LiFePO4 Batteries: Discussion Thread for Those Using Them as House Banks

[goboatingnow]

Originally Posted by T1 Terry

Li batteries are energy hungry, they have very little internal resistance so current wil flow into them very quickly. Some form of current limiting would be required if a low otput alternator was used. Possibly even charging via the start batteries instead of direct from the alternator and limiting the voltage to 13.8v would enough. Alternators will run for a long time without overheating at 50% to 60% of the their rated output.
Li batteries wil not slow their acceptance rate like lead acid batteries so another possible method is using a DC to DC charger with a capacity low enough to protect the alternator from overheating.

T1 Terry

An alternator with a proper regulator can be set so the alternator is producing as much power as it's cooling system can withstand. Like all electronics that's the defining characteristics.

Acceptance rates etc are irrelevant. All large battery capacity systems of any chemistry have acceptance rates that will destroy charging devices without control systems. Alternators with standard LA orientated regulators can charge Li tech. But you have to understand it's a sub-optimal situation.

Terry all LA chemistries also have low internal resistance too especially AGMs and Gels.

Dave

[OceanPlanet]

TRUE that belt-driven alts will eventually be replaced for efficient & fast charging.

However there are certainly alts available that have much better cooling than others and can run at high output without an alt reg de-rating it. They just cost more.

[T1 Terry]

Terry all LA chemistries also have low internal resistance too especially AGMs and Gels.

Until you have actually worked with them you can't appreciate just how fast they will take on a charge. The prismatic cells that are best suied to house battery applications will handle up to 3C without much terminal voltage rise, that 600amps into a 200Ah battery pack. I have witness then accept 5C but it was a 40Ah cell so charging only lasted for about 10 mins before the terminal voltage started to rise rapidly. At 5C (200amps) over 3Ah were being pumped into the battery every minute, over 30Ah in the 10 mins, no comparison to lead acid batteries.
i posted this on another thread but I'll post it here too, this graph shows 50 amps being charged into a heavily discharged 90Ah battery, the charging curve can be clearly seen and the rapid terminal voltage rise once the cells are full. This was to demonstrate how quickly these batteries can go into run away if the charge voltage is set too high, 3.45v per cell is the max safe limit, anything above that you are playing Russian Roulette with rather expensive batteries.

T1 Terry

[goboatingnow]

Originally Posted by T1 Terry

Until you have actually worked with them you can't appreciate just how fast they will take on a charge. The prismatic cells that are best suied to house battery applications will handle up to 3C without much terminal voltage rise,

T1 Terry

I know , I have a set that I am experimenting with ( BMS, integrated charging etc). In fact I've seen 30C into Li technology!!.

As to runaway, I have not seen evidence of it in LIfepo4. But it is possible to heat damage the casing significantly. IN fact the case suffers from movement during any aggressive charging.

There no problem in charging LIfePO4 batteries from alternators. The regulator needs modification to ensure that the alternator remains within its cooling capacity.( but thats a common problem anyway for any high current charge source).

In practice, however this argument is somewhat theorectical. With large prismatic cells, the typical bank is in the 400-1000Ah range and greater then 0.5C is actually very difficult to generate.

Dave

[Captain Bill]

I am giving LiFePo4 cells a serious look for replacing my house bank. Early in this thread there was a dicussion on equalizing the SOC on new cells. Various contributers suggested a number of ways to drain the high cells and bring up the low cells. Maybe I'm being a bit simplistic, but would not connecting the cells in parallel for a period of time bring them to the same state of charge. It would seem to me that the higher voltage cells would drain into the lower voltage cells until they reached equalibrium. I'm looking at 8 cells for my house bank and it just seems that hooking them up in parallel for a day or two after I get them would be much simpler than chasing small voltage differences. Can any of you experts comment on why this would not be a reasonable idea?

[OceanPlanet]

While the DIY approaches discussed on this forum are great for some, you might consider a full BMS system where all that balancing and HV and LV cutoffs are all taken care of. Mastervolt does current redirecting from cell to cell, however I'm convinced it's not worth all the effort they put into it. They also stuck with a single positive buss. I think the Genasun BMS is the best though-out and safest for marine house banks. The energy lost in shunt/balancing is next to nothing with matched cells.

[senormechanico]

The cutoff voltage for the HiPower cells I will be installing is 3.85 Volts.

http://www.evsource.com/datasheets/B...wer%20cell.pdf

Does that mean I should set my programmable MPPT controller to 15.4 volts, or somewhere less?
I can generate around 22 amps in full sun in the summer.
Nothing on my boat would be damaged by that voltage.

Obviously, I'll set my 100 amp alternator to a setting where it won't burn up the belt.

[goboatingnow]

Originally Posted by Captain Bill

I am giving LiFePo4 cells a serious look for replacing my house bank. Early in this thread there was a dicussion on equalizing the SOC on new cells. Various contributers suggested a number of ways to drain the high cells and bring up the low cells. Maybe I'm being a bit simplistic, but would not connecting the cells in parallel for a period of time bring them to the same state of charge. It would seem to me that the higher voltage cells would drain into the lower voltage cells until they reached equalibrium. I'm looking at 8 cells for my house bank and it just seems that hooking them up in parallel for a day or two after I get them would be much simpler than chasing small voltage differences. Can any of you experts comment on why this would not be a reasonable idea?

The problem with LI cell balancing is that the one cell reaches it "terminal voltage" before the others, if all are parallel then the charged cell will be damaged by continuing to charge it in order to bring the other cells to their terminal voltage. Without cell balancing with a parallel system you have to stop charging once the bank reaches the terminal voltage , that leaves certain cells with less then full SOC, sometimes alot. Again these cells will hit the LVC quicker then those with full SOC, so in effect a large loss of capacity in a severe out of balance system can occur.

You cant solve them by paralleling, if you let the HVC occur the damage is done, its too late to then reduce the terminal voltage by bleeding charge.

You can top balance, ie equalise individual cells to the same fully charged terminal voltage or bottom balance, ie run each cell down to the same low voltage discharged voltage . The consensus is that top balancing is better as it can be done on each charge cycle , whereas bottom balancing might not occur for a long time.

However in fractional C charge regimes and where peak power draw remains a fraction of C, cells do not tend to get out of balance in my experience. Cell Balance is a holy grail in EV development, but there the requirements are high multiple C charges, multiple C peak drain and requirements to run the battery right to the discharge limit to grant the best range. Thats a very different arrangement to the domestic requirements on a boat

Dave

[T1 Terry]

Hi Dave,
I've got to disagree with you on this one. If all the cells are in parallel and charged to 3.45v per cell then as soon at a cell reaches full copacity it can't accept any more charge without the terminal voltage climbing higher than the 3.45v setting. The charge current will simple flow across the link of that cell to the next cell. Charge voltage limiting is the key, constant current charging is for EV's constant voltage charging is for house batteries and 3.45v is a safe voltage. Even in series a 4 cell 12v battery can't have a single cell go over the deadly 4v mark without the oher 3 cells all being below 3.3v. The way the charging voltage curve maps out, that would be a huge out of balance.
Steve B,
that 3.85v per cell is the upper limit, this is used for constant current charging as a point to drop back to constant voltage charging. Constant current is only used in fast charging, 3C to 5C in these type prismatic cells, in house battery banks sticking to constant voltage and allowing them drop into float voltage of 3.4v per cell will ensure long life, these cells will overheat if held at 3.85v for an extended period, I don't even like to see them reach 3.8v, they are in runaway by that stage.

I posted this on another thread but it won't hurt posting it here too, it shows roughly what the charging curve looks like, the 4 coloured lines running in close proximity to each other are the individual cell voltages, the chart was generated using the $28 Junsi Cell Logger 8 available from hobby shops, comes with a free program for the computer to do all sorts of stuff, be even better if you can read German :lol:.
The lines I've added are roughies so don't that them too literally but it gives a good guide

T1 Terry

[Sparohok]

Originally Posted by goboatingnow

The problem with LI cell balancing is that the one cell reaches it "terminal voltage" before the others, if all are parallel then the charged cell will be damaged by continuing to charge it in order to bring the other cells to their terminal voltage.

I wish I had a better understanding of Li cell physics. However when I hear things like this I start to wonder if the problem isn't my understanding of batteries, but rather others understanding of electricity.

If all cells are in parallel, all cells are at the same voltage. How can one cell in a parallel configuration be at a dangerous voltage while the others aren't?

Also, I thought "terminal voltage" was defined as the voltage potential between positive and negative terminals of a battery. By definition a battery is "at" its terminal voltage; it doesn't "reach" its terminal voltage.

You seem to describing the risks of cells being charged in series, not parallel.

Help?

Martin

[bill good]

Martin, I would think provided you have the correct charger with the constant current/constant voltage selected for the cells you have then you achieve top balancing by paralleling. It appears if less than .5C be used then cell balance in the14v (4 cell) battery the balance may not even be needed. The low voltage cut off & high voltage cut off would be still needed as min. The monitoring of each cell with a logger could show if/when the balancing was needed. I think if you look into the auto 14v batteries being offered you will find they may not have all the bells & whistles.

Bill

[Viking Sailor]

Capt. Bill,

You are correct. Suggest you do this at a low SOC.

[Captain Bill]

Goboatingnow, Perhaps I did not make my proposal clear and maybe I do not understand the original problem. It seemed to me that people were reporting that their newly acquired cells were arriving in a significantly different state of charge and that some cells tended to charge at slightly slower rated due to variability in internal resistance. The rsult was that after hooking the cells to a nominal 12 volt charger that some cells reached their peak voltage before others had completed charging. My proposal was not to charge them in parallel, but to charge them until at least one cell reached it's peak voltage, then disconnect the cells and reconnect them in parallel. In short order the high cells should drain into the low cells and all voltages will come to equalibrium at a voltage slightly less than fully charged. In a parallel arrangement electrons will; move from the higher voltage cells to the lower voltage cells until they are equal. Given an assumed difference of only a few tenths of a volt the amps flowing between them should not be too high. In the earlier posts it seemed to me that people were just trying to drain the high cells back to the level of the low cells by throwing the electrons through a resistor. The amps wre relatively high because one was working with 3.4+ volts instead of a few tenths of a volt so one needed a big resistor.

VikingSailor, why at a low state of charge? I was under the impression that this was more of a problem at nearly a full SOC because one cell in a bank reaches the top of its charge curve first and then the voltage rises dramatically at that one cell, even to the point of causing damage. I can certainly see that they would in fact reach equalibrium at any state of charge, though the amps might cause issues if there were dramatic differences n the SOC. I could even see using small resistors between the terminals istead of wires to limit the amperage, though one might be pretty hard pressed to find one small enough at a .2 volt delta. Even a 1 ohm would only let .2 amps pass through and it would take a very long time for them to reach equalibrium.

[T1 Terry]

Won't work Capt Bill, the difference in voltage is in the hundreths of volts not tenths of volts. The out of balance only really appears at the end of charging, as soon as the cell voltage drops below 3.9v the cells are all within 10 to 20 millivolts. If you charge them all together and hold at the 3.45v no cell will get cooked from over charging
The easiest way to top balance a set in series on the conditioning charge is stop when a cell goes high to 3.8v, attach a light bulb or my favourite, a 12v coffee heater element, the cigarette lighter plug removed and alligator clips fitted. Drag the high cell back down to 3.4v and then with the load still attached turn the charger on again. Takes maybe an hr at most for the conditioning charge, once all the cells are brought up to 4v together turn the charger off and let them sit, if any go below 4.45v those cells aren’t fully charged, repeat the process at a much lower charge rate, they stay fairly close in balance after that and never need to be charged to 4v again.

T1 Terry

[Viking Sailor]

Capt. Bill,

The goal is to bring all the cells to the same state of charge. That way when charged all of the cells will end up at the same SOC. The lower the cell's SOC, the lower the energy stored in the cell. Thus, the lower the energy that needs to be transferred between cells.