Batrium + Victron + CerboGX- What is controlling my charge ramp down? (ESS)

[m1kkel]

Hoping someone can help. Ive been searching the web, forums etc. for a answer to my question.

Im charging my pack to 54.0 volts. When i hit that, at some point, the victorn system will add 0.4 volts to the 54.4 volt since im exporting excess solar to the grid.

However it will keep charging at 54,4 volt for some time, serveral hours at around 30-40 amp until my soc hits 100% in the batrium system (99% in victron) and finally it ramps down to 3 Amp.

My batrium remote target is set for 54.0 in normal/100AMP and 54.0 in limited, 1 amp. But when i hit 54,0V it keeps going and slowly ramps down to 30-40 amp, and when i hit soc 100 it will return to 3 amp. I dont understand why. DVCC is enabled, so i assume my MPPT charge values are ignored, aswell as my Multiplus-II values set in Veconfigure. So something else is controlling this ramp down effect.

To my understanding victron doesnt care about set Amp values when i use DVCC, it only cares about voltage, but what defines my absorption charge time then????

Hope someone can help

 

[m1kkel]

Settings from batrium

 

[Zwy]

It appears you are using communication between the Victron and Batrium?

You have bypass set at 3.4V which is ideal. In order to get to bypass, it will require a pack voltage of 54.4V.

I do not understand why you are using 54.0V as a limit, the Batrium would never be able to enter bypass and balance cells.

 

[m1kkel]

It appears you are using communication between the Victron and Batrium?

You have bypass set at 3.4V which is ideal. In order to get to bypass, it will require a pack voltage of 54.4V.

I do not understand why you are using 54.0V as a limit, the Batrium would never be able to enter bypass and balance cells.

Yes thats correct, theres canbus communication. The reason the limit is 54.0 volt is because Victron will add an offset of 0.4 Volts since im exporting to grid, so the pack will end at 54.4 volt. And then the ramping down starts at some point, but i cant control how long it takes, and im pushing 54.4 V / 30-40 amps into the pack still. And i want to control for how long, that keeps going.
It seems like the soc level have something to do with it, but i still cant control it. Ive set the shunt soc to 95%, but it doesnt help.

 

[Zwy]

I think it's perfectly fine the way it is as long as you hit bypass for balancing. I have my pack bulk charging voltage set at 56.5V and it won't balance a cell under 3.4V. It won't begin bypass until one cell hits 3.5V. Takes about the 30 to 40 minutes you mention to reach 100% SOC once bypass begins. I have float set for 54.4V but it never seems to enter float. The EG4 6500EX charges to bulk and once that is hit, the SCC quits charging and cell voltage drops to under 3.4V quickly.

 

[m1kkel]

Do you also see a variation in cell voltages after you stop putting ampere into the battery? As soon as charging stops, the cells seems imbalanced. But maybe it’s because it’s lifepo4.
I’m only used to lithium ion.

Anyway I need to be able to control when the cut off will happen.

 

[Zwy]

Do you also see a variation in cell voltages after you stop putting ampere into the battery? As soon as charging stops, the cells seems imbalanced. But maybe it’s because it’s lifepo4.
I’m only used to lithium ion.

Should not matter if you top balanced the cells. Mine will vary 0.02V or less when charging is finished.

I top balanced each 16 cell battery to 3.65V per cell with cells in parallel after charging with a BMS in series. All 4 batteries (54Kwh total) were assembled and only then did I do any discharge.

Anyway I need to be able to control when the cut off will happen.

Please explain.

 

[m1kkel]

Should not matter if you top balanced the cells. Mine will vary 0.02V or less when charging is finished.

I top balanced each 16 cell battery to 3.65V per cell with cells in parallel after charging with a BMS in series. All 4 batteries (54Kwh total) were assembled and only then did I do any discharge.


Please explain.

I didnt top balance - its the first of 5 Lifepo4 packs. Im actually currently balancing the next 32 cells before i assemble. Maybe thats just why, i see the voltage differences.

Ok i will try to explain:
Whenever i hit 54.4 volt, victron/batrium will keep pumping amps int the battery for a unknown amount of time, before it ramps down. The time is serveral hours. I want to control for how long and how much ampere will be charged into the battery when said voltage is reached. I know absorption parmeters are ignored when using a managed battery, Victron, ESS and grid export, and only CVL parmeter is used. THats great, but whats controlling the amps going into the battery then and the time?

Im afraid im actually charging the battery to 100% this way.

 

[Zwy]

I didnt top balance - its the first of 5 Lifepo4 packs. Im actually currently balancing the next 32 cells before i assemble. Maybe thats just why, i see the voltage differences.

Ok i will try to explain:
Whenever i hit 54.4 volt, victron/batrium will keep pumping amps int the battery for a unknown amount of time, before it ramps down. The time is serveral hours. I want to control for how long and how much ampere will be charged into the battery when said voltage is reached. I know absorption parmeters are ignored when using a managed battery, Victron, ESS and grid export, and only CVL parmeter is used. THats great, but whats controlling the amps going into the battery then and the time?

Im afraid im actually charging the battery to 100% this way.

As long as the system is in CV at 54.4V, it won't hurt the battery nor would it be at 100%. You are not close to 58.4V which is 3.65V per cell.

I think part of what you are seeing is the Batrium in bypass attempting to balance the pack. The amps you see flowing in are being converted to heat.

 

[m1kkel]

As long as the system is in CV at 54.4V, it won't hurt the battery nor would it be at 100%. You are not close to 58.4V which is 3.65V per cell.

I think part of what you are seeing is the Batrium in bypass attempting to balance the pack. The amps you see flowing in are being converted to heat.

You are actually able to charge the battery to 100% even at 54.4 volt. The soc on lifepo4 have nothing to do with voltage..

And my original question is actually, what is controlling this behaviour, when ramping down happends?

 

[sunshine_eggo]

I think it's perfectly fine the way it is as long as you hit bypass for balancing. I have my pack bulk charging voltage set at 56.5V and it won't balance a cell under 3.4V. It won't begin bypass until one cell hits 3.5V. Takes about the 30 to 40 minutes you mention to reach 100% SOC once bypass begins. I have float set for 54.4V but it never seems to enter float. The EG4 6500EX charges to bulk and once that is hit, the SCC quits charging and cell voltage drops to under 3.4V quickly.

If you're talking about your Batrium in control of a charger, there is no such thing as float. Batrium BMS does not support float. It was designed for 3.7V chemistry, and that's how it behaves. Charge to X voltage and hold it.

THats great, but whats controlling the amps going into the battery then and the time?

The remote charge voltage settings and the inability of the battery to accept current without voltage rise.

Batrium works like this:

Charge until voltage is reached and hold that voltage. That's it.

Balancing can bleed off capacity and allow a little more current to come into the battery too.

Im afraid im actually charging the battery to 100% this way.

The BMS may say 100%, but you're probably not, and there's no harm in charging to 100% especially at the lower voltages. Getting fully charged at less than 3.45V is unlikely.

You are actually able to charge the battery to 100% even at 54.4 volt. The soc on lifepo4 have nothing to do with voltage..

Oh that's very untrue as a blanket statement. Below 3.45V, charges to 100% either don't happen, or they happen incredibly slowly at very low current over DAYS. At and below 3.40V, unlikely to hit 100% - more likely to hit ~95% @3.40V after a long charge. Regardless of what the BMS reports. The BMS essentially reports whatever you tell it to.

And my original question is actually, what is controlling this behaviour, when ramping down happends?

Again, the remote charge voltage and the inability of the battery to accept additional current without a voltage rise produces the ramping.

Since you've not balanced your pack, and you should. Here's what I'd do:

Charging high cell cut out to 3.65V
Disable SoC high cell cut out
Bypass to 3.45V
Enable Autolevel:



Configure as follows:



It will work to close the gap so that all cells are within 0.02V, it will work on the worst 0.05V range of cells (from the highest down), and it will only engage in balancing above 3.40V.


Want to control current based on SoC?

Use Ramped targets in Remote:




Voltage and Limited voltage to 54.8 (assuming 0.4V gets added to that)
I've set an example of SoC and Voltage limits. The 99% and 0.1A is intended to tell the unit to effectively stop charging.

Batrium is not the best choice for LFP. It was designed for 3.7V Li-ion, and that's how it works best. LFP requires charging like lead acid with a float, and Batrium doesn't allow that.

 

[Zwy]

If you're talking about your Batrium in control of a charger, there is no such thing as float. Batrium BMS does not support float. It was designed for 3.7V chemistry, and that's how it behaves. Charge to X voltage and hold it.

You quoted me Eggo, I don't do BMS comms with inverters, so no, I don't have the Batrium controlling charging. The EG4 6500EX's on my system never seem to enter float. But I don't sit there and watch it and never looked in Solar Assistant either.

The remote charge voltage settings and the inability of the battery to accept current without voltage rise.

Batrium works like this:

Charge until voltage is reached and hold that voltage. That's it.

Balancing can bleed off capacity and allow a little more current to come into the battery too.

That's my theory on why he sees the amps flowing in, the Batrium is burning it off as heat in bypass.

The BMS may say 100%, but you're probably not, and there's no harm in charging to 100% especially at the lower voltages. Getting fully charged at less than 3.45V is unlikely.



Oh that's very untrue as a blanket statement. Below 3.45V, charges to 100% either don't happen, or they happen incredibly slowly at very low current over DAYS. At and below 3.40V, unlikely to hit 100% - more likely to hit ~95% @3.40V after a long charge. Regardless of what the BMS reports. The BMS essentially reports whatever you tell it to.

I agree, I think the OP has many misconceptions about LFP.

Again, the remote charge voltage and the inability of the battery to accept additional current without a voltage rise produces the ramping.

Since you've not balanced your pack, and you should. Here's what I'd do:

Charging high cell cut out to 3.65V
Disable SoC high cell cut out
Bypass to 3.45V
Enable Autolevel:

View attachment 145761

Configure as follows:

View attachment 145762

It will work to close the gap so that all cells are within 0.02V, it will work on the worst 0.05V range of cells (from the highest down), and it will only engage in balancing above 3.40V.

Agree 100%.

Want to control current based on SoC?

Use Ramped targets in Remote:

View attachment 145766


Voltage and Limited voltage to 54.8 (assuming 0.4V gets added to that)
I've set an example of SoC and Voltage limits. The 99% and 0.1A is intended to tell the unit to effectively stop charging.

Batrium is not the best choice for LFP. It was designed for 3.7V Li-ion, and that's how it works best. LFP requires charging like lead acid with a float, and Batrium doesn't allow that.

 

[sunshine_eggo]

You quoted me Eggo, I don't do BMS comms with inverters, so no, I don't have the Batrium controlling charging. The EG4 6500EX's on my system never seem to enter float. But I don't sit there and watch it and never looked in Solar Assistant either.

Right. Because I didn't think you did... needed to make sure.

In some manuals on the Voltronics-style inverters, I have found reference to the same Sigineer/AIMS absorption profile where:

T1 = time to charge to some voltage a little below absorption.
T2 = absorption time = 10X T1, min 1 hour, max 12 hr

That means the absorption period may be VERY long.

 

[m1kkel]

If you're talking about your Batrium in control of a charger, there is no such thing as float. Batrium BMS does not support float. It was designed for 3.7V chemistry, and that's how it behaves. Charge to X voltage and hold it.

Would i be better off, just using the Victron Multiplus'es for this, and letting the Batrium trip a shunt in case something goes wrong? Its just, i purchased a lot of nbatrium equipment for this project. damn!

The remote charge voltage settings and the inability of the battery to accept current without voltage rise.

Batrium works like this:

Charge until voltage is reached and hold that voltage. That's it.

Balancing can bleed off capacity and allow a little more current to come into the battery too.

Thanks a lot for clearing that up once and for all.

The BMS may say 100%, but you're probably not, and there's no harm in charging to 100% especially at the lower voltages. Getting fully charged at less than 3.45V is unlikely.


Oh that's very untrue as a blanket statement. Below 3.45V, charges to 100% either don't happen, or they happen incredibly slowly at very low current over DAYS. At and below 3.40V, unlikely to hit 100% - more likely to hit ~95% @3.40V after a long charge. Regardless of what the BMS reports. The BMS essentially reports whatever you tell it to.

Ok - what im seeing as i have lowered the charge voltage the last 9 days, is that it reached 54.0 pretty fast, and then im not able to charge fast, since the voltage then will rise. So im only charging at around 3000 Watt. Will that change when i get the last 4 packs online?

Since you've not balanced your pack, and you should. Here's what I'd do:

Charging high cell cut out to 3.65V
Disable SoC high cell cut out
Bypass to 3.45V
Enable Autolevel:

View attachment 145761

Configure as follows:

View attachment 145762

It will work to close the gap so that all cells are within 0.02V, it will work on the worst 0.05V range of cells (from the highest down), and it will only engage in balancing above 3.40V.

Thanks for this very detailed and well documented description. I will implement the changes now.

Want to control current based on SoC?

Use Ramped targets in Remote:

View attachment 145766


Voltage and Limited voltage to 54.8 (assuming 0.4V gets added to that)
I've set an example of SoC and Voltage limits. The 99% and 0.1A is intended to tell the unit to effectively stop charging.

Batrium is not the best choice for LFP. It was designed for 3.7V Li-ion, and that's how it works best. LFP requires charging like lead acid with a float, and Batrium doesn't allow that.

Yes 0.4 will get added when the new target (54.8) is reached. It will not happen right away, but after some time as the charge amps starts to slow down.

Thanks for your SoC charge example, but im sure it will not work with Batrium and Victron. As far as i understand Victron ignores the charge current sent by the BMS, and only listens to the Charge Voltage (CVL), so setting the ramped targets will not work. Its an issue only if ESS and "feed in excess solar" is enabled.

Se this thread, where Daniel Böekel from Victron is defending his case: https://community.victronenergy.com...-dc-feed-in-enabled-charge-current-limit.html


Another question

Since im TOP Balancing the other packs before assembly, and i have already put one pack into service. Lets just assume it gets balanced, how will i attach the remaining 4 packs which currently sits at 3.65 minus whatever they have dropped to. Should i Charge my already build battery, via solar to 58,4 and hope it settles at around the same voltage as the remaining 4 packs?

 

[m1kkel]

You quoted me Eggo, I don't do BMS comms with inverters, so no, I don't have the Batrium controlling charging. The EG4 6500EX's on my system never seem to enter float. But I don't sit there and watch it and never looked in Solar Assistant either.




That's my theory on why he sees the amps flowing in, the Batrium is burning it off as heat in bypass.



I agree, I think the OP has many misconceptions about LFP.

I also think i do, even though i've read a lot about settings, chargning etc. and also the other very long post from sunshine_Eggo where he helps another victron guy with a heavily imbalanced pack. I still have many misconceptions. Been running a 20Kwh 18650 Powerwall for 2 years now, so im not hands-on familiar with LFP until now and the last 9 days ? Im so glad you're helping here.

 

[sunshine_eggo]

I am weak in ESS. I'll read that thread in depth and comment more in-depth if needed.

Based on my recommended charge voltages, the description of the OP's concerns and Victron's direct response to him, I can't see there is a significant risk of over-current at that level.

Even with the ESS current override, Voltage will ALWAYS be honored, so if you're at or near the voltage, it's impossible to over-current the battery as more current will increase the voltage, and that will force current to pull back as voltage is maintained. The BMS is sending BOTH voltage and current limits, so I don't see how there's any danger.

 

[sunshine_eggo]

Okay, I have a better grasp of the issue. DVCC current limits are not observed, and they recommend using voltage limits. Unfortunately, I find Victron way too dismissive because even with their "approved" batteries, there is the potential for exceeding the rated charge current.

Would i be better off, just using the Victron Multiplus'es for this, and letting the Batrium trip a shunt in case something goes wrong? Its just, i purchased a lot of nbatrium equipment for this project. damn!

Definitely not. The last thing you want to do is disrupt the system by disconnecting the battery from it without darn good cause.... especially when that trigger is a buttload of incoming PV.

The 54.8V is set because I want 55.2V or 3.45V/cell as the maximum. When the battery is at 3.45V and a high-ish state of charge, it will not accept more current than is needed to maintain 3.45V.

So... there is definitely risk of over-current below 3.45V/cell and at lower states of charge; however, cells are far more tolerant in that case, and this isn't just a Batrium issue. The aforementioned Pylontechs have the same issue. It's kinda what you have to put up with if you're going to use Victron ESS/Feed in excess solar.

However, I'm confident the settings I recommended will absolutely minimize any potential for damage to the batteries. Ideally, the ability to put it in limited charge mode based on a current value would do the trick, but I can't see that it's possible.

There would be substantial risk of damage if running to 3.55-3.65V @ greater than rated current, but the risk is much lower at the lower voltages.

 

[Bitrover]

To my understanding of the victron setup the 0.4V is only added to the mppt chargers.
This to give the multi`s some leverage to pull this down to CVL again, and thus feeding back to the grid.

You can use Dbus-spy.py on your Cerbo to see the live parameters in all the devices.

I am using victron ESS with batrium and also looking to fine tune my settings for LFP
No float mode with this setup and current software.
But also not sure if it really is a problem when running ESS.
As the batteries will not be at 55.2V for more then a few hours at some days per year.

 

[m1kkel]

Okay, I have a better grasp of the issue. DVCC current limits are not observed, and they recommend using voltage limits. Unfortunately, I find Victron way too dismissive because even with their "approved" batteries, there is the potential for exceeding the rated charge current.



Definitely not. The last thing you want to do is disrupt the system by disconnecting the battery from it without darn good cause.... especially when that trigger is a buttload of incoming PV.

The 54.8V is set because I want 55.2V or 3.45V/cell as the maximum. When the battery is at 3.45V and a high-ish state of charge, it will not accept more current than is needed to maintain 3.45V.

So... there is definitely risk of over-current below 3.45V/cell and at lower states of charge; however, cells are far more tolerant in that case, and this isn't just a Batrium issue. The aforementioned Pylontechs have the same issue. It's kinda what you have to put up with if you're going to use Victron ESS/Feed in excess solar.

However, I'm confident the settings I recommended will absolutely minimize any potential for damage to the batteries. Ideally, the ability to put it in limited charge mode based on a current value would do the trick, but I can't see that it's possible.

There would be substantial risk of damage if running to 3.55-3.65V @ greater than rated current, but the risk is much lower at the lower voltages.

Hi!

I’ve implemented all
Your recommendations. My pack is almost balanced. However 2 cells are lower than the rest. Do you have any recommendations to what I can do to mitigate this?

 

[m1kkel]

To my understanding of the victron setup the 0.4V is only added to the mppt chargers.
This to give the multi`s some leverage to pull this down to CVL again, and thus feeding back to the grid.

You can use Dbus-spy.py on your Cerbo to see the live parameters in all the devices.

I am using victron ESS with batrium and also looking to fine tune my settings for LFP
No float mode with this setup and current software.
But also not sure if it really is a problem when running ESS.
As the batteries will not be at 55.2V for more then a few hours at some days per year.

My batteries reach 55.2 every day, as im feeding in excess. But not if theres less PV compared to a sunny day, when they end at 54.8v, and im still feeding in excess, just much less than "normal" so its actually a bit strange how that algorithm is applied.