Lithium batteries - true longitudinal capacity profiles?

[avanti]

How are you defining "full"? At max charge for lithium at 14.7v rested for a 12v system, or with some % headspace on the charge?

Well, that's the thing. I honestly don't know. When I get around to writing that rule, I will research the issue (or ask you guys )

 

[booster]

I think this Masters thesis paper may be of interest as it sort of addresses the original question directly. Be aware it is from 2017 so doesn't include any research since then. It is obvious from the paper that high temps and storage SOC were already points of interest in cell life. What appears to be missing is testing on charge voltage/Crate and hold times at that voltage, which didn't appear to even be mentioned the I saw during a quick look at the very long paper. The charging may have been considered a non issue on life then, but I think that may have changed since and would be nice to see. It appears to be a quite well written paper to me. It appears that he did get his Masters.

per Linkedin
Technology Solutions Engineer at Ballard Power Systems

https://uwspace.uwaterloo.ca/bitstream/handle/10012/12177/Catton_John.pdf

 

[Steve Guo]

Slow charging doesn't seem to be an issue from what I have read, and cycling in the middle of SOC is said to be a good thing. We have the same panels and had no issues with ours either. What voltages do you have set in the solar? Do you have a battery monitor on the system? Some solar controllers from that era put a full charge of like 4 hours on the batteries every day as they run on timers and don't check starting voltage to see if a charge is needed, and that can be a real issue when the van is sitting, stored, with full batteries and in the sun.



The going to 20% SOC will reduce life by a lot according to a growing number of sources I found, but you should still get something like 2000 cycles out of them. Are you also leaving some space at the top of the charge by stopping changing at 13.6-13.8v?


Of growing concern to me, at least with the drop in batteries, is the balance of the cells going off over time if you don't have a balancer, preferably an active one. Do you have a balancer on them?

Yes, slow and gentle charging is beneficial for battery life. The charging rate of 0.25C~0.5C can balance the lifespan and speed.


Charging before the battery level drops below 20% and end charging after reaching 80% will significantly improve the lifespan of lithium batteries. Charging the battery to 100% is not a problem, but keep it long time of 100% can accelerate the performance degradation of lithium batteries.

 

[RossWilliams]

Trojan defines a cycle from 0% DOD or fully charged, down to xx% DOD.
1000 cycles from 0% DOD to 50% DOD.
600 cycles from 0% DOD to 80% DOD.
So the math in my original post applies.
This is from Trojan’s AGM application guide.
Lots of other variables here like how long the battereis remain discharged, etc.

That math is based on discharging from a fully charged battery. The math changes if you compare discharging from a 70% level to 80% and from 40% to 50%. That is closer to how most of us use our batteries. We discharge as much power as we need when we need it and then charge whenever we can.

I suspect if you size your system so you are discharging a lot when your battery is near 80% you are going to substantially reduce both the life of your batteries and how much power they will produce over that life.

 

[booster]

A lot of the data is saying lithium like the middle of the state or charge range, often stating things like staying within the 40-70 percent for instance. With no need to get full that can be achieved in many cases, depending on how much you drive and how fast you charge while driving or how often you plug in. Solar can also increase times between larger charging rates.


The one thing that never seems to be discussed is the idea of constant float or even just covering use as solar might do. The whole cathode plating theory we here about is said to be caused because of the accumulation of too many ions on the cathode from voltage and current higher than the chemistry can absorb, so how does that fit into the picture of mid range charging when you aren't at the top of the SOC? I would think that any time you are at max voltage setting and the current is tapering, you would be in the excess ions state, but I am not a chemist so maybe not.


Our early testing is showing that we will mainly be able to stay midrange on our new 618ah lithium bank we only have to drive every 4-6 days to do that. We also have 300 watts of solar that we haven't even setup or tested with the system so might be even easier.

 

[Winston-ClassB]

. . . and I am curious as to what other early adopters are now discovering about their own lithium systems.

Having completed our 7th annual capacity test (actually 8th as we initially tested our lithium pack on Day 1 before placing the pack 'in-service’), we thought we'd accept InterBlog’s invitation and provide our statistics. Before responding, we undertook to re-read the entirety of this thread - - an interesting challenge. We expect to respond to few points raised in subsequent posts.

We purchased our pack late in 2016 ==> 20 - GBS 100ah cells of the type that “technomadia” initially installed. However, our Elite BMS has advanced over the 5 years since ‘technomadia’ purchased theirs - - each of our 20 cells has a printed circuit board which does provide us with the “granularity (that technomadia didn't have) to monitor each of the 20 individual cells”. The ability to "see" the voltage of individual cells has proven very important in monitoring battery performance and behavior and, especially, during the balancing process. Our cells are mounted inside, under the bed. They are not well ventilated, but neither are they tightly encased. We live in western Michigan where temperature extremes are rare. Yes, we do travel to south Texas and have experienced excruciatingly hot temperatures, but this is a rarity. We’re like Davydd, we try and follow acceptable temperatures.

Our “Ground Zero” capacity was: 516ah. This capacity was maintained through year 3.

By the 5th year capacity had dropped to: 475ah. Year 6: 425ah. Year 7: 415ah.

It is important to note that our capacity tests are run by placing a 10 ampere load on the fully charged pack and letting it ‘discharge’ to zero (we have defined zero to occur when the first cell reaches 2.7 volts).

This is a fairly ‘tame’ (low) discharge current. We think a second measure - - the ‘internal resistance of the pack’ which corresponds to the ‘voltage drop of the pack, particularly under high loads’ - - may be of equal importance. We’ve definitely noticed - - under the common 125-150 ampere loads of our induction stove and hot pot - - considerably increased ‘voltage drops’. Indeed, we’re wondering if we might soon run into the problem that MsNomer noted with her former AGMs - - that the system may not be able to maintain a sufficiently high voltage for normal operation under very high current draws. Unfortunately, not realizing that voltage-drop-under-load might become an issue, we have not undertaken tests to document voltage-drop over time.

 

[Winston-ClassB]

Not discussed thus far is the extent to which L3 [Lithium Life Loss] wreaks havoc on the reliability of the output displayed by the BMS that was calibrated to the original nominal capacity.

. . . I have largely ignored . . . [most measurements] . . . in favor of that one magic number displayed in the biggest, most convenient font - the percent SOC.

L3 destroys the accuracy of the percent SOC readout, and recalibrating it to the reduced effective battery capacity is not as easy as theory suggests. The drift is somehow far larger than what my intuition says it should be, leading to the following maddening result: Right at the point where my reduced battery capacity requires me to monitor the SOC more closely than ever, my ability to do so is thwarted . . .

We have been faced with the same basic question - - we programmed (both) of our battery monitors to 500ah . . . now the pack only has 415ahs - - What to do?

Our first inquiry was: Does it matter? “Does it really destroy the accuracy of the percent SoC” in a meaningfull way? If one knows that - - rather than 1 to 100 - - the real SoC range is 17 to 100, don't they still have a pretty good picture of the status of charge of their battery? Of course the displayed number is not the actual SoC percentage - - yet, knowing that our bottom is 17, we can ignore the fact that the displayed number is not a percentage, as we still know where we are within the battery's range.

In the end, at least so far, we’ve elected to do nothing. These monitors do exactly the same as they did on Day 1 - - they reset to 100% when the pack is fully charged and they accurately decrement the AH’s used, the Watt-Hours used, the pack voltage, currents in and out . . . everything the same, even though our actual capacity is now only 415ahs. The only thing these monitors don’t accurately report . . . is the amount of charge remaining. But, as noted, we really know how much we have left - - we know the number of AH’s used, and we know that we started with 415AHs. Or, again, we look at those SoC numbers knowing that 100 is FULL and 17 is EMPTY.

We could, of course, reprogram our monitors to this lesser capacity. It makes sense. But it didn’t seem necessary as long as we remember that when our displays says we’re down to “17%” - - we’re actually at zero.

If there’s some distortion in tracking/linearity - - we haven’t noticed it. And the fact that one of our monitors (the Victron 712) provides, in addition to State of Charge, an accounting of AH’s used - - this is an accurate measure of energy removed from the pack and is immune from discharge non-linearity.

This prompts us to respond to Davydd’s post at #33 where he comments: “I thought it [Lithium Life Loss] was imperceptible after 6 years of ownership. It was the same general percent usage overnight and charged back up in the same amount of time.”

Of course! Davydd could have had a 50% Lithium Life Loss and he would have noticed no change in his instrumentation. He’s using the same power each night. The instrument is properly ‘decrementing' his AH’s used and State of Discharge (albeit, the SoC number would have been highly inaccurate). Regardless of the actual health of his battery - - those overnight numbers wouldn’t change - - Davydd would not have noticed a problem until he tried to discharge his batteries below the 50% SoC level at which point he’d be ‘dead in the water’. Davydd, having a very similar system to ours, we’d have loved it had someone actually measured your actual capacity at the time of sale.

 

[Snowy-ClassB]

Davydd, having a very similar system to ours, we’d have loved it had someone actually measured your actual capacity at the time of sale.

Doesn't ARV buy back all of those sold used? If so, i expect that they might have a record of the battery's actual capacity when they got it back from Davydd.

 

[Winston-ClassB]

Doesn't ARV buy back all of those sold used? If so, i expect that they might have a record of the battery's actual capacity when they got it back from Davydd.

As I have no relationship with ARV, that would be a task for Davydd were he so inclined.

 

[booster]

davydd has posted in the past that ARV does a capacity test on ever unit the sell new and also when they receive it back, so they certainly have that data on hand. I had asked if he got it and he said no and had no interest in getting it IIRC.

 

[@Michael]

Here's some technical information on lithium battery life related to charge/discharge/temperature:

https://youtu.be/w4lvDGtfI9U?si=nsRRPRTtgu9kfEfA

It's focused on NMC/LiPo batteries, not LiFePo4. I suspect though, that much of this applies to LiFePo4 as well.

 

[booster]

Here's some technical information on lithium battery life related to charge/discharge/temperature:

https://youtu.be/w4lvDGtfI9U?si=nsRRPRTtgu9kfEfA

It's focused on NMC/LiPo batteries, not LiFePo4. I suspect though, that much of this applies to LiFePo4 as well.

Yeah, a lot of what I have seen in the specs for LFP also.


Not hot
Not totally full
Not very low SOC
Narrow SOC in middle of range best with frequent charging



He didn't address very cold weather issues, charge voltages, etc and glanced over the charge rate question which is a very important it IMO. Saying that rate has to be OK because the manufacturers control in the charging may or may not be true because the all sell on fast charging and range which are both contrary to best practices, it appears.