House runs off “uninterruptible” output of quattro, mains to quattero. “summer mode” export most solar, charge up on cheaper night electricity; “winter mode” mostly use solar, top up from grid cheap night.
Have tried dynamic ESS a few times in “Export mode” but this thrashes batteries to 30% charge and would likely cause problems in winter.
I do recharge on timed charge to 95%, sunday 100%. Unfortuinately the solar does a full 100% charge with balancing every day in summer, and sunny days in winter. This is excessive. Ideally I would like the solar to charge to 95% except for balancing once every 14 days (as per setting in Quattro).
I am more than a little surprised (given its in the Quattro setup) that this seems to be impossible to do in the basic (ie as supplied) system setup.
The main way it is done is capping the voltage in DVCC and then removing the voltage cap when you want the battery to go higher i.e. to the full absorption voltage.
Its not fone natively as most bms self manage and not every one wnats to do what you are wanting there. So its custom now
Since my entire system is Victron, including the BMS (VE bus BMS V2) it really should be possible to avoid full balancing/absorption daily from the MPPT/Cerbo/Quattro system. Note in the Quattro its set to happen every 14 days, but this doesn’ttr seem to have any effect.
So what you are actually saying is that with an all-Victron unless you accept dynamic ESS (which has other flaws) there is no way to prevent rebalancing every day?? This would seem to be a major flaw as it doesn’t actually meet Victrons own preferred schedule.
To suggest that a solution is to reconfigure every 14 days to manually do this is in reality totally unacceptable. We have a cerbo that should be doing this, and this should have been available for years to meet Victrons own battery spec.
That’s not what the battery manual says, and its not the default setting in the LiFePO4 battery setting on the Quattro (its 14 days) and its not recommended anywhere for this battery type. Indeed its said in many places that charging to 95% with only occasional full balancing charges is preferred to maximise battery life. You will of course be well aware of this.
I see you have some experience with node red frommother posts is there a reason why you haven’t implemented something there for this?
The 14 days absorption is the re-absorption meaning when it has been sitting on float. Which you are not doing.
But you also have repeated absorption set at 24hours. So every day.
That 24 hours is the MAXIMUM tine period allowed. When I set this from the default (something like 1 hour) it made no difference.
No, I have never used node red. I do use Home Assistant though. Sadly did that a while ago and can’t remember exactly what I did. Directly worked with MQTT I think. I don’t recall if controlling balancing is an option. I use it to pull off power and energy figures.
That belies the point though. It should be quite simple to set a balance period and not balance the cells in between. Clearly each system may have differing appropriate periods but that is for the user to decide. This is particularly important when its good battery management to use balancing appropriately.
At the end of the day no matter what you program the ve bus to, bms will override it. Unless you cap it or control it separately in dvcc where the lower of the two will be chosen.
What is doing the BMS balancing setting? Clearly not the VE bus BMS V2 as its dumb. It looks as if its the MPPT, but this claims to be on external control so its either the cerbo or the Quattro. Whichever it is it should be possible to set this to maximise battery life.
Yes set optimised with BatteryLife (!!).
So, what you are saying is that there is no way a victron system can be set up optimally.
In the battery manual, the cells start balancing from 3.3v per cell.
So to make the system skip absorption aka balancing, you would have to use DVCC (not ess) to cap the voltage at less than 52.8v.
And then lower voltages on the bus are so inefficient. I don’t think it is worth it.
But (see the graph above provided for your edification) its balancing at 3.55V/cell (56.86V) whilst 54.2V (3.4V/cell) seems to be exactly the absorption and float voltages specified as the default for their batteries in the Quattro.
Its very clear that providing you are not thrashing the battery (that is basically one shallowish charge and discharge a day) a full balancing (which is ‘absorption’ because Li need balancing but no actual absorption) then once in 14 days is sufficient. If not then shorten the interval.
I really do not think either of the two experts replying have grasped the basic principle for Li batteries.
They do not like being either overcharged or over-discharged. Both cause permanent damage (much more so discharge to be said).
In a pack of 16 cells some are better than others and will have slightly lower capacity (which doesn’t actually matter that much per se) and slightly lower efficiencies (which does) such that over a period of time and/or a number of charge-discharge cycles the ‘worst’ cell becomes progressively discharged until it falls below the save minimum voltage and the BMS detects this and trips out (bad), even though the entire pack is still within voltage limits.
To avoid this when the battery is at the “nominally fully charged voltage” charging is continued (note the sudden step in the voltages in the graph attached). So what is going on here? The SOC falls gradually during the afternoon as the battery powers my house until at 00.30 it starts its charge (cheap electricity). The charge power (blue) rockets uo and the battery voltage moved from ‘discharging (53V)’ to ‘charging (54V)’. Initially the voltage is stable but as the battery comes to full charge the voltage rockets up to 57V whilst still taking full charge power but this rapidly drops when the battery reaches full charge (for that voltage). This voltage is maintained fpr quite a long time. Inside the battery is a device (not the best now available) which essentially maintains the cell at a set maximum by bypassing current until its deemed that even the worst cell is fully charged (by tim I think). The battery is now balanced.
It does this every day (sometimes twice) which is unessecary and not great for the battery. It would be better to stop charging when the charge power falls (just vefor the vertical white measurement axis).
So I am not very impressed with Victrons battery management, to be honest. Its leading to more battery thrashing than is really required. It would also be nice to retrofit inverter balancing which can actively balance cells as these are pretty cheap these days.
I guess I will have to accept that Victron have fallen a bit behind in this area lately.
Holding it at one voltage (within operating rage) does not age it. But this is almost impossible to do in operations.
From the studies that i have seen and read - it is energy in and out that ages a battery. Not any particular voltage or depth of discharge or range of use. So sipping out the top or sipping out the middle wont make a difference in aging (but it does in efficiency of conversion)
The graph does not look different from others i have seen from other manufacturers. So i am not sure what you are referring to? It would seem then at least 4 manufacturers that i am familiar with are also behing the curve by that reasoning.
in my setup, the DVCC max charge voltage is set to 55.6 V. This equals 3,475 V per cell.
Balancing starts at 3,45 V at cell level.
100 % SOC is displayed, if the battery remains at or above 55,2 V (3,45 V x 16) for 30 minutes.
I have absolutely no concerns about operating the battery like this during the summer months and charging it to 100 % daily, because the maximum voltages are relatively low.
Maybe my setup suits your application as well.
Best regards
But I see your point: Setting DVCC max charge current to i.e. 10 A, still allows the MPPTs to charge at full power. This is unexpected for me too. In my setup, there are 2 MPPTs: One 150/60 and one 150/85 VE.CAN, with 3x Multiplus-II 5000.
yes, that is probably about as good as it gets given Victrons ability to distribute energy within the system.
Charge current control should be exactly that, maximum charge current. Not to be confused with MPPT current which many well be substantially more if its effectively directly powering grid export and/or consumption.
Do you have a reference in the manual for balancing starting at 3.45V/cell → 55.2V (or is that settable somewhere) and that balancing is considered complete after 30mins of same.
Also where are these settings made? My BMS is set to the smartshunt here as reported in Victron Connect. I am not sure any of these are actually used, although they should be. Really it all seems to be a bit of a mess with different devices doing their own thing and the battery balancing out because that’s all it can do.
Probably I should not change anything for 7 days and see how it pams out over time.
“Active” cell balancing begins when the first cell reaches 3.3V or less for severely unbalanced batteries.
“Passive” cell balancing starts when the cell voltages are 3.50V. This can happen only during the absorption charge stage, as during this stage, the charge voltage (14.2V or 28.4V) is high enough to for the cell voltages to also be sufficiently high to allow smaller cell differences to be corrected.
The cell balancing process is nearing completion when all cells have reached a voltage of 3.55V and the charge current has dropped below 1.5A. Balancing is complete when the charge current has dropped even further.
If you really want/need a programmable bms that balances or not how you want - then a custom bank with a JK or other after market bms is the way to go. For a company to warranty a turnkey product they wont allow that.
No, what I need is to know how victron are doing it, what is the spec and in what ways (if any) it can be modified. Which devices take precedence (or rather which settings in which device) when there are multiple ones and how they can be set.
For obvious reasons I have set my BMS as the Smartshunt as its the only device which actually reads the battery status accurately. However it would seem that many of the settings here are being over-ridden by something else (probably the quattro) despite the quattro not actually knowing (as you and other victron ‘experts’ have pointed out) the SOC or charge/discharge currents to basic accuracy.
It looks as if you do not know either, indeed does anyone know? My system is pretty basic. MPPT, Quattro, battery, smartshunt, cerbo. Which setting where control charge voltage, charge current, balancing time etc?? Is this always the same or do they switch about depending on how they feel?
All answers i have shared so far have been from personal experience and have worked. I dont have just 2 or 3 systems and not just 3 minutes of experience.
Dvcc. I have used the voltage cap many times. It works.
There are quite a few things. The main one is the bms. For example Pylontec settings in the ve bus are 51v and 52v however (depending on the fw version on the pylons) the system will charge to 53.4v or 52.6v. they will sometimes sit for 4 hours at absorption (bigger bank) some will be there an hour.
The second thing is grid feedback, there are things that change for example the charge current limit gets ignored this is mentioned in the ess manual as intentional. You can cap it all you like but if you have feedback on it menas nothing.
So at the end of the day, the result is a mix of everything ounched into the system.
The second thing that changes it the ve bus raises its voltage 0.4v above the max voltage which for some batteries does not go well - the voltage cap in dvcc still works here.
A third thing is this its main reason for existing is to make sure that the battery gets to full charge every day. Also in the ess manual. So you have told the system to do what you are wanting to prevent.
Nope. You have a victron battery with the bms it is the bms. None of the settings for bms and how it handles its cells are editable.
I do not believe this is correct. The internal BMS in the battery seems to be a dumb balancing circuit and nothing more. It handles cell over and undervoltage alarms (which should stop over/undercharge of individual cells by charging devices, usually by tripping them off). Its not an active one and nor can it switch the battery out of circuit. Those devices are more modern and found in later battery packs by others (they also have active balancing). The actual charging control is usually done by a quattro/multi (or MPPT) or a smartshunt. That’s because the Quattro is is a position to dump energy to grid and thus leach off excess charging power from the battery and of not it (possibly via cerbo) restricts the MPPT output current by setting max current. Indeed its possible to set the active BMS in the Cerbo (see pic). I use the smartshunt because the quattro is hopelessly inaccurate for SOC%. I have no idea what ‘automatic’ does.
