What voltage to use when estimating kWh for LiFePO4 capacity?

[krby]

Or... "How many watt hours are in a single LiFePO4 cell?"

Before I bought my first big LFE, I had a spreadsheet of various options and in order to know if a given battery was "enough" for my use case AND to compare price per kilowatt-hour, I estimated the capacity of each battery in kWh. To do this I did:
amp hours x estimated_voltage / 1000 = kWh

I derated this value by some DoD percentage (80%, 85%, 90%) give me usable kWh, depending on how hard I wanted to be on the batteries.

My question is about that estimated_voltage value above. Originally I used 3.2V per cell "estimated voltage", because many people that sell these batteries list that as the nominal voltage (12.8/25.6/51.2 for 4/8/16S) respectively. Is that the right value? This page shows a chart where LFE spends most of its time at 3.3V. I know it's going to be complicated by the discharge current (for me this is going to be between 0.25C-0.5C in the worst case), but I'm just looking for an every number to plug in.

What got me thinking about this is that I over the weekend I did a couple of discharge tests (ended up about 0.3C) on a new 24V 200Ah LFE and I got about 4.45kWh of usable AC from a Victron MultiPlus, given the inverter overhead I saw during the discharge, that's about 4.9-5.1kWh of DC from the battery, which is more than I expected, I also had the charge voltage lower and discharge cut-off higher than the manufacturer recommended.

So, is 3.2V too conservative? Should I estimate with 3.25 or 3.3V per cell?

 

[Sojourner1]

Go by nominal voltage. 12.8 @12v, 25.6 @ 24v

Usable battery watt hour is what I think your asking?
Ah x DoD x V = Wh
100 x .80 x 12.8 = 1,024Wh

Your battery... usable @ 80% DoD
200 x .80 x 25.6 = 4,096Wh

 

[krby]

Ya, I know how to calculate Wh given Ah and V and DoD. I was really asking specifically about using 3.2V in that calculation.

So you're suggesting to use 3.2V per cell, that'll definitely be a safe estimate. But the battery spends quite a bit of time delivering power above 3.2V, so if I want something more accurate, what should I use?

 

[Craig]

If you want the exact measurement I think you would need to figure out how much energy is produced at each 1/10 volt increment and figure Wh from that.


So I just did this with data from 1 cell tested Notice that the actual Wh produced is less than listed nominal Voltage x Ah

 

[krby]

If you want the exact measurement I think you would need to figure out how much energy is produced at each 1/10 volt increment and figure Wh from that.
<... IMAGE SNIPPED ...>
So I just did this with data from 1 cell tested Notice that the actual Wh produced is less than listed nominal Voltage x Ah

Nice test! What battery chemistry is it for? Is 2.3V nominal for an LTO?

The overall point isn't lost on me, I really should count on 3.2V when estimating total capacity and derate for DoD from that.

 

[Craig]

Nice test! What battery chemistry is it for? Is 2.3V nominal for an LTO?

The overall point isn't lost on me, I really should count on 3.2V when estimating total capacity and derate for DoD from that.

Yes that's nominal for LTO

 

[DLTooley]

I just did an ~80% capacity test by voltage on a ‘dumb’ Daly bms that I don’t fully trust yet. Adjusting for the rated 90% efficiency of the inverter I got approximately 14.4v x 280ah.

These are a more likely grade A from a recent Amy/Luyuang delivery. I did not top balance instead relying on the BMS. I did trip the BMS on the initial charge using the default Victron settings.

I’ll now recharge for a more (?) accurate measure of total watt hours.

 

[JoeHam]

I would stick with 3.2v per cell to compute watt hours.

Most commercial batteries are rated that way like my GYLL and others I have seen.