diy solar

diy solar

'Storage' Inverter

a619

Solar Enthusiast
Joined
Aug 28, 2023
Messages
135
Location
TX
One of my family member have an existing Enphase PV system that generates excess energy during the day that they're exporting at very unfavorable pricing. They want to add relatively cheaper battery storage (server rack batteries and not Enphase) that can charge during the day and discharge from evening-morning to power all essential home loads tied to an existing 100A sub-panel (actual max load is ~7-8kW).

I understand that large AIOs like XW Pro, EG4 18kpv or Sol-Ark 15k can handle AC coupling as well as power these load w/o any issues, but those are out of the price range. What are some cheaper (and reliable) inverters that can be used just to charge/ discharge batteries based on TOU to power these essential loads and pull-in power from grid if needed? Existing AC coupled PV would be upstream from this system, so don't need any frequency shifting solution.

Reliability is a must have, beause they're not very technically inclined and just want a set and forget system.
 
A design question is if you intend to put the loads on a critical loads panel downstream from the hybrid inverter. That could require significant rearrangement.

Alternatively, you have to backfeed the battery power and use CT's up stream to meter it.

Any hybrid can accept generation from the micros, but they won't be able to communicate to understand what is excess and what is grid power.

If the owner operator of this system, your family member, is not an enthusiast themselves who wants to be troubleshooting and tweaking the system, then they are probably better off getting an enphase battery inverter so that they can use all the built in logic in the enphase app.
 
A design question is if you intend to put the loads on a critical loads panel downstream from the hybrid inverter. That could require significant rearrangement.

Alternatively, you have to backfeed the battery power and use CT's up stream to meter it.

Any hybrid can accept generation from the micros, but they won't be able to communicate to understand what is excess and what is grid power.

If the owner operator of this system, your family member, is not an enthusiast themselves who wants to be troubleshooting and tweaking the system, then they are probably better off getting an enphase battery inverter so that they can use all the built in logic in the enphase app.
I think that you missed the part about not wanting a hybrid system.
 
One of my family member have an existing Enphase PV system that generates excess energy during the day that they're exporting at very unfavorable pricing. They want to add relatively cheaper battery storage (server rack batteries and not Enphase) that can charge during the day and discharge from evening-morning to power all essential home loads tied to an existing 100A sub-panel (actual max load is ~7-8kW).

I understand that large AIOs like XW Pro, EG4 18kpv or Sol-Ark 15k can handle AC coupling as well as power these load w/o any issues, but those are out of the price range. What are some cheaper (and reliable) inverters that can be used just to charge/ discharge batteries based on TOU to power these essential loads and pull-in power from grid if needed? Existing AC coupled PV would be upstream from this system, so don't need any frequency shifting solution.

Reliability is a must have, beause they're not very technically inclined and just want a set and forget system.
There are a lot of off grid AIO's, that would fit the bill.
I'm not really inclined to recommend any one in particular.
If they won't need to expand on the system in the future. They could use something like the Growatt 12k.
If expansion is a possibility. A stackable would be a better option.
 
I think that you missed the part about not wanting a hybrid system.
Yes, the use case here is a lot simpler. This inverter does not need to interact with upstream PV at all. It would only see "grid" power that needs to used for charging batteries during the day and use that stored power in the evening/ night based on pre-defined schedule
 
Since the Enphase system is tied in upstream (main panel I presume) of the crital loads subpanel you could use a less expensive, value brand inverter such as MPP Solar or EG4 or Sunpower or Growatt. Two things to consider.
1) This is NOT an AC coupled system and the Enphase microinverters as well as the main panel would not be powered during a utility outage.
2) Off-Grid inverters have an internal transfer switch that prevents backfeeding. Therefore every day when switching over to TOU then back to utility later there is going to be an interruption of power for a few milliseconds.
 
There are a lot of off grid AIO's, that would fit the bill.
I'm not really inclined to recommend any one in particular.
If they won't need to expand on the system in the future. They could use something like the Growatt 12k.
If expansion is a possibility. A stackable would be a better option.
I read through Growatt 12k manual and there's no mention of TOU. Most off-grid inverters would work to charge batteries off of PV or based on SOC/ Voltage levels only and not based on TOU.
 
I read through Growatt 12k manual and there's no mention of TOU. Most off-grid inverters would work to charge batteries off of PV or based on SOC/ Voltage levels only and not based on TOU.
It's been awhile since I read through the 12k manual.
I assumed that it had that feature, sorry.
I thought that it would, because my Growatt SPF-5000-ES's do.
 
If you don't find something else that fits the needs.
You can always add a timer on the input, to get the functionality.
 
I looked at the EG4-3kW off grid unit for similar needs. The critical loads panel was predictable enough that I could use a smart relay on the grid side to control charging and branch circuit transfer switches (really just smart relays with a secondary voltage-seeking relay) to control discharge.

Between my refrigerators, IT loads, living room entertainment system, and a few other things I could time-shift for self-consumption about half my daily energy. It would not work though if I needed to cover a 4-hour TOU peak though.

I didn't go that route in the end because it felt too much like a kludge.
 
To the original poster...
What you are asking for is not as simple as it sounds.

I also have an Enphase system that was originally installed as just a grid tie setup with no storage. I then added storage and spent nearly 2 years before I got it to do what I want. Which is what you also want. Store the extra day time energy to use at night.

Here are the issues you need to deal with.

First off, you need to determine how much power is the "extra power". This can be done with a smart power meter of some kind. If you just use CTs at the main panel, they need to be able to tell import from export. Many cheaper meters can't.

Then you need a charger that can be controlled to use this extra power only. How accurate do you want to just consume this extra power for battery charging? For a while, I was just manually turning on the charger at a fixed current that would use about 80% of the extra solar I was typically getting. I had to change the amount of time and max current about every 2 weeks. It would use some grid power from 10 am to 11 am. Allow some expor from 11 am to 2 pm. Then use a little grid power again from 2 pm to 3 pm. The total energy into the battery was then close to my 80% of the extra solar.

Then I would power the loads in my sub panel completely from the batteries from 4 pm to 9 pm when the grid rate is higher. That was the easy part. But that does nothing for the loads back in the main panel.

If you don't mind only supplying the sub panel, and not perfectly matching the import/export power, an inverter/charger that can be set as a UPS can sort of do this. But you do need a timer to control the full power into the sub panel, and the inverter also needs to run the sub panel full load. On a "Time of Use" system. You just cut off the grid power into the inverter/charger. It will switch to invert mode. Hopefully quick enough to not crash a PC. The batteries will then run the sub panel loads. But now the tricky part. When do you put it back on the grid? With my "Time of Use" I had it run from 4 pm to 9 pm as that was the high rate time. After 9 pm, the grid cost was the same as at 10 am. But what if the battery did not store up enough? Oops. You would also need to monitor the battery state of charge and be able to turn the grid power back on before the battery goes too low. And without smart charging control, most inverter/charger units running in UPS mode will immediately start charging as soon as the grid power is restored to the input. It won't wait until you have extra solar coming in.

Doing like @Shimmy suggested above, is similar to this. It can be made to work, but it is far from ideal.

You really need to think about what you are trying to do. And even my Schneider XW-Pro was not able to do it without external help. I had to add power meters and an external PLC controller to tell it what to do.

Since I built my system, things have certainly changed a bit. Some of the newer All in One systems are far more capable. But they are not going to be super cheap. I am still quite surprised that no one has come out with a dedicated AC coupled time of use shifting box. The closest on the market is the Tesla Powerwall2. There are a few others, but they all cost over $10,000 USD once you include batteries.
 
I don't think that perfect zero export is the goal.
If I understand correctly. Just taking back some of the unfavorable export and using it at night, or a more favorable time. So, it wouldn't have to be very complicated.
 
That is correct. I'm just looking to have a simple set up. Doesn't need to be perfect self consumption, goal is just to reduce major chunk of grid import since excess power is available during daytime
 
I am doing exactly what you are talking about with 2 eg4 6500s and using the timer function.
 
I am doing exactly what you are talking about with 2 eg4 6500s and using the timer function.
So far, I have been leaning towards 2x Sunny Island. They're available locally for ~2k each. A bit more expensive compared to 6500s, but so far whatever I have read points to these being significantly better units.
 
After trying numerous configurations here's what I found to be the simplest and least expensive:
First decide on how big of a battrey bank you'll need to get through the night or at least most nights.
Buying a battery bank larger will most likely be a waste of money.
Buying a battery bank of larger capacity than your average daily export will also most likely be a waste of money
Determine the time of day and time period when your export rate is sufficient to charge the battery bank to full capacity.
Consider electrical rates that may vary with time of day.
Set a time to turn on a charger powered right off your grid.
No fancy circuitry should be needed since the charger will draw power from the grid and eliminate or reduce export.
If your grid tie inverter can be operated directly from you battery bank you can use a simple relay/timer circuit switching input from solar to battery set to turn on at night (your chosen hours).
The system will require the typical saftey devices, but most are built into chargers/battery bms's etc.
 
First decide on how big of a battrey bank you'll need to get through the night or at least most nights.
Buying a battery bank larger will most likely be a waste of money.
Buying a battery bank of larger capacity than your average daily export will also most likely be a waste of money

This is not always true. Look at the battery specs. If you cycle 80% of the capacity every day, you will shorten the life dramatically. Double the storage will cost twice as much, but cycling only 40% each day can more than triple the life of the cells. If you plan to run over 10 years, that can save money. I am cycling about 11 KWHs every day. I started with an 18 KWH bank, and it did work, but it was running the cells pretty low each night, since I was only charging to 85%, I was down to 25%. And one night, after running that low, then the grid failed and I had little left in reserve. I doubled my battery bank to 36 KWHs. I still cycle 11 KWHs a day, from 90% down to above 50%. The cells are coasting along. And when I have a power failure, I have 50% left each night.
 
Off the top, it seems quite simple:
First step is to roughly determine load max kwh and desired excess kwh to be shifted, as this would determine battery sizing, to operate between 25-75%. Since there is no AC-coupling, nor solar input, needed items: cheap AIO inverter, battery pack, and CriticalLoads sub-panel.

For example, my "cheapy" ($800) MPP 5648gk has a setting for start/end Output-Source-Priority (in hour increment) and a setting for Charge-Time. As an example, set to use battery in the late afternoon (~6pm) and charge-time in full-sun (~noon). Set the low battery voltage at 25% to turn-back to grid and the max charge voltage to ~75%.

Of course there's the seasonal solar duration to consider. But once configured, it's hand-off operation. Of course another way if OP's family member has EV is to charge during day-time, run the heatpump to for heat/cool storage etc....

am I missing something or way-off ?
 
So far, I have been leaning towards 2x Sunny Island. They're available locally for ~2k each. A bit more expensive compared to 6500s, but so far whatever I have read points to these being significantly better units.

Sunny Island is a fantastic off-grid or grid-backup inverter.
But I don't think it has the function to preferentially store power in battery rather than letting it go to grid, or any support for CT to monitor grid current (only what goes through it.) It can operate as an off-grid inverter inverter, using grid as "generator" to avoid completely discharging battery, and will never allow GT PV to export. It should be used with a GT PV inverter that does frequency-watts.

Probably Tesla PowerWall, SMA Sunny Boy Storage, the new Sunny Boy Smart Energy would do what you want.
An AC coupled battery, monitoring current transformers (perhaps through an external Energy Meter) and charging or discharging battery <--> AC in an attempt to bring export/import near zero. And you want time based functions due to time of use rates.

Both those SMA products use HV battery such as from BYD. Not cheap, I think 2x the cost of server rack batteries. SBSE is announced, expected to have compliance certifications in December and be delivered January.
 
This is not always true. Look at the battery specs. If you cycle 80% of the capacity every day, you will shorten the life dramatically. Double the storage will cost twice as much, but cycling only 40% each day can more than triple the life of the cells. If you plan to run over 10 years, that can save money. I am cycling about 11 KWHs every day. I started with an 18 KWH bank, and it did work, but it was running the cells pretty low each night, since I was only charging to 85%, I was down to 25%. And one night, after running that low, then the grid failed and I had little left in reserve. I doubled my battery bank to 36 KWHs. I still cycle 11 KWHs a day, from 90% down to above 50%. The cells are coasting along. And when I have a power failure, I have 50% left each night.
I stand corrected.
The wear on the batteries definitely has to be taken in to account.

My battery bank was built as for power outages and has much larger capacity than needed for a single night.
I didn't need to concern myself with battery degradation.
 
Last edited:
Back
Top