question

thanar avatar image
thanar asked

BUG report: SoC jumps to "SoC when Bulk finished" value, way before Bulk has finished.

ESS system: MultiPlus-II 5kW, SmartSolar 250/60, 4kWp P-V, Cerbo GX, LFP 10kWh battery (no BMS connection to the GX device at the moment). No BMV of SmartShunt, SoC calculation based on MultiPlus, enabled through VE configure.

So, the battery was drawn to 49% during the night and early morning. Sunny day since, and the solar is ramping up. SoC and battery voltage gradually rising as expected. Bulk voltage is set at 55.2V and float voltage at 53.6V.

At the exact moment SoC reaches 73%, while the battery voltage is still below the bulk target voltage (it reached a maximum of 54.81V according to the controller), SoC percentage jumps to 95%, which is the "percent when bulk finished" value. However, the system is still on Bulk charge, and will be for a couple of hours more. Actual battery percentage is at 74%. SoC will remain at 95% until bulk charge is indeed complete and then SoC will slowly ramp to 100% during Absorption.

This has been kind of happening since the beginning. Any ideas why? Any insights? Here's the link to this installation, in case someone wants to have a look: https://vrm.victronenergy.com/installation/278343/share/f53e777c

Looks like the jump happens as soon as battery charging current reaches the maximum value I've set on DVCC, which is 45A, and I did that to keep the controller working a little bit less. DVCC was enabled on this system with no BMS-connected battery, purposely to be able to ramp down charging current, so as to lift the burden off the controller a bit.

Chiming @mvader (Victron Energy)

Wouldn't it make much more sense if the SoC calculation followed the same rules as the SmartShunt?

screen-shot-2023-04-27-at-105343.pngimg-9669.png


battery chargingESSLithium BatterySOCbug report
screen-shot-2023-04-27-at-105343.png (116.5 KiB)
img-9669.png (248.5 KiB)
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.

thanar avatar image thanar commented ·

Today is the next day. I have disabled Current Limit through DVCC and the controller would give out the full 60A to the battery. However, as you can see from the attached screenshot, SoC jumped again from 68% to 95% when the battery current hit 52.4A, and the battery voltage was steady at 54.8V.

Really can't figure this one out...

screen-shot-2023-04-28-at-131804.png

0 Likes 0 ·
screen-shot-2023-04-28-at-131804.png (124.4 KiB)
thanar avatar image thanar commented ·

Day 3, exact same behaviour... Once battery voltage hits 54.78V, SoC jumps to the "bulk finished" value. As stated on the OP, bulk voltage is set at 55.2V, and this is definitely visible by reading the rest of the attached graphs.

screenshot-2023-04-29-at-202205.png

I intend to be setting up a bunch of identical systems, and would like to know the answer, when a client ask "why the SoC jumps every day like that?"

0 Likes 0 ·
screenshot-2023-04-29-at-202205.png (197.0 KiB)
thanar avatar image thanar commented ·

I have not set "State of charge when Bulk finished" to 99% and the SoC still jumps to 99% as soon as battery voltage reaches 54.78V. Like clockwork. System state is still in Bulk, 1.5 hour later, and it will be (as it should be) until battery voltage reaches the Bulk voltage, which is set to 55.2V.

screen-shot-2023-05-05-at-132124.png

0 Likes 0 ·
screen-shot-2023-05-05-at-132124.png (156.0 KiB)
thanar avatar image thanar commented ·

It was quite cloudy lately, so I didn't have the chance to experience the jump for a few days, since the battery was running low.

Today we got some proper sun, so this time the jump was from 55.3% to 90% (new "SoC when Bulk finished" setting). At the time of the jump, the voltage was maximum 54.8V, while the "SoC when Bulk finished" value as set on the system is still 55.2V.

As usual, the jump didn't set the system to absorption mode; this happened 1.2h later, stayed on absorption for 1h and then the SoC was reset (as it was supposed to) to 100%.

Please have a look at the attached screenshot.

screenshot-2023-05-12-at-175153.png

0 Likes 0 ·
screenshot-2023-05-12-at-175153.png (269.5 KiB)
thanar avatar image thanar commented ·

Today the jump to 90% took place at 54.34V, it really doesn't make any sense...

screenshot-2023-05-13-at-123945.png

0 Likes 0 ·
screenshot-2023-05-13-at-123945.png (224.8 KiB)
klim8skeptic avatar image klim8skeptic ♦ thanar commented ·

@thanar have you enabled "use solar charger current to improve VE bus SOC"?

ac-input-4.png


0 Likes 0 ·
ac-input-4.png (70.1 KiB)
thanar avatar image thanar klim8skeptic ♦ commented ·
This option is not available on my setup and according to this thread, it was removed over two years ago.
0 Likes 0 ·
klim8skeptic avatar image klim8skeptic ♦ thanar commented ·

@thanar Still available for me, how odd. v2.94

settings > system setup > system status > enable.

Edit. looks like it is for only for showing status, not enabling.

0 Likes 0 ·
thanar avatar image thanar klim8skeptic ♦ commented ·

Even on the official Cerbo manual it says it's a "diagnostic tab", there is no way to "enable" it. However, I'm also on v2.94 on the Cerbo, and I don't even have a "System Status" menu under "System Setup", although the official manual mentions that option. Maybe disabled for ESS systems? I can't tell, got to love Victron's interfaces and documentation, it's kind of a complete mess sometimes...

0 Likes 0 ·
thanar avatar image thanar commented ·

I've adjusted "SoC when Bulk finished" value down to 85%, which is obviously way wrong for lithium batteries, but higher values end up in high virtual SoC when there is not enough solar energy available throughout the day and the battery doesn't really get to 100%.

Today, the jump happened again at 55.80V (that's 0.4V lower than the value set as "SoC when Bulk finished"), it went from 50.5% all the way up to 85%, and -as expected- Bulk isn't yet finished. As a reminder, Bulk voltage is still set at 55.20V and the battery will be over 95% when this value is reached, a few hours from now.

I've also opened a "proper" support request through my dealer, referencing this thread however it's been a week and haven't heard from them since.

Attaching screenshot:

screen-shot-2023-05-23-at-110015.png

0 Likes 0 ·
screen-shot-2023-05-23-at-110015.png (115.5 KiB)
5 Answers
Paul B avatar image
Paul B answered ·

As you dont mention what victron battewry monitor is connected, thus presuming you have a bmv or smartshunt. check the charged voltage setting under battery settings of the smartshunt

3 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.

thanar avatar image thanar commented ·

No, I don't have neither battery monitor. SoC monitoring is done in conjunction by the MultiPlus, the GX device and the SmartSolar controller (running ESS assistant on the system). Voltage and current real-time values are always share between all Victron devices on an ESS system.

0 Likes 0 ·
teloth avatar image teloth thanar commented ·

I asked pretty much identical question regarding this exact phenomenon few months back , detailing and providing values / facts on VRM data log sheets.

Did not get any sensible answers explanations that causing this exact behaviour , not even sure people kindly replied understood what I was asking ...


Anyway , short answer , get smartshunt and be done with it . Multiplus does not have a proper shunt , it only " guesstimates " anyway , so unless you can bring it upto full 100% state of charge or whatever you configure " when bulk finished , set this value " , it will never be accurate. As I could see it drifted as much as 3% per day , prior to smartshunt.

It will cost £100 , but your problems will disappear.


Yes, I know it is pointless to call a £2000 product ( Multiplus II ) and advertise it as " Has battery monitor " , where in fact what it does is nothing better than guesswork, neither the AC usage figures or charge figures are accurate enough, and add to the mix , this SOC jump, will cause you alot of aggro.


I was not keen to keep adding things to my setup as I wanted to keep it " simple " but bit the bullet = £100 , all good now , Smartshunt is super accurate and no " SOC Jumping " now.


Anyway , hopefully you are luckier than me and a proper Victron Techie looks at your question ( keep your fingers crossed ) and they get it sorted.

Link below to my dusty thread


https://community.victronenergy.com/questions/145680/vebus-system-charge-state-showing-bulk-or-absorbti.html?childToView=147382#comment-147382


1 Like 1 ·
thanar avatar image thanar teloth commented ·

Your issue is similar, however since you are not running ESS not DVCC, according to the documentation the GX device does not properly calculate the current in and out of the battery. In my case, it does. At least in theory.

0 Likes 0 ·
nickdb avatar image
nickdb answered ·

Without a shunt you won’t get an accurate reading. There is too much estimation happening. Add a bmv and configure it for your battery bank, you will get what you need.

1 comment
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.

thanar avatar image thanar commented ·

I am not looking for accuracy; It doesn't matter if the SoC estimation is off by 5 or 10%, since every day the battery will get fully charged, in sunny breezy Greece. However, according to Victron documentation, on an ESS-configured or DVCC-enabled system, all MPPTs and inverters loads should add up, so as the battery current calculation should be correct. And it is. Measuring on the battery terminal prompts the same current as calculated by the Victron system; so, it's not an accuracy thing, it's rather a bug.

The issue it that SoC jumps to "bulk finished" as soon as battery voltage reaches a very specific value of 54.78V, a value which is not configured in any place that I know of.

I am not asking to complain. I am an installer Engineer and have to give my clients a proper answer as to why this is happening. Step one is to understand the reason myself and act properly, if possible. Local distributor unfortunately was not able to help with my query.

0 Likes 0 ·
marv21 avatar image
marv21 answered ·

This is set under "State of charge when Bulk finished".

As soon as you finish the Absorption voltage time this will be set.

As far as i know you cant disable this behavior - just set the % to a near value

5 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.

mvader (Victron Energy) avatar image mvader (Victron Energy) ♦♦ commented ·
Hi, the there configured SOC value get set when bulk is finished (and absorption is starting).


Which is not the same as when the absorption timer is finished.


Its actually quite precise, and good enough for many situations. The battery voltage level is the configured Absorption voltage in VEConfigure/VictronConnect.

1 Like 1 ·
thanar avatar image thanar mvader (Victron Energy) ♦♦ commented ·

All relevant settings are set correctly. Bulk voltage is set to 55.2V, as per the attached screenshot from VEconfigure:

I have also set the Bulk and Float voltages in the Charge Controller as well, through VC, in case that was the culprit, as per the screenshot below:

screen-shot-2023-05-05-at-130315.png

screen-shot-2023-05-05-at-130730.png













The problem is not the "Bulk" - to - "Absorption" transition; that one works as expected and according to the documentation. The problem lies with the SoC calculation and display: If you examine the OP attached screenshots, the SoC jumps from (whatever actual value) to 95% (or the relevant "State of charge when Bulk finished" value) as soon as the voltage reaches 54.8V, which is a value nowhere set. That's 0.4V lower than the bulk voltage setting, and it doesn't make any sense.

This jump happens regardless the charging current; high charging currents could probable trip the BMS to disconnect, which would bring the voltage above the Bulk value. However, that would also happen only under high charging currents, and that would also start the Absorption phase as well, which on my case absorption phase starts much later, as expected.

@mvader (Victron Energy) thanx for taking a look into it. Please let me know if you need any additional information through e-mail.

0 Likes 0 ·
screen-shot-2023-05-05-at-130315.png (65.1 KiB)
Show more comments
marv21 avatar image
marv21 answered ·

I've also been dealing with this problem since I got Victron. At high charging currents, the voltage tends to go to float level early on. The battery is then not even close to the percentage specified in the VE.Config. But if you set it too low and the charging current is only low, it gets stuck there and can no longer reach 100%. It makes me want to puke.
I have now helped myself with a PythonScript and MQTT. The charging current is read out and when the maximum charging voltage (float) is reached + less than 4A charging current (52V system) flow, I write the SoC to 100% after 30 minutes at this level (provided there is enough solar to bring more than 4A).
I've been using this for 2 years and it works great!

1 comment
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.

thanar avatar image thanar commented ·
In my case, the charging current doesn’t matter; I’ve included numerous cases in my OP, under different charging currents. On every single one of them, as soon as battery voltage reaches 0.4V lower than the “percentage when bulk finished” value, the SoC jumps.
0 Likes 0 ·
uli-c avatar image
uli-c answered ·

Take manual reading at endpoints. I think that problem is voltage sag. 50A are alot and voltage sag on cable resistance could steal some voltage points, so charger output its voltage that drop before reaching battery so the charger see top voltage while it is not real. Very big cables could limit the effect.

Usage of smart shunt really close to battery will limit this effect maybe. True thing is that usually are used balance cable to read proper voltage directly in cells, whiteout load in it to get correc clean voltage reading.

Smart solar not forsees it, some extra probe cable to get proper battery voltage on battery connector.

But victron has capability to make share smartshunt close in and clean voltage to the mppt with mppt through VE Network.

also it is a shame that VE Network is very close and voted to few functions provided by victron. Otherwise it could be a great wireless network though all victron device without cables....


Moreover smaller mppt has controlled output load also of 10 15 20A max but has no SoC functionality estimation like smartshunt has, despite if I use only its output it could read everything to calculate it... so I think different product, more sels, too much niche functions...


Great thing is to have nodered. I will try to use it to get custom SoC reading... not started yet..

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.