Battery BMS not equal to Victron Smart shunt/Cerbo GX

[jerrypeters]

Good morning All. I have a battery monitoring quandary.

I have 2 Kilovault 3600HLX batteries installed in a parallel 600AH house bank. The system is measured through a Victron Smart Shunt 500AMP and further with a Cerbo GX display. The issue is that the shunt consistently records more load than the load reported on the battery BMS through their Bluetooth app. With my loads each day the total SOC difference between the battery bms and the shunt is about 10%. So after maybe a week the bms is showing 99% SOC when the shunt is showing 20%. At this point (approximately) the shunt resets itself to 100% and the process continues.

I would be tempted to believe the shunt but in reality the battery bms is correct. I believe the battery BMS is correct because the voltage is consistent with a 99% SOC and not a 20% SOC for instance and my solar controllers are behaving normally - charging bulk, absorption and float in accordance with the programmed voltage/current profile.

Any insights?

[MerMike]

It's difficult if not impossible to determine SOC from voltage in an LFP battery due to the very flat charge/discharge curve.

What peukert exponent and charge efficiency settings are you using in the shunt? Kilovault recommended 1.05 and 95% IIRC.

[Zanshin]

The 500A shunts are usually quite accurate, and I have the SmartShunt with a CerboGX as well and it works accurately (see Zanshin Victron Installation). Do you have a clamp ammeter or have a neighbour with one? I would clamp it around the lead on either side of the shunt and compare that reading with what the Victron is showing.

If the two numbers match, then I suspect you have a consumer of DC power wired in between the battery bank and the shunt. I'd also recommend doing this while charging as well.

Do you have the BMV-172 monitor attached? If so, could you do a screenshot of the "Battery Settings" page from the mobile app? Also check the "Misc" option to ensure that you have the 500A/50mV option selected for the shunt

[more]

Originally Posted by MerMike

It's difficult if not impossible to determine SOC from voltage in an LFP battery due to the very flat charge/discharge curve.

What peukert exponent and charge efficiency settings are you using in the shunt? Kilovault recommended 1.05 and 95% IIRC.

not thru

13-4 o 13,6 V no current flow or 1-2A 100%
under 13v 0-20% capacity

almost always (99%) BMS SOC is valid victron and other not
victron shunt to many data

[jerrypeters]

Thank you all for your feedback and ideas.

I understand voltage is not a reliable indicator of SOC but the difference between 20% and 99% SOC should show a voltage difference of almost 1 volt. Also the solar controllers are behaving normally with respect to voltage and current i.e. the solar controllers "believe" the battery bms SOC as indicated by their response to bulk, absorb and float settings.

The shunt peukert setting is 1.05 and the charge efficiency setting is 99%.

I do not have a BMV-172 installed.

If I had a DC load between the battery and the shunt, the shunt would not show that load and the shunt load would be less than the battery bms load. The opposite is true. I do have a clamp ammeter and I will try to measure against the shunt report. It's not easy to access and I'd prefer to do it without solar input - which is easiest at night. Perhaps I can measure tonight. (Unfortunately the shunt is under one of my guest's bunk )

[Capt.Don]

I have exactly the same situation -- my BMV 702 and KV app SOC differ by more than 10%. Kilovault support suggested I cycle the battery by fully charging to 100% then fully discharging, then bringing back to full charge. Made no difference in the SOC readings.

I don't believe I have phantom loads on the battery that are not connected to the common ground.

My BMV settings: charge voltage 14.1v, peukert is 1.05, charge efficiency 95%, tail current 4%.

I'm inclined to believe the BMV 702 SOC...
Don

[MerMike]

Originally Posted by more

not thru

13-4 o 13,6 V no current flow or 1-2A 100%
under 13v 0-20% capacity

almost always (99%) BMS SOC is valid victron and other not
victron shunt to many data

I think you meant "not true", which in fact it is, as evidenced by the data you provided. What is the voltage for an SOC between 20% and 80%? You can't say. Nobody can. If you think otherwise you have a poor understanding of how batteries and shunts work.

Having installed many Kilovaults, I can also say with certainty that the SOC as reported by the battery BMS is no more or lass accurate than any other measurement. Kilovault recommended the occasional full charge and discharge to get the BMS SOC back to a known state. They ALL drift.

[Capt.Don]

Here's the Kilovault charge/discharge voltages vs. SOC. I've watched the charge cycle sitting at about 13.5v till 90% SOC, where voltages rapidly rise to 14.1v at 100% SOC.

[MerMike]

Originally Posted by Capt.Don

Here's the Kilovault charge/discharge voltages vs. SOC. I've watched the charge cycle sitting at about 13.5v till 90% SOC, where voltages rapidly rise to 14.1v at 100% SOC.

The voltage curve will change over time with charge/discharge cycles. The curve is for a new battery in lab conditions. Your batteries aren't new and aren't in the lab.

Point is, there's no SOC measurement that is accurate enough to say that a battery is, for example, at a 52% SOC. It might be 42%. It might be 62%. The only real adjustment to be made is in the expectation that a battery monitor can provide an accurate SOC.

What a battery monitor CAN provide accurately is how many coulombs have gone in and out of the battery. Where the drift comes in is when the monitor adjusts for temperature, charge efficiency and peukert exponent. Now you're back to not knowing whether the conditions match the adjustments, and you no longer know what's in the battery. You just know a range.

[MerMike]

Originally Posted by jerrypeters

I would be tempted to believe the shunt but in reality the battery bms is correct. I believe the battery BMS is correct because the voltage is consistent with a 99% SOC and not a 20% SOC for instance and my solar controllers are behaving normally - charging bulk, absorption and float in accordance with the programmed voltage/current profile.

Discharge the bank and charge it to full. Assuming you set your charge sources to 14.1V absorption, set the Charged Voltage setting in the shunt to 14.0, or 0.1v below the absorption value. The monitor in the shunt will reset to 100% just before it reaches max absorption.

The SOC readings in the batteries and the shunt should then agree, at least for a while.

[more]

Originally Posted by MerMike

I think you meant "not true", which in fact it is, as evidenced by the data you provided. What is the voltage for an SOC between 20% and 80%? You can't say. Nobody can. If you think otherwise you have a poor understanding of how batteries and shunts work.

Having installed many Kilovaults, I can also say with certainty that the SOC as reported by the battery BMS is no more or lass accurate than any other measurement. Kilovault recommended the occasional full charge and discharge to get the BMS SOC back to a known state. They ALL drift.

i don't want install victron shunt. people call me from 100-300 km farr and my battery is empty what to do .
step down press left button what voltage,press middle what A,
13.4 25A+ ok press last button keep 10 second. voila battery full
yes middle day battery charging who care. i dont drive 200 km.
after waranty first send bill 2 hour x160€ ( maybe i rise hour BMW raise to 320€ i must check) and do some think.

in life not important what lcd show,important is how many energy is in battery. Victron have bug in shunt
simply must put
if voltage 13,4 if current 0 to 5 A battery is 99%

[jerrypeters]

I understand the voltage vs SOC comments and I
also understand that SOC is relative and drifts over time and needs to be reset periodically - preferably by discharge/recharge, etc.

Nonetheless right now I have a 99% SOC on the batteries and a voltage of 13.4. I have -.1 amp drain as recorded by the battery BMS.

The solar controllers are happily in float mode and producing minimal current despite bright sunlight. They are sensing "full" SOC by their settings.

OTOH the shunt is showing 66% SOC and -2.77AMP current.

The shunt and battery BMS were both at 100% SOC 3 days ago.

[more]

Originally Posted by jerrypeters

I understand the voltage vs SOC comments and I
also understand that SOC is relative and drifts over time and needs to be reset periodically - preferably by discharge/recharge, etc.

Nonetheless right now I have a 99% SOC on the batteries and a voltage of 13.4. I have -.1 amp drain as recorded by the battery BMS.

The solar controllers are happily in float mode and producing minimal current despite bright sunlight. They are sensing "full" SOC by their settings.

OTOH the shunt is showing 66% SOC and -2.77AMP current.

The shunt and battery BMS were both at 100% SOC 3 days ago.

reset shunt to factory and make new setting again, but 95% shunt is gone. i simply dont know why victron dont make clamp current meter

[Zanshin]

@jerrypeters - If you do measure the actual amps with an ammeter at the shunt and it doesn't match what the Victron is showing, you can adjust the sensitivity in the application so that it matches the actual amperage. I don't think that this is likely, as those 50mV shunts are usually quite accurate.

As you've seen by the responses in this thread, battery state-of-charge is almost as contentious a subject as mono vs. multi or what anchor to use!

DC electrics is not magic and the only way to diagnose is to measure coulombs going in and out, and monitors such as the Xantrex and Victron generally do a good job in that regard. Voltage is tough subject, because the published numbers always assume a battery "at rest" and on a boat you need those batteries for lights, refrigeration, etc. So the voltage isn't going to be a good indicator. And you have such a large discrepancy in amps that something aboard would be glowing red-hot because the energy has to go somewhere.

Could you post a screenshot of your SmartShunt battery settings? Did you enable the DVCC on the Cerbo GX? If so, could you post those settings as well? That might help to analyze once you've either confirmed that the SmartShunt is reporting the real amps or that it isn't.

[smac999]

Originally Posted by jerrypeters

I understand the voltage vs SOC comments and I
also understand that SOC is relative and drifts over time and needs to be reset periodically - preferably by discharge/recharge, etc.

Nonetheless right now I have a 99% SOC on the batteries and a voltage of 13.4. I have -.1 amp drain as recorded by the battery BMS.

The solar controllers are happily in float mode and producing minimal current despite bright sunlight. They are sensing "full" SOC by their settings.

OTOH the shunt is showing 66% SOC and -2.77AMP current.

The shunt and battery BMS were both at 100% SOC 3 days ago.

Measure with a clamp meter. You can zero calibration the shunt if needed. But I have install 100.’s and never had one off that needed calibration. I’d bet it’s wired wrong and you have a wire on the battery side of shunt.