owen-spa avatar image
owen-spa asked

An Alarm for SOC Mismatch with Battery Voltage?

The following is a question if such an alarm is desirable, feasible, and practical:

As state of charge readouts have become more and more common, the rate of people reporting problems with accuracy is increasing and I regularly come across the situation where a customer believes their battery bank of some years is fine due to reported state of charge being only 80 to 90% minimum but after assessing their battery voltage dynamic it's clear they are taking the battery into real 0% SOC territory, often daily, and unknowingly damaging the battery. Typically the battery monitor has been setup properly but something else has changed, most commonly due to aging and use the battery has lost capacity so the SOC values show as higher than they really are.

Originally I had thought surely there's a way that battery monitors could recalibrate and set themselves to the real capacity but this gets complicated... what I do think could be a real improvement and put forward for consideration is an alarm when the SOC and battery voltage seem too mismatched. Obviously the values would need to be considered and refined but the logic could work something like this:

If SOC >60% and Vbat < 0.96Vnom for 10mins with current of <3% of Bcap then report "SOC Mismatch - Please synchronise battery monitor or test battery capacity"

I think these conditions are clearly incorrect enough to avoid nuisance alarm, apply to most battery chemistries, and could prompt people to check out their system before further damage is done.

In the immediate era, this applies most to older lead acid installations, but as the large number of new lithium installs gradually lose capacity, the potential mismatch presents a significant threat where decisions based on calculated SOC may not come in at the right time, risking hitting very low real SOC or BMS cutout.

BMV Battery MonitorLithium BatterySOClead-acid
2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

owen-spa avatar image owen-spa commented ·

Perhaps this isn't really the place for such a suggestion @Guy Stewart (Victron Community Manager)?

0 Likes 0 ·
Guy Stewart (Victron Community Manager) avatar image Guy Stewart (Victron Community Manager) ♦♦ owen-spa commented ·

Hi @Owen SPA

It's a good suggestion and I like your thinking.

I have a list of 'machine learning' possibilities where there is some characteristics of a system that usually indicate an issue, or at least notice should be given to the user.

Finding the right opportunity in the product feature roadmap schedule can take an indefinite amount of time, and is usually a very competitive process.

I'll add this thread to that list, and I hope one day it is included (along with the many others I also have in mind!).

Thanks, it's a good idea in my opinion.

1 Like 1 ·
owen-spa avatar image owen-spa Guy Stewart (Victron Community Manager) ♦♦ commented ·
Thanks very much Guy. I have no doubt the list of ideas vying for incorporation is long an competitive!

Technical complications aside, I think the scope for inaccuracy in reported SOC is a real weak spot in modern battery electronics. Here's hoping both our ideas, including your progressive sounding improvements, find their way in and make a difference for the better :)

1 Like 1 ·
snoobler avatar image snoobler Guy Stewart (Victron Community Manager) ♦♦ commented ·
I think this is also an opportunity to bring calculated internal resistance into the mix. IR trending is a great indicator of health/deterioration.
0 Likes 0 ·
Masquelier Pierre avatar image Masquelier Pierre owen-spa commented ·

Hi, I solved this SOC issue with an assistant who garantes a minimum limit do not destroy the batterie. (2.1V/cell * 24 cells for exemple in my case. I want minimal residual energy in the batterie even in C5 or C10 discharge!)

In my case, the charge of the batterie is only made by solar energy (MPPT), and the 100% resynchronization is done at each cloud passage during the absorption phase and makes believe that the battery is full when not (critical). The SOC degrades every day and all settings have been tried as explained in the Victron's recommendations with no positive result. It is the principle of resynchronization that is in question in certain cases and should systematically be doubled by voltage monitoring with the possibility of activating something (i.e. a relay, an alarm, etc.) Voltage is a very good indicator of empty battery in all cases.
An alarm will be good but not suffisent!.
Solution: I put ON the ACin relay with general flag when the batterie VOLTAGE or SOC is less than a critical value. Then the grid can put ON the Easy-Solar's charger. it is very simple to make a double check with SOC and VOLTAGE. To stop the charger; the ACin relay is reset when the voltage reach the floating voltage + 0.1V during 15mn. The values are set for a Lead batterie and be extended to other batteries. The SOC test can be put at any limit you want. It would be nice to implement this function in other native algorithms without add an assistant and to leave the choice of an action in case of low battery voltage rather than stopping the inverter.


0 Likes 0 ·
0 Answers

Related Resources

Additional resources still need to be added for this topic

Victron BMV battery monitors product page

Additional resources still need to be added for this topic

Battery Compatibility

Did You Know - How to create a battery profile for non-Victron batteries?