LifePO4 Battery Capacity, SOC and their relationship?

Im turning my attention to my batteries now trying to figure out why the capacity readings are so far off when SOC is 100%. These are all 100ah "ESS PowerUp" LifePO4 batteries I got from watts247 a couple of months ago. https://watts247.com/product/powerup-lifepo4-5-1kwh/

They have always showed these capacity numbers from the day I installed them and got the BMS working w/ Solar Assistant.

They are wired in Parallel using 2/0 gauge battery cables, + wired at the top + terminal and - wired at the bottom - terminal. cable length from batteries to inverters is 40" max. I've tried doubling up on 2/0 cables between batteries with no change.



Solar Assistant readings:



I have max charge current set to about 80A, but it makes no difference what I set that to. Measuring DC current on each battery during recharge I get about equal amps going into each battery (ie 20A) during max charge either from PV or from Grid. They are at 96% right no bc the sun went down an hour ago and I'm on my discharge cycle.

Any help in understanding whats going on would be helpful. Watts247 says I need bigger battery cables.... I respectfully disagree.

Jeff

Am I seeing a double interconnect at the bottom right?

30 minute resting voltage (no charge or discharge of any kind) correlations to SoC is somewhat accurate.

Have you ever charged them to 57.6V?

That pic was when I tried doubling up on interconnect cables which had no effect so I removed them. Here are the params I’ve set. And no, I’ve not tried charging them to that voltage. Each inverter is set to these.

Are the batteries actually that Ah, or should they all be 100Ah?

Do they actually get to 57.0V on each battery?

What do the cell voltages look like when you're at peak absorption?

Have you considered pushing your float to 55.2V?

I suspect they're not hitting a high enough voltage to trigger 100% SoC.

They are all rated for 100ah and they all fully hit 100% SOC at about 1pm (assuming Good sunny day), at least as reported by the bms.

I'll change the charge params now and charge at 57.6v @20A and watch what happens. I'm lollygagging in bed anyways, and it should only take a couple hours to fully charge.

drain them down to 46v so the BMS can measure capacity of the pack and update the numbers

Copy that...

You are charging above Working Voltage Range and will NEVER get to where you want to be.
There is NO POINT AT ALL for charging LFP above 3.425Volts per cell... or 54.8 V per 48V Pack.

The "Allowable Voltage" allowed for LFP is 2.500-3.650 Volts per cell, that is their Max Allowable Charge & Empty state.
* Above or Below the "Allowable Range" = HARMING BATTERY CELLS ! These are LIMITS not useable power !
The REAL Working Range is 3.000-3.400 per cell with the Nominal being 3.200 Volts per cell. (THAT IS 50% SOC BTW)
This is where the Deliverable Amp Hours comes from NOT from 3.425-3.650 OR from the bottom (2.500-2.900) Those are the "cliffs.

Charge the Bank to 54.8V Absorb/Bulk, then FLOAT at 54.7 (3.41Vpc) (Variable Current) to top off the bank to saturate the cells at that voltage.

57.6V = 3.600Vpc which is in the Useless Top Curve !
You can charge the batteries to 3.600 and taking less than 1 Amp, stop chanrge & they WILL Settle to 3.3400+/-050 within an hour or so BECAUSE that is their Working Voltage. Charging ABOVE 3.425 IS NOT GOOD FOR THEM ! It is Overcharging !!!

Battery Charge Voltage 54.8V (3.425Vpc) Constant Current/Constant Voltage) takes them to 95% charge,
Battery Float Voltage 54.6 (3.412)-54.7(3.418) Constant Voltage Variable Current to finish off the last 5-7% of cell capacity.
Battery LVD cutoff 44.8V (2.800Vpc)
Battery HVD cutoff 56.6V (3.525Vpc)

Active Balancing Trigger point should be 3.375Vpc or 54.0V
Target Delta between cells set to 0.010V

CASE IN POINT, I Run 5 Packs in Parallel set in a similar config. They are in FLOAT by 13:00 on a Sunny Day. Balance Delta of <0.010 is reached within 1 hour of being in Float. All Packs are at FULL 100% SOC at that stage, and go into Idele as Float can carry running loads provided there is enough solar input. SEE HERE: https://diysolarforum.com/threads/m...ell-includes-charge-profile-bms-config.42549/ (Note that is prior to pack-5 being put into place.)

A WORD ON VENDORS, LABELS & MARKETING!
The way they spec charge profiles like yours is Marketing Shmucks doing and NOT Technical people !! It's a FOOLEY ! The WORKING RANGE is what it is and applies to ALL Lithium Iron Phosphate battery cells, regardless what companies make them (Exception Winston which uses Yttruim Doping for wide temp range use).

Ok, I'm good on all That and params have been updated. so how do I get the correct capacity readings? Discharge to 46v as @HighTech!ab stated?

On my system, that pack capacity number is set in the BMS manually. It's a setting. Not addressable in Solar Assistant. Must BT into the BMS.

Also, the field called 'Charge Capacity' below your green circles in first post is actually capacity remaining and is dependent on accurately setting previous post.

Mine is calculated. It’s not a setting.

Hmm. Ok.
Go look anyway.

So I got to the bottom of this confusing mess by doing a full capacity test on the 4 x 100Ah pack. It turns out that the batteries have a full 100Ah of capacity each, and the params in the battery are incorrect. After a full charge to 54.8v and then a full discharge down to 48v (leaving 10%) I got 365Ah, a full recharge gave me the same result. So, 365+36.5=401.5. Close enuf.


So... @TomC4306, you must be correct, there has be a way to change the capacity params in a bms, although there's no software I know of to do this. Ian at watts247 is getting in touch with the manufacturer to see if there is some sw available. This is important because Solar Assistant uses SOC to control power management, and my batteries are only reporting 68% of their actual capacity so that feature is whacked until I can fix the actual SOC values.

I'm almost of the mind to try a few different BMS Windows apps just to see if one of them works. There can't be that many BMS apps, or are there?

Jeff

The ONLY Time cells should be pushed to 3.6XX Volts is to INITIALIZE them after long-term storage &/OR shipment.
NOTE THE ELECTRICAL PERFORMANCE IMAGE !
It States NOMINAL VOLTAGE @ 51.2V / 3.200Vpc which IS CORRECT as that is the Mid Point between 3.000-3.400 Working Range.
The 5120Wh Stored Energy comes from THAT Working Range.
BALANCING should always begin before the top of the Working Range as above 3.415Vpc the cells will most often deviate due to being outside of the working range.

Below are other adjustments to make which brings the battery packs inline with specs.

So I found a copy of pbms and it worked on my Powerup batteries. The mystery is solved... at first I thought that SOH was being reported but then realize that it's the actual value of the "Full capacity" Ah's reported by the BMS.

But why are the capacity numbers coming out of the actual BMS so low on 3 of my "brand new" batteries?

I don't see any place to set these values, assuming they were incorrectly set up in the first place.







note: the cell voltages dont match SA bc the pack was plugged into my computer & not the battery pack So the data in SA is stale.

I got a surprise call today from the PowerUp Distributor and his support tech. Ian at Watts247 set that up. To say I was shocked would be an understatement, but they want to fix this, so I'm all ears.

They have the data and are currently looking into why these batteries are incorrectly reporting capacity. There actually is a place in the BMS sw to enter capacity for each pack, but it requires a password to edit the values. So strange that the design capacity number is correct, yet the full value is not.

More info. I just happened to take a screenshot when I first installed the system back in May. The capacity numbers are all close to or 100Ah.

what is causing the reported capacity to change so drastically???

I will PM you the password. Please do not share.

So the battery vendor (ESS) really came through today. Their tech spent a couple hours on the phone with me and we upgraded the PACE bms firmware to the most recent. It was on 1.0.3, and the new version is 1.0.5. After a full discharge and charge, ALL readings look correct now.

An interesting takeaway from all this is that the cool "Maintain battery state of charge" feature in Solar Assistant should really be based on actual battery voltage, not SOC. If that was the case then it wouldn't matter what soc was being reported. This is especially true if the BMS is emulated by SA. I may start another thread on this topic to solicit feedback on this.