Choosing LFP batteries for my conversion from AGM

[SteveK]

Yesterday I left the boat with the batts at 65%. I decided to try my " float in the middle" idea a way to make sure I don't go shutdown low if I don't visit the boat in time. Which I've done twice now. Set the charger for 30 amps, absorb at 13.1 and float at 13. Enough juice to run the bilge pumps if needed, the refers and the rest of the parasitic loads. Winter is coming and my visits will be getting farther apart especially since my fix it projects are diminishing.

Monday will be my next visit and that should tell the results of my "middle settings"

I see. But why? If just to see what happens great, if there is a reason I am not seeing I would like to know.
Seems like your plan is to float at 13v but stay at 65%~ SOC.

 

[Delta Riverat]

Just a test to see what the SOC at that level is. Sort of "look-see" about long term storage. Like if we were gone for a month or so.

 

[HeadedToTexas]

If the pumps never come on and there are no other loads during storage, won’t even 13 volts slowly bring the SOC up from 65% to something close to 100%?

 

[Delta Riverat]

I started out with the charger set at 13.2 and 13. The charger was inputting 2 to 5 amps. When I reduced to 13.1 and 13 the batteries were discharging at 5 amps. So I don't think they will get anywhere close to 100%.

I don't have to worry about freezing here, just experimenting with "long term storage" with a mid range charge.

 

[SteveK]

If the pumps never come on and there are no other loads during storage, won’t even 13 volts slowly bring the SOC up from 65% to something close to 100%?

You are the beta tester for this. I suppose it depends on the charger and whether just voltage is maintained or if amps are added.
I have not used a setting for SOC level as it needs an addon device which then would maintain at a chosen SOC.

 

[DDW]

I don't think I device that tried to maintain a certain SOC would be reliable. There is no way to determine the SOC of a battery except perhaps by measuring true at rest voltage. SOC meters that count AH in and out will drift out of calibration over time, rather rapidly if not reset frequently (most often done by achieving 100% SOC). Short term (say a week) this would work, but long term (months) it would drift off of reality quickly. So float voltage is probably the best way and perhaps the only way.

 

[Delta Riverat]

Of course I would 100% charge overnight after I got back to rebalance and reset the SOC. Just interested in how this idea works for now. It will be 3 days with this setting so should be accurate.

 

[SteveK]

I don't think I device that tried to maintain a certain SOC would be reliable. There is no way to determine the SOC of a battery except perhaps by measuring true at rest voltage. SOC meters that count AH in and out will drift out of calibration over time, rather rapidly if not reset frequently (most often done by achieving 100% SOC). Short term (say a week) this would work, but long term (months) it would drift off of reality quickly. So float voltage is probably the best way and perhaps the only way.

does the BMS not know the SOC?
With lead resting voltage was a close guess that was accepted.
With LFP voltage is not reliable as SOC can be 20% with a 12.8v.
However if a BMS can manage charging/discharging then it must know the SOC.

 

[Helmsman]

SIAP Don't murder your batteries, tips for winter storage of LFP batteries

This was posted on TF somewhere else, then made its way to MTOA. It is worth the read for winter storage folks who have lithium.

 

[DDW]

does the BMS not know the SOC?
With lead resting voltage was a close guess that was accepted.
With LFP voltage is not reliable as SOC can be 20% with a 12.8v.
However if a BMS can manage charging/discharging then it must know the SOC.

That seems to depend on the BMS. I have one LFP that is wildly off on reported SOC after the battery has been sitting for a couple of months. It does not display individual cell voltages. I think it probably counts amps in and out, and does not account for self consumption of the BMS or chemical self discharge.

I have another one that reports SOC seemly based on cell voltage, in fact (this is the ubiquitous JBD) there are tables set up (that you can change if you wish using something like Overkill) indicating cell voltage and associated SOC.

With LA batteries, there was a fairly steep slope of resting voltage vs SOC. There is a slope with LFP but it is very shallow, with just a few tenths the difference between full and empty, so hundredths of even thousandths of a volt count. And that is easily swamped out by even slight charge or discharge. So trying to determine SOC for a battery still in a circuit is pretty difficult. However from experience, if I set the charge float to something like 13.25 or 13.2, it keeps the cells at the voltage representing around 80 or 85% SOC, while supplying the ongoing parasitic loads both inside and outside the battery.

I've read several papers studying the longevity effect of storage at various voltages, most show not really much difference if you are somewhere in the 40 - 80% range, and any significant roll off in life occurring only at <20% or > 90 or 95%. So not super critical to hit an exact number, just don't "bang the corners" of the envelope in storage. The problem of erring on the low side (say 30%) is that if power goes off or disconnects or something, you have little reserve before you are near the low end or even into low voltage cutoff, which is not good.

 

[SteveK]

if I set the charge float to something like 13.25 or 13.2, it keeps the cells at the voltage representing around 80 or 85% SOC,

But then how do you know this if anything but fully charged or dead is for certain?
BTW, my batteries are on duty 24/7, never stored, so kept at 100% ready.

 

[DDW]

Represented by the voltage on each cell, as measured by the BMS, and as listed in the table in the BMS set up by the manufacturer (which matches the literature on LFP cells). LFP (and lead) batteries do have a pretty set relationship of open circuit *resting* voltage and SOC. The problem with using it is open circuit and resting is usually defined something like disconnect from everything, leave alone for 8 hours minimum, then check voltage and correct for temperature.

SOC is a pretty fuzzy term in batteries - very hard to deterimine with any accuracy, changeable with weather, age, etc. Honestly not a very useful term, and not something I look at much. The 80% quoted above is at best an approximation - but good enough for storage purposes.

 

[Delta Riverat]

Yeah, that's what I think. I don't want my "storage" batts (except the lead ones) at 100%. My batts are stupid, no blue, no cold protection, just basic batts. My SOC is a real external shunt right on the pack with an remote monitor in the saloon.

Somewhere in "the middle" would be great. If at 50% when I hit the boat I can get them to 90% at the dock in 2 hours easy and since my alternators now work one hour would be good enough. Not talking about an extended time to up the batts before takeoff. I don't want them to hit zero again, I don't think I missed the off time by more than a day.

I want the boat to maintain systems needed while I'm gone and not need to dump the water from the ice cube bin and trash all the frozen food -

 

[HeadedToTexas]

So you guys keep float at 13.4 volts during “normal” use, but drop the float setting to 13.2 or so volts for storage so that the SOC stays well under 100%?

 

[SteveK]

So you guys keep float at 13.4 volts during “normal” use, but drop the float setting to 13.2 or so volts for storage so that the SOC stays well under 100%?

With my charger if set to LFP setting, the volts are 13.5v~ but managed by the BMS. The LFP setting is in line with Battery manufacturer. Of course I could choose custom and choose float volts.

 

[HeadedToTexas]

So I discovered the “problem” with my Balmar 614. It only works and charges if the sense wire is connected. You can’t leave that one off. Rookie oversight.

So I’m up here to land the red sense wire. In the old configuration, the sense wire landed on the positive terminal of one of the AGMs in the house bank. My alternator is wired to the main positive bus.

Balmar is closed today. The manual offers three landing points: the positive output of the alternator, the common side of a battery switch, or the positive post of the battery being charged.

The house bank is not isolated. The positive output of the alternator, the common side of the battery switch, and the positive post of the battery being charged are all connected through the positive bus bar. Any reason why landing the sense wire on the positive bus would not work just as well as the battery, switch, or alternator?

 

[SteveK]

You could also go to the run terminal of the start switch.

 

[HeadedToTexas]

So basically anywhere that has continuity with the batteries being charged? My concern was loads on the bus bar, but that is the case anywhere on that circuit.

 

[SteveK]

So basically anywhere that has continuity with the batteries being charged? My concern was loads on the bus bar, but that is the case anywhere on that circuit.

Actually, rethinking this, a sense wire maybe should not be hot unless the engine is running IMO, otherwise why is it not attached to the positive post of the alternator which is the same as positive battery post.
My internal regulated ALT does not have a sense wire outside the ALT for instance.
Leaving it hot, would that not be a drain on battery when engine is stopped?

 

[DDW]

So you guys keep float at 13.4 volts during “normal” use, but drop the float setting to 13.2 or so volts for storage so that the SOC stays well under 100%?

Correct.

 

[HeadedToTexas]

My understanding is that the sense wire is “just” the external regulator’s voltmeter. It senses voltage so that the regulator knows where it is in its charge profile. Like any voltmeter, there is a tiny draw when it is on, but in the case of the Balmar 614 it is only on with the ignition switch. Should be no drain when the engine and regulator are off.

 

[rslifkin]

otherwise why is it not attached to the positive post of the alternator which is the same as positive battery post

Ideally you want it to read as close to true battery voltage as possible so minor voltage drop in the alternator wiring is compensated for.

 

[HeadedToTexas]

4 feet of 2/0 cable from the positive battery terminal feels plenty close enough.

 

[DDW]

Land the sense wire as close to the battery terminal as possible, consistent with the requirement that it never be disconnected from the battery (for example, though an isolation switch). The reason for the separate sense wire is to account for voltage drops between the alternator and battery. This can be up to a volt or so, in lightly wired systems. That will make a huge difference in AGM charge time, and considerable difference in LFP charging for various reasons.

You do not want it to ever become disconnected from the battery that the alternator is charging, because the regulator will then run 'open loop' which may result in what you saw, or worse, runaway high voltage. On the other hand, it should have a fuse close to its connection point as though it draws no current, a fault on the wire somewhere could.

The way the Balmar works is if it does not see a reasonable voltage on the sense line it shuts down the alternator. I think also if the power line and sense line differ by a lot it shuts down. However if the sense line is reasonable, but not representative of the alternator output post (for example, it is connected to the start battery and the alternator connected to the house) then you can get no regulation and very high voltage.

 

[Delta Riverat]

Today was pretty cool. The batts were at 54% and discharging slowly. When I turned on the lights there was more draw so I upped the charger to 13.2 absorb and 13.1 float. It was cool watching the charger pick up the DC loads and the battery either charging a little or discharging a little. I think this setting will work great. The only thing that could cause another ice melt incident would be a prolonged power outage and those are not common here. Less than 12 hours mostly and not often.

 

[Delta Riverat]

Back to the boat today after 5 days at the new (13.2 absorb, 13.1 float) and found this;

This will be my new "set it and forget it" during the winter time so I truly can leave the boat w/o worry. Just maintains a 2/3 charge on the batts.

 

[DDW]

Do you have a way to check cell balance on those? That would be something that might drift off under these conditions. I prefer "set it and monitor it"

 

[Delta Riverat]

Nope, I have "first generation stupid" batteries.

The plan will be to set the charger back to 14 and 13.5 at least once a month and leave it that way for a day or 2 and then go back to storage mode.