Power Cycles vs Depth of Discharge vs Longevity

Started by aaapilot, August 25, 2024, 05:52:23 PM

Previous topic - Next topic

aaapilot

Fine tuning the LifePo4 battery pack for optimal life.  My Rosie has AC Grid support, so I can set a specific SOC that I can drain to and/or set specific times when only the battery will be used, etc.  I also have excess PV available in the summer, so I can choose to keep batteries in a 'float' state for 8 hours a day and fully charge to 100% most days, but what is best for the batteries and what is the battery warranty REALLY based on?

Battery life is specified in Cycles (MNPowerflo are 6000 cycles), but I don't see a depth of discharge specified by most manufacturers and not from Midnite (did I missed it somewhere in the literature?)  Some sites RECOMMEND 20%, but what numbers are the MANUFACTURERS really basing these cycle life claims upon? 

I'm not totally off-grid, so I have the luxury of draining less than the **number** recommended, let's just say 20% is the number to get 6000 cycles, if I drain to only 50%, what does that do, if anything, to increase the Life Cycle count the battery will achieve? We know what it does for lead batteries, but I haven't seen any hard numbers from Lithium manufactures.

BMS's have cycle counters, so I'm going to assume that's what would be used to decide if a battery doesn't perform to the warranty. So, if I drain a battery to 20% and charge to 80-90%, will the battery count that as a cycle or no?  My experience says no, so does NOT charging a battery to 100% effectively extend the life cycle count 'life' and extend the warranty, per se. Discounting the warranty, is it better for longevity to charge a lithium to 100% every day or keep it in the 20%-80% range for a week, then charge to 100%?  My 30kw pack has been in operation since February and the 6 individual 5kw batteries have from 31-34 cycles on them.  I haven't intentionally restricted the charging, but have 2 separate PV arrays w/different controllers, so have been testing output and adding loads to the backup panel, etc. The backup panel loads are ALWAYS powered by Rosie/battery, so it can use 6-10kw from 6pm until 9am when I begin to get PV, 4 kw more in summer when the well pump runs for irrigation, more if it's cloudy.

I haven't seen much in-depth discussion on these factors from sources that I would trust, like the battery manufacturer themselves or folks that have been early adopters and have enough experience to have hard numbers.  Or White Papers. Folks on this forum are mostly in for the long haul and would appreciate some incites from what 'YOU' are doing .... or not doing! :)

Dave
_ _ _ _ _
Dave /:\

ClassicCrazy

It will be interesting to see what the answer is to this.
On my lithium batteries it seems to count a cycle when the voltage and current get up to what is programmed in the bms. And below that the bms reported SOC gets all out of whack and can't trust it. My Classic reported SOC which averages all the batteries is usually always reliable. But of course that is going to depend on what Midnite has used for a bms and what it is programmed to report if those are the batteries you are wondering about.
Every other lithium battery manufacturer or bms often has different ways they do things.
I made a video on how I observed the soc fluctuate as reported by bms . Of course I have an open loop system and not sure how that would differ from closed loop ones.
https://youtu.be/eLsq2q8eHfQ?si=tttP-VllszoCqCjJ
Larry
system 1
Classic 150 , 5s3p  Kyocera 135watt , 12s Soneil 2v 540amp lead crystal for 24v pack , Outback 3524 inverter
system 2
 5s 135w Kyocero , 3s3p 270w Kyocera  to Classic 150 ,   8s Kyocera 225w to Hawkes Bay Jakiper 48v 15kwh LiFePO4 , Outback VFX 3648 inverter
system 3
KID / Brat portable

dapdan

Gents,

This is indeed the question that is on the mind of most person using lithium nowadays. I have personally decided that I will use as much energy and cycles I can from the batteries because I currently subscribe to the idea that current lithium batteries will die (or degrade) from old age first before they do so due to overuse or cycle death. I arrived at this view primarily based on the view and experiments of Andy from the youtube channel " offgrid garage" and my own experience.

Typically they will say that lifep04 has a shelf life of 20 to 25years. I had installed some new CATL 280Ah cells at a client. The system has 30kwh installed using two parallel banks of 15kwh. I visited the installation last week and one of the BMS was showing a cycle count of 140 after a year of installation. The current specs on these batteries indicate 6000 cycle to 80% dod. On setting them up I have not limited maximum charge or discharge and have the inverter charging from grid at around 44v, so basically I am looking at as much discharge as possible. So for this discussion lets call it 80% discharge. At the current rate it will take 42.8yrs, at a cycle count of 140 per year, for the batteries to reach a state of health of 80%. I am not sure if I will be alive at that point (I would be around 90yrs at that stage), I will update this response if I am :>.  There by the time the battery gets to 80% SoH it will be well past its shelf life. This why I subscribe to the idea of taking as much energy as possible now be cause the batteries will die of old age first before it runs out of energy cycles as the batter still has thousands of cycles to go before it decrease to 60% SoH.

I can say that I am now defaulting to undersizing battery banks, use the grid as back up when the battery falls to it lowest possible SoC to cycle batteries deeper in an attempt to get the maximum possible lifetime energy extraction before the battery loses energy storage or cycles due to old age all the while minimising the upfront capital cost of battery. If there is to much storage where DoD is shallow then you will not be able to maximise lifetime energy extraction from the batteries.

Damani

aaapilot

Interesting take, Dapdan. The batteries are certainly the most expensive part of the system, so maximizing their ENERGY output for any given amount makes sense.  I selected my battery size to allow for aprox 36 hours of back-up panel operation, no more, and not running large 'utility' size loads for any period of time.  Microwave for a short period, sure.  Well pump for short periods, absolutely. My thinking was to elliminate running a generator overnight for a load of just 300-500 watts for a couple refrigerators, a fan, lights, internet, TV.  If there's no sun/PV charge for 2 days, then run the generator at full power as necessary to recharge batteries and do laundry & other load heavy devices, then back to battery for another 36 hours.
I believe/hope these lithium batteries will last as long as you describe.  What I'm a bit more concerned about is the BMS board that controls them! If the BMS fails in 11 years, I'll bet an original BMS replacement won't be available.  Then I'll have batteries that are still good, but no BMS to control them.  Technology will have moved on to the next 'better' chemistry & the electronics to support it.  This thinking supports your philosophy of undersizing the battery system and using the snot out of it NOW!

Dave
_ _ _ _ _
Dave /:\

dapdan

aa,

I would not be to concerned about the latest and greatest battery tech. Just look at lead acid it was around for decades before lithium started to de throne it in the past 5yrs or so. Lithium were out commercial for the past 20yrs and i bet they will be king for the next 20-30yrs comfortable because of how long it really take to bring mature battery tech to the commercial market. The advancement of lithium will continue to be driven by the automotive market. It is automotive market we have to thank for the sharp drop in price of lithium that it became viable for residential and commercial use. This took about 15-20yrs to happen. If you look at the automotive market there are no new wide scale battery tech being deployed apart from Lifep04 in base model EVs. So I think we are good for a minimum of 20yrs comfortable based on the maturity period for new battery tech to trickle down to us DIYs/ off-gridders. As for the BMS, there was a time bms were non existent or very expensive and guys bottom balanced and operated without them and regulated the voltage with other devices and had fair success doing so. One can still operate lithium with external regulation from a trusted reliable inverter or CC. So the community will always find a work around. Lot of brilliant intelligent engineers out there. There is nothing to worry about here.

Damani

ClassicCrazy

Quote from: dapdan on September 12, 2024, 07:48:54 AMaa,

I would not be to concerned about the latest and greatest battery tech. Just look at lead acid it was around for decades before lithium started to de throne it in the past 5yrs or so. Lithium were out commercial for the past 20yrs and i bet they will be king for the next 20-30yrs comfortable because of how long it really take to bring mature battery tech to the commercial market. The advancement of lithium will continue to be driven by the automotive market. It is automotive market we have to thank for the sharp drop in price of lithium that it became viable for residential and commercial use. This took about 15-20yrs to happen. If you look at the automotive market there are no new wide scale battery tech being deployed apart from Lifep04 in base model EVs. So I think we are good for a minimum of 20yrs comfortable based on the maturity period for new battery tech to trickle down to us DIYs/ off-gridders. As for the BMS, there was a time bms were non existent or very expensive and guys bottom balanced and operated without them and regulated the voltage with other devices and had fair success doing so. One can still operate lithium with external regulation from a trusted reliable inverter or CC. So the community will always find a work around. Lot of brilliant intelligent engineers out there. There is nothing to worry about here.

Damani
I got 48v worth of 100 ah Calb lithium batteries in 2015 to try out. At that time there were barely any bms like you say, and what was out there was very expensive and not that great. I tried to bottom balance like someone on this group said he did, and since I didn't really understand I thought I ruined my batteries because I had no low voltage cutoff and the Classic took them down too far in voltage when I was letting them just idle and not around. They just sat around for years no charge or attention and last year I was very surprised that all of the cells were in acceptable voltage range. I hooked them up and gave them a charge which they took, but I never load tested the pack. I now have a good active balancing jkbms on them but took the pack off line to make space for a newer battery. Not sure what I will do with them, but amazing they still have life in them. I was sure some of the cells were down to 2 volts years ago and that is why i neglected them. Guess they healed themselves.
Larry
system 1
Classic 150 , 5s3p  Kyocera 135watt , 12s Soneil 2v 540amp lead crystal for 24v pack , Outback 3524 inverter
system 2
 5s 135w Kyocero , 3s3p 270w Kyocera  to Classic 150 ,   8s Kyocera 225w to Hawkes Bay Jakiper 48v 15kwh LiFePO4 , Outback VFX 3648 inverter
system 3
KID / Brat portable

dapdan

CC,
Wonderful testimony. Those CALB would be nearly 10yrs old. It would be a fantastic test case to see where they are now in respect to capacity vs shelf life/age given that you have not used them. Big HINT. I bet if you hook them up they will run loads just fine.

Damani

mahendra

I have calbs they are great.
I don't know when they were manufactured but they are about 7 years .
I had reasons to buy one of those battery impedance testers to check some other batteries and decided to check the Calbs while I was at it ,turns out they are in good shape.
I usually cycle no less than 40% soc
Looks like they may out live 10years
1.5kw on Midnite classic 150(whizbang jr.) networked 0.660kw on classic lite 200 wbjr ,180ah CALB Lifepo4 48v battery bank,jkbms 150amps bms(top balanced) Outback vfx3648.
Soon to be added Sol Ark 8k with 5Dyness Bx51000 ,7.8kw pv

Brucey

Currently I have 20kWh of assorted LFP batteries connected to my Rosie. 8kW of assorted panels. Pretty shaded spot right now. I had a period of about 10 days where I wasnt hitting float due to clouds etc. My philosophy is to try and get to float so the batteries can do a bit of balancing. I've got them down to 3-5% soc at times. My understanding is that LFP cells don't like to be held at 100% for extended periods (weeks, months?). But in our ESS utilization patterns we will always be discharging below 100% during the overnight, so I don't see that as a risk at all. If only using as a UPS backup system with grid as primary and just fail over, maybe an issue?

If the batteries are good for say 4000/5000 cycles, thats over 10 years of daily full cycles at a worst case. I still have more batteries to add so I doubt I'll be venturing down to really low soc in the future unless there's a bunch of bad weather over multiple days.


ClassicCrazy

Quote from: Brucey on September 14, 2024, 08:44:05 AMCurrently I have 20kWh of assorted LFP batteries connected to my Rosie. 8kW of assorted panels. Pretty shaded spot right now. I had a period of about 10 days where I wasnt hitting float due to clouds etc. My philosophy is to try and get to float so the batteries can do a bit of balancing. I've got them down to 3-5% soc at times. My understanding is that LFP cells don't like to be held at 100% for extended periods (weeks, months?). But in our ESS utilization patterns we will always be discharging below 100% during the overnight, so I don't see that as a risk at all. If only using as a UPS backup system with grid as primary and just fail over, maybe an issue?

If the batteries are good for say 4000/5000 cycles, thats over 10 years of daily full cycles at a worst case. I still have more batteries to add so I doubt I'll be venturing down to really low soc in the future unless there's a bunch of bad weather over multiple days.


Not sure why your batteries would balance at a float setting . usually float is the voltage where it is resting without charing or discharging which is around 53.6 to 54 v depending on what an inverter needs to see to still supply loads.  . What voltage are you floating at ?

It really depends on what your bms is programmed to do to balance. Luckily my batteries I can adjust bms to where ever I want. I charge to 52.2v which is about 99% full.
Larry
system 1
Classic 150 , 5s3p  Kyocera 135watt , 12s Soneil 2v 540amp lead crystal for 24v pack , Outback 3524 inverter
system 2
 5s 135w Kyocero , 3s3p 270w Kyocera  to Classic 150 ,   8s Kyocera 225w to Hawkes Bay Jakiper 48v 15kwh LiFePO4 , Outback VFX 3648 inverter
system 3
KID / Brat portable

Brucey

I'm absorbing at 55.2V and floating at 54.4V. Taking it nice and easy on the cells.

Battery bank is currently:
Two 24V 100Ah Ecoworthy in series
Two 24V 100Ah djlb in series
Two 48V 50Ah ecoworthy pizza box in parallel
One 48V 100Ah gokwh rack mount battery.

As you can see most of my batteries are black boxes and no idea what's happening internally.

Still to finish assembling, a 14kWh battery using REPT 280Ah cells, and a non inverter jk BMS (200A, 2A balance).  Also have four Redodo 12V 100Ah batteries in series with litime 10A balancer to connect, just need some decent metal shelves to get most of my gear off the top of the washer and dryer and more room to stack things nearly.

ClassicCrazy

Quote from: Brucey on September 15, 2024, 03:13:11 AMI'm absorbing at 55.2V and floating at 54.4V. Taking it nice and easy on the cells.

Battery bank is currently:
Two 24V 100Ah Ecoworthy in series
Two 24V 100Ah djlb in series
Two 48V 50Ah ecoworthy pizza box in parallel
One 48V 100Ah gokwh rack mount battery.

As you can see most of my batteries are black boxes and no idea what's happening internally.

Still to finish assembling, a 14kWh battery using REPT 280Ah cells, and a non inverter jk BMS (200A, 2A balance).  Also have four Redodo 12V 100Ah batteries in series with litime 10A balancer to connect, just need some decent metal shelves to get most of my gear off the top of the washer and dryer and more room to stack things nearly.
wow you really have the variety of batteries there !
You sound a bit like me in getting things working first, and then doing the permanent install later.
Maybe you can find an old washer and use that as the case for batteries so that it matches the washer and dryer !
I have been trying to make my lithium battery collection here nicer than I often do things. But they are heavy and I needed to build an insulated box around them to keep them warm in winter. And every time I add a new battery I need to redo some things. But last year my heated insulated box worked very well .
I do have a rack I built that is on wheels so sort of can move it around. I also bought a hydraulic motorcycle lift that goes 36 inches or so high so when I have to take a battery out or move it , saves my back or dropping it on my foot. https://www.amazon.com/gp/product/B00CKY3J3Y/ref=ppx_yo_dt_b_search_asin_title?ie=UTF8&th=1
Larry
system 1
Classic 150 , 5s3p  Kyocera 135watt , 12s Soneil 2v 540amp lead crystal for 24v pack , Outback 3524 inverter
system 2
 5s 135w Kyocero , 3s3p 270w Kyocera  to Classic 150 ,   8s Kyocera 225w to Hawkes Bay Jakiper 48v 15kwh LiFePO4 , Outback VFX 3648 inverter
system 3
KID / Brat portable

Brucey

Here's my solar disease progression.

River 2
Couple 100W bifacials
2nd River 2
Delta 2 Max
150/35
10 300W Hyundai panels
Started buying Redodo 12V 100Ah as additional batteries for the river 2s
Victron 12/30 charger
2nd 150/35
450/100
48/1200
10 Aptos 370W bifacials.
The 24Vs and the 48 50Ah were all on Amazon or eBay sale at around $360 each. Trying to target around $700 per 5kWh of battery. The gokw rack mount was $885.

And then more recently, a Rosie and last week picked up a Hawkes Bay 90A and breaker box. Will be going with the Midnite gear going forward, just no comparison in build quality and being built in Washington.