This is where I don’t know, sorry. As far as I know the ESS assistance’s goal is to optimize self-consumption by using as much PV & battery as possible before using any grid. So for your use case you will have to do some fidgeting around either with the Victron or the loads. Short of changing these loads to Sonoff / CBI and using Node-RED to only switch them on when the battery is at 100% or a certain timestamp passed I don’t know how to achieve what you’re looking for.
In my case I have my larger loads (pool pump, heat pump) spread around based on a ‘normal’ sunny day where there’s enough PV to power them and have BatteryLife taking care of getting the battery to 100%.
As a quick test I switched my setup to Optimized (with BatteryLife) and a 100% minimum SoC. That immediately increased the Active SoC from 25% to 80% and the inverter went into a recharge state using both PV and grid to charge the battery. That seems to explain the Active SoC being 80%. Without BatteryLife it does the same thing where it also uses the grid to charge up to the SoC.
I’ll be managing everything with Home Assistant. Monitoring the SoC, PV, tomorrow’s forecast energy production, adjusting the SoC and also turning loads on and off throughout the day.
The geyser is on a smart isolator, pool pump will go on a smart breaker soon.
In the morning the geyser will be boosted. If the battery is above 30%, go ahead and use it down to 30%. Once it reaches 30%, set Min SoC to 100% 95%, effectively mitigating what Jaco mentioned where it starts using battery again once SoC reaches 8% higher than Min Soc (i.e. 38%). This prevents the morning geyser boost from draining the battery from 38% back down to 30% repeatedly by forcing AC Loads to the Grid (to boost the geyser).
When PV becomes available, it will start charging the Battery from 30% to 100% 95%.
When the geyser is at the desired temperature, turn off the geyser’s isolator and turn on the irrigation controller/JoJo/borehole pump for 30-60 min.
Batteries should be charged by 11AM.
Turn on the pool pump on for 3 hours etc.
Towards the end of the day (while there’s still sun), boost the geyser again for an hour or two.
When PV starts going down and the AC Loads exceed the PV, lower the Min SoC back down to 30% so that the battery can start taking over from PV.
I’ve switched back to with Battery Life and SSH’d into the GX to lower the Active SoC down to 30%. That worked. Thanks for the link!
I’ll pay attention to what Active SoC does when making my adjustments from now on and I’ll stick to 95% then. I’ve verified that 95% does not push Active SoC to 80%.
O my, @plonkster … imagine my surprise seeing another person seeing what I have been seeing for years.
Ok, tongue in cheek over.
It is a known issue, the SOC hitting 100% and the MPPT’s falling off. Have been using Keep Charged since forever.
To automate that, I use NodeRED to set it to Keep Charged when the SOC hits >97%. Then little, mostly no Eskom, is used.
Never made too much noise about this, cause I found other things, as per NodeRED, that I solved at the same time, the weather the main reason for this flow, the Keep Charged a “ag, kom ons doen dit sommer hier” addition.
There is an issue for it in the bug tracker (not on the public side though), to make it less confusing. An easy fix is to simply cap the max at 95%, so people cannot do this anymore. Another possibility is to simply switch to KeepCharged mode when people set 100%.
The trouble is there is a very very small number of people who use optimised-with-100% successfully, probably through sheer dumb luck, and a change like this makes that group upset Also, there’s more important things to do right now. There almost always is.
Truth be told, me too. When this battery was tested, it was one of the best I’ve dealt with. It had a nice low charge voltage (54.5V) for a 16-cell battery, it allowed LOTS of room at the top, and it individually blocks charge and discharge… which is really nice. No tricks were required to make it work well, since it never sends CCL=0. And then… somewhere along the line… something changed. And now it is a problem.
The problem is you have to decide what’s more important: Full Batts from PV only, or maximum use of PV for loads.
Something you could do with Node-Red that I would rate would work pretty well.
at 6AM when your batts are at 30% drop the ESS Minimum batter percentage to 20% but at the same time limit the maximum inverter power to say 300W and as the SOC goes up you slowly increase the maximum inverter power until you reach 100% SOC and you can have it go ham again.
I drop my maximum inverter power at night like that with Node-Red to 100W until 00:00 and then up it to 400W until 06:00, I try to hit my minimum SOC JUST before the sun comes out, it’s really helped to lower my grid usage when we had lots of loadshedding, no point in having the SOC at 70% at 8PM only to have LS at 22:00 and to recharge from the grid at 00:00.
Another “weather” adaptation I had to do, setting the inverter max wattages based on what goes into the batts. If not, then on weather days, the SOC will stay at the min setting the entire day, unless one uses Optimized with Battery Life, to get 100% SOC at least every few days.
If the SOC is stuck at say 20%, and there is LS … hence this adaptation. Only have to intervene on really extended bad weather weeks, like we have had for weeks now in Cp, where the batts will never get charged using solar.
Right now, my inverter is set to 1850w … increases as the sun comes over.
So I use some Eskom with big loads… batts are more NB to get charged. Big loads are schedules in any event for after 11am.
Not the batt … and I’m not going to “argue” that. In my non-expert way, I’ll just mention that maybe the batts accidentally just happen to run into this old issue I’ve seen for years now already, and I have had many different batts, and BMS’es, with software interfacing and with no software interfacing.
And I get it, it is a quirk that comes out in selected cases, for whatever reasons … and it does not happen all the time either.
It is a quirk and it is complicated to find, trap and fix. And IF it gets found, it could lead to other issues. So there is that too.
It would, could, be an idea to circumvent it all and just use what is there, as I had, to overcome the quirk … at least to buy more time.
Just on that … I can set whatever I want, if the batts hit 100% SOC, by themselves, and not set to Keep Charged, the MPPT’s can fall off the system. Have had hours where the MPPT’s were just sitting at zero watts on sunny days, using Eskom, when I missed that.
Hence, as you rightfully pointed out to me yonks ago, Keep Charged solves that perfectly.
Two things come to mind:
Sometimes you have to break eggs to make an omelet.
Since when does the tail wag the dog? (the very small number of people who may get upset)
Ok, I will stop now. He says with a HUGE smile on his face …
In this particular case, you are so far off you might as well be camping out on the far side of the moon. The battery asking for charge and then stopping it, a few seconds apart. Your favourite issue doesn’t even manage to get into the room
Let me add, that I have been aware of the optimised+100% issue for years now. Probably a good 6 years. We know it happens. We know why it happens. We tell people not to do that. We deal with two groups of people though, 1) those who insist that this is they only way it works for them, and 2) those who insist that really, having a particular combo that doesn’t work as expected is confusing.
I agree with group 2, and will gladly just lock it down. Here is some news though: For every installation we know about, there are ten out there that are not connected to the internet. The scale is just massive. Even a small group, might not be a small group.
Seems like we found the issue. With the new Slimline BSl it can take a few days for the cells to balance. While it is busy balancing the CCL will frequently issue a 0A charge command and fluctuate between 0 - 30 - 180 rather fast. I have seen the BMS change these commands with 2 seconds intervals. This messes around with ESS as some of the processes in ESS can take up to 3 seconds to send a command to the inverters.
Now after about day 4, the cells are balanced and the CCL command is stable and not going to Zero even once today.
I was worried for a moment but all seems okay now.
That is best. The picture of today showing 3.39V per cell, was that at 100% SOC? Typically a LFP cell is full at 3.45V, but you can squeeze them to 3.55V without danger. They just get very prone to spiking up once you exceed 3.5V per cell.
As I recall, BSL wants 54.5V. Although I’ve seen some models ask for slightly more. At 54.5V, I expect all the cells to level out around 3.4V per cell. Which is a very nice spot to keep them.
Looking at Monday’s picture. Cell 8 is quite high (over 3.5V). Cells 4, 7, 13 and 14 are low, but not dreadfully so. The rest are essentially just full, as per the spec of this battery.
Also, there’s more important things to do right now. There almost always is.
