Scott,

Thanks for taking the time to reply.  I have read your posts with great interest and they have definitely helped me in planning and implementing my own system which shares many of the same components that you chose (i.e. I have lots of blue Victron stuff in my SM!).  

To be clear, I'm not saying that what the owner of MV Tanglewood is doing is the "correct" way to do it.  It certainly is not what most LiFePO4 manufacturers recommend.  I was simply stating that there does appear to be an observable and actionable correlation between voltage and State of Charge (SOC) with LiFePO4 batteries.  I totally agree that there is not a lot of "resolution" in the middle of the curve but there is enough resolution at the tail ends of the curve to take action before damage is done on either end of the curve.

Regarding the data posted that is posted here: https://www.powerstream.com/lithium-phosphate-charge-voltage.htm  you can see from the description of the test above the data that the test was done using very small batteries (less than 2.5 Ah capacity) and was done at over a 1C discharge rate (2.5 Amp discharge on 2.2 to 2.4 Ah batteries).  That is an extremely high discharge rate that is not really indicative of how we generally use our battery banks on our boats.  My normal "sitting at anchor" discharge rate is closer to 0.01C and I can even run one of the Air Conditioner units on the boat and still stay under 0.1C.  I think this high discharge rate has a significant impact on the Voltage vs. SOC results.

The data posted at the bottom of this article is from a test that I believe is much more indicative of how we use our battery banks:  https://marinehowto.com/lifepo4-batteries-on-boats/ and even this test is done at a relatively high discharge rate of about 0.075C.  Here is the chart of Voltage vs. Ah Discharged on the  772nd(!) Cycle on his ~11 year old LiFePO4 battery bank: https://marinehowto.com/wp-content/uploads/2015/09/34-LiFePO4-On-Boats.png


  


It's impossible to read the chart above but you can see the details in the links I that posted above.  I took the data from this 0.075C discharge test to get the following approximate Voltage vs. State of Charge %:

At 3.15v per cell or (3.15*4/8 = 12.6v/25.2v) you are at around ~15-18% SOC.
At 3.2v per cell or (3.2*4/8 = 12.8v/25.6v) you are at around ~30% SOC
At 3.3v per cell or (3.3*4/8 = 13.2v/26.4v) you are at around ~95% SOC
At 3.4v per cell or (3.4*4/8= 13.6v/27.2v) you are at around ~99-100% SOC

So on our nominal 24v boats there is a 1.2v difference from ~95%SOC to 15-18% SOC (26.4v-25.2v=1.2v).  That's not a huge difference, but it is easy to observe and use in decision making.

Other things to note in those test results:

• He charges the bank to only 13.8v (3.45v per cell) with a 7.5amp tail current.  That's 27.6v on our 24v boat and is well below what most LiFePO4 battery manufacturers recommend. 
• The batteries bank reaches only 13.5v (3.37v per cell) at the start of the test with those charging parameters
• The voltage drops from 13.5v to 13.26v (3.32v per cell) after discharging only 6 Ah
• This data is an eleven year old bank of LiFePO4 cells that has been in use on a sailboat and has had 773 cycles most to 80% depth of discharge or more.
• This bank was rated at 400Ah when it was new but eleven years later still has MORE capacity than it was rated for when new.

Based on all of the above, Tanglewood's seemingly relatively simplistic scheme of charging to 3.45v per cell (but apparently only reaching 3.35v per cell or 26.8v on a 24v system), discharging normally, then charging again when he reaches 3.15v per cell (25.2v for Amels) seems pretty reasonable as it would cycle his batteries from 95% SOC to 15=18% SOC.  Only time will tell!  Hopefully he will continue to post his findings, good or bad. 

--
Mark McGovern
SM #440 Cara
Deale, MD USA