Recommendations to grid-tie a small off grid system (with backup)

[Ampster]

........ I currently have an off grid system (for backup). Considering upgrading to hybrid or AC coupled for net metering to put the solar panels to a good use.

A hybrid inverter is one that can work as an off grid inverter or a grid tie inverter. AC coupling is a concept often used by a hybrid inverter to control a grid tie inverter when the grid is down. I presume that is what you mean by putting your solar panels to good use.

The most important fact about my questions is that the batteries will be used on a backup system (will be 24/7 floating, unless a disaster strikes). That's the reason I limited my choices to AGM and lithium. For these applications, if the batteries are flooded, the best lead acid 'chemistry' is lead-calcium (telecom batteries). Unfortunately, those batteries, at least where I live, are very expensive. The other lead acid 'chemistry' is lead-antimony (deep cycle golf cart, L16, forklift, etc.). Those are readily available. But perform poorly in float applications (die early, would be best to cycle them to 10-15% daily {not an option}). That's the reason why some pro installers recommend AGMs. Not only because of the maintenance. But because they are usually lead-calcium (like telecom they like to float).

These are my doubts/questions right now. If you recommend a regular deep cycle flooded battery (lead-antimony GC, L16), what maintenance steps should I take to extend their life (prevent stratification, recover capacity lost, not because of sulphation, but because it has been floating for too long). If I should cycle them periodically, how often?

I included lithium because of the recent marketing surge. But honestly, I don't think is the right battery for the application. For off grid, maybe, honestly I don’t know, time will tell. I do not remember the web site, but a couple of months ago I read a master’s or PhD thesis about the loss of life and capacity of lithium batteries in float applications (the batteries lost more than 20% of its capacity in 5 years). Also, several, if not most, lithium battery 'manufacturer' recommends not to float charge them. If that is true, why do people use lithium in residential backup applications? I could understand the needs of an industrial or an energy company. They usually need a lot of energy in a moment notice until the generators are online. They take significatively longer (in case of utilities hours) because they must sync with current production. But that is not the case of a residential application. Time will tell, but in my opinion a lead-calcium lead-acid battery (flooded or sealed) will outperform them (in capacity if flooded/sealed and in calendar life if flooded). For regular off-grid use they are an alternative. Unfortunately, most charge controller are not 100% lithium compatible. Most charge controllers will get damaged if you connect the PV before the batteries. Precisely that is what happens if you discharge them or overcharge them. The BMS will turn them off and bye bye CC.

My comments are strictly related to your comments about Lithium. You are correct, Lithium batteries do not like to constantly have a charge current connected to them. They do not self discharge like Lead Acid so Float is not needed to keep them healthy. The important thing is to have a charge controller that can control the stages of charging. The charger is built into my Outback Skybox and I can control how it charges. Therefore I have set the float voltage to not come on unless they have discharge to a set point. In that case there is essentially little difference between Float and Absorb. They are both Constant Voltage charging. You just need some way to turn them off either by time or Amps. There are good examples of this in the stickies.
I guess it depends on how you define outperform. I am of the opinion that Lithium batteries outperform Lead Acid batteries if you follow some simple rules. They are more efficient than Lead Acid, maintain their charge at any state of charge. Therefore they do not need constant charging and in the long run one would need less solar panels to charge them. In an off grid situation that affects the economics. In a grid tie that means either less investment or more of your solar panel investment is available to sell back to the grid.
I also agree with Mike that many uninformed DIYers are going ruin their first set of batteries. That actually happened when I built my first ebike using Lead Acid. I left them unattended for a couple of weeks at 75% full and they self discharged and lost capacity. I switched to Lithium and have never looked back. Your mileage may differ.

Also, as a note, current I'm considering a third option. That is preserving the off-grid inverter and install a string inverter and replace the charge controller with a 600v model with a manual 600VDC transfer switch.

I am not sure I fully understand this option. Is it grid tied or off grid?

[josepr]

Yes. They have advantages. What I mean when mentioned that lead acid will outperform them is in capacity (in float applications). They will lose their advantage (retain capacity above 80% for a higher cycle count than lead-acid). You have a good point. I see a problem when used with a DC coupled hybrid inverter. But I guess they would be alright with an AC coupled inverter, if you can program the inverter/charger not to float charge them. Do not cheat (just kidding), the SkyBox is a true hybrid inverter. You can use it with or without batteries. You do not need to be in absorb or float in order to sell. It has all the latest bells and whistles. In the end, in my opinion, the main problem is that there is no industry standard communication interface and protocol. So, in the meanwhile, only custom implementations (SkyBox, lithium with xanbus, etc.) works. I dream of a future with one standard. Where you can select your preferred battery, inverter, CC, battery monitor, etc. and they will seamless operate and communicate with each other.

[Ampster]

Yes. ...In the end, in my opinion, the main problem is that there is no industry standard communication interface and protocol. So, in the meanwhile, only custom implementations (SkyBox, lithium with xanbus, etc.) works. I dream of a future with one standard. Where you can select your preferred battery, inverter, CC, battery monitor, etc. and they will seamless operate and communicate with each other.

Yes, all batteries are dumb except the expensive versions like SimpliPhi and others at $1000 a kWhr. Zanbus is a proprietary communication protocol. I agree with you about the need for a universal standard protocol but the market for that has not developed. In the meantime I get by with voltage controls on my Lithium batteries. Bulk (constant current) and Absorb (constant voltage) are all I use for charging. In a sense AC coupling has a protocol for controlling a grid tie inverter based on frequency and the protocols of UL1741.

In my application of load shifting I have everything I need with the exception of being able to control loads to follow my solar output. That would be how I would be able to self consume more without having to store as much energy on the grid to charge my EVs. I do hope for the day that there are more DER (distributed energy resources) signals from the grid.

[Ampster]

Yes. They have advantages. What I mean when mentioned that lead acid will outperform them is in capacity (in float applications). They will lose their advantage (retain capacity above 80% for a higher cycle count than lead-acid). You have a good point. I see a problem when used with a DC coupled hybrid inverter.

I am not sure that is an advantage to always have to Float charge a Lead Acid battery to keep them from losing their charge. To me that is wasted energy but maybe it is a nominal amount. I constantly hear on this forum people advising that a PV system should be larger or the batteries will not reach full charge and their life will be shortened. Ironically I only take my Lithium pack to 90% to increase its life.

The only DC coupled system that I know of is the SolarEdge StoreEdge system and AFAIK it only works with the LG Chem battery which runs at upwards of 400 volts. LG Chem and SolarEdge have collaborated and there is a specially designed communications board that communicates with the LG Chem high voltage battery. DC coupling is very efficient because there are small conversion losses. When the battery is finished charging the rest of the solar output goes through the inverter to the loads or the grid. I have heard that SolarEdge is coming out with their own battery and I presume it will be Lithium. Perhaps it will also be DC coupled. The original Tesla PowerWall was DC coupled but they quickly evolved to a version that was AC coupled and dropped the DC coupled version.
NOTE:
I subsequently learned that Pika Energy also makes a DC coupled system. I think it has to be installed by a dealer.