I am planning a new DIY 24 V LiFePO4 battery system for my existing Victron ESS installation.
Planned setup:
Victron Cerbo GX
Victron MultiPlus / ESS
later possibly a second MultiPlus on another AC phase
2 separate LiFePO4 battery packs in parallel
each pack: 8S EVE MB31 314 Ah
approximately 25.6 V / 314 Ah / 8 kWh per pack
approximately 16 kWh total
each pack with its own Seplos BMS 3.0, 8S / 200 A
Seplos BMS units in master/slave configuration
slave BMS connected to master via RS485
only the master BMS connected to the Cerbo GX via CAN
The cells are rated for approximately 160 A continuous charge/discharge per pack.
My current MultiPlus can draw approximately 135 A DC at full load. If I add a second MultiPlus later, total battery current could reach approximately 270 A, or about 135 A per battery pack.
My main question is about CCL/DCL aggregation.
If both battery packs are available and each pack allows, for example:
DCL = 150 A
CCL = 100 A
and the Seplos master BMS aggregates these values and sends:
DCL = 300 A
CCL = 200 A
to the Cerbo GX via CAN:
Will Victron DVCC use these aggregated values as the charge/discharge limits for the complete battery bank?
And if one battery pack is switched off, disconnected or enters protection mode and the Seplos master BMS reduces the limits to:
DCL = 150 A
CCL = 100 A
will Victron/DVCC immediately follow the reduced limits?
Ideally, I want the two physical battery packs to appear to Victron as one managed battery, with correctly aggregated limits.
I am also considering a Victron SmartShunt in the common negative connection of both battery packs.
Can the SmartShunt be used as the main battery current/SOC source, while the Seplos CAN BMS remains responsible for:
CVL
CCL
DCL
battery protection limits
Or is it better to use the SOC reported by the Seplos master BMS when using a managed CAN battery?
I would especially appreciate feedback from anyone running multiple Seplos BMS 3.0 units in parallel with Victron.
If i understood correctly, you already own a 24V Multiplus? Otherwise you should consider using 48V for ESS. Even if you already own the MP, you should consider switching. Doubling the voltage will half the currents, for the same power.
As for the actual question, i dont know the seplos BMS in detail, but if it works like others then the master BMS aggregates individual packs into a total, and sends that data over canbus. Which is the only way Victron can handle it. One Victron system can only handle one Battery, aggregation needs to happen elsewhere.
The Victron system acts upon whatever the main BMS sends. If the BMS works correctly, then the DVCC signals will reflect the actual, total battery status, so if you have two packs connected then the CCL for example will be 200A in your case. If the secondary pack is switched off, then the CCL should drop to 100A. But i dont know if that is also the case during an error or protection condition. Possible that it drops to 0A then.
As for the shunt, i cant comment on that, i havent used both a shunt and a smart battery in the same system, only separately.
For context: since 2022, I have already been running a fully installed 24 V Victron ESS system with MultiPlus, Cerbo GX, a Carlo Gavazzi meter, MPPT, DC busbars and a battery bank that has so far consisted of around 650 Ah of AGM batteries.
The AGM batteries have now reached the end of their service life, so they are to be replaced entirely with LiFePO4 batteries.
At present, converting the entire system to 48 V is not an option for me, as the complete 24 V infrastructure is already in place and has been installed properly. The MultiPlus, cabling, busbars, protection devices and other components would also have to be modified in part.
I therefore deliberately want to remain with 24 V and am looking for a neat solution with two separate 8S LiFePO4 packs of approximately 8 kWh each, each with its own BMS and connected to a shared DC busbar.
What matters most to me is that the two BMSs work together correctly and that Victron sees only one combined battery via CAN, with the CCL/DCL values aggregated correctly.
That is precisely why I am currently trying to determine which BMS system would be the most suitable for this two-pack configuration.
I’m running two 15 kWh 48 V packs in parallel here — both with Daly BMSs, connected via CAN. The Victron system is running DBus-Serialbattery and Batteryaggreggator — that then regulates things down to the level of an individual cell.
It works relatively well, but it’s also a lot of tinkering.
Would the costs really be that high if you switched to 48 V? You can keep most of what you have, surely; a second-hand MP2-24 is worth almost as much as a new MP2-48.