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

[diugo]

Originally Posted by goboatingnow

Have you seen the DataQ Data Acquisition Products for ANY Application and Budget , voltage loggers , the lower spec ones are very good value

Two of their starter kits are amazeballs inexpensive! I was going to suggest using an Arduino clone (like I do) to measure and write the data to an SD card, but they already do it at a comparable price---without the need to be a computer programmer.

[senormechanico]

FWIW, my "float charger" previously discussed here is a 0.500 amp wall wart with an adjustable voltage regulator circuit added. It has no problem keeping my 700 aH bank with a constant 300 milliamp drain (various led's and the Midnite Solar MPPT for grid tie backup system) at 27.6 volts.
I might dial it back a couple hundredths as Balqon recommended 27.4 volts.
From full charge at 28.5 v. when my Clean Power Auto BMS gave the warning beep, I've slowly bled off to the 27.6 figure. That took over 100 aH.
I still haven't discharged the bank anywhere close enough to find out the real capacity.

[diugo]

Originally Posted by Blue Skye

Question - Why are deckofficers cells not maintaining balance? It appears everyone else is reporting the cells voltage are staying very tightly balanced. Is it because they were not properly balanced or he is pushing them too hard?

Ideally, individual cells are matched by the vendor by both capacity and internal resistance. Deckofficer (and I) bought used cells from Balqon, so it's unclear how well matched they were. And even if they were well matched based on measurements taken at the factory, they may have aged differently auld lang syne.

Moreover, even very closely matched cells may have different characteristics at the top and bottom ends. This is why most EVers swear on the value of the BMS---they need automated balancing to ensure good balance at both ends, to ensure they get nearly full charge and full discharge from every cell, every cycle.

But boaters and RVers aren't nearly as demanding. We have found that as long as they don't go below 10-20% SoC (if top balancing) or above 80-90% SoC (if bottom balancing), the cells do stay pretty well balanced on their own, without the need (and expense) of automated balancing.

Deckofficer's full capacity tests down to 0% SoC required him to violate this rule, and that's why he is observing the imbalance.

[Blue Skye]

Thank you for that explanation. If the cells aren't closely matched and you kept the SoC between 20-80% would they maintain balance better? Or is getting closely matched cells imperative for cell balance? Thanks again for your help.

[Maine Sail]

Originally Posted by Blue Skye

Thank you for that explanation. If the cells aren't closely matched and you kept the SoC between 20-80% would they maintain balance better? Or is getting closely matched cells imperative for cell balance? Thanks again for your help.

Yes they would because you would never be hitting the tops or bottoms where the lower capacity cell may empty earlier than the others or fill earlier than the others.

I have found that a single top balance has been good for out bank now for 206 cycles. I need to get back on my cycle testing but have just been too busy...

EV guys often bottom balance and for off grid, like boats, with multiple charge sources, I find top balancing is a slightly better way to go...

[diugo]

I agree.

You top balance if you seldom need to contend with the low end of the discharge curve, ie you have one or more reliable charging sources you can take with you. Most motorized boaters fall into this category.

You bottom balance if you think you might find balanced cells at the low end a lot handier than the weakest cell always triggering the LVD while the other cells still have plenty of charge. Anyone who can't bring a charging source along (EVers, eKayakers) or those with potentially unreliable charging sources (solar, wind, PG&E) would probably be wiser to bottom balance.

[Joe Parklane]

One burnt up alternator. The fix required new programmable smart regulators to reduce alternator output to 80% max and is temperature regulated further by a temperature sensor mounted on the alternator frame that further reduces output as temperature rises. Also, the original single belt arrangement was slipping badly from LFP demand generating more alternator heat so necessitated a retro fit multi groove serpentine belt install.

[roetter]

Originally Posted by Joe Parklane

One burnt up alternator. The fix required new programmable smart regulators to reduce alternator output to 80% max and is temperature regulated further by a temperature sensor mounted on the alternator frame that further reduces output as temperature rises. Also, the original single belt arrangement was slipping badly from LFP demand generating more alternator heat so necessitated a retro fit multi groove serpentine belt install.

What kind off alternator?

I have exchanged my 840Ah LA for 1400Ah LFP. Left the alternators as they were. Each engine has a Hitachi 100A alternator. I am only 10 days into the new setup, but do not see any problems yet. The alternator's internal regulators quickly regulate them down to about 60A with the internal temperature sensor. I have already run several times on a single engine and on both engines for several hours at a time with no problem so far. It seems the interval regulator takes well care of overheating.

I have a programmable voltmeter that disconnects the alternators from the LFP batteries at 14.0V and reconnects at less than 13.35V. The two LA start batteries remain connected to the alternators. Nearing 14V the into the LFP drops to about 10A. The most I have started charging with the alternators was about 300Ah down from 14.0V "full".

[Joe Parklane]

The alternators are Balmar 6 series (small frame) 100 amp. Your alternators (wow internal temp sensors) seems to be working very well...jealous and you have a well regulated system as my very limited knowledge understands. I assume the starting batteries remain in the circuit to protect the diodes when disconnected while charging(my set up also). My system has been set up very conservatively...charging sources stop charging at 13.8, however I have seen some over shoot on my setup (750 amp CALBs 4s4p integrated BMS((BMS set for HVC disconnect @ 14.2) and cell boards per 4s with balance resisters @ 3.525 volt pack 14.1) with voltage climbing to 13.9+ volts briefly. This is very conservative with double redundancy to protect the pack designed and installed by a very knowledgeable electrician.

[Maine Sail]

Finally getting around to doing my yearly capacity testing on my 400Ah bank.

Last year I capacity tested this 400Ah bank consisting of 4X 400Ah Winston cells and took out 425Ah's at a load of 102-105A. I shut that test down at close to 2.8V for the lowest cell

This year, I completed 208 cycles, with most to 80% DOD. During this time there was no balancing, no BMS/shunt balancing and only charging to 13.8V and 5A accepted current, or close to that current level before shutting off charge sources and using the bank to 80% DOD per a calibrated Link-Pro using Ah consumption. Battery monitor was manually reset each time the bank hit 13.8V/5A net accepted current.

Pack Charge:
for this capacity test I first charged the bank to 13.8V and 8A of accepted current. I got impatient and did not want to wait a few more minutes for it to drop to 5A. The pack had been charged using my 50A bench top power supply. Top finishing voltage was 13.82V as measured with my Fluke 179. charging was discontinued at 8A of current flowing into the bank at 13.82V.

Discharging:
I ran this test at the same load as last year, using the same equipment, to keep everything as A/B as possible. Load was 102-105A continuous for a approx 4 hours. (load changes slightly with voltage)

Monitoring:

Ampere Hours: To track Ah consumption I used a Victron BMV-602 with the Peukert set to 1.03. The Victron was programmed then manually synched after charging to 13.82V @ 8A net accepted current.

Voltage: I used a calibrated Fluke 179 to track pack voltage and a calibrated Cell Log 8S to track cell voltage. Periodically cell voltage was double checked with the Fluke. I also have a Smart Gauge and the voltage on the Victron BMV.

I also have a Bogart Engineering PentaMetric with computer interface which was giving me fits today. Don't even get me going on the Log View software for the Cell Log 8. I could not get the USB interface to cooperate on the Pentametric and finally gave up and did the test using the Fluke, Victron BMV-602 and Cell Log....


Results:

I fired up the inverter which was driving a ceramic disc heater and also a smaller inverter driving a large light bulb to push the current where I wanted it.

It ran at 102-105A load (approx .25C) for the duration.

I stopped the test when I hit 425Ah the same exact performance I had 208 cycles ago. When I stopped the test the lowest cell was at 2.872V.

After I shut off the large load and the voltage crept back up I turned on the small inverter and a 15A load. As I type this the lowest cell is at 3.008V and the Ah consumption just surpassed 431Ah's..

Cell & pack voltages measured at -432Ah's (400Ah bank) under a 15A load:

Cell #1 = 2.993V
Cell #2 = 2.966V
Cell #3 = 3.043V
Cell #4 = 2.956V
Pack Voltage 12.045V

Keep in mind that these cells have not been balanced for 208 DEEP cycles. When they were balanced, they were top balanced not bottom balanced so the 0.087V variance or cell drift, while loaded, 32Ah's beyond "rated" capacity is pretty darn good and to be expected IMHO....

My Observations/Thoughts/Musings:

Charging Voltages:
From where I sit these cells never need to be charged beyond 13.8V. If you let the current tail to 5A at 13.8V you are going to get the vast majority of the capacity in the bank. As your charge source gets larger the tail acceptance time gets a tad longer and that is the only down side. The upside is the cells don't seem to need balancing and you keep yourself out of the danger zones.

Question: Why charge to a voltage beyond 13.8V when 13.8V yields more Ah capacity than the bank is rated for while pulling a constant 25% of capacity in load? Boat house banks rarely pull .25C and when they do it is not for very long. Perhaps a windlass, thruster, inverter etc. but these are short in duration not 4+ hours on a 400Ah bank pulling 105A continuously.......

I have come to believe that pushing your cells into either the upper or lower knee is where the cells become unbalanced.

Next Test:

I completed my .25C capacity test and the bank actually matched and has now exceeded last years numbers. This after 208 cycles most to 80% DOD. The problem is this test is grossly unrealistic for my boat.

I plan to do a 10A capacity test next. They say these cells have minimal Peukert effect but I suspect that under a 10A load I can easily exceed 450Ah of capacity. I even suspect I might hit 500Ah.

Lead Acid Musing:
A 400Ah lead acid bank would be subjected to just a 20A load to earn its 400Ah rating. I applied 102-105A load and hit 425Ah's. I then shut down the heater and kept pulling a 15A load after I'd hit -425Ah's. As I type this my lowest cell is hoovering at 2.856V and I've removed -438Ah's of capacity from the bank...

Summary:
At house loads when only charged to 13.8V and 5A net current these cells GREATLY exceed the 400Ah rating. There is no need to worry that you are short changing yourself if you choose to charge to 13.8V..

I just terminated the capacity test and removed -440Ah's of capacity. I stopped the test when my lowest cell hit 2.800V.

The resting voltage has not yet stabilized but at -440Ah's it is hoovering at about 11.71V

No load Voltages -440Ah's
Cell #1 2.965V
Cell #2 2.898V
Cell #3 2.966V
Cell #4 2.899V
Pack Voltage = 11.73V
Cell Drift = 0.067V

Note: I don't feel this is very accurate as the cells have only been resting for a few minutes. Will update again if I remember..

[deckofficer]

I'm on cycle #141. Because my bank is 1/4 the size of yours, I'm catching up to you on the cycles. I am getting the idea that your not as pessimistic for 2000 cycles as you were earlier.

[dlentz]

Originally Posted by Maine Sail

Next Test:

I completed my .25C capacity test and the bank actually matched and has now exceeded last years numbers. This after 208 cycles most to 80% DOD. The problem is this test is grossly unrealistic for my boat.

I plan to do a 10A capacity test next. They say these cells have minimal Peukert effect but I suspect that under a 10A load I can easily exceed 450Ah of capacity. I even suspect I might hit 500Ah.
[B].

How about stepping outside and trying that in your current weather? A little cold temp performance test? Just kidding, after all it's hard to sail on the hard white stuff.

Seriously though, nice info, thanks.
Dwain

[rjtrane]

I have a set of 1,000 AH Winston cells - 16 each in series for a 48v/1,000 AH bank. I'm on my second (boat is dormant half the year) - when I turned on the battery after sitting idle all summer - no charging - no discharging - no loads - it had barely lost any charge.

I have a BMS that monitors all cells and turns off loads/charging if any cell gets out of spec. I have yet to balance the cells since December 2012. Only when the battery is discharged do I notice a significant difference between highest and lowest cell - about 2.82v low and 3.20v high. When charged, all cells are typically 3.34v.

This season, I have ONLY charged with a 6kW solar array and a pair of Outback FX60 controllers.

I'll have a better handle on them once I upgrade my Steyr hybrid drives, which charge or load the battery and give them some serious loads. I typically only run a single 8kBTU A/C, 14 cu.ft. Hitachi fridge, DC/DC converters to ALL 12v and intermittently the induction cook tops and water maker. The drive motors can crank up to 140A/48v each - that should be interesting, but only .25C.

[deckofficer]

Reuben,

(2) drive motors, each pulling 140 amps from your 52 volt bank and yet only 0.25 C. That is a big bank (52 KWhr). Just the idea of a full size fridge, induction cook tops and A/C running quietly on battery power is so sweet. 6000 watts of solar must allow you to live on the hook and never run a gen set.

[diugo]

I too am doing some capacity tests on the first anniversary of the arrival of my four 260Ah Balqon Specials.

Result #1: Rudely awakened this morning by my LVD alarm (set at 2.80V) after pulling a mere 185Ah. Individual cell voltages were 2.80, 2.82, 3.13, 3.15. Obviously, the cells had gotten significantly out of balance since my last bottom balance ten months ago.

During those ten months I meticulously kept the bank between 13.0 and 13.8V---and usually between 13.1 and 13.6V---in order to maintain a SoC between 60-80%. Charge current never exceeded 23A and load current exceeded 100A only once a day for 2-4 minutes (using a microwave oven).

Since I have basically babied these cells, I am fervently hoping the 185Ah was just a vivid illustration of the importance of having balanced cells---as opposed to a capacity loss of 30% per year!

I just now manually rebalanced at 2.81V each, and am recharging them now. I will put in exactly 208Ah (80% of 260) and repeat the discharge, just to see if the bottom balance survives a single 80% cycle.

Stay tuned...