Hello there!
I use SmartShunt 300A with a 12V 50Ah LiFePO4 Lithium Batterie from eco-worthy to run an outdoor drinking water station (10A would be more than enough for my use case but that was the smallest shunt I found from Victron). The energy consumtion is pretty low, mostly standby and rare usage. I have to recharge the battery manually every 6-10 weeks. In winter I noticed that when storing the system for ~2 months unused, it lost 7Ah but the shunt showed 100 % SOC all the time.
For such use cases where batteries are not recharged very often, I wonder:
- Is the self-use energy of the smart shunt metered and accounted for?
- Is there a way to factor in physical discharges or discharges from the BMS that cannot be measured by some sort of config like x% per day?
How much power does your water station draw? The Victron Smartshunt doesn’t measure very small currents. In the Smatshunts setting the default value is 0.1A. All smaller currents are calculated as 0.
Actually, internally it does upto 24bit resolution. And with that setting set to zero (0) I suppose it will.
..
Re OP questions:
- The smartshunt standby losses are very small, and likely to be compensated as well. They are noted on the specification.
- Battery internal BMS standby losses should be small but balancing losses can be substantial over time. These cannot be factored in by the smartshunt.
I see a current alternating between 0 and 0.01A in the app. So I assume the standby usage is in theorder of magnitude of 10mA. I did not check yet if this is from my application only or application plus shunt.
and likely to be compensated as well
I.e. it is measured? Or we don’t know?
The digital resulution has nothing to do with that. All measuring instruments are less accurate at the extremes of their measurement range than they are in the middle. Furthermore, the regulation of a PV system inevitably causes current and voltage to fluctuate slightly around the control point. To ensure that this unavoidable behavior does not become a source of error in SOC determination, the Victron SmartShunt features a setting for a threshold value, below which currents are no longer recorded.
These currents fall below the default value in the Victron SmartShunt settings. I strongly doubt that changing the settings here would make any difference. The Victron SmartShunt datasheet specifies a measurement accuracy of ±0.01 A. Consequently, measured values ​​between 0 and 0.01 A are entirely random and bear no relation to reality.
Many laypeople regard the readings from electrical measuring instruments as the absolute truth. However, this is not the case; with such low currents, we face multiple sources of error that preclude any meaningful determination of the SOC. The issue here is not merely that the SmartShunt is unable to accurately capture such values, but also that these minute current flows are themselves a product of the control logic within a Victron system—and are therefore subject to inherent, random control deviations.
The only solution in this scenario is to regularly fully charge the battery, thereby recalibrating the SOC to 100%. Yet, even without this step, the battery voltage—given such low currents—is unlikely to deviate significantly from its open-circuit voltage; consequently, at least at the upper and lower extremes of the battery’s voltage curve, a rough estimation of the SOC based on voltage remains possible. Such an estimation is unlikely to be any less accurate than the SOC calculation derived by the SmartShunt from such low current readings.
I see, thank you. I am certainly not expecting accurate values here. I am just trying to get an estimation that is better than blindly swapping and charging the battery every 6 weeks. Looking into the voltage is a good idea, thanks.
I assume a real soulution would require a 30x smaller shunt (10A for charging, 3A max load). Is this something Victron or other vendors build? I really like the Victron system with bluetooth and the app.
I doubt that. Technically, the real solution would be to mount the shunt before the BMS in the negative path, so that it can account for the BMS energy use.. But that requires modifying the battery. The thread I linked to, has more details. As does the search button, this topic has been discussed before.
Okay I now measured my application. It has a standby current of 13mA which matches the 0.01A the shunt measures. I am targeting a 100 day recharge cycle with that (killing a power LED reduces the standby to 8mA, so that could work).
Given that this use case cannot be served well by a shunt which sits behind the BMS as you explained, I am migrating to the Smart Battery Sense which allows me to check the voltage every now and then. As I only really need a _low battery_ indicator this is good enough. Also the Smart Battery Sense comsumes very little evergy (0-2mA, depending on the bluetooth connection status) which helps reach that 100 day target.