Increasing capacity.

[arpolis]

Hello all!

I have a 4kw solar array and two cheap Chinese windmills charging a 24v battery bank of AGM batteries at about 15kwh of storage.

Let's say I want to increase tee storage capacity but I can't get these batteries that I have any more.

Would it be possible to build a lithium Ion pack from 18650's in a 7s configuration; so 25.9v nominal voltage, 147 cells all together for just over 2.25kwh storage. Then run both the AGM and lithium banks to separate 90 amp bridge rectifiers to prevent the batteries from charging one another and then paralleling them to my 5kw power jack inverter? Would the different voltages from the two battery banks fry the inverter? Would it matter much?

[joestue]

as long as you ensure the batteries never go above 14.35 volts (times two) you could join the lithiums with the lead acid battery.

you won't be able to drain the lithiums all the way unless you let the lead acid go down to 10.5 volts (times two) which is far too low.
and you won't be able to fully charge the lithiums unless the voltage gets to 14.35 which is typically too high for the standard 13.8 volt AGM "sealed" batteries.

so you'd have to experiment to measure the actual amp hours that flow into and out of the lithium batteries. who makes them and how old they are (if you're thinking of using scrap laptop cells for example) will have substantial impacts on this because the internal resistance might vary by an order of magnitude between a good lithium, and some mystery "sure fire's" bought off ebay.

[arpolis]

Thank you for the reply. I can have my controllers charge the two separate battery banks equally within the high/low specs of the batteries. My issue is that if I have a 2-3kwh lithium bank and a 15kwh AGM bank, the lithium bank may charge faster and before they get to full voltage of 13.8x2 or so then the voltages will be off between morning and mid day. So the AGM may be at 12.2 and the lithium be at 12.8 or whatever. The lithium would not feed the AGM batteries and waste power because of the isolating high amp diodes I would use. But the inverter is now seeing two separate voltages. I was just concerned that it would cause an issue there. Anyone have experience with that?

[DamonHD]

FWIW (and I am ill today, so not concentrating well) I combined LA and LFP for a while:

http://www.earth.org.uk/LiFePO4-battery-testing-with-solar-PV-off-grid-system.html

Rgds

Damon

[arpolis]

Thank you for that. Very informational. However I have a hard time understanding your loads. Are you running both battery sets to a single DC-DC converter and then to an inverter? Or was everything you had running off the DC? In my operation I have one set of LEDs that will run off of the batteries through individual DC-DC converters but the rest of everything is 115v through my inverter. So would I need a massive 200a dc-dc converter to smooth out my battery voltage before feeding to my inverter?

[DamonHD]

Both battery sets via diodes to the same load.  The load was in this case a tiny low-voltage DC load of a few watts.  No inverters in sight...



Rgds

Damon

[arpolis]

I see. My issue is that I don't know if having differentiated dc voltages will throw off the inverter some how. Or would the inverter only draw from the battery with the higher voltage? Or what?

[arpolis]

What about using the idea of “smoothing capacitors”. I've seen in power supplies where it is common to put capacitors in parallel with the output after a full wave bridge rectifier which helps reduce voltage ripple. I would effectively have voltage ripple by having two batteries at slightly different voltages based on load and charge values. So I would run each battery to a car alternator bridge rectifier and the - and + outputs paralleled to their respective buss bars and have a couple 30v 26000uf capacitors in parallel then the leads to the inverter. Hopefully that would keep the voltage reading close enough that it does not freak out the inverter. Or am I over complicating things?

[joestue]

Thank you for the reply. I can have my controllers charge the two separate battery banks equally within the high/low specs of the batteries. My issue is that if I have a 2-3kwh lithium bank and a 15kwh AGM bank, the lithium bank may charge faster and before they get to full voltage of 13.8x2 or so then the voltages will be off between morning and mid day. So the AGM may be at 12.2 and the lithium be at 12.8 or whatever. The lithium would not feed the AGM batteries and waste power because of the isolating high amp diodes I would use. But the inverter is now seeing two separate voltages. I was just concerned that it would cause an issue there. Anyone have experience with that?

no, the inverter will only see the 12.8 volts from the lithium, until they are discharged below the LA batteries. when they are similar in voltages, the current drawn by the inverter will not be split equally between the batteries but will be split according to the internal resistance of them (and the diode, typically diodes might start to conduct at .2 volts (or .5 for silicon diodes), increasing all the way to 1 volt at full "rated" load)

[arpolis]

 Thank you! I was actually kind of internally thinking that the battery with the higher resistance depending on the voltage would not be passing on current and whenever one battery depleted the other battery would kick in. But I was only thinking that in theory. I hadn't read anything specifically that would relate to that exactly. But that makes a lot of sense now. So I think in the near future I'm going to build a 5 kWh Lithium bank And put 1000 W of solar panels to charge it and 3000 W on my AGM batteries. The lithium cells will have approximately a quarter of the total battery capacity giver take a little so that should be pretty close.