SOK, depth discharge, cycle life

[Bluedog225]

I saw some disturbing charts regarding the correlation between depth of discharge and cycle life which made me second guess my investment in server rack batteries.

The specs on currentconnected.com are pretty robust.

I don’t know which is accurate. My plan was to run my batteries to 80+ discharged daily during summer months. About 120 days.

Maybe SOK are better cells? The difference between 2000 cycles and 7000 is significant.

Any thoughts?

 

[sunshine_eggo]

You just letting us know? Are you looking for feedback?

 

[Bluedog225]

Post is somewhat edited. I hit post before done.

I’m curious about the disparity. And probably confused about correlation between rated capacity, depth of discharge, and cycle life.

 

[Bluedog225]

Looking at the SOK info, I can run these things into the ground regularly. Other LiFePO4 charts, not so.

 

[sunshine_eggo]

That actually looks more like a Lithium NCM (3.7V) cycle life chart. I would be surprised if it actually is legit for LFP in the way LFP is typically used.

NCM has a high voltage to SoC correlation, so it's relatively trivial to keep a battery in the 20-80% SoC range. A 40-70% SoC range (30% DoD) for NCM typically has a 10-30K cycle life.

If you've been paying attention, you know that it's pretty much impossible to target a peak SoC for LFP without a lot of hassle. I really doubt that if you charged LFP to 100% and then only discharged it to 70% you would see a 20K cycle life. Maybe you would. It's also highly unlikely that SOK cells any better than any other.

Furthermore, when cycle counts start exceeding 10-20 years aging becomes a concern. Here's a little blabbering about aging:

Chasing Decades Another Way || Not Another NiFe post

Hello everyone! I understand time is valuable; what follows immediately below is some background and framing for the central question- please feel free to skip to the bolded section for the elevator pitch version. ~ First time poster here. A quick thanks to members of this community and...

diysolarforum.com

If you want to squeeze the most out of your cells, charge to 3.45V/cell with an extended absorption cycle (6 hours), float at 3.4V and discharge to 20% SoC or higher. The lower peak voltage and lower charge rate reduce cell stress and maximize life while still achieving 98% SoC. Note that if you've sized your battery capacity based on 24 hours of computed usage, that is a conservative approach, and your typical daily cycle will be significantly more shallow than the 80% you've assumed.

 

[Koldsimer]

Simple. Buy enough batteries to only run to 10% daily. Enjoy your batteries for 273.9 years!

 

[Hedges]

"Cells rated for 80% capacity after 4000 cycles, 100% DoD A 1C rate"

But the graph shows ~1200 cycles.

 

[Koldsimer]

One of the issues you may have if you are planning to run to 80% daily in summer is that just one cloudy/rainy day could have you running the generator or grid.

How many kwh were you planning on running?

Ive got 25kwh and on a 100 degree+ summer night, i use about 40-50% of that keeping the house cold. Obviously, a lot of factors go into that- how well your how is insulated, what type of ac you run etc.

 

[rhino]

Just for reference, the EVE LF280K has cycle life of 6000 using 0.5C discharge/charge (2.5V - 3.65V) at 25C.

 

[Bluedog225]

The chart is not from current connected. Random grab off the interwebs.

I’ve got 20 kWh so far. Haven’t insulated the place or installed air con yet so usage is a guess. 1-2000 watts per hour. It’s all about air conditioning. The other loads are trivial and can be time shifted to solar production hours.

Fortunately, when it’s hot here, it’s usually very sunny.

Appreciate the thoughts on maximizing battery life. Though I don’t know if I can get a 6 hour absorption phase out of solar. I wonder how real life production curves look with east, south and west arrays. Will need to noodle around with the calculator.

 

[sunshine_eggo]

The chart is not from current connected. Random grab off the interwebs.

I’ve got 20 kWh so far. Haven’t insulated the place or installed air con yet so usage is a guess. 1-2000 watts per hour. It’s all about air conditioning. The other loads are trivial and can be time shifted to solar production hours.

Sounds a bit thin.

Fortunately, when it’s hot here, it’s usually very sunny.

And those conditions also degrade solar performance. Intense sun + high ambient temps = very hot cells. Common to see 10-20% degradation due to hot cell. Check your panel's NOCT rating for a more realistic performance of hot panels.

Appreciate the thoughts on maximizing battery life. Though I don’t know if I can get a 6 hour absorption phase out of solar.

Depends on your solar, but you'll hit absorption very quickly, and the current demand is reduced over those 6 hours pretty substantially.

I wonder how real life production curves look with east, south and west arrays. Will need to noodle around with the calculator.

While you can't do it for multiple arrays simultaneously, you could run 3 configs on PVWatts, download the hourly data and import it into a spreadsheet to see your combined array. I've done it a few times, and it was a little underwhelming. It's not just about orientation. Solar intensity is substantially reduced in morning and evening hours.

 

[Bluedog225]

Which Kool-Aid should we drink?

If Victron is a reliable source, it looks like Rolls Surrette flooded lead acid are in the same ballpark.





Battle Born

 

[Bluedog225]

I realize that battle born is a different chemistry. But they are clear; run them dead, no worries.

edit-I would have sworn that battle born was not LiFePO4. Huh. Gotta look into that.

[same as above with relevant portion circled]/

 

[sunshine_eggo]

Well maintained FLA with regular equalizations can have an absurd cycle life. The new "carbon" technology FLA pretty much mirrors LFP, and that's what you've linked.

Owning about 10kWh of Rolls batteries, that I have abused the living crap out of, means I'll drink the Rolls cool-aide all day long.

 

[rhino]

What is your expected DoD for the rest of the year? You said only 120 days to 80% which is only 1/3 of the year.. even using the chart provided that would then mean it would be better than what the 80% line shows.

 

[Bluedog225]

Absolutely lower discharge for the other 8 months. And yes. It will make a big difference in life.

My surprise this morning was both the “all over the map” charts for LiFePO4 and the comparability of quality lead acid.

 

[OffGridInTheCity]

As I've written before, early on I found this Battery University page - https://batteryuniversity.com/article/bu-808-how-to-prolong-lithium-based-batteries with this somewhat famous chart This chart is lithium-ion rather than LifePo4 but the question is the same.... can one really extend life by 2000 or 4000 cycles thru control of DOD? Lithium-ion's wider voltage range makes it a bit easier to target low DOD.


I've seen reports that Tesla (lithium-ion) car batteries can last 10yrs (potentially 3,650 cycles) but we don't know the actual cycles of DOD etc.
On another site, a contributor charged/discharged 4 x 18650 cells 100% and got over 2000 cycles for more expensive / name-brand cells.

Personally, my powerwall started at 75-25% SoC line (pic above) and thru expansion, I'm closing in on the 75-45% Soc line. Does that mean I'll get 7000 cycles (e.g. 19 YEARS) and still have a functional lithium-ion powerwall? Who knows... but if I live that long I'll certainly report back

Currently at 1,429 cycles on the oldest 18650 battery with ~37% average DOD - no sign of degradation so far

 

[sunshine_eggo]

Absolutely lower discharge for the other 8 months. And yes. It will make a big difference in life.

My surprise this morning was both the “all over the map” charts for LiFePO4 and the comparability of quality lead acid.

The only way you get those numbers is with maintenance and record keeping. One MUST regularly check specific gravity, add distilled water and perform regular equalization charges. Rolls is also VERY particular about how you charge them, and some equipment can't handle it. Absorption on Rolls FLA is 60V (15V/12V battery), and they want a nice long absorption cycle until the current has dropped to 0.01C. They consume water at a relatively rapid rate. The cost of this durability is weekly SG checking of 24 cells per 48V battery and documenting the results looking for trends to catch them early and equalize. It's pretty tedious.

 

[Koldsimer]

The fla's come with a whole can o worms. Checking specific gravity and adding water weekly, cleaning cables of sulfation, cleaning battery tops, burning holes in your shirts when acid splashes out, fighting to keep them in absorb when you have heavy loads, seriously compromising life when dipping below 50% dod.... You would need a very large and expensive bank of rolls to get the kwh you need on a nightly basis.

Seriously, i know the rolls are the shiz when it comes to fla's, but lithium is where it's at! I go weeks without even checking my batteries now that i'm comfortable with their performance.

 

[Bluedog225]

Fair points. I’ve been taking care of my big trojan fla for about a decade. It is a pain.

I vaguely thought there might be some fairly automatic charge and equalization equipment. Along with an automatic waterer that would make it less of a deal.

But I hear you, not for the faint of heart or the dabbler.

Thanks

 

[wiseacre]

But I hear you, not for the faint of heart or the dabbler.

Nor the memory impaired like me.
My DYI Lithium battery bank hasn't needed my attention since it was put into use.

Also I'm about to double the bank which wouldn't be recommended to do with LA
something about mixing old and new LA that led me to abandon a Gel bank when I originally wanted to expand my battery capacity

 

[sunshine_eggo]

Nor the memory impaired like me.
My DYI Lithium battery bank hasn't needed my attention since it was put into use.

Also I'm about to double the bank which wouldn't be recommended to do with LA
something about mixing old and new LA that led me to abandon a Gel bank when I originally wanted to expand my battery capacity

The no-mixing stuff is true across all chemistries. Everybody with any liability in the matter is going to recommend that all batteries be of the same age and capacity regardless of chemistry. Even with LA, folks routinely mix and match.

 

[Goboatingnow]

DoD is the primary cause of cycle life. It’s different for different versions of the chemistries , so lower DoD results in extended cycle life. ( it’s got to do with SEI growth ), essentially charging rather then discharging kills Lithiums. The brand is irrelevant

 

[Bluedog225]

Battle Born is a honey badger. ?

 

[MichaelK]

The fla's come with a whole can o worms. Checking specific gravity and adding water weekly, cleaning cables of sulfation, cleaning battery tops, burning holes in your shirts when acid splashes out, fighting to keep them in absorb when you have heavy loads, seriously compromising life when dipping below 50% dod.... You would need a very large and expensive bank of rolls to get the kwh you need on a nightly basis.

Seriously, i know the rolls are the shiz when it comes to fla's, but lithium is where it's at! I go weeks without even checking my batteries now that i'm comfortable with their performance.

Utter nonsense.

I have Rolls, and I'm finding I need to add water every 3 months, not weekly.

I'm running my 1hp well-pump all day long, and I have no problem keeping them in aborption. This is all about proper design.

I have the 568Ah Rolls battery. I was running the air-con all last night because to was 100+ even at 5000 feet. I woke up to 75% capacity.

Yes, I've holed a shirt with battery acid. It was a new dry battery that I just bought for my ATV, and I had to pour acid into the battery to activate it. From years of lead-acid maintenance, I've NEVER had a single hole from acid damage.