Have you had any further reaction or solution to your issues?
I’m facing same frustrating issues, really stupid behavior of DESS. And it is not about a wrong prognose of nightly usage. See picture below.
Last night DESS ‘bought’ from the net for 0,21 and it ‘selling’ it back at 07.00 and 08.00 for approx 0,09
I have communicated with the Victron employee several more times by email and sent him many proofs of DESS’s bad behavior, but they are not solving the problem. They did not even promise to solve it in the next update, etc.
I have sent them about 4 illogical things that they could fix. I don’t know why they are not fixing it. It seems to me that they cannot communicate directly with the developers, or their company does not value customer feedback, etc.
Overall, Victron device works OK, these are rather marginal issues, but it freezes.
I see the same problems here. It’s or choosing for a very aggressive approach, or green mode, which is sometimes OK. But it still chooses to trade with loss. It doesn’t take the losses with (dis) charging into account. When you put these losses into the selling price, it still doesn’t take that into account.
As if it only looks at the market price? But then still it doesn’t make sense, as there is mostly around 18%loss.
And then the fact that it’s not very straight forward to let the system charge planned.. I now made something in home assistent. But basically it’s something victron should provide.
I had a look at your system and the 26th February:
DESS bought the energy in the night, because it expected way higher consumption between 7 and 8. But this day, there was not much consumption. Then the selling happened instead.
As far as I know this is caused, because the scheduler does not track actual purchase-prices of the battery content (beause this could lead to deadlocks, when buying is expensive before next day prices are known and then you have days of cheap prices). So, what the scheduler does instead:
If you have selling to grid enabled and have more energy in the battery than needed “at sunrise”, it will sell excess for the most expensive hours before sunrise.
Usually that works out fine, because energy in the battery is either deferred solar - or it was buyed cheap enough with the idea to be sold (hence the margin was expected at buy time) - just energy bought for consumption that finally didn’t happen - is unideal and may cause what you saw this day.
But, there’s a BUT:
After the consumption did not happen - selling again was the most economical thing to do:
Your battery reached 100% from solar again this day. (expected)
Selling the energy “7-8” was worth 0.09
If it would have been kept in the battery instead, you could have charged LESS solar later, causing feedin for 0.01, when the battery would have filled up earlier.
This is a typical example situation, where it is important to understand, that making purchasing decisions based on forecasted values could always have 2 outcomes:
Forecasting was right, then the (deterministic) schedule is fine.
Forecasting was wrong (or say: Reality hit different) - then the decision could be wrong and a financial drawback.
(“Not charging” despite the forecast indicates a need has the same two options: It could have saved money, or lost money in form of more expensive purchase prices - But acting against the forecast would not be reasonable)