I have a Multiplus ii parallel setup which works perfectly when using ACIN to provide all house loads. I have an issue that the grid use increases to a constant 10% of the load when I have anything connected to the ACOut1 connections.
When removed the issue goes away, the connections need to be physically removed.
This is an ESS system with ET112 monitoring the grid use.
Cable lengths have been kept idential from the ACOut1 and terminate together before providing loads.
Its almost like there is a miscalculation somewhere.
Hi, you will need to detail your environment more thoroughly and support with settings and charts.
Parallel systems aren’t easy and there many opportunities to get the install and setup wrong.
You should also start with a clamp meter checking ins and outs to see if the units are balanced.
I would expect any supplier is likely to ask for a more detailed troubleshooting.
Also parallel connection units so far between manufacturing dates is definitely part of the problem
(Looking at the serial numbers there)
So either you added at a later date and old and new model or did not mention to your supplier you wanted to parrallel them when you purchased them.
To add to what @lxonline has said, the newer unit also needs to be master, it sometimes helps, but if you are new to this, you may want to get someone to assist you onsite.
The units are 1 year and 7 weeks apart. I got mixed reviews when looking into this. Is this too far apart?
The new unit is configured as the master. The Grid code is set to G99 on both.
ESS Assistant configured on both. I went through all the victron training material and videos. I had this running as a single unit for over 1 year.
When clamping the DC side its pretty much spot on between the two units.
Clamping the AC-IN, It is pretty close.
Clamping the AC-Out-1 there is about a 75% difference between the two units.
In regards to the supplier they haven’t asked for anything yet. I have provided lots of screenshots and a description of the issue. I can provide more once they get back to me but its been 3 days.
I agree that seems excessive. The cable lengths are the same 2m terminating into a changeover switch. There doesn’t appear to be any settings in regards to the AC-OUT-1. It appears to be a on or off within the carbo.
I could try swapping the master and slave around, but I read the newer unit should always be the master.
You are describing what happened to us as well. Ran fine for over a year, added another unit, grid draw now much higher than before. I did use Nick’s nodered sketch to monitor the imbalance, but in the end just left it as it was.
HQ2207 and HQ2315 are 60 weeks apart. That’s more than 1 year. Usually max recommend time is less than 26 weeks (six months.) the hardware revisions matter as well.
You can sometimes “fine” tune by “tuning” cable lengths on AC.
Resistance is the key to balance.
Newer unit should be master.
I can never remember, does G99 or G98 limit the inverter power?
If they are balanced mostly, supplying the loads on AC IN, then it can’t be a major problem with the system age.
Your ACOUT is low usage, and deltas tend to be exaggerated, though it is still high.
If you lower the grid setpoint, does it respond?
Awesome, I disabled the flows i was using to look into this more but I will get this up and running.
As it stands I have disconnected AC-Out-1 but will connect it back up.
You need to add some increasing constant loads and track imbalances across a broader range. Low usage can skew the readings.
With a near zero setpoint mine shows huge imbalances on AC in, but it is only a handful of Watts.
Okay, this evening I connected back up ACOUT1 from both inverters combining in the change over switch. I have to physically disconnect the cabling otherwise the issue still occurs. There seems to be quite abit of imbalancer during load. Let me know what you think:
