This is an old revision of the document!
Words of Wisdom:
“LFP batteries do not need to get back to 100% SOC, ever” – Maine Sail1)
“Others will likely say you should read a few of the previous pages in this thread and a lot of your questions will be answered. I'd suggest you pick and random number between 1 and 420 and start on that page and just read 10 pages. I promise you will learn about something relating to your inquiry.” – SailRedemption 2)
The Cruisers and Sailing forum has a 300+ page LiFePO4 Batteries: Discussion Thread for Those Using Them as House Banks thread.
There is gold in the thread but it can be hard to find. This page attempts to distill the thread down to currently known best practices with LiFePO4 batteries, providing links back to the thread for context-checking.
LiFePO4
Con
Maine Sail has discussed the most common errors made by RVers vis-a-vis lithium.
LFP batteries do not require multi-stage charging like lead-acid batteries; they are charged until full and then charging is stopped. LFP does not have an absorption phase like LA7) and does not need floating in the LA sense8).
That said, smart chargers can be used to charge LFP. Maine Sail describes the thought process of utilizing LA smart chargers with LiFePO4:
With a Li bank we are in bulk 99% of the time, on most boats with inadequate charging capacity. Once we get to “limiting voltage” how long do we stay there? Is this driven by time or current? After the limiting/absorption/CV stage voltage where do we go and how do we again re-trigger “charging”…“ – Maine Sail9)
Lead-acid stages are adapted to LiFePO4 charging like this:
There are important differences between how RVers/boaters and Electric Vehicles (EVs) charge and discharge LFP battery banks. EV batteries are used aggressively: charged at high current to full and drained at high current to empty. This can cause individual cells to get out of sync, voltage-wise, with others and warrant a BMS.
In Fractional C scenarios BMS failures may cause more damage than they prevent.12)
Boat and RV house batteries are generally charged and discharged with less amperage than the bank's capacity, hence Fractional C use.
This can make LFP banks less complex and less expensive in an RV application. Assessing SOC by voltage is also more accurate with Fractional C charging/discharging than with high currents.
Although his setup does not use it, T1 Terry describes a “middle 80%” approach that works well with Fractional C systems. It can eliminate the need for BMS balancing:
This barefoot (no BMS) and Frac-C approach will be called Terrycharging below.
LiFePO4 are fully charged at 13.6v (3.4v/cell).20) Voltage rises very quickly after LiFePO4 are fully charged (faster than the internal temps rise) so voltage can be used to assess full charge.21)
There is anecdotal evidence that sustained Vfloat at 13.8v (3.45v/cell) for months may not affect capacity.22)
LiFePO4 cells have a markedly lower full charge voltage than LiPo which is 3.7v.23) The LFP gives up energy density for stability and safety.24)
Rated cycles will be easier to attain by going no lower than 80% DoD, 12.4v (3.10v/cell).25) Shallower cycling may increase both cycles and lifetime capacity.26),27)
Under lab conditions LFP banks have been cycled 2000x to 100% DoD resulting in ~5% loss in capacity. MaineSail points out that regular battery temps above 80F would affect these results significantly.28)
BANK | CELL | NOTES |
---|---|---|
16.80 | 4.20 | cell damage43) |
14.40 | 3.60 | High Voltage Cutoff44) |
13.80 | 3.45 | Terrycharging Vabs, with gentle overcharge for balancing. Max safe non-BMS voltage.45) |
13.60 | 3.40 | fully charged |
13.40 | 3.35 | Terrycharging Vfloat |
13.25 | 3.31 | 20% Depth of Discharge |
12.80 | 3.20 | 50% Depth of Discharge ![]() |
12.40 | 3.10 | 80% Depth of Discharge. Best for storage?47) |
11.80 | 2.95 | Terrycharging Low Voltage Disconnect. Min safe non-BMS voltage. 48) |
11.20 | 2.80 | 100% Depth of Discharge49). Low Voltage Cutoff.50) |
”…lithium batteries can be very hard on alternators and DC to DC chargers that aren't designed for continuous 100% output as these battery will accept all they can get till they are full, no charge tapering caused by terminal voltage rise like lead acid batteries… overheating of charging equipment can be a problem.“ – T1 Terry51)
Increased alternator load may also cause premature destruction of belts52) or throwing a belt off the pulley.53)
Banding (physical binding of cells to prevent expansion) may not be necessary in RVs:
“Normal use for a lead acid would be 0.1C charge, <1.0 C discharge, and the Winston cells would not need banding for the same parameters. If you plan to flex the muscles (pun intended) and charge at 1.0C and discharge at 3.0C, then the experts recommend banding.” – deckofficer54)
“Due to conservative a-hr ratings, nil Peukert, and being able to cycle to 80% DOD, you will discover that what ever a-hr rating worked fine for you in a lead bank, 1/2 that rating will give the same usable a-hr in LiFePO4. So if you were going to upgrade your lead bank to 400 a-hr, 200 a-hr LiFePO4 would be the same upgrade over your current 320 a-hr bank” – deckofficer55)
Unless the alternator's output voltage is dropped with an external regulator the charging voltage will probably be high enough to damage your LFP cells.
Diode-based battery isolators