LiFePO4 batteries as lead acid replacements investigation (chime in!)

I believe some of you have already chipped in this stuff, automotive applications maybe?

Since I am The Battery Whisperer I am always looking for ways to increase durability of my pack builds, namely reducing maximum charge voltage.

For automotive or UPS applications I know lead acid is cheap and common. Lead acid batteries are usually handled with trickle charging and “high” maximum charge voltages since letting them go down usually results in some form of degradation. LiFePO4 does not have this problem and its full capacity and energy can be used without worry (down to ≈2V/cell, I know).

However, how well does LiFePO4 handles overcharging or high voltage in practice? I've been tempted to build some vehicle battery packs or something smaller for an UPS, but I'd like to first make sure the battery can last for ages. In a vehicle the battery price is going to be quite high, I need to guarantee long service life.

I do not really know how high do UPS or vehicle regulators pump the voltage up but, in my opinion, if I can stick below 14V maximum the better. For this I've even though in using series diodes to drop input voltage, with low drop diodes or an ideal MOSFET diode setup for current in the opposite direction.

Thanks for listening. :-)


Yeah man, I am very surprised considering Maglite is using for Magtac rechargeable light.

This quote makes no sense to me here.

Is Maglite using LiFePO4 chemistry?


Well, he probably meant that Maglite uses LiFePO4 cells for their battery packs.

Wellp, just saw it. That ML150LR/ML150LRX with LiFePO4 battery likely built with 2x 26650 cells in series.

My above question, however, is aimed to determine how fine can be a 4S LiFePO4 battery in an environment geared for typical 6S lead acid batteries. The same applies for other 2xS LiFePO4 to 3xS lead acid conversion scenarios.


Sat, 02/02/2019 - 02:33

The nominal resting voltage of a LiFePO4 cell is 3.0 volts with a max charge of 3.6 volts. That lines up very well with lead acid, since 2x LiFePO4 would be 6.0v nominal and 7.2V full charge. 4x would be 12v and 14.4v respectively. Beyond that, I have read that LiFePO4 cells even handle overcharge fairly well, with some people claiming that they regularly use a normal Li-Ion charger with 4.2V terminal voltage to charge their LiFePO4 cells. This ability to take abuse is also possibly the reason why LiFePO4 cells are used in solar rechargeable garden and walkway lighting now, replacing NiMH.

EDIT: I vaguely remember reading a while back that some guy had built a pack for his electric bike and used “overcharge” as the way to balance his pack, since he had not incorporated a BMS or any type of charge leveling in the pack.

LiFePo4 in 4S is often used to replace 6S lead acid batteries without any type of converter or regulator in between. Although is is recommended to have a BMS.

Most UPS have some sort of LVP preventing the SLA battery to go below 50% its capacity. Not sure how that would work with LiFePo4 since the whole point is to get the full discharge of the cell for added capacity.

There are solar controllers units with AC inverter, with integrated AC passthrough with just dozens milliseconds inverter turn on time. This means practically no VA limitation as compared to regular UPS which relies on first converting AC-DC to the battery then DC-AC back, very inefficient as they’re always on and producing heat. These units are very expensive, but they have charging algorithms designed for Li-ion and LiFePo.

We use GBS brand prismatic LFP cells for our off-grid power system. These are the rectangular shaped cells, 365mm x 73mm x 312mm for the 400 Ah size. 3.2 volts nominal. 13.3 kg. weight per cell. We have 16 of those in series for a nominal 48 volt 19.2 kWh battery bank. We have a BMS system in use. It is programmed to give us about 80% of the rated capacity, which is a little more than 15 kWh.

We use a per cell minimum cut off voltage of 2.5 volts and a maximum of 3.6 volts per cell while charging. I was warned that 4.0 volts per cell was the absolute limit, but that 3.8 volts per cell was better, okay, if I wanted to enjoy a long battery life. So I tried 3.6 and found I have plenty of capacity and should have a very long lifetime.

I also have a small 12 volt LiFePO4 used on my motorcycle, selected to save weight. 860 grams instead of 5 kg and it has more cranking amps.

LiFePO4’s biggest disadvantage is they cannot be charged below freezing temperature. They can be discharged down to –20 C. Charging below 0 will result in cell damage apparently.

LFP cranking batteries have cold weather quirks. When it gets down to about 7 C or less, the bike cranks better if I turn on the headlight for 1 to 2 minutes or so, then turn the light off and hit the starter. The headlight draw warms the battery and the warmed battery releases more AH. If the headlight was halogen 20 to 30 seconds would be sufficient but it is an LED so does not have enough amp draw to warm the battery as quick. It does seems really weird to turn on the light when it is cold; that seems counterintuitive, but the bike really does crank faster and starts quicker. The bike electric charge system seems to top out at about 14.5 volts, still quite safe.

I have thought of trying an LFP as the cranking battery in the tractor, but it still has a couple years of life in the lead acid battery, plus that would be quite costly. Not sure I can actually realize any advantages with an LFP.

Wait, since when can LiFePO4 cells be charged at higher than 3,6V?

4,0V seems very dangerous.

Apparently, LiFePO4 cells can take some amount of overcharge and survive.

Yes, but dangerous to cycle life.

I mean, I’ve actually overcharged an old laptop cell to 5V once.

The CID popped though, and nothing was harmed.

We have a lottery sales business using an always on online computer terminal with closed software, this means every time there's a few minutes blackout the UPS battery dips rock bottom, receiving tremendous damage. A Riello iDialog 600 plus is somewhere lying around, accompanied with some other cheapo unit. They do not last more than a few years. Despite cheap appliances, I believe this is a waste of money which could be fixed with a custom LiFePO4 battery replacement. For example, 4S2P cells of these Heter/Enerpower 18650 3.2V 1100mAh.

However, while I see this clearly and could do the battery swap, my sister (in charge of the business right now) has not endorsed such conversion yet. Maybe I could convert an old unit and show her the results while invoicing the corresponding :-D bill.

No charging below 0°C MtnDon? I wonder how much of an actual problem that can be. In a warm climate like the one here quite unlikely, and even when cold the constant in/out current sipping to/from the battery when operating the vehicle should keep the thing above 0°C, don't you think?

Edited: just read the latest replies. Remember my aim is to ensure maximum cycle life. Overcharging is a no go. Even 3.6V/cell may be too much LoL. A friend of me bought an inexpensive LiFePO4 battery for an scooter, and it didn't last long. It has to survive operational at least till the next ;-) millenium.

Cheers ^:)

Sat, 02/02/2019 - 05:15

LFP cells are somewhat common in RVs, for what it’s worth.

Wellp, I am finally going to test this stuff in an UPS. A friend of me requested me to restore one of his De9502 drill batteries, thus I went to HobbyKing for their sub-C 5000mAh cells and battery packs. With single cells back ordered, after two 6 cell sub-C stick packs in the basket the budget already was more than $40 and the shipping quote from the UK warehouse was €15,76 or ≈$17,69 (yikes!), thus I decided to add some stuff else to reach the magic free shipping $50 figure. I completed with two bar clamp tools and 2x ZIPPY Flightmax 1800mAh 2S1P 30C LiFePO4 packs.

The following video sort of convinced me:

Have fun. :-D