LFP Batteries below 20 percent?

[Suijkerbuijk]

Those are prismatic cells by the way.

And again, unless everyone is cycling their cells at optimal 25C, you'll have degradation well beyond the different between 0%-100% and 20%-80%. With LFP, what you really need to do to get optimal life cycle is temperature control. Cycle at 40C and watch it drop in half. Why is this important in this thread? Because it gets hot where OP is located.

Reason that my battery are set in the boat space that the temp on the bms sit between 20/25 Celsius.
The spot i have use is in the water 24/7 .
But away from the engine .
And in the winter i use heater in the boat (i life on the boat) so the temperature is always around the same time.
Same for my Victron inverter its on the same spot so operation temperatuur is really perfect .
Have the max output on my Victron.

 

[Luk88]

From what I've read, and how I've interpreted it, they state 6000 cycles for 0-100% and 8000 cycles for that 20-80%

For my Higee and EVE cells they don't mention it, but I don't doubt it(higher cycle life) is true in reality. The way I look at it. Imagine one cycle is worth 100 points. And a 50% cycle is worth 50 points. At 6k cycles of 100% this means your cycling life is 600k points. Now if you only use 60 points per "your cycle". It is logical (disregarding calendar and other types of aging) it should last for 10k cycles. The number 8K probably comes from being conservative, or taking some calendar aging into account.

As for what they (Higee and EVE) say.

Specifically (Higee):

and

This is Higee LFP71173205E-280Ah type cell.

EVE for MB31 on the other hand describes it like this:

And in appendix 1.15 we have




This is what they say. Is it true if you only use a portion of capacity you get longer life? Yes, but to decide if it is worth it one would have to overlay calendar aging curve on top of a cycling curve and see if one is actually gaining anything by "babying" the cells. In my case I actually might, because my cells are being stored at 15Celsius or less (above 5Celsius). So calendar aging is hopefully going to be very slow. However, for someone living in tropics? I bet their cells will calendar age well before any serious cycle caused degradation (unless they also exceed the charge/discharge ratings). Also, what about movement and vibration. In case of home ESS this is not a problem, but imagine it on a boat or in an RV. Will the vibration do the cells faster than calendar aging or cycling? Who knows?

I don't think there is any disagreement in principle here. Different manufacturers focus on other parts of the same thing. Also there are advancements in cell chemistry going on at the same time so the actual product might be different between two manufacturers.

 

[Brucey]

Remember, OP is talking about the PowerPro pack which not made in a barn, and has a warranty of >8000 cycles @ 80% DOD 0.5C

Actually if you look at the actual eg4 warranty it doesn't specify any of that, but its pro rated for operation for 10 years.

 

[wpns]

Those are prismatic cells by the way.

And again, unless everyone is cycling their cells at optimal 25C, you'll have degradation well beyond the different between 0%-100% and 20%-80%. With LFP, what you really need to do to get optimal life cycle is temperature control. Cycle at 40C and watch it drop in half. Why is this important in this thread? Because it gets hot where OP is located.

As Jimmy Buffet said, "It's 25C Somewhere". 8*)

My batteries have been between 31 and 41 C over the last week since I started logging, and it'll only go up from there over the summer. But unless I close off the ceiling in my boathouse where the equipment is and install an AC split (another 10KWHR/day?) that's not going to change.

But my charge/discharge current never exceeds 0.2C, and I'm not cycling 100% _every_ day (though I'd like to know that it's OK to do so occasionally), and it's become clear that the estimate for the number of batteries I needed was too conservative, as it was based on my historical 60KWHR/day, and after adding AC and getting used to 'free solar power' we definitely need more.

Adding more (pairs of) PowerPro batteries is about an hour's work (and 4 months shipping and transport delays), so that's definitely in the cards, I just need to convince the CFO that "It's Only Money" 8*)

 

[wpns]

Remember, OP is talking about the PowerPro pack which not made in a barn, and has a warranty of >8000 cycles @ 80% DOD 0.5C

Except, as recent discussion of the warranty seems to indicate, they don't seem to warranty cycles or MWHR or anything but function, and if a cell dies it's 10 year pro-rated.

 

[wpns]

From what I've read, and how I've interpreted it, they state 6000 cycles for 0-100% and 8000 cycles for that 20-80%

And at 8000 60% DOD cycles you get the equivalent of 4800 100% DOD cycles, so you actually come out behind in terms of $/KWHR-transferred, or whatever you call that metric.

 

[wpns]

Reason that my battery are set in the boat space that the temp on the bms sit between 20/25 Celsius.
The spot i have use is in the water 24/7 .
But away from the engine .
And in the winter i use heater in the boat (i life on the boat) so the temperature is always around the same time.
Same for my Victron inverter its on the same spot so operation temperatuur is really perfect .
Have the max output on my Victron.

Big thermal mass for the win!

 

[wpns]

Actually if you look at the actual eg4 warranty it doesn't specify any of that, but its pro rated for operation for 10 years.

Even stranger than that. The warranty is for function, not SOH or cycles or anything. 10 years parts, but if you have to ship it back (welded cells means if a cell fails you are shipping it back) then it's 10 years pro-rated on the current retail price of the battery. And from here, shipping it back would be a real undertaking. Had a neighbor who sent an outboard engine back under warranty, and it was not an easy task.

 

[Brucey]

I like to think that my rept 280s had a hard early life, sitting in a container in the blazing sun for a few weeks on the ship, then prob another hot warehouse in Texas, a little more baking in the FedEx truck and then finally home to a sweet life of 65-75F and light cycling at low C rates.

 

[wpns]

I like to think that my rept 280s had a hard early life, sitting in a container in the blazing sun for a few weeks on the ship, then prob another hot warehouse in Texas, a little more baking in the FedEx truck and then finally home to a sweet life of 65-75F and light cycling at low C rates.

How good of you to take them on. #WhoRescuedWho?

 

[Brucey]

Even stranger than that. The warranty is for function, not SOH or cycles or anything. 10 years parts, but if you have to ship it back (welded cells means if a cell fails you are shipping it back) then it's 10 years pro-rated on the current retail price of the battery. And from here, shipping it back would be a real undertaking. Had a neighbor who sent an outboard engine back under warranty, and it was not an easy task.

Surely it should be pro rated based on price initially paid?

Otherwise, say you paid $3500 at release. Eventually midnite releases their big battery at crazy price and eg4 drops to $2500 in five years. If your eg4 fails in year 5 you ship it back and they give you $1250?

 

[Brucey]

How good of you to take them on. #WhoRescuedWho?

I have a lot of love to give so I'm sure many more will be adopted soon.

 

[Suijkerbuijk]

Big thermal mass for the win!

Wat do you mean ?
My boat its make from steel.
And the spot i use its a closet that is in the living room.
Only the steel is in te water and cool space.
No water comes thare.
And in the winter the heat systeem of the boat warm that spot up to about 17 celcius.

 

[wpns]

Wat do you mean ?

You have the thermal mass of the ocean to keep the temperature of your boat fairly constant.

 

[Suijkerbuijk]

You have the thermal mass of the ocean to keep the temperature of your boat fairly constant.

Aha that do you mean..
Its yes and no really.
The masterbed room and the living room can be hot to about 30/35 celcius.
But i can always use a airco unit to cool it down.
And yes all what is in the water its cool.
I drop sometimes drinks and beer by the engine in the summer .
If i'm on the port and stay thare .
That room spot will be cold around 20/25 celcius so its colder than in the living room .

 

[ksmithaz1]

From what I've read, and how I've interpreted it, they state 6000 cycles for 0-100% and 8000 cycles for that 20-80%

Again, if true, (and it's not AFAIK, there is no data that supports the theory) It's still stupid. The math does not work, in fact in this example it's even worse. Foregoing 40% of your charge gives you a 25% longer life. So you must buy an additional 40% capacity over charging 0-100, thus spend 40% more money to save 25% of the lifetime. I find these arguments unbelievably dumb, it's not rocket science, it's simple math. I find it sad that people struggle with basic math skills, no wonder these financial scams work so well.

 

[olivinelife]

It's still stupid. The math does not work, in fact in this example it's even worse. Foregoing 40% of your charge gives you a 25% longer life.

Normally in scientific papers investigating cycle degradation, the "cycles" is referring to "Full-Equivalent Cycles" (FEC). So going from 100-50% would be 0.5 FEC, and 100%-90% would be 0.1 FEC, which means that we are comparing actual useful throughput without having to do basic math or deal with misleading charts. Generally for lithium ion batteries you get more FECs with shallow cycling. This is very much true for high-nickel cathodes but the story is more nuanced for LFP. However the chart on website linked by @ShirBlackspots does not match most of the scientific literature on LFP DOD cycling. Based on my searches I cannot find the source of the chart on any scientific publication or the rest of Google for that matter.


Due to image feature duplications, it appears that the different curves are simply copy pasted from 1 original curve and shrunk/scaled on the x-axis. It is also highly unusual for a cycle degradation test to end at different #s of FECs, since usually the study is done for a fixed amount of testing time, and thus all curves should end at the same FECs or x-coordinate. It is also highly unusual for them all to end at the same y-axis / SOH point, which doesn't even reach the industry standard of 80% SOH. Yes sometimes the more abused cells end testing because of early failure but this is just weird.

Unfortunately, it is highly likely this chart is "made up".

 

[olivinelife]

If you want to see an actual scientific publication on LFP cycling degradation as a function of DOD, see

Capacity Recovery Effect in Commercial LiFePO4 / Graphite Cells

Figure 2: Capacity retention vs FECs as a function of cycle depth and mean SOC (c). Capacity retention with full cycles is shown in blue in both graphs for reference. Cycling with low cycle depths around 50% SOC leads to much faster capacity losses than full cycles.

Shallow cycling (ex. 40-60%) leads to lower degradation after 7000 FEC, but substantially more "reversible" degradation at <7000 FEC. This "reversible degradation" can be reversed by prolonged holds at either 100% or <0% SOC, or it can be prevented by cycling at different SOC ranges. Taken to the logical conclusion, 100-0% is honestly the best to avoid this effect while maximizing the the useful capacity of the battery because I honestly doubt any user's battery will reach >7000 FEC in real-world usage.

 

[wpns]

If you want to see an actual scientific publication on LFP cycling degradation as a function of DOD, see

Capacity Recovery Effect in Commercial LiFePO4 / Graphite Cells

View attachment 226538


Shallow cycling (ex. 40-60%) leads to lower degradation after 7000 FEC, but substantially more "reversible" degradation at <7000 FEC. This "reversible degradation" can be reversed by prolonged holds at either 100% or <0% SOC, or it can be prevented by cycling at different SOC ranges. Taken to the logical conclusion, 100-0% is honestly the best to avoid this effect while maximizing the the useful capacity of the battery because I honestly doubt any user's battery will reach >7000 FEC in real-world usage.

So in the 6000-8000 cycle range, just use them?

 

[olivinelife]

So in the 6000-8000 cycle range, just use them?

Under all the different DOD cycling conditions, the LFP batteries successfully survived to >8000 FEC, albeit with slightly different % degradations. For most residential use-cases those few % difference in degradation won't have any practical difference.

I think the only operators who should care are commercial operators doing commodity load-shifting. But many of those installs are already cooking in the desert where degradation from high temperatures will overshadow any noticeable difference in cycling DOD strategy. And even if they are thermally-controlled, operators need to contend with the opportunity cost of purposely using less capacity.

Just use LFP batteries.

 

[ksmithaz1]

I honestly doubt any user's battery will reach >7000 FEC in real-world usage.

I would beg to differ. I think this WILL happen it's more of a when. By the time we get there is it going to be relevant? Will there be a a vastly superior tech?

 

[zanydroid]

I would beg to differ. I think this WILL happen it's more of a when. By the time we get there is it going to be relevant? Will there be a a vastly superior tech?

7000 FEC is 19 years of full cycle, if you keep a reserve it’ll be well more than 20 years. If in a inverter paired setup the probability of either the BMS or the inverter failing by then is extremely high, and that would likely be a replacement with the disposable lifecycle planning a lot of these companies are going for.

In a DIY context with battery chosen for serviceability and willingness to scrounge up replacement inverter notwithstanding what the manufacturer and regulations want, 20 years is reasonable.

 

[sunshine_eggo]

Ok, whatever. You have your bias,

Everyone has their bias. Mine is based on experience, data presented on this website and cell manufacturer's published data.

Your bias is clearly strongly influenced by your 21700 battery build. It would help you to understand that your experience with 3.6V/3.7V chemistry does not translate to LFP in any way.

and obviously can't be convinced of the legion of data out there supports what I have said..

No one can. The data you presented and the "legion of data out there" does not support what you have said. You have incorrectly interpreted the data.

When I am finally able to buy a DIY 16S 280Ah battery box, my intention is to keep my batteries between 20-90%

When you figure out how to do this reliably, consistently and without manual intervention, you will win the internet and be the LFP god of all. Targeting a peak SoC on NCM/NCA/LMO (3.6/3.7V) chemistry is an absolute trivial matter due to the strong voltage to SoC correlation. On LFP, this is essentially impossible. You can't just charge to a peak voltage and expect to hit a specific SoC with LFP... unless you're shooting for 100%.

 

[wpns]

Surely it should be pro rated based on price initially paid?

Otherwise, say you paid $3500 at release. Eventually midnite releases their big battery at crazy price and eg4 drops to $2500 in five years. If your eg4 fails in year 5 you ship it back and they give you $1250?

You’d think so, but that’s not how it’s written.

Your scenario is entirely plausible. How they’ll actually start handling warranty claims has yet to be seen. PowerPro hasn’t been out long enough, and I’m sure they’ll make a business decision when the time comes to either agressively pursue customer satisfaction or get a rep for poor warranty performance. . Ask me again in 5-10 years.

 

[time2roll]

The trouble with taking them down to zero by design is loss of power. That matters to me.