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

[Jd1]

Originally Posted by nimblemotors

It is also good to test each cell to make sure they are relatively balanced capacity, since you only get what the weakest cell gives when connected in series.

That will be readily apparent at the knee voltages

[ColdEH]

Originally Posted by Maine Sail

In order to accurately test these cells to "spec" you would need a 350A load. A tad large.... They are rated at 700Ah at a .5C load (50% of capacity)...

For example on my own 400Ah bank I test them at a .25C load (100A) and get 425Ah's. If I test them at .5C I get about 400Ah. If I test them at 15A I can extract 440Ah's from them.....

On my next capacity test, at 550 cycles, it will be done at a 40A load and future tests will be done there because my capacity testing equipment for lead acid only goes to 40A and I can easier automate it than I can the 100A load.. All my capacity tests from the beginning have been done using the same equipment and the same load to remain as consistent as possible.

Lead acid batteries are rated at a 20 hour load which is Ah capacity ÷ 20. So a 400Ah lead acid bank would be tested at a 20A load to the cut off voltage at 75-80F. There is a huge difference between a bank rated at .5C vs one rated at .05C

Hey Mainsail

How close are you to your 550 cycle amp hour test . I think we would all be very interested in those results . Bob Ebaugh was getting a 10% loss of capacity at 330 cycles, which I know you are aware of .

Looking forward.

Regards

[OceanSeaSpray]

Originally Posted by Maine Sail

...on my own 400Ah bank I test them at a .25C load (100A) and get 425Ah's. If I test them at .5C I get about 400Ah. If I test them at 15A I can extract 440Ah's from them.....

Very interesting. 425Ah at 100A then gives a Peukert's exponent k=1.09, and 440Ah at 15A gives k=1.04, reasonably close. Also the value can be expected to change a bit over time.

[Maine Sail]

Originally Posted by OceanSeaSpray

Very interesting. 425Ah at 100A then gives a Peukert's exponent k=1.09, and 440Ah at 15A gives k=1.04, reasonably close. Also the value can be expected to change a bit over time.

IIRC I have my battery monitor set for 1.04 (My note book is on the boat). Interestingly enough I can draw it at 100A to the cut off voltage and get 425Ah's, I can then let the voltage rebound and hit it with a 15A load and get 438Ah's or an additional 13Ah's after a .25C discharge test. When I first got these cells I believed it was a 1.00 - 1.01 Peukert but that is pretty impossible so after some tests I realized that it was closer to a 1.04, for the use I would be drawing it at....

Originally Posted by typhoon

Hey Mainsail

How close are you to your 550 cycle amp hour test . I think we would all be very interested in those results . Bob Ebaugh was getting a 10% loss of capacity at 330 cycles, which I know you are aware of .

Looking forward.

Regards

Unfortunately Bob never got a baseline so it is really tough to say what his capacity loss has been.

The minute I finished top balancing I ran a discharge to 80% DOD then charged back to my chosen charge voltage and ran a capacity test. It yielded 425Ah's. This was my baseline.

I am now only at 504 cycles. I did a capacity test at 500 and got 423Ah's but this was done on-board and I did not let the current fall of to my normal tail. Still, when new I got 425Ah's and at 500 cycles I got 423Ah's...

I am very happy with the results and will continue capacity testing every 50 cycles. It should be noted that I don't float, I don't let them sit at 100% SOC and I stop charging when full. My max charge voltage has been 14.0V but for the last few hundred cycles it has been a max voltage of 13.9V..

The cells have not been balanced in 504 cycles and don't need it..

Many ways to skin a cat and this is what has been working for me...

[ColdEH]

Maine Sail , You lost 2 amp hours after 500 cycles ? !!! Thats it.... im gonna throw mine overboard LOL .

Im not going to do a capacity test on mine, maybe I should . I going to top balance them as you did and put them into use . I'm about a week away from firing up my generator and finishing the cabling but all the hardware is installed . Maybe i will charge them up and let them sit parallel until the generator is ready to charge them .

Regards

[Maine Sail]

Originally Posted by typhoon

Maine Sail , You lost 2 amp hours after 500 cycles ? !!! Thats it.... im gonna throw mine overboard LOL .

Im not going to do a capacity test on mine, maybe I should . I going to top balance them as you did and put them into use . I'm about a week away from firing up my generator and finishing the cabling but all the hardware is installed . Maybe i will charge them up and let them sit parallel until the generator is ready to charge them .

Regards

Yes at a .25C +/- load test. Unfortunately that test was done on-board and I prefer to do it on my test bench so temps can be identical etc. etc..

I suspect it would still be close to 400Ah at a .5C load test but would not bet on it, cause I have not tested at .5C, except for once..

Regardless the capacity on our vessel is drawn out at sub 10A levels anyway.... We rarely if ever see the voltage dip into the high 12's..

Most of those cycles were to 80% DOD or as close to it as we could get. Sometimes timing for charging, coinciding with no wind and a known motor run, dictated when to charge even though we were at 40% SOC etc......

I think fractional "C" use is pretty easy on these batts and that combined with a BMS that provides "insurance", a good initial top balance and sane charging voltages (sub 14.0V) mean an easy life for these batteries.... I suspect age degeneration will get these cells before our usage does but only time will tell.

Once I hit 500 cycles I have slowed down on my cycle testing. Accelerated cycle testing is very time consuming and getting to 500 cycles used a LOT of energy and burned a lot of time...

When the batts come off the boat I will do a more controlled capacity test, even if not at 550, just to confirm my on-board test. I primarily bought this bank for my own experimentation and education but from what I have seen and learned I will not go back to lead....

[ColdEH]

Originally Posted by Maine Sail

Yes at a .25C +/- load test. I suspect it would still be similar at a .5C load test but would not bet on it, cause I have not tested at .5C, except for once.. Regardless the capacity on our vessel is drawn out at sub 10A levels anyway.... We rarely if ever see the voltage dip into the high 12's..

I will be driving my Victron inverter , so probably pulling them down at higher rates then you , I expect .5c out of my 600Ah bank on occasion.

This is kinda getting interesting, and I am curious , maybe I will do a capacity test . What kind of load can get me a .5c draw down . Now Im thinking . hmmmmm

I wonder if anyone did a test on Sinopoly's

Regards

[Jd1]

The batteries have arrived. Balqon packed them very well but the lack of a proper invoice either with the shipment or via email caused a bit of an issue when trying to clear through customs.
I was hoping to find a manufacturing date on the cells but did not find anything.
The 700 AHr cells sure are big boys! I have not tested the initial voltages yet. The next step, after getting the initial voltage readings, will be four days of hooking them up in parallel.

[evm1024]

Originally Posted by Jd1

The batteries have arrived. Balqon packed them very well but the lack of a proper invoice either with the shipment or via email caused a bit of an issue when trying to clear through customs.
I was hoping to find a manufacturing date on the cells but did not find anything.
The 700 AHr cells sure are big boys! I have not tested the initial voltages yet. The next step, after getting the initial voltage readings, will be four days of hooking them up in parallel.

Glad you got them through customs. Take a look on one end or the other at the top. There is a pencil mark that may be date codes on serial numbers. Other may know better than I

[Jd1]

Just checked - no serial numbers, no pencil marks but one cell was obviously installed someplace as there are indents from allthread rod as would have been used for strapping.
Voltages are 3.33, 3,33, 3.34 and 3.34

[OceanSeaSpray]

Originally Posted by Maine Sail

IIRC I have my battery monitor set for 1.04 (My note book is on the boat). Interestingly enough I can draw it at 100A to the cut off voltage and get 425Ah's, I can then let the voltage rebound and hit it with a 15A load and get 438Ah's or an additional 13Ah's after a .25C discharge test. When I first got these cells I believed it was a 1.00 - 1.01 Peukert but that is pretty impossible so after some tests I realized that it was closer to a 1.04, for the use I would be drawing it at....

When a cell runs out of voltage at high C-rate, it still has ions to displace to supply more energy, it is just that it can't do it fast enough to sustain the current. Lead-acids behave the same.

Trying to apply Peukert's law at high currents is very suspicious anyway, because this also starts heating up the cell internally and lithiums are very temperature-sensitive. Just using your two measurements (440Ah@15A and 425Ah@100A) and ignoring the rated capacity suggests k=1.02 and it is probably much more relevant for low-current applications.

[glassbead]

I think I have read enough threads on this topic to make me a new mainsail though admit I have not read the entirety of this particular thread....

I have an older 30' racer cruiser that I recently acquired, that is in need of an electrical update. The 100 amp house battery needs replacing, the regulator should have gone to the scrap heap along time ago, and the alternator is showing signs of age. I have a single bus bar and separate on/off switches for start, house and combine. I don't use lots of current, and only use one charging source - the alternator. I'm not doing extended blue water trips, just shorter coastal stuff.

Taking the LiFePo4 path is appealing if it can be kept relatively simple which I would like to think is doable in my situation. Here's what I am considering:

- Keep my existing lead-acid start battery for now
- Replace the house battery with a 50amp "drop-in" style LiFePo4 - for example Smart Battery's product.
- Keep using my existing 60 amp alternator until it needs replacing
- Using a Balmar MC-614 regulator programmed to some yet to be determined voltage below 13.9
- Victron BMV 700 monitor for the house battery
- Keep the existing single bar set-up - cuz that's what's there.

While I think I understand the concerns with the "drop-in" batteries, with my small system needs, seems like using one makes sense? BMS and HV/LV is built in, and the MC 614 regulator gives an added layer of protection with the battery and alternator temperature sensors and programmable lower charging voltage.

I was considering the BMV-702 to allow for swapping the start battery to LiFePo4 as well at some point, but not sure if there is much advantage in this. Also, I think the lead-acid start battery provides additional protection to the alternator if for some reason the HV cut-off on the house battery is triggered. I still might go with the 702 to allow for more flexibility in the future though.

When running the motor, the "both" switch would be on, so charge current would go to both batteries. I think this is ok at some yet to be determined voltage? and this gives a place for the current in case the HV cut-off on the house battery gets triggered. It is relatively easy for me to take the start battery home and leave it on a good multi-stage charger overnight every month or so to avoid undercharging damage.

For me KISS = keep it small and simple.

Fire away!

Thanks

[Jd1]

I am in the process of checking out a Cellog 8 to monitor the cell voltages of my pack and it seems to behave as if there is zero hysterisis built in into the logger alarm functionality. I wonder if other people can confirm this to rule out that my test setup gives me false readings.
Does anybody use the alarm functionality to turn on/off power relays for charge/discharge paths? If so and if there is in fact no hysterisis built in, how are you dealing with this issue ?
It seems strange that there is so much packed into that little gem yet no hysterisis provisions are made .....

[campr]

Jd1,
using a contactor, rather than a relay in combination with the cellog8. Once it triggers, manual action is required to flip the contactor back to normal operation. I would not want it to automatically switch everything on again. I am using two cell-logs. One for the consumers, and one for the generators. If a cell-imbalence is detected, both will trigger. I have one set up for HVC and one for LVC connected to the generators/consumers respectively.

[NahanniV]

Originally Posted by campr

Jd1,
using a contactor, rather than a relay in combination with the cellog8. Once it triggers, manual action is required to flip the contactor back to normal operation. I would not want it to automatically switch everything on again. I am using two cell-logs. One for the consumers, and one for the generators. If a cell-imbalence is detected, both will trigger. I have one set up for HVC and one for LVC connected to the generators/consumers respectively.

Campr,

I would be interested to know what contactor you are using, also the details of how you have it connected to the CellLog.

Thanks,
JM.