Battery soc drift

[Tariq]

I have a Deye inverter with a Volta battery, the cut off for soc is set at 20% for the 0100 to 0500 time slot, however, the battery keeps draining below 20%, any ideas as to the cause

[TaliaB]

52 minutes ago, Tariq said:

I have a Deye inverter with a Volta battery, the cut off for soc is set at 20% for the 0100 to 0500 time slot, however, the battery keeps draining below 20%, any ideas as to the cause

I won't blame the inverter as it gets the data from the battery bms. Soc drift is a complex problem in lfp. It all depends on the complexity of the bms. For ev applications the soc is much more important hence more expensive bms than for domestic or industrial applications.

Open Circuit Voltage (OCV) is a measurement of a battery cell’s voltage at a known state of charge when at resting equilibrium. The OCV curve is the mapping of the OCV from 0-100% SoC. A simple but imprecise method of estimating State of Charge is to use the manufacturer-defined relationship between voltage and charge level to look up the SoC based on a measured OCV. OCV can also be used to improve other charge-tracking methods like Coulomb Counting. Both techniques can be unreliable in many cell chemistries, especially Lithium Iron Phosphate (LFP), where the OCV relationship to charge is flat and there are significant hysteresis effects. It also fails to take into account other factors such as age, temperature, and time since the cell has last been used.
SoC estimation problems are well documented within LFP cells. Today's most performant SoC algorithms combine coulomb counting with temperature and voltage-based compensation. Coulomb counting cannot be used effectively alone because the inevitable errors in each measurement accumulate over time until the SoC estimate is unrelated to the state of the battery. What’s more, as cells age, they lose capacity, creating additional error's because coulomb counting cannot track the changing health of the cell. Voltage compensation can be effective in correcting for both coulomb counting drift and capacity fade, but the strategy breaks down quickly for LFP. The coulomb counting with voltage compensation approach can still yield SoC estimate errors over 20% in LFP, even on brand-new batteries operating at normal temperatures.

[zsde]

Are you able to access the BMS data of the battery, perhaps with PBMS tools or an equivalent software?
Perhaps the BMS settings themselves may not be set to shut down at 20%? 
 

[Bobster.]

3 hours ago, Tariq said:

I have a Deye inverter with a Volta battery, the cut off for soc is set at 20% for the 0100 to 0500 time slot, however, the battery keeps draining below 20%, any ideas as to the cause

What's your grid voltage like? I had a case similar to what you describe, and it turned out to be low voltage. The inverter would disconnect if the voltage dropped too low. In the day when there was solar power the inverter drew from PV, but at night it could only draw battery. 

[TaliaB]

@Bobster.it is a possibility but for Sunsynk and Deye F42 "Ac line low voltage" would  come up on fault code.

@Tariq see if F42 appears on your fault code list.

[Bobster.]

51 minutes ago, TaliaB said:

@Bobster.it is a possibility but for Sunsynk and Deye F42 "Ac line low voltage" would  come up on fault code.

@Tariq see if F42 appears on your fault code list.

You are right. The other thing I learned from the episode I talk about was to start checking error logs. 

[macafrican]

The BMS prob has in its literature what its factory setting protection shutdown is, sounds like you hit that when inverter thinks it is not there yet.    It might be a math and reporting issue?   Some inverters show you what they calculate SoC at based on charge and discharge numbers, in others the BMS knows exactly per cel, module and bank and tells inverter what SoC is.   To resolve, maybe set at 30% for a while and work your way down over couple of days, and see whether still get an unplanned shutdown.   Eventually you’ll figure out that for example 23.5% setting keeps you above protection shutdown.

If your BMS data is accessible, can you see millivolt max and minimum per cel?    In stable state, that should be small range between the highest and lowest cell.    Eg 40 millivolt.   If you see a range in the hundreds of millivolt, you may have cel problems that vendor must sort out.

What is your max charge and discharge set at?    Despite claims by some batteries I stick to ½ C