Looking for accurate battery state of charge table

[SilverNodashi]

Hi, 

Can anyone please supply me with an accurate battery state of charge table? The reason I'm asking, is that I have found quite a few of these charts, with conflicting values. For example, most of the charts say a 12V battery is full at 12.7V (or 12.73 / 12.79?). Some background: A remote solar powered electric gate has recently started giving problems, where it doesn't keep the charge I thought it would.

It has a GSM controller and I can query the device status via SMS and at 18:00 the battery level was at 13.00V. At 20:00 it was at 12.9V. At 00:00 it was 12.77V And this morning at 07:00 it was 12.58V (85%?). Yet, on Sunday morning the battery was too flat (10.6V) to open the gate. According to some charts, 0% is at 11.2V, others say 10.9V. 

This is on a Lead 27A Crystal battery, connected to a 50W solar panel + PWM charger. I'm am considering adding a 20W LED flood light and day/nigh camera on the gate as well and need to know, more accurately how the battery performs - for future planning. 

This led me to think, how do I know the exact level of my 48V bank (on an Axpert inverter + 3Kw PV), or our 600A/48V bank + 16Kw inverter + 9Kw PV on the farm? 

So, which chart is correct?

Where do I draw the fine line between running the batteries at 50%, or 40%? Or 60%?

For example, 

This table (common table which is used widely) shows that 50% is 12.06V: 

 

This table shows 25% is 12.06V: 

[Guest]

From my experiences, and what I have read over time, all these numbers are for batteries that are "rested".

You cannot accurately determine the SOC with volts ... on top of which, each manufacturer may have different volts for their batts.

Both these differ. One is Trojan other one generic. Both state Open circuit.

[SilverNodashi]

The samples I had are also from "rested" batteries, or so the author claims. Your examples, at least, show both to be 12.10V @ 50% or 48.4V for 48V banks. 

I do realize these figures are derived from taking the full state, which most companies claim to be 12.7V (let's use a common voltage as example), and then the flatline which is either 10.5, 10.7, 10.9 or 11.1, and then calculate 10 steps (0%, 10%, 20%, etc) or 100 steps (for more accurate readings). But are the empty values guessed, or somehow accurately measured?

[Guest]

Measured in a lab I presume.