I have a SOC reporting issue my grid charger SOC % is never balanced with the BESS SOC level
Grid SOC can shows 70% and BESS show 90%, other times. On other occasions grid reports 20% SOC while BESS shows 70%. The problem then is that the grid cuts in at 25% so I am paying for power before it is required from the grid.
How do I manually adjust the Grid reference to match the Bess reference – assuming that the BESS reference is actually correct? How do I determine what is the correct %
Welcome aboard, John @Victron1
Can you tell us something about what your system includes? eg Multiplus II, Cerbo GX, what you have for batteries and BMS, that sort of thing.
Are you getting the info from a display, VRM, or where?
That might help us understand what’s going on.
CPS_SP298_proposal-Specs 4-51kWp SOLAR-130kWh_day_North30deg.pdf (416.9 KB)
Thanks Buzzy this is the specification list does that supply the information your require.
Redei system had to be updated as they had failed to fully calculate the load draw and the system kept tripping. So they took two inverters out and put in 3 direct to AC inverters in their place in order to match load. This has improved the system but I still have to manage load and shed when required.
The problem is worse in poor solar conditions and I have to tell the Battery charger to charge off peak rate between 11pm to 7am to appropriate set points.
I change the SOC level depending on what power we will need the next day and what the sun light is likely to be.
I also set the battery charger to charge in peak rate period 7am - 11pm this is often varied but generally set to cut in at 25% SOC.
The problem arises when the battery SOC charge shows say 20% while the battery BMS shows perhaps 40 or 50 % the battery charger will cut in.
Hope this is the info you require
There’s a lot in that shopping list and a price tag to match! Is there a system schematic diagram? It’s hard to know where to start without it.
Some thoughts - maybe not the best ideas, and you’ve probably already thought about this:
Honestly, for the $200k you spent, I’d get the REDEI folks to pay a visit - even if they won’t do it for free, I’d be inclined to get them out there, as it would likely be a modest incremental.
Good luck mate!
Whats the reason to do the double conversion?
I guess the Battery system is connected to the GX via Canbus? I guess theres a GX each for the charging and inverting system side? Since the SOC is vastly different i also guess the BMS Canbus is not daisy chained from one GX to the other.
One issue here is also that both systems lack the DC side. So both would need “Has DC system” activated. Since for each system, the other one is a DC load/source. Thats for example where the additional 3kW would go in the charging system, 6.7kW come from the mains, 3.6kW go to the battery, 3kW currently vanish into thin air according to the system. If you enable “Has DC system” then the difference will be counted as DC loads. Which is kind of true.
But, the main point would be to daisy chain the BMS Canbus, so that both systems can read the batteries SOC and not need to rely on their own internal energy counting.
Im wondering why two GXs have been chosen for this, since an extra VE.Bus system can be attached to a single GX by using an Mk3 adapter. It will then follow the ones GX DVCC, making it possible to use additional MPs as chargers only.
Sometimes the rwali time data misbehaves.
Screen 1 - no inverter box at the bottom.
Battery data not correct compared to the secon and it should be impossible to jump SOC like that.
This leads me to thing old data or the vrm is reverting to a previous something or something is not right when you compare it to the bottom screenshot.
So a streaming issue possibly.
If that was happening on the GX which is where it would be preferable to confirm if you have odd data; then i would say that the batteries or whatever is sending battery data is dropping data packets or not sending them.
Either way i would not view that as a problem until the local data confirms it.
Dear Buzzy thanks for your reply very kind to share your time and thoughts. Getting Redei out is the issue. There has been a litany of problems, two auto change over switches from Bess to Grid causing switchboard fires, the last at the house on Tuesday 12 Jan 2026 15:57. I am still waiting for Daniel to come and repair hance my constant power management as I cant switch from BESS to Grid. We had to invest in this as the Nursery is on the end of a SWER with a regular drop out and only 25kva supply and we needed to add irrigation pumps. I will try and find a circuit drawing if you have time to look at it. Thanks again for your help.
These are some of the dwgs I have if they cover all you need to know if not I have more.
Please not they changed to grid charger rather than generator set(fortunately) Also change 2 of 4 dc inverters to AC as per previous picture.
Yes when it get 100% charged both come back close to matching SOC. But by the next day long way apart again with consequential grid use
CPS Comms Connections-005.pdf (692.5 KB)
CPS AC Wiring Diagram-002.pdf (268.9 KB)
Thanks John for the drawings. Some thoughts:
Batrium battery management system:
Drawing 005 shows REDEI batteries are built around a Batrium system: Getting Started and Batrium is known to work with Victron. The Batrium Watchmon device is responsible for communication between the batteries K9 Cell Monitor control boards and the Victron Cerbo GX “brains” that controls everything and sends data to the VRM portal.
I’m not sure what version of the WatchMon you have. Some info on Wiki: https://wiki.batrium.com/
It looks like most of the WatchMon units have some kind of data logging capability - perhaps getting access to that would help diagnose what is going on.
Perhaps if you don’t get satisfaction from REDEI, you could reach out to Batrium for advice or info on accessing the system to see what’s happening.
Missing Data: Both @lxonline and @chrigu point out that something is odd where the information from the DC side of things is not right/missing/unreliable. My guess would be communication or cabling problem between the Cerbo GX and the Watchmon, or the Watchmon and K9.
How many GX devices: Why do there seem to be 2 GX devices in the mix, instead of 1. eg the VRM portal should normally have everything from 1 system. Drawing 005 only shows one Cerbo GX, so where’s the other one, and why have 2? There could be a valid reason.
Any chance there’s one for the nursery and one for the office/house ?
Accurate SOC: I don’t see a battery shunt in these 2 drawings. e.g. either a Victron SmartShunt or a Batrium ShuntMon. The current to/from batteries can be measured more accurately if there is one. Maybe there is one on one of the other drawings, or maybe there was a valid reason to not have one.
If there’s a simple communication problem with what you do have, then that could also account for SOC problems.
Other than these few puzzling items, it looks like REDEI’s diagrams show nothing weird or concerning.
Thanks Buzzy I will check the dwgs to check as suggested.
I have forwarded your thoughts and words to Daniel to see what he thinks. He will know for sure, he just has so many pressures on him so your help will help him help me.
Regards
