question

Hi, to measure the capacity of a battery, don’t look at SOC. Not to see if its full, not to see if its empty, and not to see what its capacity is either

I would:

1) fully charge the battery. Fully charge means continue charging until it doesn’t accept any significant current anymore while voltage is at absorption level.

2) write down value of kWh discharged meter in BMV.

3) discharge it with approx C2 current (= 1/5th of rated capacity) until empty. Ie voltage at 10V orso.

4) compare new value of kWh discharged meter with old value.

Ps. to measure it in amps, look at what amperage you draw and multiply with amount of hours it took to discharge it.

Ps. once empty, charge it: never leave a lead battery in discharged state; as doing that will reduce its life time due to sulfation

I imagine another procedure ( lead batteries useful capacity.pdf), as completely discharging is not significant for lead batteries that are used at about 1/3 of their nominal capacity.

Contribution welcomed

Hi everybody,,
The preceding procedure is biaised as it uses a SOC value based on the maximum battery capacity which is what I want t0 evaluate. I shall study whether Voltage would not be better as electrolyte density is difficult to measure

Futher reflexion...

To evaluate the power drawn from of a lead battery at a given SOC (66% for me), measuring Amps and Volts give AH to be compared to the nominal capacity. But this nominal capaciy diminishes as the battery gets older and this method is not accurate.

Using the voltage could be a solution, but it depends upon the SOC and the Amps drawn during the test. All the more, it varies very slowly at the beginning of the discharge: from 12.67 to 12.53V on Victron's graphic for SOC from 100% to 66%. One could imagine an arbitrary protocol down to a voltage corresponding at SOC=66% on a new battery. Yearly comparison would have to be done with exactly the same material and a very accurate voltmeter.

This procedure could be supplemented by periodical measures of the electrolyte density to establish a relation and possibly a better estimation of the 66% SOC. One can hope that the density varies simply.

For an important useful capacity (10kW for me at 33% discharge at nominal capacity), with current use from 367 to 1118W per hour, discharge can be done with a 1000W (or a little less) resistor which will take about 10 hours. Computing will use the mean value of V and I between beginning and end of the period.

These measures cannot help estimate the total capacity which, anyway, is not significant to a common user. Should there be some kind of linear correlation, one could use the former protocol (stopping the test at 66% SOC) to modify proportionally the total Ah value set in the machine, and thence have a better estimation of the useful capacity.

I shall try all these next spring when I can reach my home, presently under 50cm of snow!

Comments welcome!

last version of my protocol.

Contributions welcome. It is NOT a personal affair!Batterie au plomb CAPACITE UTILE V6.pdf

Syncronizing in the BMV manuel is not clearly defined: does it set a new capacity in BMV's memory or only zeroe the variables that will be used after to measure the output?

Hmmm!

How is SOC calculated? i.e. what is the capacity used, certainly not the given capacity introduced in BMV!

Otherwise, Iwas absolutely wrong with my computations:

- correcting SOC with realcapaacity/nominal capacity: should be the reverse nominal/real!

- As SOC is NOT calculated from the nominal capacity no hypothesis stands! Waiting for some accurate SOC calculation to go on!