Added battery_state_of_health to capacity sensors

[BrettLynch123]

Wondering if the state of health sensor should be used when calculating capacity of batteries.

[Gnarfoz]

Is there any indication that multiplying the battery capacity with the state of health value results in a useful number?

[BrettLynch123]

Is there any indication that multiplying the battery capacity with the state of health value results in a useful number?

I was just thinking it more accurately calculates capacity.

If the state of health is for instance Is 98 then the actual battery capacity would be 98% of the battery capacity in theory.

And the change I made basically multiplies the battery capacity by the state of health percentage.

I noticed with mine the battery capacity is not dropping as the state of health of the battery drops.

[Gnarfoz]

What I meant was: all of that is based on the assumption that the state of health value represents "overall capacity is at X percent of when it was brand new". I'm not sure that is accurate.
Maybe we need someone whose battery has degraded enough to verify this via the daily battery charge / discharge amount. That is, if their battery is at 80% SoH, then daily battery charge, if starting at 0%, should be 80% of the battery's rated capacity.

[BrettLynch]

Oh ok I did think that was the definition of state of health.
That is what it means with my Tesla powerwall and Tesla model y

Maybe sungrow is different.

[Gnarfoz]

The conservative option then would be to add this as a separate entity instead of replacing the regular one. Thoughts? :-)

[BrettLynch123]

Just had a look at my data. When my battery was new for the first 30 days I averaged discharging 9.35 Kwh a day out of the battery.
and my last 30 days I averaged discharging 9.21 Kwh a day out of the battery.
My state of health reports as 98. 9.21 / 9.35 = 0.985 %
So its close to 98. But if people think a separate entity is better then that is ok by me. My personal system will be replacing the regular one.

[mkaiser]

hey @BrettLynch123,

I searched the web a bit - My guess is that you are right with the SoH reflecting the usable energy.

I am not sure how to integrate it. Spent a while now and am very confused what was commited 1-2 years ago. Need a fresh mind... Here are my notes. The sensors really require some more documentation :)

Read from inverter: 
Battery capacity
Battery level             0..100%  |  min_soc..max_soc
Battery state of health   0..100% calculated internally by the BMS load balancing



templates: 

# If min_soc is set to 15 and max soc is set to 90,
# this "nominal battery level" will be between 15% and 90%
# while the "battery level" is between 0 and 100%
Battery level (nominal)
uses:  level min, max 


# If min_soc is set to 15 and max soc is set to 90,
# this "nominal battery charge" will be between 15% and 90% of the capacity value, 
# reflecting the limits set by min_soc and max_soc
Battery charge (nominal)
uses:  capacity, level_nominal 


# this is the actually **usable energy ** of the battery
# with respect to min/ max soc. 
# If min_soc is set to 15 and max soc is set to 90,
# this value is 0.75 * capacity 
Battery charge
uses:  capacity, level, min, max 

Since I am not good with naming - does anyone have some suggestions how to name a new sensor with SOH?

Otherwise, I don't mind putting the SOH into "Battery charge (nominal)" and "Battery charge" as you suggested.

[Gnarfoz]

Battery charge (health-rated)?
Perhaps the other way around is simpler. Modify the existing sensors and add a new Battery charge (raw)?