I’ve got a Victron VM-3P75 energy meter at my grid supply and 2 Multi RS units that sit after it. Both use the same battery. Grid feed-in is disabled. Both units are on the same System instance, Both are set to “stand alone”. System is single phase.
If I set one of RS units to use the energy meter, I can power devices that sit between the energy meter and the RS and everything is fine.
If I set the second RS to use the energy meter as well, then the following happens!:
Both units start off supplying the devices between the energy meter and the RS units at around 50%/50% (happy days).
Slowly, The wattage going out of one RS starts rising and the other starts falling to the point where one will eventually go negative!
If I leave it long enough this gets to the point where one unit is putting out say 3000W into the gap and the other is consuming it back into itself.
the Batteries are being drainded by one and charged by the other, despite BOTH being in “Optimized without battery life” and not wanting to charge due to their current charge levels normally.
I could understand one unit being unbalanced with the other to the point wher they’d be supplying different percentages, but why is it goiing below 0% and feeding backwards and even charging the batteries when it is clearly not set to do so?
Does anyone have any suggestions? I really would like this to work?
I’m happy to set up the grid codes on the units and set the feed in to 0 if that helps but I think i tried that before and it didn’t make any difference.
@It has been covered elsewhere, but parallel operation is not supported, the device is not certified e.g. in the UK for connecting it to the grid, neither is running two systems of a single battery or sharing of energy meters so not much chance of success if you ask me.
You are right there, I’m getting the impression my chances are slim of getting anywhere with this. I did think the unit had finally got it’s UK approvals but you are right, it’s only the EU so far which is disappointing. As for the Parallel, it is now supported but only on AC out and only with AC in specifically disconnected.
Why it is doing what it is doing when using a single energy meter with two RS’s beats me because this is something that isn’t an unreasonable ask, even if it was a generator as the source and not the grid, or this was the EU, someone must want to do this, and if they do they’ll get the same issue I get.
What I can’t understand is why there doesn’t seem to be a mechanism to report bugs apart from here where we just get to tell each other it doesn’t work and don’t hold your breath?
If I set the battery charge current limit to 0A in the RS units, then it doesn’t feed energy backwards from one into the other, so it is knowingly doing this in a bizarre loop. This doesn’t really help though as I can’t then have any solar connected or charge the batteries without some fancy NodeRed going on, which I do already have for charging but not that could ever accomodate the solar generation that is comng in.
I could get this to work if I bought a separate MPPT and didn’t use the internal MPPTs at all but that makes no sense.
The other thing that makes this issue worse is if you have any load on the AC out of either unit, it then is much more likely to get this weird inbalance and start feeding energy around in a loop. If you have nothing on any of the AC out connections then things are more stable but not perfect and one of the units on occasions will drift to charging the batteries with a few 100W and the other will discharge the batteries with the same energy and pass it around the AC in connections for your entertainment!
This is clearly a software issue as I’ve said as neither unit should be charging the batteries at all.
Change the one RS to a different can instance. The only catch is that the bms lost will appear on the one that is moved. I am not sure how the energy meter is affected on the moved one either.
Are you trying to feed loads before the inverters as well?
You are trying to do something that is specifically documented as a no right now. What you are experiencing is (part of) the why.
Getting them to evenly supply will also probably be unrealistic right now, but maybe we can avoid the charge and discharge over the bus.
Jambos situation is slightly different to yours as his is configured as one system.
Thanks, I tried that and indeed it does isolate the issue.
It does however as you eluded to, remove the unit on a different instances ability to power devices on the AC in side of the RS between it and the energy meter.
This does look like a simple firmware issue with energy accounting with multiple RS units on a single energy meter rather than something more complicated and hopefully they can resolve it if there is a way to pass this on to support.
As for balancing the loads, I’m generally less worried about the accuracy of that as long as they are at least all going in the right direction.
Thanks for your insights on this, that was really helpful.
When they are one complete system, it seems to be working. The issue is appearing on systems that aren’t the prescibed set up. i.e one system on a GX (in your case it’s actually 2 individual systems which is documented as one is controlled the other is monitoring only usually)
So your system is not longer an expected plug and play set up and will require some custom set up. (So maybe node red is more your new friend)
As in your case and the other poster that has 2 of the three inverters.
The fundamental issue with your set up…
Each Multi RS sees the same meter reading and independently tries to correct to a zero setpoint. Neither unit knows the other exists in any control sense. So if unit A is supplying slightly more than unit B, unit A sees “grid is importing” and backs off, while unit B sees exactly the same reading and does the same — they end up hunting against each other. And then settling where you describe as ending up.
Do you have the output parallelled or just the input?
I think you said individual circuits somewhere, so is the issue that there is mot enough solar to cover the load before with one inverter?