System:
Quattro as ESS
Generator connected to AC in1
Grid connected to AC in2
Standard loads connected to AC in2
Fronius Symo connected to AC in2
Critical loads connected to AC out
No DC MPPTs
When the generator runs under load once a month for testing, the Quattro switches the power to the critical loads. The grid continues to supply the normal loads. The current generated by the Fronius Symo (AC in2) charges the batteries and feeds into the grid if necessary.
This is how it should be.
However, when the batteries are full, the Fronius Symo is throttled down, even though it is connected to the grid input AC in2 rather than the generator input AC in1. If the control is switched off by disabeling “sunspec,” the battery voltage rises above the set value until the BMS shuts down.
For a Multi with only one input, this behavior would be correct.
DC MPPTs, if present, would also have to be throttled down.
With a Quattro, however, the system does not take into account that it has two separate inputs and that these can be configured differently.
An interesting situation. Ithink the geneal assumption (in controls) in running is if the generator is running and connected there is no grid. But this is not the case with a maintenance run.
Do you have a grid meter before your loads and fronius? And is grid feed in allowed or not allowed?
How long is it running for? Usually a for testing scenario is warm up, short run, cool down. Not usually a full battery charge type thing. So possibly a 15 or 20 minute run.
Yes, grid feed is allowed and works fine when the batteries are’n full yet. (when there is more solar power than the Quattro can put into the batteries)
No, there is only one grid meter at the input of the grid. The normal loads an the critical loads are monitored by the Quattro, the Symo tells his power by “sunspec” or api.
We conducted further tests yesterday.
Therefore, I need to clarify my previous statement a bit:
Even if the batteries aren’t fully charged yet, the Quattro reduces the output of the PV inverter at AC-in1 as soon as a generator is running at AC-in2.
However—and this is where I really don’t understand the Quattro anymore—increasing the “Grid Setpoint” causes the inverter’s output to increase again.
So the Quattro “knows” exactly that the inverter is still connected to the grid, even when the generator is running. Why it then reduces the inverter’s output is beyond me.
This information should be of interest to people here who use the Quattro to charge their batteries using V2G.
The grid set point is global. So its not that it is differentiating at all, but the control is really about the one that is connected to the victrons input.
A workaround would be to do the generator test in not in solar production time.
But why is the Quattro able to feed the generated power into the grid only when the grid set point is raised, but not otherwise? After all, the grid set point is a lower limit; feeding in more power should always be possible.
Testing the generator when no solar power is being generated would be possible, but ultimately that’s not what Victron is known for.
It would be a better solution if Victron modified the software so that both inputs on the Quattro could be used without restriction — that is, without any interaction between them.
As you are using the set up, it is not a normal use case scenario.
ESS current control algorithm was designed with a single-input model in mind because the inverter can only actively synchronise + connect to and use one input at a time
I do hear ehat you are saying.
When the generator is active on AC-in1, the ESS logic throttles the AC PV on AC-in2 to prevent overcharging the batteries — but it fails to account for the fact that on a Quattro, AC-in2 is still a live, separate grid connection where excess PV power could simply be exported.
The software doesn’t treat the two inputs as truly independent, so the grid feed-in capability of the AC PV is suppressed during generator operation.
But again this is a niche / abnormal - less than 30 minute use case. Which can be approached differently with an adjustment in how the maintenance run is implemented.
Thank you very much for your prompt response and your understanding.
Please try to understand our situation as well.
The Quattro is installed exactly as intended and in our view, this is a normal use case scenario if you want to have a generator.
Nevertheless, the customer contacted us because his batteries were overcharging during every generator test until the BMS shut down. (The tests were usually conducted in the afternoon when the batteries were already full.)
Fortunately, he had a Fronius Symo inverter, so by activating “sunspec” communication, we were able to ensure that the inverter is now downregulated.
With an inverter lacking communication capabilities, even that would not have been possible.
Understandably, the customer is not thrilled with the current solution either.
Yes, admittedly, this scenario does not occur frequently, but it has serious consequences if the inverter is not capable of communication via AC-in2—which, strictly speaking, it is not required to be.
The ve bus logic is acin 1 overrides ac in 2. In other words ignoring that ac in favor of.
Since feeding back into a generator is not desirable and the position of the fronius as on an ac input not differentiation between 1 or 2th3 control is global.
Priority for control is on ac1 which is why the generator connected there will ovveride the grid connection. The system has been told that is what it should be paying attention to above anything else.
On the ac input, the system is using modbus for control not the actual ac voltage connection.
The frounius is a symo so three phase the inverter system you don’t mention as being such?
The reason why node red and other custom integrations are open on victron is for systems like yours which need some specific functions to happen at certain times. The solution is in your hands.