LiFePO4 battery capacity decreasing

[ShoreFun]

So a thought:

What is the error in the electronics???

You said this: "Charged to 13.8, stopped charging. Discharged to 11.9 V, measured 396 Ah. Not great but very close to the last capacity test which measured something around 400 Ah IIRC."

That is 10% off, depending on the electronics and temperature of the measurement that could be considered the same. You need to look into how accurate the device measures and how does temp affect the measurements. The 10% accuracy is a common error value found in consumer electronics. Inside the box you have capacitors, resistors and such whose value changes over time and temp. Some can change a lot and some are fairly tight.

I read all these things about batteries and see all these 'accurate' numbers and wonder just how accurate they really are in actual measurement. Plus you have to throw in accurate over time and temp.

I must point out I do not have any experience with boat electronics. Just I work with aircraft electronics doing research so I always have to know what my errors are in measurement. I also do some minor fixing of whatever comes along and see the 'quality' of various devices.
So take my comments for what they are.

[Q Xopa]

Originally Posted by john61ct

What is your referent for "it"? Maybe you don't care about that factor, I personally consider it important.

> The cell is full when it cannot absorb and store more charge, and this is at 3.65V.

Flat out wrong, in many ways, energy storage does not act like a tank holding a fixed volume of liquid.

You can get to exactly the same SoC point from a myriad of termination profiles.

The only way to tell how much of your continued attempted input the bank is actually **storing** as energy you can get back **out** is through precise load testing.

Continued input past that point is mostly dissipated as heat, and harms the battery (as in, lost lifetime cycles). As you try to get that last few Ah in, only a small fraction of the flowing current is actually getting stored.

There is **no** reason in daily cycling to try to get all the way to **any** definition of Full, only downside risk.

Yes for SoH capacity / load testing, you want to define a precisely standardized "Full" stop-charge point, and

that can safely be defined a bit higher than the one you'd want to use for daily use since such tests are only occasionally performed.

But something like "3.5Vpc held until current tapers to .03C (3A per 100AH)" is certainly a precise enough definition to use for that purpose.

Of course if you want to use "3.65Vpc and hold that Absorb until .005C" as **your** definition of "SoH testing full" that's fine, but

I'd bet the **actual usable** Ah difference between the two is well under 1% SoC, and

if you're using something like that as your daily cycling stop-charge point, I'd bet you're sacrificing hundreds if not thousands of lifetime cycles, maybe only getting the vendor rated lifetime, which too me would be a shame, given your investment of many thousands of dollars.

Not that I disagree with you suggesting not to try and total capacity to be conservative and increase cycles life.

But it does sound contry to you have a reserve battery is a wasted capacity 'dead lead'.

So pardon me for being confused where you stand with keeping some capacity in reserve for safety, cyclelife etc, or not?

[john61ct]

Originally Posted by ShoreFun

What is the error in the electronics???

You said this: "Charged to 13.8, stopped charging. Discharged to 11.9 V, measured 396 Ah. Not great but very close to the last capacity test which measured something around 400 Ah IIRC."

That is 10% of.

396Ah is 99% of 400Ah.

IOW 4Ah is well within measurement tolerance outside of lab conditions.

Even if true, over that time period, few would be upset, most lose capacity at a much faster rate.

[john61ct]

Originally Posted by Q Xopa

But it does sound contry to you have a reserve battery is a wasted capacity 'dead lead'.

Redundancy is IMO critically necessary, not an option.

When reliably cranking the engine from multiple sources becomes routine without dedicated Starter batts, the savings helps facilitate the Reserve idea.

Portable jumpstarter powerpack is the belt / icing on the suspenders / cake.

[Delfin]

Originally Posted by Maine Sail

If you stop at 3.45VPC, once you hit 3.45 VPC, the cells will be below 100%. If you holder 3.45VPC, and allow current to taper, the cells will reach 100% SOC even at just 3.45VPC or 13.8V for a 12V pack...

There may not be a definitive answer to this question, but is there a practical difference in terms of longevity/performance in charging at 3.45v with a taper vs. 3.6 with no taper, that is, disconnection from the charging sources when current acceptance drops below 3%?

[senormechanico]

Originally Posted by Delfin

There may not be a definitive answer to this question, but is there a practical difference in terms of longevity/performance in charging at 3.45v with a taper vs. 3.6 with no taper, that is, disconnection from the charging sources when current acceptance drops below 3%?

I have used "float" at 13.2-13.4 for several years with no ill effects.
13.2 when boat is not being used and 13.4 when aboard for extended time.

[john61ct]

Originally Posted by Delfin

There may not be a definitive answer to this question, but is there a practical difference in terms of longevity/performance in charging at 3.45v with a taper vs. 3.6 with no taper, that is, disconnection from the charging sources when current acceptance drops below 3%?

There are many profile variations that will all arrive at the same SoC point.

SoC as defined by precise load testing, useful Ah extracted.

My current understanding is that,

**within the context of avoiding the maker's maximum spec'd voltage**

it won't matter which path you use to get to your desired SoC point.

Charging with CC only, **to** a voltage and stop, will get to wildly different SoC depending on the charge current rate.

Lower the current, higher the SoC.

While using some CV, holding Absorb V until an endAmps taper is reached, will give a reliable SoC point at different current rates.

It is the pushing to the highest possible SoC point that makes the difference, not the profile used to get there. Internal chemistry reactions generating heat as the resistance is climbing, indicated by the voltage-rise curve getting steeper.

Past that point, the ratio of energy actually getting usefully stored, vs dissipating as heat, is getting lower and lower.

Actual SoC as defined above, is already within a couple percentage points of the theoretical maximum you could only get to only by continuing to hold Absorb CV or by increasing to a harmful higher V level.

Yes the long-term longevity consequences of pushing too high are not yet precisely quantified.

But since there is no significant benefit to doing so, I figure better safe than sorry.

So my more general guidelines:

Stop at 3.45Vpc / 13.8V for 4S, for amps rate of .2 - .3C.

At higher rates, to shorten ICE run-times, it is safe to go to 3.50Vpc / 14.0V.

At **very** low charge rates, as with many solar setups, back off to 3.40Vpc / 13.6V

[hellosailor]

"I have used "float" at 13.2-13.4 for several years with no ill effects."
Well, here's the thing. No *visible* ill effects, but a lithium bank should be good for what? 2000 cycles? 5000 cycles? 8000 cycles? Until you have put the first five or six thousand cycles on your pack, you can't really say if you've created any ill effects from float charging, can you?

Just recently, Tesla, who have to worry about warranty replacement costs on maybe a million expensive battery packs, started advising owners to only charge to something like 85-90% capacity on the everyday basis. This despite all the "range anxiety" issues, because Tesla found they were spending a lot of money replacing battery packs when owners were putting in heavy use--and kept charging them to 100%. Surprisingly, Tesla hadn't heard from any battery engineers in the computer industry, where laptops have had similar issues and have long offered a choice between "full charge" and "best longevity" with the latter typically charging to 96% and allowing the battery to fall below 90% before they kick in and charge again.

I also keep seeing references that say lithium should be a binary charging system: On or off, no bulk/absorb/float nonsense. (Which is obsoleted even for wet lead, compared to smart charging, for years now anyway.)

[nebster]

Originally Posted by hellosailor

Well, here's the thing. No *visible* ill effects, but a lithium bank should be good for what? 2000 cycles? 5000 cycles? 8000 cycles? Until you have put the first five or six thousand cycles on your pack, you can't really say if you've created any ill effects from float charging, can you?

You're right, and even if he had 6000 cycles, it wouldn't be a controlled study, so the conclusions would still be mildly suspect.

But conducting such a test is very expensive (in time alone, not to mention money), so we don't seem to have any documentation of anyone doing it in the lab.

What we do have is some physics intuition, and I suspect that's what most of us who are floating are applying -- perhaps mistakenly! But, nevertheless, the advantages of a sustaining float voltage are huge in practical terms for many of us living off these packs. So even an unnoticed consequence for cycle life is probably worth it, as long as it's small enough that we still get years and years of real usage.

Just recently, Tesla, who have to worry about warranty replacement costs on maybe a million expensive battery packs, started advising owners to only charge to something like 85-90% capacity on the everyday basis. This despite all the "range anxiety" issues, because Tesla found they were spending a lot of money replacing battery packs when owners were putting in heavy use--and kept charging them to 100%. Surprisingly, Tesla hadn't heard from any battery engineers in the computer industry, where laptops have had similar issues and have long offered a choice between "full charge" and "best longevity" with the latter typically charging to 96% and allowing the battery to fall below 90% before they kick in and charge again.

I owned an original Tesla, and even my relatively ancient pack was managed in a way that kept the SOC well south of 100%. Modern Teslas surely still have default charge and "range mode" optional charge for the occasional push to closer to 100%, don't they?

Do you have any links to this news? Because I find it surprising, but then I don't keep up with the details these days, and I'd be interested to learn about this new phenomenon.

As an aside, the group-sourced trend data on Tesla pack behavior substantially disagrees with your statement: it looks like most of these batteries are going to outperform their planned (predicted) lifecycle.

[senormechanico]

Hellosailer,
The solar pumps the bank up to 13.8 in the morning, then goes to float.

I have a 24/7 load on the bank of somewhere around an amp.
When it's on float at 13.2, the bank is generally being slightly discharged, even during the day and it rarely goes to bulk unless I've hit it with a big load such as inverter making hot water.
Sometimes it will go back to bulk as the sun is going down, so life is good.
If they die, I'll get new ones. Lead is SO DEAD.

[CatNewBee]

Same for me.

Batteries are consumables and there to be used. If you need something to pamper all the time, ger a pet or a tamagochi.

My charger is set to absorbtion 14.2V for 5 minutes and float at 13.5V to usethe energy directly for the fridges and other gadgets, hot water, water making etc. when the battery is full..

[senormechanico]

A couple other things to pamper:


Bonsai

and

https://www.youtube.com/watch?v=fCiTAJi1yRk

[hellosailor]

No, I don't recall where I heard the Tesla advice on charging routines, except from a video posted by a small company using their cars commercially, racking up the miles. They were running to 100% range and doing a lot of miles, and their packs were failing prematurely. Tesla told them to cut that back and not use 100% all the time, which was news to the company. Don't know if they failed to read the owner's manual or if it wasn't really put out there when they got their cars.

The average customer, doing much lighter charge cycling on an everyday basis, wouldn't see the same issues.