Hi I was wondering what or who sets the state of charge (SOC) of the batteries
State of Charge is a calculated number between 0 and 100%, and is always just an estimate.
For lead-acid batteries, there is a fairly reasonable conversion from measured rest voltage to state of charge. For LiFePO4 batteries, there isn’t such a great correspondence.
The State of Charge can be estimated by some battery BMSes, particularly for LiFePO4 batteries. It isn’t typically a very good estimate, and needs to be reset frequently by charging batteries to 100% occasionally.
The preferred Victron approach is with a shunt in-line with the battery negative terminal. It measures Coulombs (charge in and out) and works well with both lead-acid and lithium batteries.
The 500A SmartShunt is what I use in my system to measure State of Charge. Inverters are able to generate an estimate of State of Charge estimate as well, but this is not as accurate as using a shunt.
2 Likes
In my view and experience, information about the SoC from a BMS is at best a rough estimate. And that’s putting it positively and diplomatically. Typically, the SoC values ​​from a BMS cause more confusion than clarity. Small charging or discharging currents aren’t recorded at all, causing the values ​​to become completely unrealistic within a very short time. I operate one shunt for several storage blocks in my system. While the shunt records all charging currents to the battery and all discharging currents from the battery, the currents are divided among the individual blocks and their BMSs and are sometimes completely ignored. This leads to deviations in the SoC calculations of 20-30% within 3-4 cycles. That’s just crap. Forget this miserable SoC calculation by the BMS.
@Sarowe1990 Your comments should be prefaced by the warning that you are talking about the integrated BMS boards often found in the cheaper end of the drop in batteries and many people have experienced this and I agree with your view.
However, once you are talking about the higher end of the lithium battery / BMS spectrum then your generalisation no longer apllies. The Victron Lynx BMSs for example include a shunt in the BMS providing completely reliable SOC calculations
Yes, but that’s an external BMS.
Try to do an imaginary Victron AIO battery (cells, balancers, BMS) and nobody will buy it because of the price.
Somewhere something must give up…
For most of the cases, where the charging/discharging currents are over 2A, the actual, non Victron BMSes are quite OK.
I am just adding this in case people who DO NOT understand the nuances see this and take it wrongly, plenty of people out their with higher end BMSs given this is a Victron forum.
I think that 90%? 95%? 99%? of all these SoC discussions are not about a Victron BMS. I have been following the entire discussion in various forums for several years now. This is precisely the reason why I speak out so clearly. If I discredit individual products, then I am sorry. But I think it applies to the vast majority. A BMS is simply not designed to collect any data for other applications. It is designed and built as a backup instrument to protect individual battery cells. It disconnects the battery from the rest of the system when critical parameters are present. This is what BMSs are designed and built for, and most of them do this very reliably. I would like to give another small analogy. A car is designed to transport people from A to B. Even if you can use it to transport certain loads, no one would think of using a car permanently to transport large loads or to use it in agriculture.
Once again you are 99+% correct, I am not disagreeing with you, rather just stating that when you come onto a specific product forum that offers something that is designed to overcome the issues you raise, then you can not state all. If you were on a generic forum I would have let it ride.
To take your analogy, cars can not transport heavy loads, but not all vehicles are limited to light loads.