How do you operate a Li Ion battery bank between 40 and 80% SOC?

softdown

 If you did so the cycle life theoretically would be well over 20 years - according to charts. I've seen lithium fans floating 40 years or more. Keeping it over ~ 40% is not hard, just choose the proper size bank after establishing loads. But how does the CC quit charging at ~80% SOC? 

Not trying to be negative. Thinking my next bank has a good chance of being Li Ion. 

Dave Angelini

Electronics are used for Soc.

BB.

Li Ion cells tend to have very low internal resistance. So, unlike Lead Acid batteries where you let the battery "rest" for 2-3+ hours before measuring voltage to estimate state of charge... With Li Ion batteries (as long as you have "reasonable" current for the cell capacity), a simple voltage check can estimate the SoC at any time.

Also, Li Ion output voltage is not temperature dependent, unlike Lead Acid cells (hot battery, depresses output voltage).

And there is simply measuring the current into and out of the cell (precision power resistor, possibly a Hall Effect transistor) to do real time measurements and use the voltage to prevent/correct drift.

-Bill

softdown

Dave Angelini said:

Electronics are used for Soc.

So Schneider has a CC that can be programmed to do this apparently. Having an Outback MATE system, I'd like to use an Outback CC. Seems like an accessory would have a hard time intervening with the traditional charge controller like the FM80.

One could "maybe" jerry rig a super low absorption and float. Another problem is that Li Ion like a high charging voltage. But perhaps it doesn't matter too much? 

LA and Li Ion are opposites is so many respects. 

Dave Angelini

The Outback battery monitor would be added to an Outback system to do similar with a Radian or similar Outback networked devices.

The really nice part of the Schneider XWP based system is it allows other manufactures (like battery companies) access into the charging parameters if they can write CANBUS or MODBUS instructions into the Schneider Gateway. This gets the Soc directly from the BMS in the battery into your network closed loop.

 If this sounds complex is does not have to be. One of the systems I use just needs the time of day to be set and everything else is programed in for you. Outback is not doing this yet.

mcgivor

Without using a closed loop as depicted in the Schneider example, one would choose the SOC by voltage per cell then multiply by the cell count to obtain the charge / discharge setpoints of the charge controller and inverter to regulate the point at which each disconnects. The settings depend on the particular chemistry used, but using LiFePo4 as an example the fully charged state would be 3.600 VPC , multiplied by the cell count,8 in the case of 24V nominal would be 28.8V. 

To limit the charging to a predetermined SOC value, the charging voltage would be limited to 3.400V for example or 27.2V, this would result in the charging being terminated somewhere in the 90% regon, similarly the inverter LBCO would be raised to cut the discharge at a higher voltage than completely discharged  3.200 VPC or 25.6 V would be somewhere in the 20%SOC regon. These values can be adjusted up/down to achieve a narrower useage band, the cell voltage in a ballanced pack remain extremely close to one another in the normal operating voltage regon, typically ~20 mA, at the extremities the drift is more significant. Having programmable equipment will greatly simplify the process as will having a sound undestanding of the charging requirements.

Some people operate without a BMS relying strictly on equipment setpoints, though it's my opinion that this is not good practice as a BMS provides significantly more safeguards than simply SOC by monitoring the pack at a cellular level. Having the settings of the controller and inverter set below/above the BMS limitations will not affect the protection provided by the BMS but would provide protection should the equipment settings fail for whatever reason.

When dealing with lithium ion  batteries it's best to forget the theroy related to other chemistries as the characteristics are significantly different, many think it complicated but in fact it's actually not, once the fundementals are understood and the battery installed the maintenance is practically non existant.

Using the integrated closed loop equipment will save you in the education or anxiety departments  but will cost significantly more to purchase, as will using pre assembled drop in type such as Battleborne, prysmatic cells,, prysmatic CALB  LiFePo4  cells can be had for close to the cost of AGM, my latest purchase which arrived 2 days ago cost $1063 for 200Ah @ 24V nominal, add $100 for a BMS.

softdown

mcgivor said:

Without using a closed loop as depicted in the Schneider example, one would choose the SOC by voltage per cell then multiply by the cell count to obtain the charge / discharge setpoints of the charge controller and inverter to regulate the point at which each disconnects. The settings depend on the particular chemistry used, but using LiFePo4 as an example the fully charged state would be 3.600 VPC , multiplied by the cell count,8 in the case of 24V nominal would be 28.8V. 

To limit the charging to a predetermined SOC value, the charging voltage would be limited to 3.400V for example or 27.2V, this would result in the charging being terminated somewhere in the 90% regon, similarly the inverter LBCO would be raised to cut the discharge at a higher voltage than completely discharged  3.200 VPC or 25.6 V would be somewhere in the 20%SOC regon. These values can be adjusted up/down to achieve a narrower useage band, the cell voltage in a ballanced pack remain extremely close to one another in the normal operating voltage regon, typically ~20 mA, at the extremities the drift is more significant. Having programmable equipment will greatly simplify the process as will having a sound undestanding of the charging requirements.

Some people operate without a BMS relying strictly on equipment setpoints, though it's my opinion that this is not good practice as a BMS provides significantly more safeguards than simply SOC by monitoring the pack at a cellular level. Having the settings of the controller and inverter set below/above the BMS limitations will not affect the protection provided by the BMS but would provide protection should the equipment settings fail for whatever reason.

When dealing with lithium ion  batteries it's best to forget the theroy related to other chemistries as the characteristics are significantly different, many think it complicated but in fact it's actually not, once the fundementals are understood and the battery installed the maintenance is practically non existant.

Using the integrated closed loop equipment will save you in the education or anxiety departments  but will cost significantly more to purchase, as will using pre assembled drop in type such as Battleborne, prysmatic cells,, prysmatic CALB  LiFePo4  cells can be had for close to the cost of AGM, my latest purchase which arrived 2 days ago cost $1063 for 200Ah @ 24V nominal, add $100 for a BMS.

Comparable to a 400aH 12 volt set up. Such a set up normally costs about $4000 for four good 100aH batteries. How did you do it for $1163?  One BMS? So - one very large battery? Perhaps I misunderstood something? 

I've been thinking that a good charge controller does a lot of normal BMS work. Except for high and low temp cut outs I think? 

How did this become "In the Weeds"? This is 100% about using lithium off grid. Is lithium still considered some type of conspiracy theory? Fringe energy movement? I can almost feel Mcgivors pain. 

BB.

Softdown--I did not move the discussion (at least that I remember  but I am getting old)...

Moved to New Battery Tech forum.

-Bill

RCinFLA

Using cell voltage on LiFePO4 batteries is very poor way to estimate SOC. The only voltage to SOC point you can be confident of is fully charged (3.65v) equals 100%, or totally discharged (which do not recommend doing).

The only way to keep in a desired range is with a coulomb counter battery monitor that is taken to full charge once in a while to allow battery monitor to re-reference 100%.  This is very doable.  Making sure you take to 100% at least once every month or two should be sufficient.

Attached is a no-load curve for LFP taken from several sources as well as simulations.  It takes a very accurate DVM to make the measurements.  When you add in battery equivalent series resistance voltage drop under loaded conditions it can swamp out the ability to project anything. Applying 50 amp draw on a 100 AH cell will appear as dropping from 80% SOC to 50% SOC instantly based on cell voltage.  That is just the resistive losses.  If you leave the 50 amps on for another 3 to 5 minutes the localized ionic depletion losses will double the voltage drop giving the appearence the battery has drops to near 30% SOC based on voltage..

mcgivor

Voltage is a poor way of determining the SOC of any battery however in order to program  setpoints for LiFePo4  one would need to establish an approximate value which would determine the individual cell voltage and theroritic SOC. Keeping the maximum voltage below 3.650V is often recommend for solar applications though this prevents the BMS from providing cell balancing if applicable, this for the most part is irrelevant in a well balanced bank, assuming of course all cells are ballanced and  healthy.

Establishing a maximum setpoint somewhere in the 90% SOC  regon along with a LBCO  in the 20% SOC  will keep the cells within the knees of the charge/discharge curve or the "safe zone" where the majority of the useful capacity is, making.the actual usable capacity 70% of nominal. Operating between  80% & 40% SOC will result in extended cycle expectancy at the cost of lower overall capacity therefore a larger bank may be required, this is what SimpliPhi has done, they revised the setpoints thus reducing the capacity from 3.5 Kwh to 3.2 Kwh of their battery.

@softdown What I purchased are 8 individual 200Ah cells each costing  $134 which will be series connected for 24V nominal, one BMS is required. When building a DIY LFP bank to increase capacity you parrallel first then series to required Ah capacity, then series the cells or cell blocks  to achieve the required voltage, the opposite of LA. The 12V nominal Battleborne batteries are are made up of hundreds of small cells in the same manner, paralleled then series each battery with its own BMS. You're paying for the convenience of having someone build the battery, personally I prefer the larger prysmatic cells for convenience, the pre assembled 12V types usually have limitations in series connection, pre manufactured batteries can be custom made to any required voltage and capacity with on board BMS   at very competitive prices if you shop around.  The previous cells I purchased, 32×100Ah  were cheaper no name but have been in service for almost 2 years without issues these new ones are top of the line CALB with testing reports and hardware.

softdown

Looks like an easy configuration for a DIY BMS. Daly?