Charging by PV instantly resets battery SOC to 100%

[DM23322]

I pulled my RV out of the garage for the first time to test my new solar panel install. I have tested it to this point inside only through discharging and then recharging by a 50 amp shore connection.

Last night I ran some devices and pulled the battery bank down to 67% so today I could test the PV recharging. Shortly after pulling the camper outside the battery SOC went from 67% while in the garage to 100% within 3 minutes of being in the sun. No way this can be accurate...

I'm perplexed as to why this happened and what I can do to correct it. I'm using the Victron 500A Smart Shunt and Cerbo GX. Both MPPT controllers are Victron 150/45 and I'm running 2 arrays of 215w panels each for a total of 10 panels (5 in each array). I have 8- 12V SOC 100AH LifePo4 batteries in series at 24V .

Any suggestions or comments as to what is going on here are appreciated!

 

[rmaddy]

Have you verified the battery settings in Victron Connect for the two charge controllers and the smart shunt?

 

[sunshine]

Shortly after pulling the camper outside the battery SOC went from 67% while in the garage to 100% within 3 minutes of being in the sun. No way this can be accurate

3 Minutes or 3 Hours?

 

[rmaddy]

3 Minutes or 3 Hours?

The SOC graph shown in the first attached picture seems to indicate 3 minutes.

 

[DM23322]

3 Minutes or 3 Hours?

About 3 minutes...

 

[DM23322]

Have you verified the battery settings in Victron Connect for the two charge controllers and the smart shunt?

Well, I have, but its possible I have something incorrect but don't realize it...

Here are screenshots from Solar Array 1 and Array 2. Then the shunt settings... Any revelations from these screenshots?

 

[rmaddy]

Your charge controllers are not setup for LiFePO₄. Go to Battery Preset and select "Select preset" and pick Lithium Iron Phosphate from the list. Then you can tweak the absorption and float voltages to suit your taste.

Your shunt settings seem OK but I can't access mine at the moment to compare.

 

[DM23322]

Your charge controllers are not setup for LiFePO₄. Go to Battery Preset and select "Select preset" and pick Lithium Iron Phosphate from the list. Then you can tweak the absorption and float voltages to suit your taste.

Your shunt settings seem OK but I can't access mine at the moment to compare.

Wow! Looks like I know just enough to be dangerous... I just fixed those settings... Makes me wonder what else I've missed?

Thank you

 

[sunshine]

About 3 minutes...

Wow! Looks like I know just enough to be dangerous... I just fix those settings... Makes me wonder what else I've missed?

Thank you

That maybe is just the graphical interpretation of 'instant' due to the SOC being determined by a voltage reading ( V is the only measurement possible. All other readings are configured by this...amps, ohms SOC..) and in this case it is measuring the V 'in' not the actual battery V.

A human interpretation of the other graphs would give you a better understanding of the actual SOC...15amps over 4 hours

 

[Blackhat]

https://www.victronenergy.com/upload/documents/SmartShunt/Manual_-_SmartShunt-en.pdf

The voltage in the screenshot is configured for 26.4V, and the battery graph shows the voltage crossing that threshold briefly after charging started. But current was still below the 16 Amp threshold implied by 0.04*400, and after 3 minutes the full battery reset conditions were met.

This did not happen on the shore charger because the charge current is 50A which is of course greater than 16.

The only thing that doesn't add up is that the chart shows at least 30 minutes of slight charging before the voltage crosses the 26.4 threshold. Did you maybe park the RV in the driveway and wait for the sun to hit the panels? Is this delay what you meant by "shortly after"?

To test, change charge voltage to 27.4 (equivalent to 13.8 on 12V) and try again.

 

[DM23322]

Interesting... Thanks for the suggestion.. I'll try the testing method.

My best guess...The 30 minutes prior is most likely after I opened the two 14' doors on the front of the garage. That opens up most of the entire front of the storage building to the rising sun directly in front of it. Just a thought.

 

[chrisski]

@snoobler helped set my VIctron up. Turned out it was the tail current was off by a little bit. Factory setting is something like 8%. This caused my reading to go from 85% to 100%,similar to what you’re saying.

I dialed it back to 2% and it worked.

I just updated to 24 volts, and I also needed to re-enter the new amp hours For the batteries, but kept the 2% the same, and it works with that.

 

[rmaddy]

Your shunt settings seem OK but I can't access mine at the moment to compare.

I have my settings now. The following relevant details are different:

Charged voltage: 28.3V (your 26.4V is really low. It should be 0.1V lower than your SCC absorption voltage. Mine is 28.4V)
Discharge floor: 5%
Tail current: 5.00%
Charged detection time: 3m

I think you main culprit is the "Charged voltage". Looking at your graphs the SOC shoots to 100% shortly after your battery gets to 26.4V which is how you have it set at the moment.

 

[DM23322]

I have my settings now. The following relevant details are different:

Charged voltage: 28.3V (your 26.4V is really low. It should be 0.1V lower than your SCC absorption voltage. Mine is 28.4V)
Discharge floor: 5%
Tail current: 5.00%
Charged detection time: 3m

I think you main culprit is the "Charged voltage". Looking at your graphs the SOC shoots to 100% shortly after your battery gets to 26.4V which is how you have it set at the moment.

That makes sense...

What if I changed the Charged Voltage to a higher value at 28.7. because my current absorption is set at 28.8? SOK says charged is at 14.6 per 12V so at a little under that I should be okay right?

I think I originally set the Charged Voltage at 26.4 from what I read in the Victron manual.

 

[wholybee]

Well, I have, but its possible I have something incorrect but don't realize it...

Here are screenshots from Solar Array 1 and Array 2. Then the shunt settings... Any revelations from these screenshots?

View attachment 76932View attachment 76933
View attachment 76934

That makes sense...

What if I changed the Charged Voltage to a higher value at 28.7. because my current absorption is set at 28.8? SOK says charged is at 14.6 per 12V so at a little under that I should be okay right?

I think I originally set the Charged Voltage at 26.4 from what I read in the Victron manual.

View attachment 76991

This. The charged voltage needs to be .1 or .2 below absorption voltage. Either adjust the absorption voltage down or the charged voltage up. I would adjust the absorption voltage down but it makes little difference.

I would also bump the tail current up to 5% or 6% but that isn't your issue.

 

[rmaddy]

That makes sense...

What if I changed the Charged Voltage to a higher value at 28.7. because my current absorption is set at 28.8? SOK says charged is at 14.6 per 12V so at a little under that I should be okay right?

I think I originally set the Charged Voltage at 26.4 from what I read in the Victron manual.

View attachment 76991

I have a pair of SOK batteries in series for 24V and I use an absorption of 28.4V and a charged voltage of 28.3V. You can set your absorption to just about anything in the 28.4V - 29.2V range based on the SOK specs. Just set the shunt's charged voltage to 0.1V less than your absorption and your problem with the SOC jumping to 100% will be gone.

I have the BMV-712 so I hadn't seen that section of the SmartShunt manual. I doubt the 26.4V charged voltage recommendation applies to LiFePO₄.

 

[sunshine_eggo]

I have a pair of SOK batteries in series for 24V and I use an absorption of 28.4V and a charged voltage of 28.3V. You can set your absorption to just about anything in the 28.4V - 29.2V range based on the SOK specs. Just set the shunt's charged voltage to 0.1V less than your absorption and your problem with the SOC jumping to 100% will be gone.

I have the BMV-712 so I hadn't seen that section of the SmartShunt manual. I doubt the 26.4V charged voltage recommendation applies to LiFePO₄.

Accurate for any chemistry provided that's it's float voltage. It has to do with AC charging (reliable output) vs. solar charging (inconsistent). For solar charging, one must use absorption. Setting a charged voltage as the float voltage when charging with solar will trigger false syncs.

Worth the 5 minutes:

 

[DM23322]

That video was definitely helpful... I think the problem was the charged voltage was too low as previously suggested. I'll have time to test it under full sun sometime next week.

I did enjoy how she pronounced "Parameters" as (Parameeters). That was the best part...

 

[sunshine_eggo]

As stated in the video, the battery needs to be fully charged "a couple of times a month" for accuracy. More often is better, and what you may find is that you infrequently hit the "charged" criteria, but you end up pumping enough Ah in that the unit will count back to 100%.

Your "charged" voltage near the float voltage is definitely the cause for your premature 100% syncs.

Personally, I'm really happy if my BMV shows 100% within ±45 minutes of it entering float, and with the GX/VRM, I can actually see what's happening every minute during every absorption cycle in the last 6 months, so that adds confidence.

This was after a longer than normal absorption charge (bumped up to 4 hours in lieu of an EQ cycle):





It didn't actually meet the "charged" criteria, but it counted back up to 100% based on input.