LTO as buffer

drbacke

Member
Joined
Apr 3, 2019
Messages
60
I asked myself, if it would be possible to add a lto battery Bank before (parallel but physacl. before the 18650 Bank) my 18650 Bank.
The idea is to increase the lifespan of my 18650 bank with few lto batteries. Just like using capacitors to get rid of the voltage spikes.
So i currently use a 7s60p. the voltage is between 23.1V and 28.7V
With 11s1p lto (30-40ah) the voltage range should be covered.
The difference to add just more batteries would be the different infernal resistance. (Lto is mich lower) so they should be more loaded than the 18650s. What do you think?

I ordered currently 1 yinlong lto ti check the quality of the cells
 
drbacke said:
I asked myself, if it would be possible to add a lto battery Bank before (parallel but physacl. before the 18650 Bank) my 18650 Bank.
The idea is to increase the lifespan of my 18650 bank with few lto batteries. Just like using capacitors to get rid of the voltage spikes.

I too am interested in making my 18650 battery bank last a looooong time. But I can't picture the concept - could you explain a bit more, maybe a wiring diagram?
 
The bulk of the stable energy in the LTO cells (40Ah ones) is between 2.1V and 2.3V while in 18650 cells it can be between 3.5V and 3.8V.

So, if you created a 14s pack with 18650 (49V to 53.2V) and a 24s LTO pack (50.4V to 55.2V) it could work and would avoid the 18650 cells from getting whacked with high discharges.

The real issue is the voltage drop on the cells when under load and the difference between discharge and charging voltages.

Wolf is doing some interesting cell testing you might want to loot at, particularly the behaviour of cell re-balancing in a pack.
https://secondlifestorage.com/showthread.php?tid=7873
 
See the test data in detail for the parallel pack behaviour and some of the reasons become a lot more clear.
https://secondlifestorage.com/showthread.php?tid=7873

Also see this thread as to how much you should really be buying cells for so that you take into account cycle life
https://secondlifestorage.com/showthread.php?tid=4737

Another variable is the larger the pack you build the less use the cells actually see and calendar life ageing becomes more of a factor for degredation


Another option, wait for the next set of cells comming out from Tesla next year as they is the real game changer for EV's, however they will be like gold dust for a few years. These cells will be better than LTO (cost / energy density).


image_jimupo.jpg
 
Adding new cells and a hybrid system is simply not the same. From the amount of work and also from the behavior. 10kWh are enough for me and I just don't want to see notebook batteries any more ^^

So why not use and protect my 10kWh with a relatively small high-quality buffer.
With a small number of 18650s you can achieve a very long lifetime. That is my hope.
 
I think for such a system, you want to have 2 completely separate batteries (LTO / LiIon) and DC/DC converters to transfer energy back and forth as required.

The LTO would obviously be connected directly to the charge controller and inverter.
The 2x DC/DC would move energy LTO->LiIon eg. when LTO is charged to over 80%. Then back to LiIon->LTO when LTO charge goes below 20%.
So effectively, the LTO would filter out the shallow short term charge/discharges.

I'm happy to be proven wrong, but I also think it would be more effective to double the LiIon cells to reduce the dod% and amps/cell, instead of adding complexity and a small amount of LTOs.
 
I had thought about the whole mixed LTO + other cell combinations for a long time.

The ideal option would be to have a small LTO pack as the default pack, connected to the inverter and then the other lithium cells providing a slower re-charge to the LTO (with DC-DC). The charging would simply be direct to the other lithium pack. The LTO pack is always aimed to be a full charge to handle the loads.

The big win for a mixed setup is the other lithium cells can be given a very easy time and never see any real surge requirements. The LTO can then be hit with some large discharge requirements.

The issues I came up with was a combination of things.

Losses : DC-DC is around 97% maximum, optimistically 95%, so you loose an additional 5% on a lot of energy in the pack
Discharge : Continuous high discharge would need a DC/Buck from the other pack to the LTO of equivalent capacity
Cycle Life : The other pack would still end up with a lot of cycles, unless large and then the whole cost increases
Pack size : The LTO pack would need to be large enough to cope with your largest continuous load less the input from the other pack, i.e. say you have a 3000W oven and a 1000W boost converter to charge the pack. The LTO cells would see the 2000W discharge requirement and if your LTO pack was only 1kWh then the oven goes cold after around half an hour (ok longer due to warming up, switching off once up to temperature, etc.)...
Components : Going 72V or higher is quite limited on choice, while boost converters over 1000W that can be run in parallel are rare or expensive.

24V / 48V combinations could work, the inverter/LTO being the higher connected voltage pack.

The mixed pack setup would allow for some drastically different options, like the other lithium pack to be at a completely different voltage to the inverter conected LTO pack. I was looking at using 72V or 96V initially and struggling for charge controllers at this voltage, while DC boost converters were ok for 90V at about 1000W. The setup would allow for a larger higher voltage, higher capacity (and cheaper) inverter to be used.

96V version of one of these would allow the transformer to be smaller and cheaper (less copper due to half the amps), however at 96V other issues occur, like death by accident. Around 120V and higher flashover and sustained DC fault arcs create some real issues and risks.
https://www.aliexpress.com/item/33035311978.html
 
As long as LTO and 18650 batteries are physically hooked together (in parallel) there is no way to get power to come from one or the other - as they will simply equalize with each other.

There would have to be an 'intelligent switchs' between LTO and 18650 to manage the charge/discharge scenarios - correct? (or am I missing it).

Assume we are talking about intelligent switching - are there any existing examples of different battery banks (maybe legacy lead-acid vs newer lithium-ion) that do this in the world that could serve as a model?

The idea is such a great idea (to me) to preserve my 18650 battery bank - but I don't picture the specifics yet / much less understand if I can purchase something or wire this up.
 
image_ukjtyg.jpg
 
completelycharged said:

Thank you for the specifics! I think I see what you are thinking. Bothbatteries would provide the full 'battery part' of theconsumption for a 24hr cycle -its just that the LTO could provide short bursts (high consumption during parts of the day)and the18650 would have a smoother cycle - i.e. steady power to recharge the LTOover a 24hr period.

This would mean
1) The LTO would have to be large enough to provide the required 'battery consumption' for the cycle
2) The 18650 would still need to provide the 'battery consumption' - its just that it could do a shallower DOD since its spaced out over 24hrs.

In my case, the LTO would have to be huge (20kwh or more) and that defeats the purpose $ wise for me.
 
Goodmorning,

I have been reading a lot on Victrons site, i can almost memorise all there inverter/charger ect.
It did not suit my needs so i forgot about this one,
But there is an inverter, that can limit your peakpower (most of them)and connect back to the grid or to a other alternatief source.
So if your 18650 cells are comfortable with 25 ah you can limit it to that.
Anything above that will come from a second source, short period of time(lto's)
If both are depleted, it will switch back to the grid.

You can program everything, the way you want and like.

Still reading and learning on Victrons possibilities, probably i am not totally wrong.
Feel free to correct me or sendme into the right direction.

My two cents, best
 
Back
Top