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

[deckofficer]

Cycle #29 discharge voltage 12.98 volts (3.24, 3.25, 3.23, 3.26).

100 a-hr cells, load 94 amps, time 67 minutes.

[T1 Terry]

It will be interesting to see how many cycles you get Bob, you were pushing them well beyond the specs the maker uses. they don't discharge to a voltage but they watch the voltage at the end of discharge, if that makes sense.
they discharge 100Ah from the 100Ah cell while measuring the voltage under a 0.5C load. 2.8v is the cut off, under load, not rested voltage.
the LFP cells you have do have a shorter rated cycle life, the LYP cells have a 3000 cycle life to 80% capacity but only a 1000 cycle life to 100%, but that is 100Ah from a 100ah cell, not till the voltage drops below 2.8v rested, that is seriously discharged.
I will stop testing my cells to a loaded 2.8v and rather look to see I can still get the advertised capacity out of them, they will have 1000 cycles up by the time I do the next test and they have been seriously tortured during those 1,000 cycles so I'm not expecting to see the full 720Ah, just too many trips outside their comfort zone, a few hrs at 1.3v and the last week at least 4hrs at 4.17v, yet they are still going.

T1 Terry

[deckofficer]

T1 Terry,

I know my test at 0.94 C is more demanding and pulling 94 amps for 67 minutes is 105 a-hr, but my resting voltage after the load is removed is about 3.25 volts per cell, above the knee. Maine Sail's cycle tests are removing 425 a-hr from a 400 a-hr rated bank, so I'm in line with that, just a bigger C for my small cells.

You would have to agree that if my bank does 2000 cycles, then like the amp-hr rating, the manufacturer's claims of performance is conservative.

[senormechanico]

Kind of like early fiberglass boats, built heavily because nobody knew how long it would last (nearly forever).

[ColdEH]

Terry mentioned a good conductive anti corrosion grease to use to mate the terminals together , can't remember what it was for the life of me. What is everyone using. Just about to get some. .

Regards

[deckofficer]

Originally Posted by senormechanico

Kind of like early fiberglass boats, built heavily because nobody knew how long it would last (nearly forever).

Good analogy. I had a Cal 40.

But I have to wonder how LiFePO4 early sales would have been if they assigned their a-hr rating the same way lead does, the easy 20 hour 0.05 C test? I plan on doing a 0.05 C test and it wouldn't surprise me if my 100 a-hr cells deliver 200 a-hr. The lead test is 0.05 C to complete 100% DOD while the LiFePO4 test is 1.0 C to 80% DOD. Huge difference. I'm pretty sure there is some Peukert effect with lithium, a fraction of that of lead, but at a low amp draw it should show up.

[Maine Sail]

Originally Posted by deckofficer

T1 Terry,

I know my test at 0.94 C is more demanding and pulling 94 amps for 67 minutes is 105 a-hr, but my resting voltage after the load is removed is about 3.25 volts per cell, above the knee. Maine Sail's cycle tests are removing 425 a-hr from a 400 a-hr rated bank, so I'm in line with that, just a bigger C for my small cells.

You would have to agree that if my bank does 2000 cycles, then like the amp-hr rating, the manufacturer's claims of performance is conservative.

Bob,

My test for CAPACITY took the cells to 2.9V IIRC and I shut it down at 425Ah for a 400Ah bank. This was to test the CAPACITY at a 100A load or .25C.

My CYCLE testing is different and cycles to only 80% DOD or less based on Ah capacity removed. The bank is then recharged to 13.8V and 5A of accepted current then cycled again but NOT to 100% DOD...

[Aikiguy]

Where are people buying their cells these days?

Thanks,
Guy
:-)

[deckofficer]

Maine Sail, this low C would be some future test. I'm trying to mirror what you are doing with your 400 a-hr cells, whereas you pull 425 a-hr from 400 a-hr cells, I'm pulling 105 a-hr from 100 a-hr cells.

[Maine Sail]

Originally Posted by deckofficer

Maine Sail, this low C would be some future test. I'm trying to mirror what you are doing with your 400 a-hr cells, whereas you pull 425 a-hr from 400 a-hr cells, I'm pulling 105 a-hr from 100 a-hr cells.

Bob,

Yes they produce more capacity in a CAPACITY test.

However there is a BIG difference between a CAPACITY test which drains them flat, and a CYCLE, which is where you try and hit the CLAIMED cycles at 50% DOD, 80% DOD etc..

A CYCLE should not be to 100% DOD it should be to 80%, 70%, 50% etc. to see if you match the manufacturers claims of XXXX cycles at 80% DOD, 50% DOD etc. etc....... Yes the manufacturers I believe rate them at XXXX cycles to 100% DOD but this is nowhere near 3k cycles....

I am only doing one CAPACITY test per year and aiming for 200 or more CYCLES per year to 80% DOD or close to it...

[deckofficer]

That is what is amazing about these cells. Pulling 105 a-hr from a 100 a-hr rated cell does sound like a capacity test but my resting voltage is 3.25. When I first started I was going to the low battery alarm on the inverter, and pulled 94 amps for 83 minutes or 140 a-hr, and that I would call a capacity test. After 100 cycles of 94 amps for 67 minutes, I'll see if it can still go for 83 minutes at 94 amps.

[darylat8750]

Man Bob, you are really hammering your cells. I hope that casual readers recognize that if your cells don't make it to as many cycles as you are hoping for. I would hate to see your test turn into fodder for the "nay Sayers " .

[hellosailor]

bob-
Bear in mind that all capacity ratings are "plus or minus". While it is nice to see a 100A-rated cell can put out 105A when new...that's still only a 5% difference from the rating, and there's still the question of whether that capacity may actually drop 10-20% as the battery ages, so starting out 5% over rating may still wind up 10% UNDER the rating, during the service life of the battery.
You can also buy solar panels, routinely spec'd in the tiny print as "plus or minus ten percent" of their rating. So a five percent excess? OK, way better than 5% under, but no reason to break out the champagne, yet.<G>

[deckofficer]

I would agree that I'm hammering these cells if after pulling 105 a-hr left the cell voltage at 2.7 volts but they are at 3.25 after the cycle. Gathering everyone's test results it seems that all ratings are based on a 1 C load with at least 20% remaining. At 3.25 volts I feel that 20% is remaining, thus these cycles are 80% DOD. T1 Terry mentioned that the tests are preformed at 0.5 C for ratings but it appears that the cells handle 1 C without noticeable degrading.

Either way, if I get 2000 cycles, it should put cycle life to bed on these cells.

[Opie91]

Originally Posted by deckofficer

Either way, if I get 2000 cycles, it should put cycle life to bed on these cells.

It would be neat if someone took a new 100 ah lead battery and put it to the same test. I wonder how many cycles it would make. Ill pitch in $20.