LiFePO4 and FLA: a match made in heaven?

[SV THIRD DAY]

Witch....where is the Fire and stake?!?

[john61ct]

Originally Posted by rgleason

Has your thinking changed a little and what are your current "not above" and "not below"?

Your "summary" is not my words, never heard of a knee wall, where did that come from?

Best to just ask a specific question, in context, one at a time.

This is mine
>I define my 0% SoC as 11.9V (2.975Vpc) in a low-amp discharge context.
Yes, at higher amps, going a point or two lower on the LVD is equivalent, help reduce early cutoffs due to sag.
Setpoints should be easily adjusted to account for changing circumstances.
Look at any OTS BMS LVD cutoff, and you will see they are way way lower.
So lots of leeway there, I just choose to be more conservative.
And there is no significant usable AH capacity left between those two zones.


> Also if MaineSail has previously identified

Again, quote his actual words, with enough relevant context.

Giving your "summary" with different numbers from different use cases is misrepresenting others' recommendations.

> What voltages are generally accepted as the Manufacturer's range for full capacity 100% SoC?
Check with each manufacturer directly, to me that is not relevant to actual usage patterns.

If a vendor says the batt holds 80AH, and I find that charging to my 100%

3.45Vpc (holding Absorb tapering to .03C only when precise calibration required, not as in daily use where I "charge **to** and stop)

then discharging 80AH down from there

leaves voltage way above my 0%,

say 3.15Vpc after resting

then I'm happy.

Some arbitrarily higher / lower voltages that IMO may be pushing destructively into the shoulders, and any few AH capacity gained that way is just theoretical, have no utility IRL and thus are not of interest to me.

Or rather the areas above and below that vendor-defined 80AH are only of interest as no-go zones, to be avoided in daily use.

But say I found a batch of cells that required "a BMS" to do active cell-level balancing, and that functionality only kicked in at 3.5Vpc or greater.

I may then have periodic maintenance sessions

as one does equalizing FLA

where I increase my top-end setpoint to 3.55Vpc, and hold Absorb until each pack is re-balanced to spec

This higher-than-100% SoC has nothing to do with capacity IMO, and in fact is not in the least significant wrt usable AH, just surface charge and dissipated heat.

In doing a capacity load test down from my calibrated 100% above

if discharging to 2.975Vpc only used 75AH rather than 80, I would use that to re-calibrate my BM setting for bank capacity, and log the lost capacity with an eye to learning from that in order to try to avoid the same in other bank purchases or sub-optimal usage patterns, perhaps adjusting to a shallower DoD if possible in that install's context.

[john61ct]

Originally Posted by Cpt Pat

LFP can be overcharged at 13.65 volts when the charge current is very low.

I assume you mean pushing with a held Absorb time, or even Floating for a long time at that voltage?

If just charging **to** 3.42 and then isolating fron the charge source, hard to see that being harmful.

> a SOC range regime of 20% to 80% SOC

See, to me that implies you are sacrificing 40% of rated capacity, leaving 50AH out of your nominal 125AH never used?

I can understand oversizing a bank a bit, using an on-average shallower draw in order to have a reserve pool for exigencies, and also getting maybe double the lifetime cycles as a bonus.

But the difference in practice of pushing **higher into the top** shoulder and stopping with no Absorb at 13.8V, is very low single digits of actual AH capacity.
http://www.cruisersforum.com/forums/....php?p=2731306If your "usage zero" is actually 12.7V, I'd bet,

compared to a usage zero of say 11.9V, which is ~10%

Your 75AH usage is really ranging something like 36% - 96% (arbitrary WAG of course)

[rgleason]

Thank you Cpt Pat and JohnCt,
So John, for your batteries you are using
0% SoC as 11.9V (2.975Vpc) in a low-amp discharge context
100% SoC 13.8v (3.45Vpc)

Generally, do you then find that your measured amphours in that range are close to the manufacturers rated ah?

Yes, I was reading Cpt Pat's 80% - 20% figures and got concerned that LiFePo4 use range at 60% rather than 100% was too close to the way LA work and not worthwhile doing. Cpt Pat's writeup for how he uses his batteries is what got me trying to really start to understand how it works. His writeup in the other thread was very helpful.
in

Also, since we are being specific about use here, perhaps it would help if we knew what the battery manuf, cells and size were?

Thank you both for being explicit in the face of in-explicit questions.

[john61ct]

I have stated many times, yes, with banks bought new, rated AH are right at or a bit lower than what I measure within the usable range.

Plenty of lower quality no-name and second-hand sets that is not the case, but that just means - unless returned for that reason - just adjust the BM capacity setting.

Same if there is a decline in SoH over time, but have seen none so far.

[kirklkjb]

Originally Posted by transmitterdan

Just because you didn't explode your LA battery doesn't mean it can't happen. Lithium batteries don't have any current limit so big ones can deliver more current than a LA battery is rated to take. This can destroy the LA battery and cause an explosion. It won't happen every time as it depends on how discharged the LA battery is as well as the internal resistance of both batteries.

Your theory about Lithium topping off LA is missing the point that to top off an LA you need a voltage higher than is safe for Lithium. And you cannot safely bulk charge Lithium at 14.3V. Leaving Lithium in parallel with LA will ensure the early death of the LA due to chronic under charging. Charging the LA to 100% will destroy the Lithium if they remain in parallel.

Do this instead: LI-BIM 225 Recommended by BattleBorn.
http://precisioncircuitsinc.com/wp-c...i-BIM-Rev1.pdf

Its a battery isolator. Given the description should fit the bill perfectly. It's waterproof to boot.

[T1 Terry]

Originally Posted by transmitterdan

Just because you didn't explode your LA battery doesn't mean it can't happen. Lithium batteries don't have any current limit so big ones can deliver more current than a LA battery is rated to take. This can destroy the LA battery and cause an explosion. It won't happen every time as it depends on how discharged the LA battery is as well as the internal resistance of both batteries.

Your theory about Lithium topping off LA is missing the point that to top off an LA you need a voltage higher than is safe for Lithium. And you cannot safely bulk charge Lithium at 14.3V. Leaving Lithium in parallel with LA will ensure the early death of the LA due to chronic under charging. Charging the LA to 100% will destroy the Lithium if they remain in parallel.

I know this is a very old post but ..... umm..... WTF?
A fully charged lithium battery is at 3.45v per cell, so 13.8v for a balanced 4 cell 12v nom. battery.
A fully charged lead acid battery resting voltage is 12.8v or 12.9v (2.15v per cell x 6 cells for a 12v nom. battery) To push any more current into each cell the voltage must rise to 0.11v above its resting voltage, so x 6 = 0.66v + 12.9v = 13.56v, rounded off 13.6v.

The lithim battery will not over voltage charge a lead acid battery. If the current rate passing from the lithium battery to the lead acid battery is higher than the acceptance rate of the lead acid battery cell, its voltage will rise to the point that the current reduces to the charge acceptance level.

The problem with leaving a lead acid battery permanently connected to a lithium battery is the parasitic drain of the lead acid battery continually attempting to keep itself fully charged drains the lithium battery until it is fairly deep into its discharge cycle. This simply wastes the stored capacity in the lithium battery and as the lead acid battery ages the rested terminal voltage drops. The lithium battery will hide a dead or dying lead acid battery until it finally kills the lithium battery as well.

Use the lead acid battery as the device that matches the charging curve of the normal charging regime by all means, just use use a DC to DC charger between the lead acid and lithium battery that can be turned off by the BMS if a lithium cell goes over 3.6v and holds it turned off until the cell voltage drops below 3.6v by what ever amount of hysteresis you want to use and also turns off when the charging stops so the lead acid battery can't drain the lithium battery.

T1 Terry

[T1 Terry]

Originally Posted by john61ct

Your "summary" is not my words, never heard of a knee wall, where did that come from?

Best to just ask a specific question, in context, one at a time.

This is mine
>I define my 0% SoC as 11.9V (2.975Vpc) in a low-amp discharge context.
Yes, at higher amps, going a point or two lower on the LVD is equivalent, help reduce early cutoffs due to sag.
Setpoints should be easily adjusted to account for changing circumstances.
Look at any OTS BMS LVD cutoff, and you will see they are way way lower.
So lots of leeway there, I just choose to be more conservative.
And there is no significant usable AH capacity left between those two zones.


> Also if MaineSail has previously identified

Again, quote his actual words, with enough relevant context.

Giving your "summary" with different numbers from different use cases is misrepresenting others' recommendations.

> What voltages are generally accepted as the Manufacturer's range for full capacity 100% SoC?
Check with each manufacturer directly, to me that is not relevant to actual usage patterns.

If a vendor says the batt holds 80AH, and I find that charging to my 100%

3.45Vpc (holding Absorb tapering to .03C only when precise calibration required, not as in daily use where I "charge **to** and stop)

then discharging 80AH down from there

leaves voltage way above my 0%,

say 3.15Vpc after resting

then I'm happy.

Some arbitrarily higher / lower voltages that IMO may be pushing destructively into the shoulders, and any few AH capacity gained that way is just theoretical, have no utility IRL and thus are not of interest to me.

Or rather the areas above and below that vendor-defined 80AH are only of interest as no-go zones, to be avoided in daily use.

But say I found a batch of cells that required "a BMS" to do active cell-level balancing, and that functionality only kicked in at 3.5Vpc or greater.

I may then have periodic maintenance sessions

as one does equalizing FLA

where I increase my top-end setpoint to 3.55Vpc, and hold Absorb until each pack is re-balanced to spec

This higher-than-100% SoC has nothing to do with capacity IMO, and in fact is not in the least significant wrt usable AH, just surface charge and dissipated heat.

In doing a capacity load test down from my calibrated 100% above

if discharging to 2.975Vpc only used 75AH rather than 80, I would use that to re-calibrate my BM setting for bank capacity, and log the lost capacity with an eye to learning from that in order to try to avoid the same in other bank purchases or sub-optimal usage patterns, perhaps adjusting to a shallower DoD if possible in that install's context.

Ummm..... 10 yrs plus hands on with LFP and LYP cell batteries says you still have a bit to learn.
We load test house battery systems for our customers every 3 to 4 yrs, depends when they are through this part of the country. Every time the **"memory effect" of minimum cell voltage and maximum cell voltage must be cleared to regain the full capacity. Those that baby their cells the most suffer the most loss of capacity and usually schedule to be back around our area within the 3 yrs because they are seeing regular out of balance occurring and then re-occurring within a few mths
Those that work their batteries hard but still stay within the safe (for long cycle life) voltages, 2.8v under load, 3.6v while charging, often only come in at he 4 yr mark just before they are heading off into the never-never and just want the peace of mind.
We capacity test at the manufacturers test load, 0.5CA or C2 rate, discharge from 100% capacity to 0% capacity over two hrs. System now past their 10 yr cycle life are still returning better than 100% of the manufactures advertised capacity. To me, that says our system works. Discharge using coulomb counting, an alarm at 20% SOC to let you know so you can determine if dropping high load items is required to get through to the next charging period ... usually the next time the sun comes out. A total load isolate if the battery drops to 5% SOC or a cell drops to 2.8v under load (the coulomb counting can get a bit out of whack if it's been a while since a true 100% SOC has been reached so the counter resets its true 100% SOC) We also isolate the battery if a cell voltage reaches 3.8v because that means something in the charging system has failed, usually a mosfet punched through, so the charging is no longer controlled.

No problem with letting the built in charging regime for each peice of equipment to act as the primary control ... as long as it is set at sensible voltage levels, but the cell voltage monitory as the secondary level of control is essential if you want long cycle life.

T1 Terry

** I used the term "Memory effect" because people have heard of that, there is a different cause in LFP and LYP cells that that associated with memory charging effect seen in NiMh cells, but the effect is very similar, a voltage that is not truly representative of the cells SOC occurs falsely tripping an end of charge or end of discharge.

[hzcruiser]

Over at

https://www.cruisersforum.com/forums...201795-55.html

the question popped up which relays to use for a mixed chemistry installation. The BlueSea batt switches are a good option, however, they are quite expensive:

https://www.bluesea.com/products/770..._-_12V_DC_500A

yet they'll disconnect even high Amps.

I prefer smaller latching relays in the 40-60A range and putting one in front of every battery. Then each batt can be disconnected for diagnosis, resting, or a "chemistry switch".
Arcing is not a big issue with low DC voltage (below 48V) as I've tested that more than 50,000 times with currents of 30-40A. That's the electrical lifespan of those relays and two independent tests confirmed it.

Check out Aliexpress for cheap ones, e.g.:

https://www.aliexpress.com/item/32969469685.html

I then monitor the currents going in and out of each batt separately, log that in a database for long term analysis and graphical representation.

[GrowleyMonster]

When you have to apply a workaround to something, then that something isn't right. It is not commonly accepted good engineering practice to mix batteries. Nothing wrong with having them in separate banks. I would not parallel them, though, and that includes setting the batt switch to "BOTH". Catastrophes are not guaranteed, but possible. Shortened battery life is not guaranteed, but probable. There is no need to further complicate the battery situation. I have no problem with someone doing so, on his boat, where it is none of my beeswax, but I couldn't possibly approve of recommending mixed batteries in parallel.

[HankOnthewater]

Most B2B (battery to battery, or DC to DC)) chargers have a limited capacity (ie around 30 Amps and my big alternator is rated at 250 Amps) and I do not want to have an additional LA battery parallel to my LFP.......

Hence another take on this 'match in heaven'.

What about:
- have one AGM or LA starter battery
- have some LFP batteries in a different (house) bank
- have alternator(s), connected to starter battery
- if wind generator is present, that too to starter battery
- if solar panels are present they can go to LFP bank, assuming that solar-controller can be set with appropriate voltage of ie 14.1 volt

The two banks are connected to each other with 2 things in series that I happen to have already:
- a VSR (voltage Sensitive Relay), this starts conducting as soon as voltage on starter battery is greater than 13.8 Volt. Capacity of the VSR to be around 250 Amp.
- a large capacity diode ie 250 Amps, only conducts from starter to house bank and isolates house bank to starter battery.

The (regulator of the) alternator is set around 14.8 Volt or thereabouts.

The idea is that the diode has a voltage drop of 0.7 Volt, and then provides the LFP housebank 14.1 Volt. If the LFP requires different values, then alternator to be set around 0.7 Volt above that. There are some losses (ie in heat of diode, and it will need a heatsink), not sure how big such loss is.

If the BMS kicks in with a sudden HVC (high voltage cutout), alternator or LFP bank won't be damaged; same for wind generator.

Would something like that work? I still have all AGM in my boat, and considering LFP.

[goboatingnow]

Originally Posted by HankOnthewater

Most B2B (battery to battery, or DC to DC)) chargers have a limited capacity (ie around 30 Amps and my big alternator is rated at 250 Amps) and I do not want to have an additional LA battery parallel to my LFP.......



Hence another take on this 'match in heaven'.



What about:

- have one AGM or LA starter battery

- have some LFP batteries in a different (house) bank

- have alternator(s), connected to starter battery

- if wind generator is present, that too to starter battery

- if solar panels are present they can go to LFP bank, assuming that solar-controller can be set with appropriate voltage of ie 14.1 volt



The two banks are connected to each other with 2 things in series that I happen to have already:

- a VSR (voltage Sensitive Relay), this starts conducting as soon as voltage on starter battery is greater than 13.8 Volt. Capacity of the VSR to be around 250 Amp.

- a large capacity diode ie 250 Amps, only conducts from starter to house bank and isolates house bank to starter battery.



The (regulator of the) alternator is set around 14.8 Volt or thereabouts.



The idea is that the diode has a voltage drop of 0.7 Volt, and then provides the LFP housebank 14.1 Volt. If the LFP requires different values, then alternator to be set around 0.7 Volt above that. There are some losses (ie in heat of diode, and it will need a heatsink), not sure how big such loss is.



If the BMS kicks in with a sudden HVC (high voltage cutout), alternator or LFP bank won't be damaged; same for wind generator.



Would something like that work? I still have all AGM in my boat, and considering LFP.

Just connect the alternator to the li and fit a balmar alternator protector.

[hzcruiser]

Originally Posted by HankOnthewater

Most B2B (battery to battery, or DC to DC)) chargers have a limited capacity (ie around 30 Amps and my big alternator is rated at 250 Amps) and I do not want to have an additional LA battery parallel to my LFP.......

[...]

That's a mighty big alternator! Makes me very jealous
Where did you get it from, and do you use a ribbed belt with modified pulleys? How much do you pull out of it "in real life"?

[...]
The idea is that the diode has a voltage drop of 0.7 Volt, and then provides the LFP housebank 14.1 Volt. If the LFP requires different values, then alternator to be set around 0.7 Volt above that. There are some losses (ie in heat of diode, and it will need a heatsink), not sure how big such loss is.

If the BMS kicks in with a sudden HVC (high voltage cutout), alternator or LFP bank won't be damaged; same for wind generator.

Would something like that work? I still have all AGM in my boat, and considering LFP.

Your idea is pretty good! You might want to keep in mind that the diode voltage drop is not constant but dependant on the current going through. Have a look at their datasheet or almost any diode's datasheet.

I'm not sure if you want to replace your whole AGM house bank with LFP or move some AGMs to the starter bank or, maybe, having two house banks, the AGM (always on) and an LFP bank that can be connected as required?

That's how I started, anyway. Adding a few smaller LFPs (40-100Ah each) in different spots, one close to the inverter, one close to the windlass etc. . Then there are breakers and a latching relay in front of every single LFP batt, so they can be added manually or automatically, to prevent over- and undercharging them. The LFPs then don't even have a BMS, just 4 cells in series and you're done.

You can PM me if you need more info.

[HankOnthewater]

Hi HZ, thank for the offer for contact via PM. The alternator is driven by a 2 cylinder Yanmar (AP750).
I did not realise that voltage across a diode is dependant on current going though it.
My AGM bank is really 'cactus' now, and were on their way out the last 2 years or so. My plan is replace all with LFP in one hit. Just pondering now what is the best way.