What about hybrid powerwalls?

Hello all,

don't know about your "solar history", but I started my system with sealed lead-acid batteries, because -at that time and my state of knowledge then- they're the cheapest storage solution.

So, before I started collecting 18650s for my Li-Ion powerwall, I used 8 "conventional" batteries put together in 2s4p for a 24V battery pack with 940Ah. Worked fine for about a year and a half.

I then thought about how to use the capacity of Li-Ion batteries without replacing the older lead-acid ones. And of course, I also didn't want to replace my charge controllers and my inverter. I then came to the conclusion that, from the technical specs of the chargers and the inverter, it shouldn't be any problem to enhance the capacity of my old batteries with AN 18650-powerwall:

The chargers are designed to linearily charge lead-acid batteries up to 28.8V. For a 7s setup with Li-Ions this means each series will have a maximum voltage of 4.11V (28.8/7).

My inverter automatically switches one phase from grid to battery and back to grid depending on battery SOC. There're 3 settings, I chose the one that switches to battery as soon as possible (@27.6V) and back to grid as late as possible (@23.6V). That's a voltage range for my 18650s in 7s from 3.37 (23.7/7) to 3.94V (27.6/7).


So, my Li-Ion batteries will always work in a voltage range between 3.37 and 4.11V, which is almost an ideal range for them - and which means that this setup in fact acts like a BMS: It prevents the 18650s from overcharging and overdischarging, because the inverter and charge controllers take care ofthat: Charge controller stops charging at 28.8V (cell voltage 4.11V) because it thinks the batteries are full, inverter switches to grid at 23.6V (cell voltage 3.37V) because it thinks the batteries are empty.

The only BMS functions that's still missing is balancing. My solution for this are 1s single cell balancers (link upon request), that can be connected together for as many batteries in series as needed and provide a max. balance current of 10A between the seriesses.

I have this setup up and running for a few months now while step-by-step enhancing my Li-Ion powerwall capacity. And it turned out to work much better than expected: As soon as I added the first 5kWh with 18650s (700 pieces, 100p7s), I could also power device that consume 1000 to 1500W. This didn't work before with lead-acid batteries only, because the voltage drop on loads that high caused the inverter to switch back to grid. Now, with actually 1400 cells and 10kWh, it's doesn't make any problems to power devices with 3000W consumption for a (quite long) while - which results in a remaining grid consumption of 0.5 to 1kWh per day, when running on battery. And I never experienced a voltage difference in my Li-Ion powerwall of more than 40mV.

Phew, that was a lot of text and partially very theoretical, but I would be very interested what you all think about this setup, especially concerning usabillity and safety (I know, thermal monitoring is missing, but I'm already working on that) and if anyone else has done something similar...

Thanks to all and have sun!
Oliver

Sounds like you have bases covered and good to see you're considering the safety aspect of the project. can't wait to see some videos and photos of it all running. Have you added to the cell count yet ?

Hi Owitte could you provide the one cell balancer link please

I always thought that mixing lead-acid and li-ion batteries was a bad idea, because of different charge curves. It is true if you want to get full efficiency from your li-ion, you charge it at 4.11v max and most of they can take 4.2v.

Reading about that matter I see is not as bad as I thought:

http://www.battcon.com/papersfinal2005/ngpaper2005.pdf

Because li-on heps lead-acid chemistry while discharging and lead-acid help the other in the charging process. What it makes a securest system.

You already notice of benefits with high current loads, li-on maintain voltage drop that with lead-acid only can't achieve.

After reading that the only matter that still worrying me is the DoD (Deep of Discharge ) having such amount of li-on on your system if you are treating your system as li-on, discharging until a voltage such for your lead acid can be too much, reducing life of them.

But watching your numbers you have a very conservative config, you switch to grid at 23.6v what it is 3.37v for each li-io serie.

So that confirm that having your system optimizd for lead-acid protect your li-on fluctuating deom 3.37v to 4.11v, so you gain security and li-on life (cycles) but you lose the capacity that li-on can give from 3v to 4.2v.

So if you don't need more capacity of your system to avoid take from grid probably you have a good overall config.
Am I miscalculate you have 1/3 of your power is with lead acid?

Regards

Jesusangel. If you cycle your lion between 4.11 to 3.37 you get like 90% out of them easily! Note that there is not much left below 3.4V

Combining battery chems is doable but in the end you will need to remove parts of the advantages of the other bank. For instance one thing to note with lead acid is that those need to be bulk charged and topped of and then you need to back off the current.

This will not happen if you have a li bank hooked up to the bank because when the charger have backed off the lion bank will start to put current back into the lead acid bank when that bank settles it voltage to resting.

lion settle its volt with couple of mV... Meanwhile lead acid does drop alot more.

Example:

48V lead acid that is bulk charged to 14,5*4 = 58V
Float: 13*4 = 52
That means that you should only top it to 58 and then leave it going down to 52.

lithium that handles have to do that voltage is 14s = 4,14V per cell
And when the lead acid reached its float voltage = 52/14 = 3,7V

Between 4,14 and 3,7 there is ALOT of energy depending on cell = The lion bank will give all its energy back to the Lead acid bank.
So instead of dropping the bulk charge of to float that those batteries need the lion will feed back energy untill equalization is done.

Im not saying above doesnt work because it does! But everyone reading need to understand that the different charging types to create issues and they can degrade the useable energy of the cells.

How much does this then mean for the batteries? If you are conservative perhaps a couple of % in total. if your not... you can loose alot of cycles.


Also worth noting is that Liion bank will take all the punches kind of. Since Lead Acid generally sag more than liion. So if you havent sized it all correctly you will have very uneven currents flowing

Recomendation is to meassure the currents.

Btw I am working on video that do compare running Lead acid with Liion and Liion with LiFePo4

hbpowerwall said:

Sounds like you have bases covered and good to see you're considering the safety aspect of the project. can't wait to see some videos and photos of it all running. Have you added to the cell count yet ?

Click to expand...

Hi Pete,

cell counter is done, of course, and you can find some pictures here. But since I have a full-time job, volunteering at a maker space/FabLab at the weekend and doing sports, I don't have time to make, edit and upload videos right now... I'll add some more pics to the thread under DIY powerwall builds from time to time.

Have sun!
Oliver

---

jesusangel said:

I always thought that mixing lead-acid and li-ion batteries was a bad idea, because of different charge curves. It is true if you want to get full efficiency from your li-ion, you charge it at 4.11v max and most of they can take 4.2v.

Reading about that matter I see is not as bad as I thought:

http://www.battcon.com/papersfinal2005/ngpaper2005.pdf

Because li-on heps lead-acid chemistry while discharging and lead-acid help the other in the charging process. What it makes a securest system.

You already notice of benefits with high current loads, li-on maintain voltage drop that with lead-acid only can't achieve.

After reading that the only matter that still worrying me is the DoD (Deep of Discharge ) having such amount of li-on on your system if you are treating your system as li-on, discharging until a voltage such for your lead acid can be too much, reducing life of them.

But watching your numbers you have a very conservative config, you switch to grid at 23.6v what it is 3.37v for each li-io serie.

So that confirm that having your system optimizd for lead-acid protect your li-on fluctuating deom 3.37v to 4.11v, so you gain security and li-on life (cycles) but you lose the capacity that li-on can give from 3v to 4.2v.

So if you don't need more capacity of your system to avoid take from grid probably you have a good overall config.
Am I miscalculate you have 1/3 of your power is with lead acid?

Regards

Click to expand...

you're right, me too, I was sceptical in the beginning. I know that I can't make user of the full capacity of both battery systems (lead-acid batteries will only be discharged to 50%), but who cares, I just add more harvestest Li-Ion cells to get more capacity .

Concerning the DoD issue: Of course I left my chargers and inverter in lead-acid battrey type settings.

Details on my batterie mix: I have 8 sealed lead-acid battries with an overall capacity of 940Ah (@24V). My 18650 powerwall consist of 1400 cells in a 200p7s config with 420Ah. so it's almost 50/50.

Have sun!
Oliver

---

Darmok said:

Hi Owitte could you provide the one cell balancer link please

Click to expand...

the balancers I'm using can be found here. From my experience, they do a good job, the voltage difference in my system is always <40mV.

---

daromer said:

Jesusangel. If you cycle your lion between 4.11 to 3.37 you get like 90% out of them easily! Note that there is not much left below 3.4V

Combining battery chems is doable but in the end you will need to remove parts of the advantages of the other bank. For instance one thing to note with lead acid is that those need to be bulk charged and topped of and then you need to back off the current.

This will not happen if you have a li bank hooked up to the bank because when the charger have backed off the lion bank will start to put current back into the lead acid bank when that bank settles it voltage to resting.

lion settle its volt with couple of mV... Meanwhile lead acid does drop alot more.

Example:

48V lead acid that is bulk charged to 14,5*4 = 58V
Float: 13*4 = 52
That means that you should only top it to 58 and then leave it going down to 52.

lithium that handles have to do that voltage is 14s = 4,14V per cell
And when the lead acid reached its float voltage = 52/14 = 3,7V

Between 4,14 and 3,7 there is ALOT of energy depending on cell = The lion bank will give all its energy back to the Lead acid bank.
So instead of dropping the bulk charge of to float that those batteries need the lion will feed back energy untill equalization is done.

Im not saying above doesnt work because it does! But everyone reading need to understand that the different charging types to create issues and they can degrade the useable energy of the cells.

How much does this then mean for the batteries? If you are conservative perhaps a couple of % in total. if your not... you can loose alot of cycles.


Also worth noting is that Liion bank will take all the punches kind of. Since Lead Acid generally sag more than liion. So if you havent sized it all correctly you will have very uneven currents flowing

Recomendation is to meassure the currents.

Btw I am working on video that do compare running Lead acid with Liion and Liion with LiFePo4

Click to expand...

you're right, I'm losing some % of the Li-Ion power in this setup. But who cares, if I want more, I just add more cells .

And of course I meassured the current flow at different points in my setup: I checked what comes in from the chargers, how the current devides between the lead-acid and the Li-Ion banks and what happes when running on high loads. For now, I'm doing that manually from time to time (plan to do that permanently with high current sensors and log everything) and I didn't notice anything disconcerting. The "draining ratio" is always around 1/3 lead-acid, 2/3 Li-Ion. On a sunny day, with ~1kW coming in from the panels (~40A charging current) and on almost full load (inverter takes 120A), the Li-Ion bank gives ~60A and the lead-acid bank adds another ~20A. After cutting the load, there're still flowing 5-8A from the Li-Ion bank into the lead-acid one for a few minutes in order to balance the voltage.

Cool m8! Thats good to hear. This information i think is important for new guys and those thinking of it.

For instance i run LifePo4 with normal laptop cells and its not easy. it works but i loose alot from the LiIon. But on the other time i get alot more cycles out of em.

These balancers you're using sure look interesting, they also have a pre-build 7s module for a reasonable price; 10A balancing current is high enough for high-power applications using non-balanced modules. Anyone else using them? How does the 10A balancing current compare to batrium's BMSs?

thanar said:

These balancers you're using sure look interesting, they also have a pre-build 7s module for a reasonable price; 10A balancing current is high enough for high-power applications using non-balanced modules. Anyone else using them? How does the 10A balancing current compare to batrium's BMSs?

Click to expand...

Don't know about batrium, but these balancers do their job quite good. And I can confirm that they really do balance with up to 10A, I personally measured 6-8Amps flowing when connecting unbalanced packs together. And I'm also in contact with the producers for the 7s-Modules. But keep in mind these products just balance, they don't do anything else like batrium...