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

[hellosailor]

Terry-
"Why, every time the cell reaches 3.6v it is already 2v over full voltage "
Is that a typo? You are saying the full voltage of a 100A cell would be 1.6V. That's nearly half of what CALB shows for one of their 100A cells, they say 3.2V is the nominal full voltage and 2.5 would be the absolute minimum to let it go to.


Big differences between 1.6 (?!) and 2.5 to 3.2.


Or are those metric volts?(G)

[T1 Terry]

Originally Posted by hellosailor

Terry-
"Why, every time the cell reaches 3.6v it is already 2v over full voltage "
Is that a typo? You are saying the full voltage of a 100A cell would be 1.6V. That's nearly half of what CALB shows for one of their 100A cells, they say 3.2V is the nominal full voltage and 2.5 would be the absolute minimum to let it go to.


Big differences between 1.6 (?!) and 2.5 to 3.2.


Or are those metric volts?(G)

You must have posted while I was posting my correction, it should read 0.2v not 2v, dyslectic fingers or the brain not fully engaged, or both :lol:

T1 Terry


By the way, that 3.2v is nom. voltage, not fully charged voltage, just like a 12v lead acid battery the 12v in the nom. voltage, not the fully charged voltage.

[SaucySailoress]

Originally Posted by hellosailor

Terry-
"Why, every time the cell reaches 3.6v it is already 2v over full voltage "
Is that a typo? You are saying the full voltage of a 100A cell would be 1.6V. That's nearly half of what CALB shows for one of their 100A cells, they say 3.2V is the nominal full voltage and 2.5 would be the absolute minimum to let it go to.


Big differences between 1.6 (?!) and 2.5 to 3.2.


Or are those metric volts?(G)

Originally Posted by T1 Terry

You must have posted while I was posting my correction, it should read 0.2v not 2v, dyslectic fingers or the brain not fully engaged, or both :lol:

T1 Terry


By the way, that 3.2v is nom. voltage, not fully charged voltage, just like a 12v lead acid battery the 12v in the nom. voltage, not the fully charged voltage.

Guys, upon request, I just corrected the post in which the mistake was made - since although T1 Terry did post a follow up immediately, that was clearly missed, and caused confusion.

Hope this helps

[hellosailor]

I'm relieved, Terry. I never did get the hand of the new math ("Just push the button with the picture of the fries on it, dude") or metric volts.(G)


Technically, "2 volts" is the problem. Engineering folks would always write that as "2.0" volts, the same way that "0.2" volts is written, wouldn't they?

[ColdEH]

Originally Posted by typhoon

Just got my system up and running this weekend. 8 Sinopoly 300 AH cells in a single 12v bank of 600 ah . Im using a Link 10 monitor with a Housepower BMS by Clean power Auto.

Going to run them steady and see how many Ahs I get out of them before I need to recharge .

What a difference! When anything switches on , nothing flashes from dropping voltage , I can't tell when my Furnace comes on anymore (Webasto 2010).

Getting the lead out

Regards

Just got the generator up and running for the first time with the Eco-Tech alternator. All I can say , is awesome! That alternator was pumping 295 amps into my LifPo4 batteries at 13.85 volts !! When temperatures got up to 199 F, took about 15 minutes, it throttled back to about 240 amps . Filled my batteries very fast . Current tailed off to about 40 amps when it hit the high voltage cutoff point . Measured cell voltage after an hour resting and was holding at 13.45 volts.

This alternator seems to be up to the task , now what to do with all that power.

Regards

[donradcliffe]

Originally Posted by typhoon

Just got the generator up and running for the first time with the Eco-Tech alternator. All I can say , is awesome! That alternator was pumping 295 amps into my LifPo4 batteries at 13.85 volts !! When temperatures got up to 199 F, took about 15 minutes, it throttled back to about 240 amps . Filled my batteries very fast . Current tailed off to about 40 amps when it hit the high voltage cutoff point . Measured cell voltage after an hour resting and was holding at 13.45 volts.

This alternator seems to be up to the task , now what to do with all that power.

Regards

Congratulations!

It sounds like you have a good LiFePO4 setup for a typical cruising boat. Would you mind giving those of us contemplating the switch from Pb a some approximate planning numbers:

1. How much money have you spent?.
2. How many hours/days labor was involved?
3. How much does the new bank weigh?
4. How long did it take to get what you needed delivered?



Thanks in advance.

[ColdEH]

Originally Posted by donradcliffe

Congratulations!

It sounds like you have a good LiFePO4 setup for a typical cruising boat. Would you mind giving those of us contemplating the switch from Pb a some approximate planning numbers:

1. How much money have you spent?.
2. How many hours/days labor was involved?
3. How much does the new bank weigh?
4. How long did it take to get what you needed delivered?



Thanks in advance.

I have been at this for awhile, I was originally going to use AGM batteries but LifeP04 came along and changed a few things to accommodate them.

1 Spent around $12,000, or probably around there, but that includes a full generator install which I did myself . Hose cables for these things add up . Theres a lot to it .

2 Again this includes the gen install and building the thing , you can look at my profile for those pics and I started a thread on "DC generators the smart way to go" . So Im probably looking at around 200 hrs , with all the engineering and sourcing parts,

3 the batteries weigh 200 lbs . I took out three large AGM,s batteries that weighed 200 lbs each !!

4 I got my batteries from Canadian Electric Vehicles Ltd.
Got them in about 8 weeks and cost me 3400$ to my door .
I had to wait until he put in his quarterly order. This was really the only source I could find up here in Canada, short of importing them myself , and thats a lot of trouble .

I have to give Maine Sail, T1 Terry and Deck Officer a lot of credit for all there help and of course every one on this forum for asking all the questions and the people that answer them .

Regards

[diugo]

Greetings. It is hard to believe that my four used 260Ah ThunderSky cells arrived from Balqon nearly two years ago----way back on page 66 of this thread!

My cells were manufactured on December 6, 2010 and I found it ironic that on the exact fourth anniversary of that date, one of them developed some kind of internal problem that tripled its internal resistance overnight.

Around midnight each night, a microcontroller automatically calculates the internal resistance of each cell by subjecting them (in series) to a C/10 load for 100ms and measuring the current and voltage drop of each cell. This is presented in a digital display I view each morning.

Up until this week the resistances of all four were remarkably constant at around 0.6mΩ each (at 20°C). But last week an unusual period of several days of cloudiness---I only charge using solar---presented an excellent opportunity to perform a bottom balance. I had done the last one eleven months earlier.

When the cells slowly discharged down to the 2.80-volt low voltage automatic cutoff, voltages were { 2.80, 2.83, 3.08, 3.17 } --- quite an unexpected imbalance. In fact it took several hours to drain cells 3 and 4 down to the other two.

It was not until I partially recharged them that I noticed cell 2's voltage was 20-30mV above the others. I briefly put a C/100 load on it to bring it down thinking that would correct it, but it did not. Then I got the internal resistance report the following morning and this morning. Cloudiness persists so I have not yet fully charged the battery.

The cell does not appear swollen, so my thinking is that a copper dendrite has punctured a separator. So my inclination is to charge until cell 2 reaches 3.6V, then subject the battery to the heaviest load I have---90A or C/3.

Any other ideas would be appreciated.

[hellosailor]

Were they sold to you as new or used? With any warranty?

[T1 Terry]

Is the cell voltage only high while charging but drops low under load? If this is the case, remove all the cell links, clean off the cell tops and straps till you see an even bright metal finish on both surfaces, a heavy coating of Almonix and bolt the links back on, pay particular attention to the tightening tension on all bolts and try to determine if any where loose when you stripped out the cells.
Now recharge all the cells till they reach 3.6v, stop when the cell reaches 3.6v and let it sit for 1 hr, if the voltage drops to less than 3.5v, charge it to 3.6v again, repeat the process until the cell will hold 3.5v after 1 hr.
Now do the resistance test to see if the cells have returned to normal.
I am yet to understand the logic of bottom balancing a house battery, I can see the logic for an EV battery as the cells are dragged flat each cycle, but house batteries are not. I see it as an unnecessary strain on the cells and serves no practical purpose.

T1 Terry

[OceanSeaSpray]

Originally Posted by diugo

...
When the cells slowly discharged down to the 2.80-volt low voltage automatic cutoff, voltages were { 2.80, 2.83, 3.08, 3.17 } --- quite an unexpected imbalance. In fact it took several hours to drain cells 3 and 4 down to the other two.
...
The cell does not appear swollen, so my thinking is that a copper dendrite has punctured a separator. So my inclination is to charge until cell 2 reaches 3.6V, then subject the battery to the heaviest load I have---90A or C/3.

Any other ideas would be appreciated.

Increased internal resistance is usually related to growth of the SEI layer. Damage to a separator would rather promote higher self-discharge rates and should not be measurable as internal resistance (if anything it should lower resistance).

Unless the issue is external to the cells, you might have to live with higher resistance in that cell from now on. This is always going to happen over time and I am not aware of any way of reversing it with Li chemistry.

There is a curious progression from 2.80V to 3.17V in your low-end voltage readings. You may want to check that your protection circuit does actually draw equal (and hopefully very small) currents from all cells.
It almost looks like something is being powered from the mid-point between cells 2 and 3, because 1 & 2 are much lower than 3 & 4.

Eric

[diugo]

Thanks Terry, you suggested it might be a terminal resistance issue---and that is all it was! I had removed and cleaned all the bolts and straps during the bottom balance and absolutely swore I had thoroughly tightened all eight terminal bolts, but lo and behold, one was only hand tight. Doh! Now all four cells are within millivolts even under moderate load.

To answer your question about bottom balance, since solar is my only means of charging, under extended cloudiness, it is conceivable that I could draw my house bank down to as low as, say 12.00 volts. In that case, I don't want an imbalance like { 3.1, 3.1, 3.0, 2.8 } triggering the LVD prematurely. Proper balance might buy me a whole extra day.

Hellosailor, the cells are supposedly covered by a 5-year warranty by Balqon, but I was hoping to explore other more convenient avenues first.

Eric, you are right sir---a dendrite should manifest as lower resistance, and significant sudden capacity loss should also be noted; neither was present in my case. That voltage progression you saw is indeed interesting, because the particular voltage divider resistor network does actually draw 4x the current from cell 1 as compared to cell 4. But we are only talking milliamps---hard to fathom that would have such an effect on large 260Ah cells.

[OceanSeaSpray]

Originally Posted by diugo

...
To answer your question about bottom balance, since solar is my only means of charging, under extended cloudiness, it is conceivable that I could draw my house bank down to as low as, say 12.00 volts. In that case, I don't want an imbalance like { 3.1, 3.1, 3.0, 2.8 } triggering the LVD prematurely. Proper balance might buy me a whole extra day.

... That voltage progression you saw is indeed interesting, because the particular voltage divider resistor network does actually draw 4x the current from cell 1 as compared to cell 4. But we are only talking milliamps---hard to fathom that would have such an effect on large 260Ah cells.

Don't discount the effect of milliamp differences in the sensing circuit. Each milliamp drawn unevenly will put that cell out by almost 9Ah over a year, it is quite important to use well balanced sensing currents.

The capacity of the bank is always determined by the smallest cell, whether top or bottom balanced, it is just a matter of choice. The EV-guys introduced bottom balancing because they didn't have individual cell voltage monitoring and they were losing cells at the under-voltage threshold when driving based on pack voltage.

I personally prefer top balancing because it reduces the risk of seeing all cells somehow ever dropping down below the fatal 2.0V threshold. The LiFePO4 chemistry starts decomposing below this voltage and deeply discharged states make me a little nervous.

Great to see the problem gone in all cases!

[mcarling]

The above illustrates one of the reasons why I would never install a cell-level BMS on a LiFePO4 battery bank. In my opinion and experience, it is sufficient and safer to do a top balance and then monitor only overall bank voltage for LVE and HVE cutoffs, and ensure the bank is never overcharged. I see no reason to check cell-level voltages other than before a passage or otherwise once per year. Adding a cell-level BMS is just adding voyeurism and additional points of failure, in my opinion. Snake oil.

[diugo]

All right, I replaced all eight resistors in the divider network so that all four cells now draw the same 2mA. We'll see if that makes any difference a year from now.

I agree that a BMS is generally unnecessary, but I obviously believe in per-cell monitoring for both LVD and HVD. Without it, I would not have known about the loose bolt so soon---and I guess i like being a "voyeur" wrt individual cell internal resistance

Another reason I use bottom balancing is that I try to keep the cells at most between 80-90% SoC---to extend their longevity.