question

johnparvic avatar image
johnparvic asked

Quattro / ESS Not fully charging from grid

Hi All

I have a Quattro 48/10000/140 in an ESS system with Cerbo GX, Smartshunt, DIY 550Ah LiFEPO4 battery (2 x 16s in parallel each with BMS for protection). ESS is set to use the Smartshunt as battery monitor.

There is 6.4kW PV on the DC side via a couple of SmartSolar MPPT 250/60

There is also 4kW PV on AC1 out with an SMA SB4000TL.

The PV is all working very well and the battery cells are very well balanced.

The problem I have noticed is that now we are in the middle of winter with little or no PV generation, the nightly charge (setting to "Keep Batteries Charged" via NodeRed) from the grid is not fully charging the battery - each day the SoC gets a little lower.

At first I thought that maybe I was not giving it enough time - but digging into the data I can see that is not the problem.

What is happening is that the Quattro completes the Bulk phase of the charging (voltage hits 55.2) but when it switches to Absorption the voltage drops to around 53.6V instead of maintaining the chosen Charge/Absorption voltage of 55.2V. You can see this in the VRM Advanced graphs below. The period from 02:38 to 03:37 is the Absorption phase but clearly the charge voltage dropped.

screenshot-from-2023-12-28-21-40-22.png


So, the question is, what did I miss in the settings, or is this a bug? Why did the Quattro drop the charge voltage? That must be what is happening because the behaviour when the DC PV is charging looks correct.

In the graphs above, the Bulk phase is terminating when the SoC is about 90%.


Software versions:

  • Quattro: 502
  • Cerbo: v3.00

Quattro is set to 55.2V Absorption, 53.6V Float, 1hr Absorption time, Fixed curve. Battery monitor enabled, State of charge when Bulk finished = 95% .


Many thanks

Kind regards

John

MultiPlus Quattro Inverter Chargerbattery chargingESS
2 |3000

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

7 Answers
Alexandra avatar image
Alexandra answered ·

@johnparVIC

What I have noticed is in an ESS system SOC (in case of grid faliure) to 100% is very different from keep batteries charged.

Which one do you have it set to? Switch to keep batteries charged if you have not already.

1 comment
2 |3000

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

johnparvic avatar image johnparvic commented ·

Hi @Alexandra

Yep - that is how I run the nightly recharge - I should have mentioned that!

I have Node Red set the ESS to "Keep Batteries Charged" during the cheaper rate tariff (E7).

Kind regards

John

0 Likes 0 ·
johnparvic avatar image
johnparvic answered ·

Happy New Year to all!

I've been monitoring this over the last couple of weeks and it is still the same problem, Quattro is not fully charging the battery from the grid because it is skipping the Absorption phase

Any ideas?

Kind regards

John

3 comments
2 |3000

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

Alexandra avatar image Alexandra ♦ commented ·

What is set as the system battery monitor in system set up? In DVCC?

Don't use the VE bus voltage when comparing data, rather use the shunt volt and amps to se what is happening.

What are your shunt settings? That will determine SOC triggering etc

0 Likes 0 ·
Show more comments
johnparvic avatar image
johnparvic answered ·

Hi all

I've been reading and Googling but clearly there is a significant gap in my knowledge.

In this ESS system, with Cerbo GX, Quattro, MPPT, SmartShunt, and Batteries, surely the Cerbo should be in charge of charging, telling the Quattro what to do. That seems to be the case when charging from the PV via the MPPTs - they are clearly in "External Control" mode and using the SmartShunt measurements.

When I set the ESS to "Keep Batteries Charged", the Cerbo should tell the Quattro what to do and use the measurements from the Smartshunt.

But clearly the Quattro is using its own measurements because it is terminating the bulk when the Quattro sees 55.2V (at which point the Smartshunt is only at 55.0V). It then skips the Absorption phase and goes straight to float for some reason.

Is there any way to see diagnostics from the Quattro directly? Or details of the communication between the Quattro and Cerbo on VE.Bus?

@Guy Stewart (Victron Community Manager) would you be able to assist?

Many thanks

Kind regards

John

2 |3000

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

johnparvic avatar image
johnparvic answered ·

Huge thanks to Mike Davidson at Midsummer Energy for helping me find the culprit, and discovering something weird about the way the Quattro/Cerbo/ESS combo works!

It turns out that despite having DVCC turned off, the Seplos BMS is still controlling the charge current from the Quattro.

At the point where the charge voltage drops, actually just after the Absorption phase starts, the BMS is telling the system that the max charge current is only 10A so the Quattro dutifully drops the charge voltage until the charge current is under 10A.

screenshot-from-2024-01-05-15-53-57.png

I've not found the reason why the BMS is doing that yet, but from what I've read it is likely to be that one of the cells has hit one of the limits set in the BMS.

The fact that DVCC is turned OFF but the Quattro is still respecting the current limit sent by the BMS is very odd. (DVCC was off because the Seplos logic is a bit wrong).

I've also just updated the Cerbo to v3.13 and the behavior is the same.

So, I'm now looking at what is happening on the batteries to see what is triggering the behavior and will update here when I have some answers.

Kind regards

John


1 comment
2 |3000

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

Hi @johnparVIC,

Thanks for following this with the update.

0 Likes 0 ·
johnparvic avatar image
johnparvic answered ·

So, after extensive investigation and experimentation, I can confirm that the strange behaviour regarding the Quattro not fully charging, was due to the CAN connected Seplos BMS throttling the charging even though DVCC was turned off and the Seplos BMS was not selected in DVCC or System Setup.

However, the Seplos BMS was actually doing the right thing so I decided to try turning on DVCC.
That opened a whole new can of works (pun intended) so I'll start a thew topic for that!

So, beware! If you have a CAN connected BMS connected but not configured as Battery Monitor, it can still affect the operation of the system.

(VenusOS v3.13 on Cerbo GX)

Kind regards

John

2 |3000

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

richoff avatar image
richoff answered ·

@johnparvic, are you using the newer Seplos version 3 BMS?

2 |3000

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

richoff avatar image
richoff answered ·

To follow up on my question above; I experienced a similar issue with my "CAN connected" Seplos v2 BMS... Current dropping to half every 2 or 3 minutes (only on one of my Split Phase Legs) when the Seplos is connected, usually starting about 4-5 hours into my nightly battery charge from the Grid... the logs show this clearly. Something I think Victron should be looking at, yes?

2 |3000

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