Hi everyone 

 

I have recently installed a deye 12kw 3phase hybrid inverter with 3x greenrich 4.96kwh batteries. My understanding is that when the batteries are fully charged, they should go into a static or standby state until they are required to work. What i have noticed is that the batteries are in a constant state of charge and discharge. 100-99 and 99-100 , constantly. It looks like the grid is feeding the batteries, pauses while the batteries discharge slightly, and then grid resumes to charge the batteries again. Everything is properly setup and works perfectly during loadshedding. Deye did login from China and upgraded the firmware, and the batteries were in standby mode for 1 day before this issue returned again. I'm just concerned that this charge/discharge issue is bad for the batteries and my electricity bill. I would hugely appreciate any feedback or advice on this issue. Thank you

 

 

On 2023/08/05 at 7:14 AM, ZA Moose said:

Hi everyone 

 

I have recently installed a deye 12kw 3phase hybrid inverter with 3x greenrich 4.96kwh batteries. My understanding is that when the batteries are fully charged, they should go into a static or standby state until they are required to work. What i have noticed is that the batteries are in a constant state of charge and discharge. 100-99 and 99-100 , constantly. It looks like the grid is feeding the batteries, pauses while the batteries discharge slightly, and then grid resumes to charge the batteries again. Everything is properly setup and works perfectly during loadshedding. Deye did login from China and upgraded the firmware, and the batteries were in standby mode for 1 day before this issue returned again. I'm just concerned that this charge/discharge issue is bad for the batteries and my electricity bill. I would hugely appreciate any feedback or advice on this issue. Thank you

 

 

Hi, 

The batteries definitely do not go into a standby state. The electronics on board uses electricity, so a constant 20W drain per battery is normal. The SOC that you see oscillate between 99 and 100% is also normal. The SOC is the result of an algorithm that the BMS on the batteries use to guess/estimate what the SOC is. This algorithm is different on all batteries. The true SOC is almost impossible to calculate and thus a small oscillation between 99 and 100 is perfectly normal.

Also, with many inverters and that includes Deye, what it displays on its screen is also not reality. If you use external meters you will notice as I did with my system is that the inverter actually uses around 180W just for being switched on even if it shows 0 from the grid. So I am guessing your system is using around 240W (inverter + 3 batteries) just in idle.

Thanks for taking the time to respond. Much appreciated.

What you have said makes perfect sense. My takeaway here is that inverter behavior is governed by the specific BMS of the battery ?

Do you have any explanation for why I have previously seen the batteries in static/standby state after after the Deye firmware upgrade ? (this lasted 1 day)

Is it not bad for batteries to constantly charge/discharge 100-99,99-100  ?

Thank you,

Brett

 

 

24 minutes ago, ZA Moose said:

Do you have any explanation for why I have previously seen the batteries in static/standby state after after the Deye firmware upgrade ? (this lasted 1 day)

Is it not bad for batteries to constantly charge/discharge 100-99,99-100  ?

It could be that the batteries were slightly overcharged, so it looked like they were static. Not sure without a chart showing battery voltage, current, etc. If the constant charge/discharge bothers you, you could unplug the comms cable to the battery to see if that helps. But in general I would not be bothered by it. In electronics nothing is 100% accurate or stable. Everything is signal + noise. So the slight noise on the SOC signal is normal. If you want, you could add a chart showing the Battery Voltage and Current and I will check that for you. The chart that has those parameters is a bit hidden under the device specific menus on Solarman.

Edited August 7, 20232 yr by HendrikBigChief

To prevent the constant charge, discharge cycle, set your batteries to discharge to e.g. 5% when on-grid?

This is good for the battery...

So here's something quite interesting that leads me back to the BMS.

If I configure the system to use "Lithium" under the battery settings tab, and set the battery parameters. The inverter obviously reads from the BMS and the batteries stay in a constant charge/discharge state. 

If I set the battery configuration as % or use Voltage, the batteries stay static/standby state when fully charged and only discharge when grid is down, and recharge when it returns. I am assuming that when using % or voltage, the BMS is completly bypassed ?

The million dollar question here is - If I used different batteries with a different BMS algorithm; would I see the same result (charge/discharge) ?

I don't like the batteries charging/discharging constantly. I think it uses more power and must have some effect on battery lifespan. Any further comments or advice are appreciated.

 

 

 

 

 

17 minutes ago, ZA Moose said:

So here's something quite interesting that leads me back to the BMS.

If I configure the system to use "Lithium" under the battery settings tab, and set the battery parameters. The inverter obviously reads from the BMS and the batteries stay in a constant charge/discharge state. 

If I set the battery configuration as % or use Voltage, the batteries stay static/standby state when fully charged and only discharge when grid is down, and recharge when it returns. I am assuming that when using % or voltage, the BMS is completly bypassed ?

The million dollar question here is - If I used different batteries with a different BMS algorithm; would I see the same result (charge/discharge) ?

I don't like the batteries charging/discharging constantly. I think it uses more power and must have some effect on battery lifespan. Any further comments or advice are appreciated.

You will see something different with a different BMS I suspect. 

Are you sure the battery is charging and discharging? As I said before, it is possibly just some noise on a voltage line that is causing the SOC to fluctuate. Any chance for a chart showing Battery Voltage and Battery Current + SOC?

We've seen this in a few other batteries (including my own); in my case it seems to be the BMS'es way of denoting cell balancing, yours might be the same.

 

In the case of my own battery (a Hubble AM-5), 99% is the default “full state;” but after the battery SOC drops (which can happen many times during the day - even when the battery is being told to stay full - when the inverter dips slightly into it to assist with PV hand-off to grid), the battery hits 100% at the end of each charge (presumably indicating battery cell balancing) for a little while before reverting to 99% and staying there.

 

In the case of the AM-5 this is also in line with Hubble's warranty documentation which requires a full charge to 100% at least once a week to maintain the unlimited-cycle warranty; which they mention is what guarantees balancing.

Here's the relevant clause from the warranty documentation

Quote

The Installer/Client is required to ensure that that the Batteries are fully charged at least once every 7 days to enable the BMS cell balancer to
activate and perform cell balancing. This is a necessary step to prolong and protect the cells life. The cell
balancer only activates on 100% charge. If the balancer does not activate for an extended period, then the cells
could eventually become unbalanced and can cause undercharged cells to degrade in performance and
couldresult in eventual cell cycle life being greatly reduced. Cell cycle life or performance in these scenarios
cannot be warranted and the warranty will be void

 

We have no control over this (it's decided purely by the BMS) but if this is the case, then it's unlikely that it significantly increases the BMS cycle counter.

 

A few more batteries that exhibit this same behaviour were mentioned in this thread:

 

54 minutes ago, JayMardern said:

We've seen this in a few other batteries (including my own); in my case it seems to be the BMS'es way of denoting cell balancing, yours might be the same.

 

In the case of my own battery (a Hubble AM-5), 99% is the default “full state;” but after the battery SOC drops (which can happen many times during the day - even when the battery is being told to stay full - when the inverter dips slightly into it to assist with PV hand-off to grid), the battery hits 100% at the end of each charge (presumably indicating battery cell balancing) for a little while before reverting to 99% and staying there.

 

In the case of the AM-5 this is also in line with Hubble's warranty documentation which requires a full charge to 100% at least once a week to maintain the unlimited-cycle warranty; which they mention is what guarantees balancing.

Here's the relevant clause from the warranty documentation

 

We have no control over this (it's decided purely by the BMS) but if this is the case, then it's unlikely that it significantly increases the BMS cycle counter.

 

A few more batteries that exhibit this same behaviour were mentioned in this thread:

 

It is possible that 100% indicates cell balancing, but not sure. My money would be on slight noise on the voltage line. I also have the AM-5, and I don't see the 99 to 100% cycling that you see. Mine is solid on 100%.

Edited August 8, 20232 yr by HendrikBigChief

44 minutes ago, HendrikBigChief said:

It is possible that 100% indicates cell balancing, but not sure. My money would be on slight noise on the voltage line. I also have the AM-5, and I don't see the 99 to 100% cycling that you see. Mine is solid on 100%.

 

Yeah could be. SOC gets reported via CAN (which is digital so not very susceptible to noise) but the cell voltage measurements could certainly be a bit noisy.

 

I'm suspicious though because it only seems to happen on this 99-100% percentage range, just like in OP's case. For example when my inverter tells the battery to maintain 85%, it's then rock solid with no fluctuation at all. Without knowing the BMS algorithm I guess we can't be entirely sure!

 

Here's an example from a day or two ago of the 99%-100% fluctuations on my AM-5; the SOC target for that time-slot is 100% - it happens during PV use as well as post load-shedding; both scenarios dip (at least very slightly) into battery; it hits 100% briefly and then settles on 99% the rest of the time (despite the 100% target); this does look like cell balancing to me:

 

Edited August 8, 20232 yr by JayMardern

Thanks for all the advice and feedback. 

I do not have any solar installed yet, This is just a straight up backup power system at this stage.

I have spoken to the supplier and they are going to check the battery firmware version for an update and see form there.

Funnily enough, I just noticed that my inverter seems to be pushing power back to the grid! 

Can this be due to the CT clamps being the wrong way around ? Where should they point ? Inverter or utility ? Seems to be mixed options on this...

My SOC is sitting at 100%, but battery status is discharging. 

Really weird.. 

 

12 minutes ago, ZA Moose said:

Thanks for all the advice and feedback. 

I do not have any solar installed yet, This is just a straight up backup power system at this stage.

I have spoken to the supplier and they are going to check the battery firmware version for an update and see form there.

Funnily enough, I just noticed that my inverter seems to be pushing power back to the grid! 

Can this be due to the CT clamps being the wrong way around ? Where should they point ? Inverter or utility ? Seems to be mixed options on this...

My SOC is sitting at 100%, but battery status is discharging. 

Really weird.. 

 

So my battery is usually at 100%, then in that Battery Status field it says Static, whatever that means, but it still draws around 20W, presumably to feed the BMS electronics. Not sure why you are seeing discharging. That -239W to the grid is problematic. It is probably due to your settings, check if it limits feeding back to the grid. Even if the clamp is the wrong way around, I would not expect the inverter to say -239 as the inverter would try to stop any negative currents.

I agree it's completely weird..

I have recorded a couple of short clips which I have uploaded to you tube to accurately convey what happens when you setup the batteries as lithium vs % or voltage. % or voltage looks like completely normal functioning, however they register as lead acid which they are not. I assume setting them this way bypasses the BMS? See links below. Appreciate all the wisdom gained and shared here.

Voltage or % mode
https://youtube.com/shorts/ATPQApg15OI?feature=share

Lithium mode
https://youtu.be/5utuTnG-lCA

 

 

So it turns out that this issue is actually a BMS firmware problem. 

I will update the firmware an post the results in the next few days

I have a similar issue with the 3.686kwh Greenrich battery in an off-grid scenario with an Axpert King inverter (Kodak branded). If using the lithium settings, as it approaches 100% charge it gets into a loop of charging and then stopping, every 15 to 30s endlessly.  In a small cabin it is very annoying as the inverter fans spin up and down constantly.  If using the USER settings (voltage driven), this was happening as well if using the bulk and float charge voltages which were originally recommended (57v) but not if using lower voltages. Unfortunately when in BMS coms mode for Lithium or Pylontech, the BMS communicates the voltage to the inverter and even with the BMS upgrade it is still 57v.  I eventually used a serial cable and monitored with PBMS tools while it was exhibiting the stop/start behaviour, and could see there that over-voltage protection was kicking in aborting each attempt at resuming charging.  

The battery distributors have advised that we stick to USER settings with manually configured bulk and float voltages of 56v and 55v instead of the 57v the BMS communicates to the inverter.  The downside is that we lose out on the more accurate SoC reported by the battery compared to the voltage-based estimates from the inverter, which in my case are very inaccurate.  

Not sure ZA Moose if the discharging / charging you are seeing in BMS mode is related in any way to different voltages compared to voltage-based mode, but if you are able to find a cable and monitor the battery directly, if might assist.

Plwased to report  i have upgraded the firmware on the greenrich batteries and the system is functioning 100% normal as it should. Battries are now static /stanby mode when fully charged. No more charge/discharge loop. Thanks for all the input and advice everyone. 

Hello, everyone!

We have the same problem with our UPS station consisting of three Deye SUN-12KSG04LP3-EU inverters in parallel mode and eight Deye SE-G5.1 Pro-B batteries.

The problem is that the Inverter cyclically charges and discharges some batteries from the assembly, reducing their service life and wasting electricity. Because of this, the batteries gain cycles unevenly, in fact some batteries wear out more that twice as much as others. This happens if you set the SOC level of the batteries to 100%. At the SOC level of 98%, the battery becomes static, but the calibration and balancing of the batteries of the assembly does not occur.
As we managed to find out, the inverter  charges the battery to the  voltage 58.4V, then it begins to discharge it to the voltage below 53.0 V. At the same time, the SOC of the batteries starts to drop, and when it drops to 99% on one of the batteries, this battery starts charging again to the 58.4V level, and so on in a cycle.
Is it possible to do something about this problem?
We think that the problem can be solved by making a change to the firmware, for example, by lowering the threshold for triggering the additional charge to 98% or by changing the logic in some other way so that there is no wasteful cycling of the batteries. But 
[email protected] does not see the problem with it. 

PS/ The firmware of the units is up to date. maybe using Butt V or % will solve the problem, but we can't disable the connection with BMS for security reasons.

Edited January 28, 20251 yr by mikerin

On 1/28/2025 at 11:39 PM, mikerin said:

Hello, everyone!

We have the same problem with our UPS station consisting of three Deye SUN-12KSG04LP3-EU inverters in parallel mode and eight Deye SE-G5.1 Pro-B batteries.

The problem is that the Inverter cyclically charges and discharges some batteries from the assembly, reducing their service life and wasting electricity. Because of this, the batteries gain cycles unevenly, in fact some batteries wear out more that twice as much as others. This happens if you set the SOC level of the batteries to 100%. At the SOC level of 98%, the battery becomes static, but the calibration and balancing of the batteries of the assembly does not occur.
As we managed to find out, the inverter  charges the battery to the  voltage 58.4V, then it begins to discharge it to the voltage below 53.0 V. At the same time, the SOC of the batteries starts to drop, and when it drops to 99% on one of the batteries, this battery starts charging again to the 58.4V level, and so on in a cycle.
Is it possible to do something about this problem?
We think that the problem can be solved by making a change to the firmware, for example, by lowering the threshold for triggering the additional charge to 98% or by changing the logic in some other way so that there is no wasteful cycling of the batteries. But 
[email protected] does not see the problem with it. 

PS/ The firmware of the units is up to date. maybe using Butt V or % will solve the problem, but we can't disable the connection with BMS for security reasons.

I think this issue is due to the inverter is pulling power from the battery to be ready to switch rapidly to it if the power went off ( UPS functions)