Last days when was not enough solar power, were not shown in DESS GREEN MODE schedule any battery charging from grid. Battery SOC goes to low limit setting and system consume power from grid. I expected that in case of forecast of low solar power will be scheduled battery charge from grid at grid low price…
How does should work? And how does it work to others?
I’m seeing Dess charging from grid, also scheduled to cheap hours - but it’s not like Dess is deciding to charge 50,60% soc at once, it’s more like “charging 3 kWh here, then idling the battery 6 hours there will be sufficent throughout the day”.
I think dess is taking into account, that the battery roundrip causes costs + extra losses of about 30%, hence direct consuming for 20ish cent may still be more cost efficent than charging the battery for 12 cents.
What battery costs per cycle €/kWh have you set in DESS settings?
If I set it higher than 0.03 €/kWh, my DESS also doesn’t charge from grid at the actual buy-prices.
With every cent less, it charges more.
And it’s correct as dognose said, DESS will only charge as much as necessary for coming through the rest of the day, what’s OK in my opinion.
Yes, 3 cent is a critical boarder for mine as well - from where I concluded the ~30% conversion-loss-assumption.
With 3 cent costs, dess will exactly prefer 19 cent direct consumption over 12 cent charge-ahead it seems. (Cause (12 * 1.33) + 3 = 18.96 I assume)
In addition to these battery costs, however, a charging/discharging efficiency of 80% is also taken into account as standard, as far as I know. At least that’s what Dirk Jan Faber once wrote.
However, this efficiency can also be changed via Node-RED, I have set it to 77%. Additional 0.02 EUR/kWh battery costs.
The DESS will always charge energy from the grid when it is profitable enough to do so. The system will first try to cover the most expensive times with the stored electricity. If the price difference between direct use and purchased electricity + loss of efficiency + battery costs is not large enough, the system will leave it at that. Only when the electricity stored in the battery is not enough to cover all expensive times and expensive times leave a large price difference to the cheap purchase will the system buy electricity. You can actually observe all this from the very beginning in the DESS. It also makes economic sense.
You can see it for yourself in your VRM analysis. My battery costs are set at 3 cents. If you then take into account the efficiency for the AC-DC-AC conversion, I’m somewhere around 7-8 cents at which point it’s worth storing electricity. In the VRM graph you can see when the DESS is using the electricity you’ve already stored from PV… blue. Direct use is red. If there are times left where the difference exceeds the 7-8 cent loss compared to the cheapest purchase price, the system will purchase additional electricity. Otherwise it uses the stored electricity at the most expensive times and draws it directly the rest of the time. It’s interesting to watch when it gets new prices for the next time interval.
Yes, that’s what I casually described as “can’t get through the day”.
But you’ve now described everything in great detail, thank you very much for that!
I had battery cost set 7 cents/kWh
So I changed it to lower discussed price
There are many discussions about battery costs. Some are of the opinion that the battery is there anyway and now it should be used optimally. Some enter the absolute costs, taking into account how many cycles it really lasts and how high it will cost to replace it. Opinions range from to and everyone should find a real value for themselves. As mentioned above, I work with 3 cents per kWh and cycle
You do not need to calculate the conversion losses yourself, DESS calculates them automatically
Battery costs refered to cycle amount doesn’t matter if battery is charged by solar power or grid power. Only additional conversion looses, more input grid power with lower power back from battery than charged. Make sense to charge battery if DESS calculation result is that actual battery level will not be sufficient for house consumption over night or day. Price 0.7 cents/kWh is real price from baterry purchase price. It is similar like car using… it costs money and by using you calculate life time, as we calculate life cycles in battery. Battery should be used as much as possible, in ideal state 1 cycle per day.
Absolutely right. However, it is a little more complicated with batteries. First of all, many people (including me) do not trust the cycle information provided by the manufacturers. If I calculate with the 6-8000 cycles promised, I may not replace my batteries because they will outlive me. The second point is cost degradation. Not so long ago, costs of €1000-2000 per kWh capacity were the norm. This year I added less than €100 per kWh capacity to my system. I calculate machine costs every day for work and I am no stranger to correct business practices. But with battery costs, it is not quite so linear. That is precisely why I calculate a certain mixed calculation and have decided on 3 cents.
By the way, whether the daily average cycles are 0.75 or even 1.25 is completely secondary to the other factors.
I empirically tested several variants from 0 to 5 cents/kWh within a few hours and observed the respective impact on the DESS planning.
In my opinion, charging at 0 and 1 cent/kWh was too frequent or at prices that were too high.
From 4 cents/kWh mostly no charging at all, with 3 sometimes too little.
2 cents/kWh gave me a good result.
But I have set the efficiency to 77%. If you leave it at the default setting (80% according to Dirk Jan), 3 cents/kWh might be a good fit.
But then you have to say that you control the sensitivity of the DESS in relation to the purchase and not for the real lucrative use of the battery. From my understanding, it is intended for the real calculation of the battery costs
That is of course correct. But as you have already written, there is a passionate debate about whether to calculate battery costs at all > 0 cent/kWh (because the battery has already been fully paid for anyway) or to calculate them with purchase price / (capacity * cycles).
And in the latter case, again, how high you set the cycles.
If I calculate with 6000 cycles, I get approx. 3 cents. But I will never reach 6000 in 10 years, more likely only 3000 cycles. But then I would get about 7 cents, which means that the battery would never be charged from the grid.
I have to keep testing whether I use 2 or 3 cents.
And you mustn’t forget that the efficiency is taken into account anyway. Whether you then add additional costs (which are no longer incurred but have already been paid) is the subject of discussion…
Everything is fine…I just wanted to describe the tension in which we are moving with battery costs. I also fluctuate between 2-3 cents, regardless of whether it is from an economic point of view or functionality. Ultimately, even a decimal place would be good, but that is secondary to other factors. Ultimately, each user has to decide for themselves what is good about Victron’s flexibility
Today again little PV production, battery is now charged cheaply by DESS at lunchtime. Prices for tomorrow are already available.
I have now tested 2 and 3 cents/kWh and compared the resulting DESS planning.
I have now also ended up with 3 cents/kWh 