question

marv2097 avatar image
marv2097 asked

ESS Charging state misleads the SOC value

Hi,

My SOC increases even while the battery is discharging.


I have noticed that when using ESS the Inverter is always showing charging in the VBUS state similar to this post: https://community.victronenergy.com/questions/32283/ess-ve-bus-state-charging-behaviour.html

I have scheduled charge at 00:30 for 4 hours, so at 4:30 the MPII starts inverting to supply the house load. However the Charger side of things seems to continue the Bulk-Absorb-Float cycle. This causes the SOC to jump to 100% even though its been discharging for nearly an a hour. I suspect this is due to the 1 hour Absorption timer.

Is this expected operation? It doesn't always happen and is not consistent and causes different depths of discharge.


This can be seen in the graphs below with the jump in SOC happening at approx 05:30.


Many Thanks


vrm-1.jpg

ESS
vrm-1.jpg (156.6 KiB)
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

1 Answer
Alexandra avatar image
Alexandra answered ·

@marv2097

You will only be in inverting mode when you have no grid connected. You can check this by disconnecting your multi from power.

I can see though at your end of scheduled charge window the batteries are still drawing a heavy current suggesting that they need way more charge than you are giving them. Do you have solar to supplement the charge?

Basically what seems to be happening here is State of Charge Drift.

What batteries do you have? How is your battery SOC being measured?

If it is VE bus then you need to keep them fully charged for at least a day to sync. if it is the Victron BMV then you also need to make sure they have a good charge. If it is measuring its own SOC and reporting tot he GX device then it may also have had a SOC drift.

The different charge stages are usually triggered by the target voltages reached that you have set. But if there has been SOC drift then the graph will be inaccurate.

10 comments
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

marv2097 avatar image marv2097 commented ·

Hi @Alexandra

Thanks for your reply. The batteries are a DIY Nissan leaf pack which consists of 14S 6P. With their current health they are configured as approx 52.5v and 260Ah in the multi.

At the moment there is no solar, they just charge as much as they can in the off-peak 4 hour window and then discharge through the day until they reach 10% SOC. At 70amps they get to roughly 85-90% most nights.

All the SOC calculation is done by the multi and no BMV as I have no other DC sources of power or loads.

From what I can tell if the charger is still in the Bulk phase at the end of the scheduled charge window then its fine, the SOC decreases as expected. However if it makes it to Absorption stage then we have the same problem. Batteries start discharging as expected but regardless of this the SOC will always be set to 100% 1 hour after the start of Absorption even though it hasn't been actually charging during this hour.

If the batteries are being discharged then why would the multi think its correct to sync to 100% as we cant go from Absorption to Float if not actually charging?

0 Likes 0 ·
Alexandra avatar image Alexandra ♦ marv2097 commented ·

@marv2097

What bms are you using on the DIY pack?

What you may have to do (and I know it will cost more) is make your pack fully charge by letting it sit on charge untill it is no longer drawing the huge amps.

I hear what you are saying with the incorrect soc reporting but letting it get up to the full soc and sit for a few hours like that should solve the issue you are having.

And then I would recommend getting a shunt based system like the bmv for your DIY pack it works better for partial charge and discharge cycles like you are doing.

0 Likes 0 ·
Alexandra avatar image Alexandra ♦ Alexandra ♦ commented ·

@marv2097 I just wanted to further add on is usually when installing a battery we fully charge it then (discharge it at 2c or whatever the inverter can handle) down to 50% then fully charge it once more. It give the ve bus a good sync.

If you did not start from a fully charged battery the algorithm has a hard time working out what soc you are in.

Not had an issue with the systems we have down this with. Even with the diy batteries. With diy batteries we do recommend a bmv or other shunt monitor as the multi does not have a way of measuring like a shunt built it. So if you are hovering your soc like you have been doing there are bound to be issues like this, especially if you did not start at 100%

0 Likes 0 ·
marv2097 avatar image marv2097 Alexandra ♦ commented ·

@Alexandra For BMS I am using a diybms v4. Its not able to communicate with victron but I can use a relay to disable the remote input of the multi in case of a cell voltage or temp issue. I believe they are adding shunt and canbus in the future.

Its a good point you make about charging up fully. I had done a couple of full cycle tests to get a better idea of the usable capacity etc so the multi has had a see it full a few times.

I have now upped the charge window so that it will have 5 hours which should always means it gets to 100% which should mean I avoid this issue.

I think i will also look at adding a smartshunt to this setup to see if that helps.

Thanks for your help!

0 Likes 0 ·
elel avatar image elel marv2097 commented ·

Hi @marv2097

I am new to Victron product and encountered the same jump-to-100% problem similar to what you encountered. This is the related thread I created:

https://community.victronenergy.com/questions/113860/soc-jumped-to-100-while-the-battery-was-being-disc.html

Can you please help by explaining in more detail how you solve this problem? Do you mean that you set the “State of charge when Bulk finished” to 100%?

In my case,

  • the capacity of my battery pack is 277Ah
  • bulk charge (and absorption starts) ends at 55v
  • “State of charge when Bulk finished” set in VEConfig = 90%
  • “Minimum SoC” set in ESS = 10%

ESS was able to discharge my battery from 90% to 10% (i..e. around 221Ah) during day time and recharged it to 90% (i.e. up to 55V) within the 4hr window at night. However, the system resets the SoC to 100% (and changes to Float state) similar to what you encountered when the voltage reaches around 53.1 to 51.3v. However, it did not do it an hour (the scheduled absorption time) after the start of the absorption as what you encountered.

0 Likes 0 ·
nickdb avatar image nickdb ♦♦ elel commented ·

There could be many different reasons why this behaviour is occurring. Your setup and this older post may have absolutely nothing in common and could just confuse things further.

Especially since you already have a topic open for your issue.



0 Likes 0 ·
Show more comments
marv2097 avatar image marv2097 elel commented ·
Hi elel,

I didnt ever fully resolve this but managed too work around it a little. In real use limiting the minimum charge to 10% seems to provide enough buffer for this issue to not cause a problem in my setup.

It maybe more coincidence but it only seems to happen if the charge gets to 90% or more at the end of the charge window. I've not tried setting this to stop charging at 85% as I can live with it getting to nearly full most of the time.

I also think having it charge for an extra couple of hours (say on Monday morning) to make sure that its fully charged and has a good sync once per week. You could also raise the minimum to 20% but then this would limit how much capacity you have. During the summer this is not an issue as the solar handles all the charging and easily gets to 100% during the day.

I do think in my case, and maybe yours that the size of the battery bank is just on the big side of what can be charged in the 4 hour window. Another multi or a larger multi might solve that but at additional cost.

I also thought about adding a smart current shunt to see if this improved the situation. To me it seems like a bug or strange behavior in the sync on the multi and maybe the smart shunt wouldn't have that, but as yet I haven't gone down that route.

Cheers


0 Likes 0 ·
Show more comments