victronenergy / dynamic-ess

MIT License
76 stars 5 forks source link

[Bug]: Noooo don't take from battery when prices are very low... ;) #130

Closed AbelLykens closed 2 months ago

AbelLykens commented 3 months ago

Versions

v3.30~7 image

What happened?

Very odd/unexpected. Prices currently are very low. So I started my washing machine and dishwasher.

But it seems to take power from the battery. That does not make a lot of sense, does it?

image

image

When did it happen?

20240305T1301 CET

Relevant log output

No response

Screenshots

No response

dirkjanfaber commented 3 months ago

Just make sure. You are using the latest flow? The one that also sets restrictions and strategies?

AbelLykens commented 3 months ago

Yes pretty sure :) Upgraded a week ago or so :)

image

AbelLykens commented 3 months ago

image

It's currently delivering energy back, and next hour it plans to take from the net... Not ideal.

image

dirkjanfaber commented 3 months ago

We did have some updates on this from 3.30\~13 (which you are running right now) to 3.30\~17 (latest right now). Can you update and check if the problem persists?

AbelLykens commented 3 months ago

image

Thanks let's see. Any way I can get notifications about updates?

Will run v3.30~17 for a bit, and report back if odd things keep happening or if this is fixed 😃 .

AbelLykens commented 3 months ago

image image image image

System is now getting in more solar than expected. Prices will rise in a bit, and the battery is not full. Would be nice if it charged rather than fed back 👀

dirkjanfaber commented 3 months ago

Did you notice the new releases option to add extra (most of the time solar) energy into the battery to reduce the grid usage during the first part of the night?

image
AbelLykens commented 3 months ago

I must admit the b_goal does not (yet?) make much sense to me. Have not used it yet. But will set it to 50%/0 (as per your screenshots) and see what it does 😄

AbelLykens commented 3 months ago

Req: please update the README for these b_goal parameters (including rationale) please. Because it feels like you are solving my problem, I just don't understand how/why it all works.

dirkjanfaber commented 3 months ago

I will do that, but in short: b_goal_SOC is the state of charge you want the system to have at timeb_goal_hour. So, if you want the system to have an 80% state of charge at the end of the usable sunlight. Right now that would be around 19:00 (probably an option for setting it at sunset for your location will come later; depending on how all tests turn out), you need to set b_goal_hour at 19 and b_goal_soc at 80%.

This will result in the system to try to reach that 80% at 19:00. It does this by any means it has at its disposal, preferring the cheapest way (solar), but it might be charging the battery from grid too, unless you set a restriction on that.

Hope this makes it a bit clearer, if not, please ask. I'll put some more text in the README, probably tomorrow morning.

AbelLykens commented 3 months ago

Right I see. I think that makes sense. Let's set it and see what happens. Would be useful if goal_hour would dynamically change based on expected sun I suppose. But this is cool already. Let's see what happens.

AbelLykens commented 3 months ago

image image I updated to 0.1.17. And I set the variables. No luck yet, it seems. I kinda expect the blue line to go to at least 80% at 20:00?

Will now update the cerbo to v3.30~21 and report back.

dirkjanfaber commented 3 months ago

That is what I would have expected too. It should already work with the previous firmware.

AbelLykens commented 3 months ago

No luck for me yet on v3.30~21 with 0.1.17.

image image

dirkjanfaber commented 3 months ago

It appears that the parameters were case sensitive. We fixed that on our side, so now your system aims for the 80%

image
AbelLykens commented 3 months ago

Something happened! It's getting energy now from the net like crazy. Cheers.

dirkjanfaber commented 2 months ago

I'll close this issue now as it seems to be resolved.