DESS does often a great job. I love the idea and the basic concept. In the past sunny days, I do not understand the target SOC any longer.
The situation:
DESS, green mode, Min SOC 15%, optimize without battery life
Multiple sunny days
Plenty of surplus PV
Grid feed-in starts often at noon
Battery was used as expected in full range in cloudy conditions with little sun
Battery target SOC now stays at around 60%
At some days, grid to consumption or even grid charge occurs without any need
Why is DESS setting my target SOC to 60% while we have multiple sunny days risking unnecessary grid consumption? Even battery charging at high costs right before PV starts in the morning with 60% battery left?
The days remain sunny and the DESS target SOC is climbing up to 68%. I buy from the grid although my battery has lots of power. Do I really have to switch off DESS in sunny conditions?
Where is the constantly high configuration item “ESS battery life SOC” coming? Is ist part of DESS or is it set somewhere else? Maybe that is the reason for the strange behavior. I am in DESS green mode. ESS is set to Optimize without battery life. Min SOC is set to 15%.
DESS sets a high target SOC (50% or more) at night for no obvious reason. Although my battery is still over 50% in the morning, the high target SOC lead to grid consumption at night and even grid to battery charging. Since days, my battery never drops below 50% but I import from grid, even in times with high costs.
Are you able to post a sample price schedule as well?
I’m fairly certain the ESS battery life SOC limit is ignored if the system is set to “without battery life” — and what you noticed above is just a coincidence.
As you noted in the other thread, it’s probably a worthwhile first step to try the 3.60~47 beta. I’ve also found looking at the day ahead is helpful to understand today’s behavior (depending on time of day).
Ah I realized you posted your VRM link earlier too.
So for today and tomorrow this seems correct to me (though let me know if you are surprised by anything).
For the day you posted about initially (March 17)…
On March 16 (the day before) it seems like your AC loads were unusually high vs the day before / after, and this high load occurred during peak pricing.
It could be the case that DESS tried to avoid peak charges, at the expense of “over”-discharging your battery vs the available PV during coming day.
Then also it’s possible that PV was forecasted to be higher on March 17 and the forecast changed in the morning – causing DESS to charge at an unfavorable hour (March 17 @ 7:00). If you look at the actual production, it did seem there was some strong variations.
The price later in the day (18:00) was higher than (7:00), which could also play a factor (e.g., avoiding the need for grid use later in the day, at an even higher cost). Though the system could probably also charged itself mid-day at a lower cost…
I also believe that if the PV forecast vs actual production drastically differs, the PV projection algorithm temporarily changes to a more simplistic algorithm (there’s a post by a Victron employee recently that said that).
Certainly an interesting case…curious if others have thoughts.
I agree to your thoughts regarding the DESS decisions for target SOC. Nevertheless, taking the high battery SOC and the inverter power of the MPII 5000 into account, there never really was a need to take anything from the grid. Especially not in times with low load in multiple continues hours at night.
The battery never dropped below 50% and the next day was always sunny again.
Let’s see … I updated to 556 (MPII) and V3.60~49 two hours ago. Let’s see if anything changes in the upcoming days: B51 Small DESS
Agreed. I’ve been having an issue where the power draw of the system itself throws off the projections enough to cause issues (4x 10k Quattros). It doesn’t seem the scheduler takes this into account (even on the beta). Not to derail this thread though!
Unfortunately, the update does not seem to solve the issue. The system still seems to decide not to go below a DESS target SOC of 60% in times without PV. As a result, at night, the system starts to consume from grid, although plenty of battery is left. It even charges from grid sometimes to reach that artifical 60% DESS target SOC.
Settings: ESS without battery life, ESS SOC min 15%, DESS green mode.
Any idea? Maybe I would try next switching to ESS only for a few days and reactivate DESS to see if it would make any difference. Or should I reset the complete data history of DESS to start from scratch? What could trigger DESS to stay always above 60% target SOC at night?
Update: I switched end of March from DESS to ESS without battery life. SOC Min remains 15%. You can clearly see the difference. The system immediately starts to use the battery to its defined extend.
Will keep it for one full month now in April with ESS, switch back to DESS in May for a full month and see, how the system would have behaved.
As of now, my grid consumption in April is close to zero. That was different with DESS, even in a complete sunny week with much PV and SOC surplus.
