Observations of my New LiFePO4 batteries

[jssailem]

In October 2024, I received an advertisement for what looked too good to be true: LFPo4 batteries at a cost that rivaled the FLA batteries I have been using in my House Bank. With all the internet hype about these new energy storage systems, I decided to take the plunge. I purchased three 12V 100AmpHr batteries. They were shipped promptly and arrived with a 40-45% SOC. I set up a charging station and proceeded to charge the batteries until the BMS stated 100%. This took less time than expected (about 2 hours per battery). The Bluetooth app connected to the BMS and provided cell charge information, showing that the internal controls were charging and balancing the cells. This was easier than I had been led to believe. I chatted with @Hayden Watson about the process. He was helpful and supportive during this, my first experience.

This charge action was on 12 November. I recorded the BMS data.


Now, on 01 March 2025, after about three and a half months of storage at a 55ºF temperature without any charge, they have stabilized at a 38-40% SOC.


While this is much less than I would have guessed compared to my FLA batteries, due to the LiFePO4's ability to discharge to 80-100% of their capacity, they still have an adequate charge to provide 1 to 1.5 days of power for my house needs had they been resting on the boat all this time.

While preparing for a cruise, I often arrive a day or two before departure and turn on the charger. With the ever-changing Puget Sound weather conditions, I usually start a cruise under power until the middle-of-day winds develop. I can see the probability of charging the house bank from 40% to 100%, yet I have an adequate stored charge in the house bank when I get to the boat.

Your experience may vary. I wonder what others experience.

 

[dlochner]

What kind of load did you put on the batteries to cause them to discharge so deeply while in storage? Do your batteries have a storage setting which limits self-discharge? I ask, because my LFP do not discharge so deeply when in storage.

Remember, unlike LA batteries, LFP batteries should not be stored fully charged and should not have a float charged.

 

[smokey73]

What changes did you make to your charging equipment (generator, solar, charge profile, etc.) What was the cost of the additional changes needed beyond the purchase price of the battery? Was your existing battery charger able to handle the programming changes necessary.

If the LFP batteries don't like to be left at a full charge, how do you manage that when you leave the boat. If you've had to motor in for several hours before you leave the boat and are at nearly 100% what do you do? How do you get them to the state they need to be in for leaving the boat for a week or month. Just asking how that is managed?

 

[heritage]

Most of the lfp batteries are tested for up to 6000 cycles and will still retain 80% of their capacity. That's 100 years if you cycle the batteries 60 times a year on a sailboat. Leaving the batteres at 100% for a while is said to degrade them somewhat, but is the impact enough to take a big slice out of the 100 years? I think that the batteries will die of old age or obsolescence before that is a factor.
Having said that, my solar panels usually take the batteries to 100% every day, but I leave a low wattage fan running 24/7 that pull them down some overnight. Also keeps the mold from forming on humid days.

 

[Johann]

If the LFP batteries don't like to be left at a full charge, how do you manage that when you leave the boat. If you've had to motor in for several hours before you leave the boat and are at nearly 100% what do you do? How do you get them to the state they need to be in for leaving the boat for a week or month. Just asking how that is managed?

Good questions. It depends somewhat on your situation and how nice you want to be to your LFP batteries. First though, charging to 100% and then leaving your batteries alone (off the charger) doesn’t damage them, it only accelerates the natural capacity loss due to aging. This loss is heavily dependent on temperature, but as far as storing at 100% SOC versus 30%, the loss is about double initially. So if your loss over the first year at 100% was 3% it would be about 1.5% held at 30% SOC. But from what I’ve seen it’s not a linear loss over time, rather more logarithmic, so long term I don’t think you can say the lifespan is halved by storing at 100%.

For us we leave the fridge and other loads on while away from the boat, so I just set the chargers at 13.15V absorption and float and that allows the SOC to drop to around 45%. But if you don’t have loads on and are just leaving it for a week or so I don’t think it’s a big deal. What you don’t want is continuing to apply a charge voltage (>13.6V) for extended time after you have reached 100%.

 

[Hayden Watson]

If I understand what you are saying, these batteries were at home not connected to anything for the past 3.5 months. I also think that you have only charged them up during commissioning when you first got them and we found that they were very well balanced.

If this is correct, I have a couple of guesses.

• 1. With LFP you have parasitic drain even with no loads connected due to the electronics in the BMS. On a larger battery like my 560Ah and 460Ah batteries, that load is carried by a much larger battery. In your case, the total BMS load is carries by 100Ah so it will be pulled down 4-5 times faster than what I see.
• You have not really cycled the batteries at all yet and the BMS needs to learn where it is in the remaining capacity and SOC so it could be that you have more remaining power than is indicated by the App.

As you said, you will typically show up to the boat at least several hours before departure if not days. I intentionally set my batteries so that the rest at a lower voltage (13.2v) and when I am preparing for a cruise, I force it to rebulk to 14.0v to top them back up prior to departure.

 

[MFD]

I presume you have these in parallel for a single bank.
If it were me, I would go sailing for now, but when I get back and have some time...
A) Charge up each battery to 100% individually. Watching the battery cel voltages themselves and not necessarily the BMS that may say '99%' or '100%'. In your first set of pictures, one battery is substantially different than the other two? Yes, millivolts matter with LifePO4. The manufacturer will have specs, but commonly you want to get the cels to ~3.55v (safe side) or 14.2v-14.4v on a 12 volt battery.
C) They will settle to a lower voltage after charging, probably around 13.6v, and after they have all settled connect them in parallel again.
C) Give them a full cycle down to about 20%. Again keeping an eye on the individual cel voltages.
D) Top them back up.

The above is the process I have seen from a couple manufacturers for initial install and local talk with some ABYC certified marine electricians. Gives the batteries a full cycle after long term in storage - and gets the BMS in sync for what a 'full charge' means.

Also, while charging you should see the BMS try to 'balance' the cels. Basically a weaker cell will charge/discharge faster, and the BMS will shutdown charging to that cel to let the others catch up to keep them level. Some BMS do this in pairs of cels. This is called 'top balancing'. During this you can also document any cels that 'charge fast' for future reference as potentially weaker cels.

EDIT: It is also possible to go pretty far down the rabbit hole with this stuff, but being detailed during at least the initial install is probably worth it. Checking the cels once a year while doing a full charge and noting any odd cels is probably prudent as well. For myself, I avoid the full 14.6v charging as a runaway cel can get ahead of the rest. I noticed a lot of manufacturers are speccing 14.2v-14.4v charging voltage nowadays as well.

Here is an example chart - as always, get the data from your manufacturer direct. Should be nearly the same.
14.6v => 3.65v per cel, 14.4v => 3.60v per cel, 14.2v => 3.55v per cel, etc.

EDIT 2: Also looking at your pics again, the three batteries in parallel sorted themselves out while discharging so they are all pretty equal in voltage. So probably not worth it to charge each individually. You do want to do a full charge/discharge cycle, and watch the cel voltages while at the top of the charge cycle for a while. You can see in your pics as well, that BAT1 and BAT2, show 97%, but much lower voltage than BAT3. Another thing that can come into play when in parallel is if you have wildly different cable lengths or fuse sizes to each battery.

 

[MFD]

Back to the original question since I drank too much coffee this morning...

For the discharge over that time period.
For BAT1 and BAT2, although the BMS is saying 97%, the voltage is only like 3.32 in the cels or so, which would correlated to around an 80% charge level using the simplistic 'volts<=>soc' method? Assuming they had been left to rest for a couple hours.

So doing a full charge with all cels into 3.55v area should help the BMS get synced up to a proper baseline.

BAT3 is up in that area, but has pretty wide differences between the cels. It looks like it is balancing - with cel #3 being told not to charge at the time of the screenshot? Would it be the situation that the cables to BAT3 are substantially shorter than those to BAT1/BAT2, so it is getting charged faster?

Still seems like a lot of parasitic loss though, more than just the battery BMS. I know a lot newer boats consume juice even just sitting. My boat for example has three CO detectors, factory wired in so they are on even with the battery switch off. They are 37ma each, tiny, but adds up over a few months time. 0.037 * 3 * 24hrs * 90 days => ~240AH ?

 

[colemj]

Also, while charging you should see the BMS try to 'balance' the cels. Basically a weaker cell will charge/discharge faster, and the BMS will shutdown charging to that cel to let the others catch up to keep them level. Some BMS do this in pairs of cels. This is called 'top balancing'.

In balancing, the BMS does not shut down charging to a cell because that would be harmful. It either passively balances by discharging current from a higher voltage cell through a resistor to ground, or actively balances by shunting current from a high cell to a low cell. The type of balancing used is dependent on the BMS. It doesn't do anything in pairs or other combination - it just either bleeds current from any cells that are too high, or shunts current from the highest cell to the lowest.

Top balancing is putting the cells in parallel and charging them to full so they all have the same voltage at the top of their charge. A BMS doesn't do this, but it effectively does the same thing if it is allowed to fully balance the battery at the top of the charge. If the cells have a lot of imbalance, the BMS will struggle to do this, as its balancing capacity is limited (particularly if using a passive balancer).

Using voltage for determining SOC is fraught with error. LFP batteries can have different SOC at the same voltages, or the same SOC at different voltages, depending on how they are charged. For example, a 100Ah 12V battery charged at 100A can reach 14.6V before it is at 100% SOC, while that same battery charged at 3A can reach 100% SOC at 13.5-13.7V. The tables assume manufacturer's standard method and charge rates, which are generally constant current charging at 0.3C. Few boat systems are capable of a constant current charge at this rate, and pretty much impossible with solar, wind, or alternator charging.

Mark

 

[walt]

I may not have this right but it looks like the batteries went from 97 percent to 38 percent in 3 months (or so). That is a discharge of 59 percent which I believe is way higher than you would expect with LI batteries.

Some rough calculations regarding leakage current..

Battery capacity is 100 amp hour so battery lost 59 amp hour. There are approx 2196 hours in three months.

So the leakage is 59 amp hr/2196 hr = .026 amps. Ie, if we ignore actual self discharge (which should have been only a few percent), there was a parasitic load on the battery of 26 ma. Maybe the BMS itself is why this parasitic load? Seems high?

edit.. is the battery temperature different over the time frame since this is a storage case? If the battery was colder for the second reading and the battery state only considers voltage, might explain at least a little.

Not really a bad thing, just something to keep in mind when storing those for a long time.

 

[jssailem]

I presume you have these in parallel for a single bank.

Sorry if my message miscommunicated the initial setup.
I initially charged each battery, and the Nov 12 images were taken at the end of that process. There was a time difference of about 2 hours between each charge cycle. Bats 1 and 2 were resting when the images were taken, and Bat 3 had just finished its charge cycle. My goal was to have a reference point from which to start observing the batteries.

When placed in the boat, I will have the batteries in parallel and the cabling as close to the same lengths as possible.

Thank you for your observations and insight.

 

[Hayden Watson]

Sorry if my message miscommunicated the initial setup.
I initially charged each battery, and the Nov 12 images were taken at the end of that process. There was a time difference of about 2 hours between each charge cycle. Bats 1 and 2 were resting when the images were taken, and Bat 3 had just finished its charge cycle. My goal was to have a reference point from which to start observing the batteries.

When placed in the boat, I will have the batteries in parallel and the cabling as close to the same lengths as possible.

Thank you for your observations and insight.

So, can you clarify. Are all three batteries on the bench and not connected to anything?

 

[jssailem]

That is correct. The batteries have been stored in their shipping containers since I charged them in November.

Their SOC has varied 2% over the past 45-60 days

 

[MFD]

Makes sense that the screenshot is right after BAT3 was charging and hence still that higher voltage showing.

I have a pair of Epoch 460AH - the heated/essentials model, not the smart/CANBUS model.

They sat on the cabin floor at around 50F for about six weeks after they arrived. I do recall checking voltage when they arrived, wondering about self-discharge, and checking again as part of installing them. Unfortunately I have tossed my notes away since then, but I do remember there was no noticeable/significant loss.

My guess on your batteries, and at a very affordable price point, that maybe the BMS itself just has a lot of parasitic draw? I mean, sitting in the box with nothing attached, the BMS would be the only thing? The cels themselves will also lose voltage, but only a tiny tiny amount?

You may want to search/check over on some of the home solar forums. There are way more people doing this stuff for their houses than for their boats. There is an app - "Overkill Solar" - that works with nearly all BMS over bluetooth that can get you a bunch more information about the batteries. In particular - the manufacturer, model and version of the BMS itself.

 

[heritage]

I would not trust the state of charge displayed by the bms until it has cycled several times. I have three Overkill bmses on three batteries in parallel. They rarely ever agree, although they are getting closer after a year of use. A good shunt, such as Victrons Smart Shunt, is the best way to determine soc.

 

[MFD]

I would not trust the state of charge displayed by the bms until it has cycled several times. I have three Overkill bmses on three batteries in parallel. They rarely ever agree, although they are getting closer after a year of use. A good shunt, such as Victrons Smart Shunt, is the best way to determine soc.

Yes. Also it is my understanding that the shunts on the cheaper batteries (my earlier model Epoch's included) don't capture very low loads. I think on mine the cutoff is like 700ma. So tiny boat loads (like my CO detectors, 47ma each and still on even with battery switch off) will slowly drain the battery without the BMS itself noticing.


Also - in OP's charged battery screenshots. I don't think they got to a true 100%. At least doing the simplistic volts<=>SOC, they seem more around 80%. My understanding is that if charging at a very high rate - the battery volts will get up into the 14.x range, but the cels aren't really 'full'?

 

[Davidasailor26]

I would not trust the state of charge displayed by the bms until it has cycled several times. I have three Overkill bmses on three batteries in parallel. They rarely ever agree, although they are getting closer after a year of use. A good shunt, such as Victrons Smart Shunt, is the best way to determine soc.

I’d usually agree, but a long term parasitic draw caused by the BMS itself won’t show up on a shunt. The shunt just counts amps in and out of the bank. With a situation like this there are none but the bank still discharges.

 

[MFD]

BAT1/BAT2 show about 3.32v when charged and after resting, which corresponds to 80%?
Then later, they show about 3.27v, which is about 60%?
So maybe that overall loss while sitting there in the boxes isn't as much as the BMS state of charge shows?

100AH battery, sitting about 100 days, loses about 20% (20AH)...
100 days => 2400 hours
20AH / 2400hrs => about 0.008A, or 8ma.
So could just be the BMS, plus whatever tiny loss from the chemistry itself?

 

[MFD]

I’d usually agree, but a long term parasitic draw caused by the BMS itself won’t show up on a shunt. The shunt just counts amps in and out of the bank. With a situation like this there are none but the bank still discharges.

Correct.

 

[jssailem]

Thank you all for the insight.

A big part of this exercise is to familiarize myself with the batteries during the winter months before I install the system in the boat.