tried to export via forced discharge and things went a little strange

[ptryk]

so I had some excess battery power and thought this morning I would finally try to do some export which I hoped to achieve through forced discharge of the battery - when I did it, it seemed to show instantaneous discharge so I turned it off .. and has since bounced back - wondering how I should really do this?

we have an SH10RT-v112 and SBR096 9.6kw sungrow battery - inverter is now connected via LAN

not sure really what happened but my settings included
battery minimum soc : 25
forced charge discharge power (w) : 4000
max battery discharge power (w) : 4000
Battery discharging start power (W) : 0

then turned EMS to : Forced
then battery to : Discharge

in 5min i went from 45 to 25% !?

i turned it off quickly (ems to self consumption and battery to stop (both default))
went to isolarcloud, showing same 25%!? but .. most strange, no export (? i believe what I see is only the solar coming along now)

went back to home assistant / sungrow (this integration) and found it was still at 25%... looked around and after about 10min went back up and got to 41%?

isolarcloud has now also bounced back

so.. is this normal? how should one do this properly and hopefully gain some export?

[Gnarfoz]

Your process looks correct, I'm a bit confused as to your result.
Setting "EMS mode" to "forced mode" and "Force command" to "Force discharge" and "Force charge discharge power" to some useful value (e.g. 4000 W) should be all you need to do to discharge the battery to "somewhere" (meaning if you had 5 kW load in your home then of course this would not result in exporting anything).

Do you maybe have an export limit active?
Not sure how that interacts with forced mode.

[ptryk]

ahh.. maybe that is it
Export power limit mode : disabled

? thought it would just flow then at full gas but if I enable it (it is set to 10000 from I believe the installer(?) as I have never touched it)

I had like a 0,5kWh load at the time.. so I suspected 0,5-4 = 3,5 export, right?

It was the immediate drop of the battery SOC that spooked me honestly.. I was glad to see the energy come back, but it seemed to set the value somehow to what I thought was the minimum value it would work towards at the 4000 rate. (before I set it to 25 it was at 0)

-edit-v
just logged into the installer account and saw that in advanced>energy management parameters>12 battery first : disabled
it says in a popup the following:

Battery First
Set to "battery first" for fast discharge in compulsory mode. Used for maintenance purpose only.

also, you had suggested in the thread where I was getting errors (due to WiNet-S) about version and maybe it is due to the version but it seems to be within the bounds:
Module Name > Module Current Version
SDSP > SUBCTL-S_04011.01.01
MDSP >SAPPHIRE-H_03011.95.01
LCD > SAPPHIRE-H_01011.95.01
BAT > SBRBCU-S_22011.01.13

[Gnarfoz]

Disabled or enabled + set to maximum should not make a difference, I think.
Yes, if no PV power was present, using force discharge @ 4 kW while you have a 0.5 kW load should result in a roughly 3.5 kW export power.

Definitely don't enable "battery first" -- as the comment (and Sungrow installer training) says, it's for maintenance only and should only be used if you know exactly what you're doing.

95.01 is an excellent version, your battery version is slightly out of date, but I don't think it should matter.

[ptryk]

ok with that I have given it a 10min test go again now and it worked fine!
half way through i even bumped it up to 3kw and it reacted as I suspected.
it also shut down as expected.
this time it did NOT react in that strange way.

I did upgrade the script for sungrow-modbus to the latest version but from what I gathered it was only some minor things on restart (!great, I think this has fixed the issue I had of sensors going 'unknow' for up to 10min) and a registry value on emergency battery.

I did NOT do do much of anything else, not toggle anything more at all since this morning except lowering the discharge to 1000 to start and then ramping to 3000 AND reducing the min SOC% to 5 where I think it was before this morning which I left as I ran the export cycle... hmmm maybe this had something to do with it... edit 3? - i just checked this as well and tried to repeat exactly what I did this morning and moving the min SOC % before or during had no effect on the perceived battery charge. seems like ghosts.

nonetheless, thanks for your help! I was about to write as well earlier this week to thank you for the help with the WiNet-S errors, really very helpful and thanks again! This integration/configuration is really great!!

--> If someone else runs into this I guess maybe upgrade (?) and maybe just wait a bit and run it again!