question

Vangelis Beligiannis avatar image
Vangelis Beligiannis asked

BYD battery not charged by the Quattro - output frequency 52,7 Hz

img-20210404-152821.jpgHi!

I have recently purchased a BYD LVS 20.0 battery and a Quattro 48/8000

My system is off-grid and has a Quattro 48/8000 with a FRONIUS PRIMO 3.0 on AC out, an MPPT 150/70 and a CCGX .

All setting are done according to your guidelines :

Victron & BYD [Victron Energy]

Here is the problem :

After the battery is fully charged TO 100% (around 13:00-14:00) the battery starts to discharge although there is plenty of sun to keep it charged. The MPPT keeps giving power but the Quattro shuts down the FRONIUS to 0W!!!

I can understand this happening if the battery voltage is above 55 V but this goes on even if the battery voltage goes down to 53V.

I am positive that this is not a BYD BMS issue. I tried connecting the AC GRID that I have as a back up and then the quattro started charging the battery as well immediately because it dropped the frequency back down.

I believe that the quattro should keep the fronius awake as long as the voltage is 55V or lower. Even if it has the frequency locked probably because of an overcharge insident (no error appeared) it should unlock after the voltage drops under the rebulk value. This value is 54.2V for my system but the frequency is still locked even reaching 53 V.


Why does it not unlock the frequency?

I am loosing power and the battery starts to discharge from 14:00 for no reason.


MultiPlus Quattro Inverter ChargerBYD
img-20210404-152821.jpg (323.2 KiB)
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.

Daniël Boekel (Victron Energy Staff) avatar image Daniël Boekel (Victron Energy Staff) ♦♦ commented ·

Hi @Vangelis Beligiannis

I'm suspecting your system locks to 53Hz due to the battery lowering the CVL when fully charged.

Can you try raising the charge voltages slightly in ve.configure? this won't actually create a higher charging current (as the batteries BMS controls charging), but it does affect the PV inverter assistant (included in ESS assistant) and should prevent a 'frequency lock'

0 Likes 0 ·
Vangelis Beligiannis avatar image Vangelis Beligiannis Daniël Boekel (Victron Energy Staff) ♦♦ commented ·

Hi! I am not Using ESS. My system is almost totally offgrid. Using the Ac in is extremely rare. I Will try your solution but the inverter does not start even if the voltage drops to 53V with the float voltage set at 55V

0 Likes 0 ·
Show more comments
4 Answers
Vangelis Beligiannis avatar image
Vangelis Beligiannis answered ·

Resuming all the above, the solution to the problem is inside the DVCC menu. Enabling the "Limit managed battery charge voltage" option and setting the value to 56,8V solves the problem. The DVCC is "faster" than the BYD BMS and prevents the locking of the frequency.

I would expect a solution provided from Victron stuff on the issue but "if you want something done, do it your self" :)

2 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.

Daniël Boekel (Victron Energy Staff) avatar image Daniël Boekel (Victron Energy Staff) ♦♦ commented ·

Hi @Vangelis Beligiannis (and @ThomasW.)

Thank you for your feedback, I wasn't aware of the latest replies you wrote further down.

We now know what happens here:

On the new types of BYD batteries, BYD has made a change in the algorithm vs older batteries:
CVL (charge voltage limit) is now 58,4 volt, (before: 56,4 volt) and when the battery is (almost) fully charged, this lowers to 55,1 volt. (before: 54 volt)

I first thought this drop caused the frequency lock, but it turned out it actually is the height of the voltage that causes it.

Our PV inverter assistant (also in ESS assistant) uses the configured absorption voltage (in the Multi)
As a safety feature, when battery voltage is above absorption voltage * 103%, the output frequency is locked until voltage lowers to float-voltage -3,2 volt.

So to prevent this lock: raising the absorption voltage in the Multi / Quattro helps, and lowering the actual maximum charge voltage in DVCC menu also fixes it (and that is easier to do)
Both solutions don't prevent a full charge, both are safe.

It took me too long to find out what was happening, because I wasn't aware of the change in the BYD batteries and didn't know the exact cause of the frequency lock.

0 Likes 0 ·
Vangelis Beligiannis avatar image Vangelis Beligiannis Daniël Boekel (Victron Energy Staff) ♦♦ commented ·

Thank you very much for the reply. I am happy to know that my solution is also working and also easier to apply.

It took me many days to figure it out as well but it was worth it.

I now also understand what is going on with the"rebulk". So what is mentioned on the Pv assistant about rebulk at 54V for lithium voltage does not apply in this occasion because the locking is caused from the 103% overvoltage. Thank you for clearing this out as well.

0 Likes 0 ·
thomasw-1 avatar image
thomasw-1 answered ·

Hi Vangelis,
I can confirm your observation. In my 3-phase system I have the identical behaviour. In island-mode instead of using primarily the AC-OUT-connected inverter power gets converted from MPPT and if that is not sufficent battery gets discharged.

@Izak (Victron Energy Staff) and @mvader (Victron Energy): I would also appreciate, if ESS would be modified in that way, so that AC-connected inverter is used before MPPT and before battery gets discharged.

In my case the MultiPluses even go in overload instead of activating the AC-OUT connected inverter. In 1 case system even switched off because of current overload!

According to a recommendation from @Guy Stewart (Victron Community Manager)

Guy Stewart (Victron Community Manager) avatar image

Guy Stewart (Victron Community Manager)

Folgen
you should make your Victron dealer aware about that contraint, so importance gets raised on such requirement. I am going to forward that link to mine as well.

16 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.

Daniël Boekel (Victron Energy Staff) avatar image Daniël Boekel (Victron Energy Staff) ♦♦ commented ·

Hi @ThomasW.

ESS has nothing to do with islanding / off grid mode.

0 Likes 0 ·
thomasw-1 avatar image thomasw-1 Daniël Boekel (Victron Energy Staff) ♦♦ commented ·

Thanks for jumping on that topic, Daniel.
I have no insight about Victron's delelopment terminology. As you confirmed yourself the functionality is at least included in ESS-assisitant. Anyhow independent where the reason is caused, it is limiting the overall efficiency.

An increase of the battery voltage limit would be most likely not able to discover the current power requirements. It takes a while until voltage would be decreased to a level, where MultiPluses would initiate frequency decrease to wake up the inverter. During that phase battery is unnecessarily discharged (oberserved situations, where battery went down to 90% before AC-inverter was restarted) and especially not full PV-capacity is available to internal grid (leads to potential overloads of MultiPlus). From my point of view AC-inverters should have not the purpose to charge battery only.
I am looking forward seeing news from Vangelis.,

In a complex setup I am asking myself, why in island mode AC-OUT connected inverters are not also controlled by GX via modbus-TCP-based limits, when "Victron MPPT charge characteristics are automatically configured & governed by the GX device." (see BYD references on Victron webside). GX would be capable to detect power requirements much more efficient than a battery voltage based detection by MuliPlus.

0 Likes 0 ·
Daniël Boekel (Victron Energy Staff) avatar image Daniël Boekel (Victron Energy Staff) ♦♦ thomasw-1 commented ·

Hi @ThomasW.

The problem is that BYD batteries lower CVL from about 55 volt to 51 volt, this is not needed and creates unwanted effects. (in ESS mode the system will actually discharge to grid)

We hope a feature update of BYD's firmware will improve this behavior, in the mean time it seems that tweaking the Victron inverter's charge voltage settings is a work-around to prevent the 'frequency lock'

please try raising voltages with 0,5V and report back.

0 Likes 0 ·
Vangelis Beligiannis avatar image Vangelis Beligiannis Daniël Boekel (Victron Energy Staff) ♦♦ commented ·

screenshot-20210405-131829.pngscreenshot-20210405-131747.pngscreenshot-20210405-131732.png

Same thing today. I have not changed the charging voltage yet.

But here it is clear that the CVL is high and not causing the problem.

Finally the reason of locking the frequency is clear. It is because the Fronius inverter does not react quickly enough to prevent the overgharge that triggers the frequency lock.

The question is :

WHY IS THE INVERTER NOT ENTERING THE REBULK PHASE EVEN THOUGH THE VOLTAGE DROPS TO 52.5V?

I will change the charger voltage tomorrow and keep you posted.


0 Likes 0 ·
screenshot-20210405-131732.png (228.4 KiB)
screenshot-20210405-131829.png (190.3 KiB)
screenshot-20210405-131747.png (214.4 KiB)
Vangelis Beligiannis avatar image Vangelis Beligiannis Daniël Boekel (Victron Energy Staff) ♦♦ commented ·

Hello! I changed the absorption voltage to 55.7V (55.2+0.50) as you mentioned.

It looks ok up to know. The fronius switched to 0W when the voltage of the battery reached 57.1V and came back on when the float voltage was reached (55V). I will check it out the following days to make sure it is ok.

There is another problem now. When battery is at 100% and 55V float volatge, the MPPT is almost closed (5W) and the fronius is giving the loads. This is correct and wanted, but the measurement of the AC LOADS on the CCGX has gone crazy. It is constantly jumping up and down for no reason. The frequency of the AC out is 51HZ which means that the fronius power reduction is active. But the AC load has gone crazy. If the ac load increases a lot and the frequency gets back to 50 HZ the problem stops.

Any hints?

Is it necessary to instal an energy meter on the AC load cables?


0 Likes 0 ·
thomasw-1 avatar image thomasw-1 Vangelis Beligiannis commented ·

Hi Vangelis,

have you changed absorption voltage only or also the float voltage?
With that setting, I haven't discovered any change on island mode.

1617796464450.png

This afternoon I performed an additional test with 55,7 absorption and 55,0V float voltage.
I have observed following behaviour:

  • Voltage increased up to 57,1V before inverter and MPPT switch off.
    It seems that the voltages defined on MultiPlus are completely ignored and only DVCC is active.
  • Below At 54,8V MPPT returned and Fronius inverter was on idle (frequency at 52,8Hz.

@Vangelis:
What GX-release are you running? In case it is lower than 2.65 do you have DVCC activated?


Update after 1 hour:

Interestingly the behaviour changed now. In the moment main energy is provided by Fronius, Slightly throtteled at 51,4Hz, Battery remains at 100%

0 Likes 0 ·
1617796464450.png (4.1 KiB)
Vangelis Beligiannis avatar image Vangelis Beligiannis thomasw-1 commented ·

Gx release is the latest. DVCC is on (forced). I am not sure if I have to activate the "limit managed battery charge voltage" tab.

0 Likes 0 ·
thomasw-1 avatar image thomasw-1 Vangelis Beligiannis commented ·

We seem to be on the same OS level.


My "limit managed battery charge voltage" ist also deactivated.
In the moment it seems to run quite smooth (55,7V absorption and 55,0V float). Contrary to you I don't have any bouncing AC-loads being indicated. The only limiting factor is, when load deviates activation of inverter and MPTT capacity is quite slowly adapting.

However compared to before the behaviour is much better ;-)

0 Likes 0 ·
Vangelis Beligiannis avatar image Vangelis Beligiannis thomasw-1 commented ·

One Last question thomasw-1. What GX device are you Using?

0 Likes 0 ·
thomasw-1 avatar image thomasw-1 Vangelis Beligiannis commented ·

Hi Vangelis,
I use a Venus GX with version 2.65 currently.

0 Likes 0 ·
Vangelis Beligiannis avatar image Vangelis Beligiannis thomasw-1 commented ·

are the ac loads appearing correctly when Fronius is regulated by the frequency?

0 Likes 0 ·
thomasw-1 avatar image thomasw-1 Vangelis Beligiannis commented ·

Hi Vangelis,
I have not observed strange deviations during frequency regulation so far.
Since Fronius uses different thresholds than recommended in "AC-coupled PV with Fronius PV Inverters" for frequency regulation, I use 51,0Hz for start, 52,7Hz for minimum and 52,80Hz for disconnect.
Maybe that changes scenario for you.

In a different forum I recognized that weird power measurements can occur, when you have a thyristor based full wave power modulation (e.g. to control a heater).
Do you have any thyristor modulation in your network attached?

0 Likes 0 ·
Vangelis Beligiannis avatar image Vangelis Beligiannis thomasw-1 commented ·

Thanks for the reply. I have a CCGX and was wondering if a cErbo GX woild solve the problem because it has a faster processor. I Will try it. I have no thyristor installed

0 Likes 0 ·
thomasw-1 avatar image thomasw-1 Vangelis Beligiannis commented ·

Hi Vangelis,
I don't think so. Have a look to my latest post (see below).
Save the money.

0 Likes 0 ·
Vangelis Beligiannis avatar image Vangelis Beligiannis Daniël Boekel (Victron Energy Staff) ♦♦ commented ·

@Daniël Boekel (Victron Energy Staff) Please follow up with the situation.

0 Likes 0 ·
thomasw-1 avatar image thomasw-1 Daniël Boekel (Victron Energy Staff) ♦♦ commented ·

Hi Daniel,

My euphoria was a bit early.
After 2 days expected behaviour the system unfortunately returned to standard behaviour. AC-OUT attached inverter is ignored after battery charging has been completed.

Snce the the modification of the absorption voltage definition does not substantially solve the problem, I would appreciate, if you could address the topic with your support department.


1617970794293.png

Thanks
Thomas

0 Likes 0 ·
1617970794293.png (27.6 KiB)
thomasw-1 avatar image
thomasw-1 answered ·

Hi Daniel,
I performed some tests this morning. Increasing the charging parameters on MulitPlus by 0,5V does not change any behaviour. The mentioned CVL is defined at 57,2V by BYD. Contrary to your statement it is definitely not the limiting factor. Maybe the described value of 51V was used in previous firmware versions, but not in recent or current versions of LVL BMS.

I made following observations.

  • Charging curve and AC-OUT inverter limit seems not directly controlled by MultiPlus. Defined charging current limits throttled my frequency shift follow directly the parameters of BYD CCL setting, even if voltages defined in MultiPlus are still not reached.
  • Frequency shift tested with Victron recommended voltage values and Fronius defined thresholds > no difference.
  • CVL is always higher than current battery voltage. Charging current adaptation and A-OUT inverter is only limited by CCL.
  • Even with 0,5V higher charging voltage definitions on Multiplus, the PV does not return, even if loads are >5.500W and battery voltage falls down below 54V.
  • Even if BYD provides information with CCL of 243A and CVL is above current voltage the AC-OUT inverter does not start-up immediately. It takes at least until battery is discharged below 95% SOC, sometimes even below 90%.


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.

Vangelis Beligiannis avatar image
Vangelis Beligiannis answered ·

Ηello Daneil. Unfortunately the problem appeared again. The Quattro locks the frequency above 52.7 Hz and the Fronius stays out even if the voltage drops under 53V. So my system is not working.

Your PV Assistant manual clearly discribes that the Quattro Will enter rebulk phase at 54V. This Never happens. The Quattro is probably faulty and I Will have to return it.

4 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.

thomasw-1 avatar image thomasw-1 commented ·

Hi Vangelis,
3 days ago I increased the absorption voltage to 56.5V (float still at 55,0V). With that config my system runs 3 days in a row perfectly.

Interestingly the Victron-Live-Page includes a section about that. When I setup my system I haven't still not seen that ammendment around off-grid config (perhaps I was blind).

Setting the Absorption voltage for off-grid systems to 56.5V is to ensure the PV Inverter Assistant works properly. Setting this voltage to 56.5V, ie a bit higher than the default 55.2V, prevents the Frequency shift algorithm to lock at 52.7 or 53 Hz, and thereby disabling AC-Coupled PV Inverters. The setting has no effect on the actual charge voltage as used by the inverter and other components in the system, since these are directed directly by the battery via DVCC.

So the recommendation of Daniel to increase the voltage by 0,5V was too low.

0 Likes 0 ·
Vangelis Beligiannis avatar image Vangelis Beligiannis thomasw-1 commented ·

This has been lately updated! It was never in that page!!!! I am not crazy either!

And Ok. That may work. But the problem with the Quattro is not the frequency lock.

The problem is that it does not enter the rebulk phase under 54 V as clearly described in the Assistant manual. This still does not work.

0 Likes 0 ·
Vangelis Beligiannis avatar image Vangelis Beligiannis thomasw-1 commented ·

This is a cashed copy before note(2) was entered. And the offgrid absorption line was also missing a month ago. I am crazy about following the rules. So I would not have missed that. 2021-04-16-2.png

0 Likes 0 ·
2021-04-16.png (209.0 KiB)
2021-04-16-2.png (84.3 KiB)
thomasw-1 avatar image thomasw-1 Vangelis Beligiannis commented ·

Thanks for confirming that I am still with my right senses ;-)

Potentially that change was added because of our escalation. However I am disappointed that Daniel has not come back with any additional feedback in this thread.

0 Likes 0 ·