the sky.
 
Raising off-peak rates so span between peak and off-peak gets compressed isn't a problem for those PV users who produce 100% of their consumption, because the higher credits go further toward paying peak rates. It is for those with undersized GT PV systems. Having peak rates shifted to after the sun goes down does hurt.
It’s actually a double-whammy: credits earned daylight production hours decreases (by ~25% in my case) while costs associated with peak evening consumption increases (increasing daily consumption costs by over 6% in my case).
If we can produce enough kWh off-peak to get credits sufficient to pay for peak rate consumption, that is another way around the problem, without storage. If off-peak is $0.15 and peak is $0.45, it just takes 3 kWh of production off peak to get 1 kWh of consumption on-peak in return.
The problem with that approach is that you will need to ask permission (since you are adding more than 1kW) which means you’ll be bumped to NEM2.0 and soon to NEM3.0.

NEM2.0 would cost me another ~10% of annual electricity bill in taxes, so I’m not interested in any ‘solution’ that loses my NEM1 grandfathering…

Given how cheap PV is and how expensive batteries are, I think that is break in capital cost even compared to the LiFePO4 cells bought from China. Reliability, lifespan, and hassle factor of PV is much better than batteries.
Except that if you don’t want to export, your only battery-less choice is to add panels and a zero-export inverter to offset daytime consumption only.

I consume an average of 350W during the day, peaking at about twice that level when fridges turn on. So if I added a 1kW array powering a zero-export inverter, I’d be able to offset close to the full 3.5kWh of consumption during summer months and an average of 1.0kWh of consumption during winter months. Call it 800kWh per year or about half of what I’d need to offset loss of production credits alone (without accounting for increased peak consumption costs).

The break-even would be decent at less than 6 years, since that offset self-consumption would save me ~$160 per year and material cost of that system would only be ~$900, but that system would only offset about half of my increased electrical bill.

I could add another 2kW of panels to that same system, which would ~triple wintertime self-consumption offset from 1.0kWh per day to 3.0kWh per day, but that only translates to another 730kWh per year at a value of $146 for a material cost of over $1000 (so a break even of more than 7 years).

It’s not a crazy idea and would offset over 75% of my rate increase, but without a battery, that DC-coupled system is now maxed out (since pretty much all daytime self-consumption is being offset).

You are making me realize it might have made better sense financially to put in a full 3kW array first and only add a battery when I needed to expand the system second, but I wanted a battery for backup power during PSPS outages, so I invested $1500 in that instead of a larger array.

Break-even on that more expensive system is further out but once the full system including 3kW array plus battery is completed several years from now coupled to increased consumption tied to a planned-for EV, break-even is back down to 5-6 years and in the meantime, I’ve got the benefit of backup power.

What I like about the design I’ve implemented is that it is a full-blown DC-coupled platform to offset self-consumption. Today, it’s only offsetting enough of our self-consumption to offset the increased costs we’re facing associated with TOU changes.

But the platform makes it exceedingly easy to expand by merely adding more panels. So once we have an EV and/or CPUC decides to make further changes imposing additional costs on us poor NEM1 customers, it’ll be trivial for me to add more panels to keep up with things (and hopefully panel $/kW will be even lower by then ;)).
 
The problem with that approach is that you will need to ask permission (since you are adding more than 1kW) which means you’ll be bumped to NEM2.0 and soon to NEM3.0.

More panels of a different orientation on existing inverter? More (or replace with more efficient) panels so clipping middle of the day, but more hours of production? Capped-export inverter so added PV & inverter feeds 100% of house loads, allowing existing system to export all it makes to grid?

Once you've got a contract for a particular kW of export, if you could export at that rate continuously for 8 or 10 hours every day, would that cover the hours you can't produce?

Some battery is needed for uninterrupted power. I have one night's worth, but need to shut off unnecessary loads. Oversize PV compared to battery so modest loads are covered during moderate overcast.
 
More panels of a different orientation on existing inverter?
Yes, with a string inverter you could spread out your production by adding a differently-oriented new array without exceeding your export limit, but in my case, since my NEM array is Microinverter-based, that wouldn’t be an option.

And since I already overpanelled my AC-coupled array o that I’m saturated at max output levels for the peak 2-3 hours of the day during summer, even with a string inverter, I’d really only get any meaningful increase in full-day output levels during wintertime (meaning the ~2/3 of annual production during summer months wouldn’t increase much).


More (or replace with more efficient) panels so clipping middle of the day, but more hours of production?
I know you and I have different opinions on this, but PG&E tells me any changes to my existing array needs prior review and approval from them and I don’t want to take any risk of losing my NEM1.0 grandfathering (if/when they change the rules to make that NEM1.0 grandfathering worthless, I’ll have a different attitude).
Capped-export inverter so added PV & inverter feeds 100% of house loads, allowing existing system to export all it makes to grid?
Yes, that was the battery-less option I outlined earlier (and the GTIL-based DC-coupled system I’ve implemented, though with a battery).
Once you've got a contract for a particular kW of export, if you could export at that rate continuously for 8 or 10 hours every day, would that cover the hours you can't produce?
Once I expand my system, I will either need to add an additional SCC for 1-2 panels or for the same cost, I could add another Microinverter.

I’m playing around with some ideas to feed an additional Microinverter controlled by a CT sensor to switch off once total export power reaches my cap. That would give me more morning and late afternoon production year-round and more all-day production in winter.

That idea would allow me to ‘present’ a relatively natural daily production profile so I’d be comfortable with the idea (though it’s more complicated than I’d like).

Attempting to use non-sun-based power (ie: battery power) to export a ‘square-wave’ export pattern during daylight hours would not make me comfortable (too unnatural looking).

Offsetting consumption is the most comfortable solution - you are just consuming less and exporting more during daylight hours, so no difference than if you just decided to economize on energy consumption…

Some battery is needed for uninterrupted power. I have one night's worth, but need to shut off unnecessary loads. Oversize PV compared to battery so modest loads are covered during moderate overcast.
True UPS is expensive and not really a priority for me.

PSPS events are well-announced early-enough that I have no problem ‘getting ready’ for them and if there is no power for the few minutes or even hours between when the grid goes down and I switch the house over to limited off-grid mode for a few days, no biggie.

Taken to the extreme, my full battery powering nothing but our primary refrigerator and some lights can run for close to 3 days straight even if the smoke is so bad there’s no solar production at all.

That’s what my priority is for backup / off-grid power - just never want to lose all of our refrigerated and frozen food during a extended PSPS event ever again…
 
My NEM is 2.0 since I only went solar 2.5 years ago. My agreement states I can't export more than 16 amps max, and no more than 900 kilowatt hours a month. As long as I stay under those numbers, I am "in compliance" with my agreement. My existing 16 panels with Enphase iQ7 micros will make that 16 amps for about 3 hours in perfect conditions in spring. I am 25% over paneled on the inverters with 300 watt panels on the 240 watt iQ7 micros.

My NEM 2.0 agreement is based on this system, and it will max it out if I don't consume any of the power at all. My peak production day was just over 31 KWHs. The whole month hit 865.5 KWHs. Before I added my battery system, I typically use about half of that power while the sun was up, and I exported the other half. Then I would buy back that other half, and maybe a bit more, sometimes a bit less, from sun down to sun rise the next morning.

With the battery, I have it AC coupled right now, and I charge 10 KWH into the batter while the sun is shining, and then use that 10 KWH all evening. During 4 pm to 9 pm, I have it set to "export" at 700 watts. But it is not really exporting all of that. That is just how much is going back to my main breaker panel to make sure I cover all of the loads in my house during that time. If my central air conditioning comes on, I am still consuming 2,000 watts, but if all the TV's are off and such, I could export up to about 500 watts. During the summer, when the solar is still producing past 4 pm, the XW-Pro is smart enough to not invert from battery at all, until the "export" back to the main panel falls below that 700 watt setting. When it is hot out, and the A/C runs a lot, I have cranked that up as high as 1,500 watts to almost zero out my 4pm to 9pm peak time of use consumption.

The plan with adding solar panels is that they will only charge the battery bank. DC coupled with a Schneider charge controller, so it works seamlessly with the XW-Pro inverter. I am on the fence about also adding the battery monitor, and will likely also add the WattNode power consumption meter. With that setup, my existing solar will go back to exporting a lot more during the day, as the XW will not go into charge mode. The batteries will only charge from the DC solar. During the hot summer, I always exceed 5 sun hours here, so 2,000 watts of panel should give my the 10 KWH's now that I cycle into the batteries each day. With the WattNode, the XW-Pro will be able to monitor my true grid consumption and export rate, and automatically adjust the grid sell power to cover my main panel without actually exporting out to the grid from the batteries. It will still shut down the inverting if the Enphase is meeting the house needs. My only grid sell will once again be just what I don't use from the Enphase solar.

I can easily produce over 50% more power without violating my NEM 2.0 agreement. I am not making any change to my approved Enphase solar installation. As far as So Cal Edison will be able to tell from my power meter. I am just using a lot less power each day, and probably even exporting a little more when the sun is up, as if I have everything in my house turned off. With my system as it is now, I have hit zero electric bills for 3 months of the year. With the 50% more production, managed properly with the WattNode, I think I can zero out at least 6 months, and cut the other months in half.

The split AC/DC coupled system has a lot of other benefits as well. It becomes a true "Best of both worlds" system.
With my current AC coupled battery system, I am actually throwing away about 1 KWH every day because of the losses from th AC to DC charging, and the DC to AC inverting in the evening. With the DC charge controller, the efficiency will go from 0.97 from the iQ7, and 0.90 from the XW charging circuit = 87%, losing 13% of the solar energy to the battery, and then inverting from battery again at 93% = 81% total round trip efficiency. The DC solar charge controller is 97% efficient from solar panel to battery. And still 93% battery to AC, so my night time energy will be 90% from solar to AC outlets. That is about the efficiency I would get in an all DC coupled system. Not too bad. But still not as good as my day time iQ7s giving my 97% from solar panel to AC outlets. So as long as the sun is up and the iQ7s are making enough power, I am at that efficiency, and all the DC panels are only charging the batteries while the XW inverter sleeps. And as the sun does start to go down, it gets real interesting. The WattNode will see my export fall to zero, and it will tell the XW to turn on and start producing power to keep my grid power at zero. So late in the day, the sun is still making 30% power on my panels. The iQ7s are meeting 70% of my needs, the XW will be inverting the DC buss to meet the other 30%, but since the DC panels are still making power, that solar is actually powering the XW, and the batteries might even still be charging, not draining.

My system is sort of "True UPS" but only 4 x 20 amps 120 volt circuits into my house. 1) My furnace, and some lights. 2) My refrigerator, the microwave, and some lights. 3) My garage door opener, the WiFi extender, and a few other garage outlets. And finally 3) my main PC and the internet gear, the bathroom lights, my dish network box and a TV. The main panel does go dead, so a few things go off, but enough is still on, we didn't notice the last power failure.
 
My NEM is 2.0 since I only went solar 2.5 years ago. My agreement states I can't export more than 16 amps max, and no more than 900 kilowatt hours a month. As long as I stay under those numbers, I am "in compliance" with my agreement.
My NEM1 agreement is limited to 3.5kW and I was not aware of any monthly kWh cap, but your making me think I should go back and review the terms more carefully.

At any rate, it contains the Clyde stipulating that any changes to the system or system components requires pre-review and ore-approval, so that’s the main thing I’m concerned about.
My existing 16 panels with Enphase iQ7 micros will make that 16 amps for about 3 hours in perfect conditions in spring. I am 25% over paneled on the inverters with 300 watt panels on the 240 watt iQ7 micros.
My 12 250W NEP Microinverters are overpanelled by 34% with 335W panels but I only cap out during the brightest 1-2 hours of the day and only on the clearest days of June and July.


My NEM 2.0 agreement is based on this system, and it will max it out if I don't consume any of the power at all.
My inverters support peak output levels of 300W but only for a short time, so the rated cap on my system is 3.5kW rather than 3kW. That’s the ‘headroom’ I’m thinking of exploiting with another 250 to 500W of Microinverters.

But in any case, that’s without considering any self-consumption (which adds additional headroom). I’ve never capped out at 3.5kW of export.
My peak production day was just over 31 KWHs. The whole month hit 865.5 KWHs.
I’ve had a few days reach 26kWh but am generally hitting 22-24kWh on peak days (often lose a bit to morning fog in summer months). 660kWh is my monthly peak (76% of your peak month with an array which is 84% the size of yours feeding Microinverters capped at 78% of yours).
Before I added my battery system, I typically use about half of that power while the sun was up, and I exported the other half. Then I would buy back that other half, and maybe a bit more, sometimes a bit less, from sun down to sun rise the next morning.
I’m generally generating credits the entire time the sun is shining.

My average consumption after the morning ‘rush’ and before the dinner-time peak is 350W or ~3.5kWh between 8am and 6pm, so ~84% of summertime production goes to export but that drops to only ~68% by depths of winter (pretty much now ;)).


With the battery, I have it AC coupled right now, and I charge 10 KWH into the batter while the sun is shining, and then use that 10 KWH all evening.
So you can time-shift your grid-tied production, nice.

During 4 pm to 9 pm, I have it set to "export" at 700 watts. But it is not really exporting all of that. That is just how much is going back to my main breaker panel to make sure I cover all of the loads in my house during that time. If my central air conditioning comes on, I am still consuming 2,000 watts, but if all the TV's are off and such, I could export up to about 500 watts.
I’d be worried about any export after sunset. Too easy for the utility to detect something has changed. So my grid-tied array just starts exporting when the sun starts shining and stops once it goes down.

I’m using a small new DC-coupled array for time shifting and self-consumption offset.

During the summer, when the solar is still producing past 4 pm, the XW-Pro is smart enough to not invert from battery at all, until the "export" back to the main panel falls below that 700 watt setting. When it is hot out, and the A/C runs a lot, I have cranked that up as high as 1,500 watts to almost zero out my 4pm to 9pm peak time of use consumption.
The HW-Pro is a nice Hybrid and one I nearly jumped on before going another route. I believe it required the inverter to be positioned in between powered loads and the grid, or does it support a CT sensor to supply loads on the grid side?

At any rate, I’ve achieved something similar with my new DC-coupled array and battery.

During the day, DC array output is used to charge the battery.

Come peak TOU hours, the GTIL inverters kick on and use stored battery energy (along with tensing DC charging energy) to power home loads. I’ve rigged up a modified sensor consisting of two CT sensors in parallel onto each leg, so that ‘grid_consumption + AC-coupled_generation’ controls GTIL output.

This way, when home consumption is exactly enough to fully or partially consume AC-coupled generation, the GTIL inverters will supply that power from stored battery energy instead, preserving full output of AC-coupled array for export credit.

The GTILs continue to offset self-consumption until the battery is drained, whether before 9pm or after.

My Epever SCC also allows me to set a second battery threshold to activate a second relay I call ‘excess’ so that anytime the battery charged up full enough that it might get full before peak period starts at 4pm, the GTILs can be turned on early to use that excess DC-coupled energy to offset pre-peak consumption.

So far, I’ve managed to consume 100% of solar energy generated by my DC array without ever having it move out of bulk charge mode (though the true test will be next summer).

The plan with adding solar panels is that they will only charge the battery bank. DC coupled with a Schneider charge controller, so it works seamlessly with the XW-Pro inverter.
Yeah, so that’s similar to the setup I’ve built, except I’m using a $250 60A Epever SCC and 2 $275 GTIL inverters instead of the more expensive Schneider hardware.
 
(Second of 2 - didn’t know there was a 1000 character limit ;))
I am on the fence about also adding the battery monitor, and will likely also add the WattNode power consumption meter. With that setup, my existing solar will go back to exporting a lot more during the day, as the XW will not go into charge mode. The batteries will only charge from the DC solar.
Yes, that is precisely the capability I’ve achieved with my cobbled-together setup. AC-coupled solar serves loads and exports until DC-coupled solar has charged battery to a sufficient level, then DC energy and battery energy is used to offset self-consumption so that close to 100% of AC-coupled energy is exported (or as much as possible during peak TOU window on low-production days).
During the hot summer, I always exceed 5 sun hours here, so 2,000 watts of panel should give my the 10 KWH's now that I cycle into the batteries each day. With the WattNode, the XW-Pro will be able to monitor my true grid consumption and export rate, and automatically adjust the grid sell power to cover my main panel without actually exporting out to the grid from the batteries.
Precisely what my GTIL inverters do (only invert enough battery power to offset consumption without exporting any to grid).

It will still shut down the inverting if the Enphase is meeting the house needs. My only grid sell will once again be just what I don't use from the Enphase solar.
During peak window, I want to export as much power from my grid-tied array as possible, so my GTILs invert as much battery energy as needed to cover self-consumption (without any additional export).
I can easily produce over 50% more power without violating my NEM 2.0 agreement. I am not making any change to my approved Enphase solar installation. As far as So Cal Edison will be able to tell from my power meter. I am just using a lot less power each day, and probably even exporting a little more when the sun is up, as if I have everything in my house turned off.
That is precisely the objective I have with the DC-coupled GTIL system I built - make the house look as much as possible like all breakers have been turned OFF and export as much of the energy being generated by my NEM-approved AC-copped array as possible for credit.

With my system as it is now, I have hit zero electric bills for 3 months of the year. With the 50% more production, managed properly with the WattNode, I think I can zero out at least 6 months, and cut the other months in half.
This is the difference between NEM1.0 and NEM2.0. I don’t pay the tax on 100% of imported power that you do. So I have no incentive to try to minimize consumption on a month-to-month or day-to-day basis. As long as I generate enough export credit over the summer to cover the grid power I need to make it through winter, things are as good as they can be.

I’ve only had my system running since September but managed to run the meter back slightly that month, pretty much kept it flat through October, ran it forward a bit past where it was in late September by the end of November, and I’m clearly now going to be running it further through December, January, and February.

But assuming March comes out pretty close to even, my meter should be back under where it started after my last true-up in August.


The split AC/DC coupled system has a lot of other benefits as well. It becomes a true "Best of both worlds" system.
With my current AC coupled battery system, I am actually throwing away about 1 KWH every day because of the losses from th AC to DC charging, and the DC to AC inverting in the evening. With the DC charge controller, the efficiency will go from 0.97 from the iQ7, and 0.90 from the XW charging circuit = 87%, losing 13% of the solar energy to the battery, and then inverting from battery again at 93% = 81% total round trip efficiency. The DC solar charge controller is 97% efficient from solar panel to battery.

Yes, that’s exactly why I elected to go with a new DC-coupled array to get my battery charged every day…
And still 93% battery to AC, so my night time energy will be 90% from solar to AC outlets. That is about the efficiency I would get in an all DC coupled system. Not too bad.
That is my one gripe about the GTILs - ~80% efficiency when powered from a 24V battery. So. ~78% efficiency from solar to outlet…
But still not as good as my day time iQ7s giving my 97% from solar panel to AC outlets. So as long as the sun is up and the iQ7s are making enough power, I am at that efficiency, and all the DC panels are only charging the batteries while the XW inverter sleeps.
And that’s the reason I’m considering putting one of my 3 strings on another Microinverter rather than adding a second SCC when I expand my system. 97% or 96% efficiency from solar to outlet/grid is like having another 23% to 24% of array power…
And as the sun does start to go down, it gets real interesting. The WattNode will see my export fall to zero, and it will tell the XW to turn on and start producing power to keep my grid power at zero. So late in the day, the sun is still making 30% power on my panels. The iQ7s are meeting 70% of my needs, the XW will be inverting the DC buss to meet the other 30%, but since the DC panels are still making power, that solar is actually powering the XW, and the batteries might even still be charging, not draining.
If I understand correctly, I’ve got a small leg up on you there. My GTILs will be offsetting close to 100% of consumption using remaining DC-coupled power complemented with battery power, reserving 100% of AC-coupled power to be exported for maximum credit during peak TOU window.

My system is sort of "True UPS" but only 4 x 20 amps 120 volt circuits into my house. 1) My furnace, and some lights. 2) My refrigerator, the microwave, and some lights. 3) My garage door opener, the WiFi extender, and a few other garage outlets. And finally 3) my main PC and the internet gear, the bathroom lights, my dish network box and a TV. The main panel does go dead, so a few things go off, but enough is still on, we didn't notice the last power failure.
Yeah, that’s the one thing I don’t have. When grid goes down, I’ve got to turn off the main breaker and turn on the 3kW PSW. That provides 120V power to the full house but none of the 240V stuff can be powered (which is fine since it would swamp that little 3kW inverter anyway). Power to the fridge(s) is the main priority and it’s got plenty of juice for that…

Similar objectives, different approaches (but with more similarities after you invest in a DC-coupled array and I invest in a true hybrid inverter ).
 
Does anybody know if the popular Growatt and MPP "hybrid" inverters (or others) that have grid feedback have the ability to do BATTERY grid feedback?

Updates:
According to MPP solar, only their MPI series has the battery-to-grid export feature. That's only 220V single phase so it doesn't work for me but FYI.
Sol-ark and Deye based inverters seem to be able to do that. Unfortunately, Deye doesn't make split-phase inverters anymore and Sol-ark is $6,000.
Schneider is also capable of battery export, at least some models. $3,000+
Growatt likely does NOT carry this feature but I have not been able to confirm that for sure.
not quite true on the mpi series.
they have a 10 and now 15 kw , 3 phase version
 
fafrd,

That is one of the issues with the Schneider universe.

Each component is another box to buy. If you don't need all the functions, it can actually save money over something like a Sol-Ark, or even the SkyBox. But once you add in the battery monitor, Gateway, and MPPT controller, it is right there. The main reasons I chose the XW-Pro was due to the Cal Rule 21 export grid profile, full UL listing, adaptability, and very robust build quality. The thing is a beast. The 6,800 watt inverter will handle any load I throw at it, and the 12,000 watt surge capacity is insane. I have been running my little Hobart MIG welder off of it so I am have a solar/battery powered welder. And it even runs my air compressor without complaint.

The WattNode is a separate box made by Continental Controls. The required Mod Bus version runs about $400 with the two CTs. The XW-Pro is able to read it to add the true grid export control. You can set it to limit to any level. So you could still allow it to export up to a chosen level. Without it, the XW-Pro can still export back to the main panel, but it won't know if it is not meeting the loads, or is exporting. I have to just manually choose a grid sell current level. It will always feed about the same current to both legs in the main panel. As long as the grid connection is stiff, it won't see any imbalance. On it's output side, it has no problem driving very unbalanced loads. I had well over 20 amps on one leg, and nothing on the other leg, and it didn't care at all. I think it can safely run over 5,000 watts on just one leg. It has a huge toroid output transformer that also acts as an auto transformer to balance the two legs. When it is on grid though, it does lift the neutral with a contactor, so it is not fighting the grid, if there is a voltage balance issue.

You do have an advantage to my system without the WattNode. If I was able to put all of my loads after the XW, it would not be an issue at all, as it does measure it's grid input and load output currents. But when you have additional loads in the main panel before it, it has no way of knowing how much power it needs to push back to power those loads. For that reason, I may end up getting the WattNode even before I add the DC coupled panels. But I am still getting the maximum AC coupled back feed export. The Enphase micros are essentially connected back to the grid anytime the grid is "qualified". The XW has a 60 amp contactor that just ties the input to the output. But it does have separate CT's on them so it knows if power is going to the input (grid selling) or if it is just going to the output loads. I have thought about just remoting the internal CTs out to my breaker panel main lines, as that would do what I need, but it would totally confuse it during a grid outage, or any other time it needs to open the contactor and isolate from the grid.

The last Schneider issue is purely software/firmware. When running in AC coupled only as I am now, there is no way to have the internal software/firmware start a charge cycle automatically. It stops inverting when the battery voltage is 0.5 volts above the set recharge volts threshold. I either need to send it a bulk charge command, or pull the battery down a little more than 0.5 volts to trigger the recharge cycle. A few people have a timer take the XW off grid just long enough to pull the battery down so when it goes back on grid, it will charge, but it is risky. If it happens to pull too low, and then you have a real grid outage, you could get stuck as it can't "dark start" from AC coupling alone. Basically, AC grid tie inverters need a stable grid to be able to go online, but if the inverter shut down from low battery, it won't be able to make the grid so the solar can turn on to charge the battery bank again. Even a small amount of DC solar can fix this as it will be able to charge up from DC solar to then firs up the battery inverter. Then it can make a grid and get the AC solar back up.
 
fafrd,

That is one of the issues with the Schneider universe.

Each component is another box to buy. If you don't need all the functions, it can actually save money over something like a Sol-Ark, or even the SkyBox. But once you add in the battery monitor, Gateway, and MPPT controller, it is right there.
Yeah, that was one of the issues that turned me off on Schneider - it’s a ‘walled garden’ and any Schneider box you need to add is going to cost you a pretty penny more than if you could go open market.
The main reasons I chose the XW-Pro was due to the Cal Rule 21 export grid profile, full UL listing, adaptability, and very robust build quality.
For sure, Schneider, Magnum or Victron deliver a far more reliable and robust product. I was just not happy that nine of the current-generation products fit my requirements and this market is changing so rapidly, I’m convinced it’s just a matter of another few years.

I’ll eventually want that level of quality and will be buying a proper UL-Lusted Hybrid solution when I upgrade my system for a new EV in a couple years.

In the meantime, I’m ‘camping out’ with cheap components that are doing a remarkably good job achieving what I was aiming for (though who knows for how long).

With Schneider, I also wasn’t confident I could make it work with a DIY battery, and that’s an absolute requirement for me (open market).

The thing is a beast. The 6,800 watt inverter will handle any load I throw at it, and the 12,000 watt surge capacity is insane. I have been running my little Hobart MIG welder off of it so I am have a solar/battery powered welder. And it even runs my air compressor without complaint.
I’m comfortable paying for surge power from the grid and as I said, whole-house backup isn’t a priority. So 4kW or even 3kW will be plenty for my goals.
The WattNode is a separate box made by Continental Controls. The required Mod Bus version runs about $400 with the two CTs. The XW-Pro is able to read it to add the true grid export control. You can set it to limit to any level. So you could still allow it to export up to a chosen level. Without it, the XW-Pro can still export back to the main panel, but it won't know if it is not meeting the loads, or is exporting.
OK, so it can support ‘zero-export-to-CT-sensor’ using a(nother expensive) external box like the Elgris Export Controller: http://www.elgrispower.com/zero-export-solar-controller/

Separating the sensor and brain from the power electronics is probably the best architecture as the market continues to mature…

You do have an advantage to my system without the WattNode. If I was able to put all of my loads after the XW, it would not be an issue at all, as it does measure it's grid input and load output currents.
Yes, that was my issue with the XW and most of the hybrid inverters currently in the maker (before Solark) - they required being wired in between the grid and the loads they could offset. That’s pretty much of a showstopper for older wiring.

Solark / Deye was the first Hybrid to do it right, but they are larger and too expensive than makes sense for me.

The last Schneider issue is purely software/firmware. When running in AC coupled only as I am now, there is no way to have the internal software/firmware start a charge cycle automatically. It stops inverting when the battery voltage is 0.5 volts above the set recharge volts threshold. I either need to send it a bulk charge command, or pull the battery down a little more than 0.5 volts to trigger the recharge cycle. A few people have a timer take the XW off grid just long enough to pull the battery down so when it goes back on grid, it will charge, but it is risky. If it happens to pull too low, and then you have a real grid outage, you could get stuck as it can't "dark start" from AC coupling alone. Basically, AC grid tie inverters need a stable grid to be able to go online, but if the inverter shut down from low battery, it won't be able to make the grid so the solar can turn on to charge the battery bank again. Even a small amount of DC solar can fix this as it will be able to charge up from DC solar to then firs up the battery inverter. Then it can make a grid and get the AC solar back up.
Keep an eye on what Huayu is doing. Looks like they are delivering a smart charger that will monitor both AC-coupled solar generation and grid consumption using separate CT sensors so that AC battery charger can consume exactly the excess to avoid export. Combine that with PLC-based communication to throttle-back Microinverter output the way Enphase does with their Envoy ‘brain’ and you start to have the right solution at hand.

DC-PV charges battery through MPPT SCC (self-throttling) integrated into hybrid or stand-alone.

Hybrid inverter monitors AC-Microinverter output and grid consumption and consumes excess generation to charge battery with AC charger.

Anytime consumption exceeds generation, hybrid inverts to offset consumption to maintain zero-export zero-consumption.

If the battery ever depletes (at night), you’re back to consuming from grid until the sun comes up.

If the battery ever charges up to full, DC-coupled array is throttled-back by MPPT SCC and hybrid throttles-back Microinverters through PLC communication to maintain zero-export.

That’s the kind of hybrid system I’m hoping exists by the time I upgrade…
 
Back
Top