Hi,
I have setup dynamic ESS.
I notice that sometime in the evening ESS system try to consume energy from grid .
Battery it is fully charged ~ 98%
Any idea how to prevent this?
Best Regards,
Gabriel
Hello Gabriel,
I have a similar behavior in my DESS system. DESS - Target SOC 60%, sun, surplus PV, grid consumption - why?.
You are not alone 
So far, I could not find out, why it behaves like that. Maybe we get improvements with the new OS version 3.6x, which is still in Beta. It mentions DESS improvements: Venus OS v3.60~49 available for public testing - #96 by Maxs.
v3.60~47
Dynamic ESS
Improve how Dynamic ESS reacts when actual consumption or solar production differs from forecasted. For example when there is more solar then expected, the updated algorithm will in many cases use that extra solar power to charge the battery, even if that would exceed the planned SOC level for the current hour. Instead of feeding that excess solar back into the grid for a low price, to later have to buy it back at a higher price.
I might give it a try within the next days in combination with the new MultiPlus firmware v556.
The beta is worth the try if you can afford a few potential bugs here and there.
What does the projection for the following morning look like? For our system, DESS usually waits until early the next morning â before PV production â to discharge excess PV from the day before. I think this makes sense given thereâs more certainty on the upcoming expected solar production at that point.
It also depends on your price schedule / expected consumption.
I have two units MultiPlus-II 48/5000/70-48 configured in parallel; with Firmware version 555.
couple minutes ago update of Cerbo GX to âv3.60~49â
My intention is to upgrade multiplus II 556 in two weeks.
Prediction for today remain the same.
For tomorrow is not very good also.
This looks like the correct behavior to me.
- At 1:00 the system is planning to start discharging the battery to supply your loads (stored from the excess PV production from the day before).
- At 7:00 it expects PV production to begin, and will use all of that PV to supply loads and charge your batteries (not sell to grid, given your sell price is low).
- At 13:00 the system is projecting PV to decrease and consumption to increase, so it begins supplementing with grid power.
- Cycle restarts at 22:00, based on tomorrowâs expected consumption / PV production
DESS will not discharge the battery to your set âminimum SOCâ on a daily basis, like normal ESS does. It takes into account the round-trip efficiency of your system, the expected PV production, expected consumption, and grid / sell prices to minimize costs.
Of course I could be misunderstanding something.
Thank you for sharing
-
at 01:00 Red mean âGrid to Consumptionâ : in red portion it will supply power from grid to load and not from battery . (battery should be full at that moment).
âBattery to Consumptionâ is represented in Blue and it will not be discharged. -
It is as per my expectation too .
-
At 13:00 system forecast to consume from grid (from solar 2.60 kWh + 1.47 kWh from grid): why !? this is confusing me because in my case during day time after battery is full at 13:00 consum usually is around 1 kWh. (and that load can be covered by solar production
-
yes, at 22:00 system should run on battery : Sell price (0.25) is more than 3 time less than buy price (0.80) and it is constant no variation day/night or weekend
I will monitor tomorrow to see this behavior
How long as your system been running for? The behavior youâre seeing at 13:00 could partly be a result of DESS needing more historical data to accurately project consumption (itâs recommended to have 4 weeks of history).
In reality, if the consumption ends up being lower than planned at between 13:00-18:00 (when you have simultaneous PV production and grid use planned), the system should minimize grid use in favor of solar.
For excess PV stored to battery, given your constant buy prices, it economically wonât matter whether the system discharges from the battery at 20:00 or 2:00 (for example). Rather, DESS seems to favor discharging at in the early morning.
As mentioned, this makes sense to me given the PV projections for the coming day should be more accurate. This should allow the system to better calculate how much SOC it will gain during the day (and therefore better calculate how much SOC to discharge before PV production starts).
I setup DESS 2 weeks ago . When I notice this behavior I disable-it for one or two days .
I have 14.3 kW battery : enough for entire night (SOC is set to 20% , usually in the morning battery is at 25- 31%) ; so there is no reason to consume from grid âproactivelyâ in order to preserve battery.
If it consume grom grid , battery will be in the morning at 55% .
While you may have enough battery to go an entire night, do you have enough PV to recharge the battery and go a full day without grid use?
I have a 8kWp; those are charging battery every day.
usually full day consumption will be ~ 20KW. (10 kw during day + 10 will fill up battery and will be consumed in the night)
Hmm â is there a reason youâre using DESS if your grid prices are fixed? Normal âESSâ should suffice.
Based on the charts its not fully clear to me thereâs enough solar to cover your daily usage (since weâre looking at projections mostly).
Do you use âbattery Lifeâ or something similar in DESS?
yes. Until yesterday I use âOptimized with BatteryLifeâ ;
I have set now âOptimized without BatteryLifeâ and enable DESS.
look forward to see what is happening at 13:00 ,
âOptimized with BatteryLifeâ does not work well with DESS and its own âbattery balancerâ
2 Likes
