question

paulo-m avatar image
paulo-m asked

Multiplus II AC IN vs AC out ramp up speed

I have my ESS system on a simple grid parallel setup with all the loads on the AC IN of the Multi, with an ET112 as grid meter. When there is a change on the loads, the ramping up/down of the inverter takes a few seconds to match it, resulting in overshoots of the grid setpoint. Heard someone mentioning on an youtube video that if the loads are instead installed on the AC outputs the ramping up/down is almost instantaneous resulting on much smaller overshoots, but found no info on it. Does anyone know if this is the case?

grid parallel
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

4 Answers
Jason - UK avatar image
Jason - UK answered ·

@Paulo M you are correct. The ET112 is not very responsive. I have the Victron 100A CT sensor on the incoming mains supply to allow my Multiplus to power assist the loads connected before the multi (loads on AC in) then I have everything else connected to AC out1 which makes the ESS very responsive.

I did have a ET112 but I removed it for that very reason, in that it is unresponsive. The Victron 100A CT sensor is not as accurate bit provides a improved response time from the inverter when in ESS mode.

2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

schollex avatar image
schollex answered ·

I have an ESS-system with MP II 3000 on L1 and another on L2.

Grid meter measuring has a latency of ~1 sec.
This means that chages on Watt-measurement on the gridmeter takes about 1 sec to arrive on venus/debus. Thus, my ESS must have a latency >1 sec.

Now, I wonder, because:

ESS balanced the grid-meter to 0 W.

When I turned on or off a 1600W resistive load on the AC-in side, the grid-meter get's a spike auf 1500W and then 4-6 seconds are needed to rebalance to 0W.

When I turned on or off this 1600W load on the AC-out side, the gridämeter just get's a spike ouf 300W and rebalancing to 0 W is done within 1 to 1,5 sec (my gridmeter updates Watt 2x pro sec).

I've read

I know about ramp-speed - see https://www.victronenergy.com/live/ess:ess_mode_2_and_3#response_times_and_ramp_speed

and switched to ESS mode 3 for manually setting consumption respectively supply of power.
But there it's the same: when I set a change on e.g. com.victronenergy.vebus.ttyUSB0 > Hub4/L1/AcPowerSetpoint, the Multi needs about 5 sec to achieve this setpoint.

My conclussion:

The Multis can (physically) adjust the setup extreme quickly - this can be seen on load-changes on AC-out.

But adjusting load changes on AC-in is considerably limited via software-control.
Of course, power adjusting has at least the latency between my gridmeter and my venus-device, but this latency im measering is just 1 sec, my ESS-latency is ~5 secs (much to high!).

My Multis are setup to grid code Germany internal NS.

Has anybody tested grid code Europe to achieve better ESS-rebalancing (see link above).

Bye

Alex

2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

k-k avatar image
k-k answered ·

Afaik out of my memory:

You should reset the gridcode to none/other for a test -> not having AR4105 implemented. But you need a password for that. If you do so your loosing the validity of the certificates. If you are uncoupled (off grid) as an island-system in DE , thats not a problem, if official (DE: Stammdatenregister) you loose.

DE: irgendwo hier im forum wurde darauf schon eingegangen. Im zusammenhang mit pulslasten wie induktionskochplatten. Z.b. 1kW anfo bei 2kW nennleistung in 50% PWM im sekunden bereich umsetzen. Es gibt steilheitsvorgaben deltaV in ??sek die dem genau entgegenlaufen. [Just runtergeregelt kam 0,5sek vorher die lastspitze mit 2kW]

1 comment
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

k-k avatar image k-k commented ·
1600W ohmisch -> wasserkocher mit thermostat?
0 Likes 0 ·
k-k avatar image
k-k answered ·
1 comment
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

schollex avatar image schollex commented ·
Als Elektrolaie bin ich bei den VDE-Papieren raus.

Mein Fazit lautet:

Der Multiplus kann extrem schnell ESS nachregeln, macht dies aber nur, wenn die Nachregelanweisung vom Multiplus selbst ausgeht, wenn er selbst die Leistungsabgabe an den AC-out-Ausgängen misst.
Wenn dagegen die Nachregelanweisung von außen kommt (durch eine Änderung am Netz-Stromzähler oder via dbus im ESS Mode 3) dann wird nur langsam (~400 W/sec) nachgeregelt.

Diese Einschränung ist wohl erforderlich aufgrund irgendwelcher Normen (durch wahl der Grid Codes), die der Multiplus dann einhalten muss.

Und nun kommt mein großes Unverständnis: egal ob die rapide Laständerung an AC-out oder AC-in auftritt, im Regelbetriebsfall sind diese Anschlüsse doch alle miteinander verbunden. Wenn also der Multiplus rapide AC-out nachsteuert, dann bekommt dass das öffentliche Netz doch "mit". Warum ist dies zulässig, aber eine rapide Laständerung an AC-in darf nicht umgehend nachgesteuert werden? Kann ich, mit meinem physikalisch Verständnis, nicht nachvollziehen.


0 Likes 0 ·

Related Resources

Additional resources still need to be added for this topic