Help with thinkering/configuring - Victron/Grid/SSR

when I’m bored my brain goes weird places and I just can’t help my self :smiley:

So, as many - I am grid tied. Using Victron ESS and parasitically pulling 30-50w from the municipality in order not to trip the meter.
I saw in Venus GX there is an option to have a relay function under settings so I was thinking :slight_smile:

Firstly I am not sure what current is on the relay output - DC or AC? from googling it looks like DC but there is no tech spec to be found.
And then - where do I configure how will this relay work?

So then I thought that I can keep city grid off at all times and connect the grid only when I need it, say when my power demand is more then for an example 3000W

Here is the two diagrams I put together :slight_smile:

Please comment / criticise / break it


To be with Solid State relay in place:

Yes, I have started reading stuff :exploding_head:

So I got my victron config into VE Configure and added a relay assistant.
So will this initiate the relay function from Venus GX Relay 1 connection (I would use NO and Com).

Any guidance would be appreciated :slight_smile:

And this should work in conjunction with ESS - it will have grid to pull from it if required load is well - example on screenshot is 3kw but I would move that to say 5kw for over 5 seconds as I know that Multi can output more for limited amount of time

It would be easier to do this using a Relay Assistant on the Multi itself. It’s a bit tricky, you have to configure a general flag assistant, then configure the flag assistant so AC input is ignored (it pretends it isn’t there), and then you configure TWO relay assistants that sets the flag according to the load.

You can also look into the old deprecated Hub-1 assistant, which also has a feature to disconnect from the grid.

And finally, there might be a better firmware option for this in future, but again, this is merely rumour now.

My grid setpoint is Zero, thank you CoCT lTron PAYG meter. Yes, I’m charged for watts going back, but per annum, it is cents on the rand. Methinks a larger ah bank will reduce that, will see.

Furthermore, disconnecting the inverter from Eskom would stop the grid-tied feedback to the house and that will cost me a lot more than the teeny bit fed back, or the 50w used 24/7/365 as is normally set.

And if I was to use AC-Out2, to power the entire house, the inverter will switch off as I will exceed 4kw.

Unless I misunderstand the plan … then correct me, please?

Furthermore, disconnecting the inverter from Eskom would stop the grid-tied feedback to the house and that will cost me a lot more than the teeny bit fed back, or the 50w used 24/7/365 as is normally set.

Yes, I guess you make a good point with this…as title states, I was merely thinkering :smiley:

There is some work ongoing to improve the pull-back. Again, no timeline on this, but the requirement from Germany is that you need to pull back 90% of the “durch den Sprung verursachte Fehler” within 3 seconds.

(gotta love German, the “through the jump caused mistake”).

This won’t stop a Conlog meter from tripping though!

so they basically demanding that the excess power be “contained” within 3 seconds…:face_with_monocle:
just not too sure what to do with it as from what I know, that excess energy is simply not needed anywhere in that moment :roll_eyes:

It describes the requirements of a limiting device. If a load switches off and there is a sudden “Sprung” in the amount of power fed into the grid, you need to pull back 90% of the way within 3 seconds. You also need to be under the line within any 3-second rolling average after that.

Are they basically talking about “integrating ziehl” in every inverter for their market or?

There’s some confusion here. At least I am confused by the question.

Victron inverters (when using ESS) already have a grid limiter. It is fairly slow, which is why people think up all sorts of ideas to prevent those accidental feed-in events. There is a German requirement to make it faster, which means in future it will probably be faster than it is now. That is all I was saying really.

The Ziehl device is an anti-islanding switch. Anti-islanding is about preventing grid feed-in while the grid is OFF. What we are talking about here is preventing excessive feed-in while the grid is ON.

Ok, then I misunderstood the purpose of this device completely… :roll_eyes:

But another question, will they be able to do it via software?
Isn’t the reaction speed of handling excess load hardware limitation?

From bits I picked up over time, and I stand to be educated more, it starts with the software on the Venus/Cerbo to process the calls and action faster, and not throw the baby out with the bathwater doing that.

Partially. Some of it has to do with the actual speed of the VE.Bus, which I will not get into the details. But with later hardware you can run that faster (aka, if you have a Cerbo). A bigger limitation is that the grid meters normally used (Carlo Gavazzi) only give you a new value every 600ms to 750ms, which takes a big chunk out of your time budget. But there are meters that are faster.

Another limitation is that the older CCGX doesn’t have the processing speed, and most of the older ones also only have an mk2 chip built in (you need an mk3). So there is enough room to make it faster, but it will be hardware dependent.

Also, note that if all the loads are on AC-out, in other words it is not a system where you have loads on the grid side that is being compensated for by feeding back, then you already have a 3-second response.