Hi all,
We had our PV system installed yesterday and had a query about the recommended maximum charge capacity on the LI-ION battery.
My understanding is it is not recommended to charge LI-ION batteries to 100%, but instead between 80-90% to prolong the battery life.
Is this the case with the Giv-Energy batteries, or is it recommended that they're charged to 100%?
Many thanks all
Midlands_Solar
The Givenergy batteries are LiFePo and can be charged to 100% without issue.
Midlands_Solar Welcome to the forum.
80% is the traditional advice for EVs, but they have tended to be NMC chemistry. GivEnergy batteries are LFP chemistry which tend to be more forgiving. In addition, the full capacity isn't used: 100% indicated doesn't correspond to the maximum voltage tolerated by the cell, so the 0% to 100% reported by the software isn't the true 0% to 100% of the cell. (That tends to be true of cars too.)
The voltage curve for LFP chemistry is very flat, so the firmware has trouble tracking the true SoC over time. Visiting 100% regularly gives the BMS a reference point. And, like EVs, it may do cell rebalancing at 100%.
DD Never seen ANY rebalancing at 100% on G2 batteries & in fact the firmware (3013 but same on 3015) states it only happens below 80% SoC.
In practice it's extremely unlikely to occur at all unless the battery is idle for more than a couple of hours.
Only way I've found to force it to rebalance (internal calibration routines) is to leave it idle for at least 3 hours with the reported SoC below 20%, although below 10% and idle will usually start the internal calibration after an hour or so.
In short its a total mess, don't expect SoC to be in any way accurate unless the batteries are run flat every single day. Any PV charge to the batteries is pretty much guaranteed to result in a (wildly) inaccurate SoC within 48 hours or so unless you run them flat and leave them that way for a couple of hours. Expect over and undervoltage warnings in the portal from time to time. Not impressive, other manufacturers of LFP batteries do much better.
In terms of the cells used in the G2 battery, they're rated for 5000 full charge/discharge cycles (80% capacity remaining) - and I do mean FULL, they take them all the way from 2.5V to 3.65V. For the voltage in the G2 that equates to 40V-58.4V as there's 16 cells.
Basically the tl;dr from the battery cell datasheet in terms of maximising lifespan are -
1) Keep the cells as near as possible to 25 celsius, probably 20 celsius ambient is about right given it'll be warmer inside the case. Low/high temperatures can reduce lifespan by 50% in terms of number of cycles;
2) Charge them below 0.5C (2500W) rates, ideally charge at 0.2C (1000W) to maximise lifespan;
3) Discharge them below 0.5C rates if possible although higher discharge rates don't have quite the same effect on reduced lifespan as higher charge rates do.
There's not a lot you can do really in terms of managing the battery using SoC as if you keep it in what is the sweet spot (20-80%) for all lithium ion battery chemistries then after a few days the SoC reported by the GE BMS/inverter will be out by at least 10%, probably more. If you use some local management via GivTCP/HA then you can usually work it out for a given load but this isn't straightforward. Only other option is to leave the battery idle for a couple of hours a day when its below 20% reported SoC and hope internal calibration kicks in - not easy once its summer.
It is what it is. No point bothering about it as all of the above is on GE, not the end-user. There should be nothing we can do which would adversely affect the lifespan/warranty of the product - barring local environmental conditions - and if there is then its on them again as the controls weren't fit for purpose 🙂
Edit - note that figures above relate to the G2 9.5kWh battery. I have no idea about the AIO....
Vestas
Only way I've found to force it to rebalance (internal calibration routines) is to leave it idle for at least 3 hours with the reported SoC below 20%, although below 10% and idle will usually start the internal calibration after an hour or so.
Ah, okay... How can you tell that it's balancing ..? (Despite lower threshold of 4%, my battery seems to pause when it reaches 6%, and as it idles, the SoC can increase up a few %. Perhaps that's it.)
As we move towards sunnier days, there may be opportunities to leave it idle at low charge at the start of the off-peak period, and still have enough time to charge it up a little for the morning.
DD That's it exactly. That's internal rebalancing/calibration routines operating.
It'll usually run through two cycles if you let it. On the battery graphs you'll see the SoC blip up 1-2% and it'll supply load for a little while then it'll go idle again. Give it another 30 minutes or so from going idle and you might see the same again, but much shorter supply of load before idle again. IMHO that's about as close as you're going to get to a true SoC figure without getting GE to run a full calibration.
I set a cutoff value of 6% on the battery after v3013 f/w just in case GE mess up like that again & that works fine in winter but I've run an automation which idles the battery at 15% reported SoC and that worked OK. Probably going to have to look at that again really 🙂
Edit - also you do know there's two settings for the lower limit? One is reserve and one is cutoff. Visible in portal->inverter->remote control. Reserve is supposedly for EPS so if you don't have that then use cutoff.