question

Mark Frencken avatar image
Mark Frencken asked

Why is ESS (and dynamic ESS) not taking the AC side solar production into account

As for some reason the bug report system on Github is not functioning for me at this moment I'm posting the 'issue' filed here, perhaps the issue I have is recognized by someone since it seems like a very basic thing that I likely missed somewhere in setting up the system.


In short; in the Dashboard the minimum power flowing into the AC side loads is never below 0. This seems to have the effect that the DESS system is never taking the AC side production into account and dumps all energy produced there to the grid. AT the same time all energy generated by the solar connected to the DC side is routed to the battery.


This happens every day, but I'm now posting a bug so you can look at really any date.. The issue is the following: there is ample solar power, the viltron charger connected to the battery buis charges as expected (prices at lowest, battery SOC low) but the energy from the AC side solar is completely dumped into the net. This happens every day. If I look at the schematic overview menu, I see that the system is not able to calculate that the AC side is generating power. In my opinion, it should be able to figure out that energy cannot occur out of thin air. To be more precise: I have 5 kW going into the grid, I have ~0 kW going into the viltron ESS from AC and I have ~700 W going to the critical loads. This does not add up. The issue is that the system limits the numerical value of the AC side loads to 0, it apparently does not allow this number to go negative. This then leads to erroneous predictions and calculated values. As a consequence, in DESS mode the energy is not used.



DESS
2 comments
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

Matthias Lange - DE avatar image Matthias Lange - DE ♦ commented ·
What components do you have?

Can you please add a wiring diagram and some screenshots?

You might have to upload the pictures somewhere else and share the link here.

0 Likes 0 ·
nickdb avatar image nickdb ♦♦ commented ·
To add. If you are using any betas, there are separate topics to report issues.
0 Likes 0 ·
4 Answers
prein avatar image
prein answered ·

Solar is only correctly taken into account if it is metered directly. That can be done by either place another ET122 at your solar output. Another way is to have the solar hardware directly communicate with the cerbo (fronius and solaredge can do that).

If the solar is not seperately known in VRM, it will never take that into account.

5 comments
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

Mark Frencken avatar image Mark Frencken commented ·
Hi, I would like to comment on this... As I have 6 solar inverters connected on the AC output side and one solar converter on the battery side (Victron device) I think there should be a smarter solution as to buy 1000 EU worth of metering devices and draw god knows how many cables under the pavement etc just to do what the viltron device could do internally, subtract with limit two numbers that are available to it....
0 Likes 0 ·
dirk-s avatar image dirk-s Mark Frencken commented ·

Hi, I would also like to comment your reply: Its impossible. Nobody knows, which part of AC out is inverter production and which is consumption on AC out. Its a mix usually. Its your duty to give the Venos OS the information about production either by a separate smart meter or by direct data connection between Solar inverters and Venus OS.

0 Likes 0 ·
Mark Frencken avatar image Mark Frencken dirk-s commented ·

Nope I cannot agree with you on that. Please refer to my bug report with clear pictures: https://github.com/victronenergy/dynamic-ess/issues/153 It is a very simple calculation and it does not care whether or not the exact raw production is known, it only matters (in this case) that the surplus is measured/calculated which is a very simple a-b calculation. And yes, that can be easily implemented in the DESS algorithm. Just to be clear (again), if I have a system delivering 14 kW solar on the AC output side instantaneously, AC output side connected stuff uses say 4 kW and Victron has measurements of what is going out of the combined AC out of the multi plusses and gets data what is flowing out of or into the grid, the calculation is really very basic and shows instantly that there is 10 kW production somewhere between AC out and gris connection. From there, it should not be difficult to decide that when prices are low and the battery is empty to NOT dump the 10 kW onto the grid but to use it to fill the batteries. I really absolutely fail to see how someone can defend this obvious bug.

0 Likes 0 ·
nickdb avatar image nickdb ♦♦ Mark Frencken commented ·

Lack of desired functionality isn't a bug, your expectations are beyond what the system was planned, and documented to do.

The PV inverter configurations that are supported are sufficiently detailed, venturing away from what is tested will only deliver unexpected results.

You are welcome to add a "feature request", but considering the current todo list, I wouldn't get my hopes up.

Add a single meter at the combined output of the PV inverters, and the system will operate as designed.

0 Likes 0 ·
Show more comments
daza avatar image
daza answered ·

How are you measuring the AC solar input ie ET112 VM-3P75CT or similar, I could be wrong here but I’m sure the multi can only measure without a meter on the AC output. I take it ESS has been setup with where the AC solar output is on the multi a system diagram would help though

2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

sharpener avatar image
sharpener answered ·

Sounds a bit like what happened before I put an ET112 on the PV connected to AC Out. ESS worked but the dashboard shows incorrect figures as the AC load cannot be displayed as negative.

Maybe you have an ET112 on the PV but it is wrongly configured?

2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

Mark Frencken avatar image
Mark Frencken answered ·

Not sure about the picture upload yet so I'll try to explain in wording.\I have a ET122 connected between the main switch from the input and the house, so it will measure power to/from th net only. Then I have several loads and solar sources connected to the AC side as well as a 3 ph connection to 3 multiplus inverters. The outputs of the multiplus inverters go to the loads I want to have running incase of a blackout. The 3 multiplus units are connected to a 48 V bus to which also some batteries and a viltron solar inverter is connected (2x 2 kW)\The Dynamic ESS (released version, no beta) recognizes the solar output from the DC side, but it cannot figure out that (like today) when there is 9 kW going back to the grid (as measured and reported to the 'brains' of the viltron system ) that it cannot figure out that it has to come from the AC side. Mind you, during install it somewhere asks whether or not there are inverters connected to the AC side where I of course answered 'yes'. I did in the end figure out how to file a bug in GitHub, where I was able to upload screenshots. There it is shown that when disabling DESS, the system actually routes excess power to the batteries, and when enabling DESS, the excess power is sent to the grid...

2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

Related Resources

Dynamic ESS - manual

Additional resources still need to be added for this topic