Hi, we are based in South Africa and connected to Grid.
Occasionally (not always) when grid drops, inverter trips with Low Battery Alarm - last occasion, battery was at 100% late afternoon. My setup:
Quattro 10k, 48v; Firmware: 2653559
ESS DC Grid Feed in disabled.
Any ideas appreciated - Thanks
@grahamS Have a look at the battery graphs on vrm and make sure to use the cog to get a more detailed view of actual as opposed to averages of data to looks for possible related drops. How big is the storage and what type? Loads at the time? What are the BMS Charge and Discharge limits charts show at the time? What is the data capture interval on VRM? Even at the min 1min intervals it could have missed an event, but still useful for problem solving.
Yes, common when the grid browns out which happens often in SA, and the battery is undersized and struggles to meet the short transient power demand.
Also more likely to be seen on DIY or generic BMS batteries.
You failed to mention what the battery configuration is, size. Models etc.
Thanks for the reply @ejrossouw and @nickdb
Battery is a Dyness PowerBrick: 280ah, Max Charge/Discharge - 200A
VRM update interval is 1 min - can see when grid browned out and a few secs later, dropped - trip occurred at the brown out. Loads at the time approx 1200W with about 800W from PV.
Sadly, I don’t have access to the BMS but i’m using DVCC with settings below the Dyness recommendations.
Thanks
Check what DCL the dyness is sending.
VRM will not catch a brownout, I get them in Jozi regularly but my battery has no issue with it, just manifests as a noticeable dip in lighting and occasional sensitive gear restart.
Dyness can access the pack, would be worth a support call, but at 200A that is on the edge of suitable for a 10kVA inverter which can pull way more when it has to.
Smaller packs are fine when there is grid, but battery requirements increase in all guidance when off-grid, and people forget dealing with grid-loss, or bad grids, is effectively the same thing.
My own pack on 10kVA is rated to twice that.
Looks like a classic case of the BMS not tolerating the sudden surge from the dirty grid failure.
You will find it behaves perfectly on a clean shutoff if you drop mains.
I hear you re: pack size - have a 2nd PowerBrick which I still have to connect, As for DCL, also waiting for cables so can’t answer that - Dyness spec is 140A
You are spot on re: a “normal” grid drop, no problem at all and I’m inclined to agree with your theory that the BMS is shutting down.
All loads are powered from the Inverter and at the time, loads were 1200W - please forgive my ignorance but why would a dirty grid fail cause loads to increase and trip BMS?
Thanks
When running any load on grid, the mains connection provides the “buffer”.
Without a grid this is the batteries role. Peaks and surges often go unnoticed.
In a brownout the grid voltage tapers and the inverter tracks this downwards (the two are synchronised and the inverter is trying to remain in sync).
Until it hits the cutoff threshold and now has to switch back to its configured voltage.
The multis and quatros are low frequency inverters, their reaction time is longer to that of a hybrid/transformerless device like the RS.
This transition puts extra power demands onto the system which the battery must provide, it is a momentary blip but some BMS’s just don’t tolerate it and it is most common when the packs have been optimistically sized.
We saw plenty of these during load shedding, always resolved by more battery.
If you look at the battery guidance for sizing for on grid and off-grid/backup most manufacturers spec significantly more packs for the same size inverter (Dyness is unfortunately a bit sparse on this detail).
System design for a first world grid and cooler climates is very different to that for Africa. Our hot, harsh climate means systems derate so need to be over-specced, and our grid is at best poor. We need to adopt practices that are more aligned to off-grid than ideal operating conditions.
As you have another pack planned, you should be fine.
I suspect you will seem BMS events if you get access to the current battery.
Hi @nickdb , thank you for the detailed info, appreciated. Will speed up adding the 2nd pack
