Dear All,
Was there any progress on the power oscillating topic? Very annoying and probably something would break eventually.
I experience kind of same (not yet failed parts).
My Setup:
1 x Miltiplus-2 GX 48/3000
1 x 16k batteries
1 x MPPT 250/70 (6 x410 WP panels)
1 x EM540
House net: 3 x 230V without neutral (Belgium)
1 x PV (SMA) system injecting in AC input network. (10 x 330 Wp panels)
Nothing connected in AC2 or at AC out of MP-2.
Issue:
In low load, total power nicely balanced.
In high load, MP2 is overshooting and back with a frequency of 1-2 sec.Making energy to run to and from grid, in high values. Also battery is struggling to cope with that fluctuation.
When the load is removed, it does calm down.
ESS is set for grid monitoring, single phase, as total phase would be even worse. No DESS set.
According Victron Belgium, it should work. Yes, it might not be easy implementation, but I did not get a clear “NO”. To bring them to my place could be an option, but it would also cost. I will keep that as last option.
Thanks.
For units in 3 phase configuration: Our products have been designed for a star (Y) type three phase configuration. In a star configuration all neutrals are connected, a so called: “distributed neutral”. We do not support a delta (Δ) configuration. A delta configuration does not have a distributed neutral and will lead to certain inverter features not operating as expected.
As the topic is back from 2018/2019, I am kind of surprised that there is no fix from Victron on this one.
However, one question.
Where are these (wrong) calculations made and based on what input? Reason of my question: if the calculation is in Victron, based on Power, Voltage and currents given by EM540, then I do not see the issue. I can implement a “filter PLC” in between, to tell to Victron what it needs “hear”, base on what I want to get as result over full power.
Its not primarily wrong calculations but measurments made in a delta grid but being used in a system that only works in a star/wye grid.
Thats what i tried to explain in the linked post, you would need to recalculate the delta measurments to get their wye equivalents, in order for the ESS to regulate on those.
However, since im not in a delta grid region myself, i cant test this myself. I have essentially no idea if the ESS works fine or not even if the recalculations are done correctly
If you install an MP on L1 and L2 in a delta grid, feeding back with it will change the current in both of those phases, which any meter will report correctly, but the Victron system is not expecting that, it only expects a change in L1 since it expects the MP to be connected to L1 and N.
An EM540 can be set to 3Ph without N, but that doesnt change that the Victron system is still expecting to be used in a 3x230/400V star grid with N.
Three phase delta grid without N is not the same as star/wye with N. Its a single MP but its still connected to two phases, because thats the only possibilty there is
It is not, because thats the voltage that you get in a delta grid to earth.
If that is enough for the ESS to work, then all you need to do in a delta config is arithemtically sum up the three power figures of L1, L2 and L3, and only transmit that to the virtualDevice gridmeter, the rest you need to omit
not necessary as there is no neutral that is connected to earth
the 130V in the other thread, this comes from the meter - because there is no ground connected on the meter - it “builds” its own internal neutral through the resistance of the voltage measurement
yes, but where do you take these three power figures from?
The right way would be to use a delta meter
here is something interesting: he connected a shelly, used one leg for neutral connection:
There does not need to be a neutral in order for there to be a voltage to earth.
In a three phase delta grid, theres 3x230V between each combo of phases L1-L2, L2-L3 and L3-L1. Therefore, the relation to ground is (squareroot of three) or about 1.73. The same realtion there is between 400V and 230V. But this time its 230V and 133V. Each combo of Lx-PE measures 133V.
The meter in the other thread is an EM540 which can be set to 3ph delta grid mode, but Victron still shows 130V as the grid voltage probably because they internally recalculate the Phase-Phase voltage by dividing it by 1.73. Which is correct if you are in a star grid, which is offcially supported.
From a grid meter like an EM540 or some other meter that can measure 3x230V.
But you cant use the measured figures directly. So you either recalculate them like i mentioned in the other thread, or you can simply use the total power. I was not aware that thats enough for an ESS to work. So if that true then the solution would be quite simple.
Below you see the control for 3x230V (one single MP-2). (Even you do not see all 3 phases (but one only), it is a 3 phase system with no neutral. And actually there is the trick.) All is having as reference total power. Bingo!
It deals fine with load variation from the house.
Remaining extra energy is directed to a boiler to make hot water (another project of mine. Now combined and they do the job.)
What I could see is that EM540 and the official meter reports, are not in line. Especially on high exported power. But that does not bother me.
Victron / Venus always uses only total power for ESS, they can´t do much more as there is inly one transformer in the MP2 which only could deliver sinus power
The problem was / is the EM540 which doesn´t measure correctly in a delta system
It would be interesting how a shelly would behave with one pjase connected to N - like in the link i posted
system unstable. (It was hunting like crazy +/- 1000w for indefinite time)
Reading unreliable.
End / actual point:
System stabile (happy days) (After a load step, in 4-5 oscillations it is back to the setpoint.)
Reading compared with official meter in deviation of tens of w. To me, good enough. (In my case, the deviation is even on positive side. EM540 zero is kind of 100w export from the perspective of legal meter. (I’ve built a program and hardware to read al 3 phases Voltage, Current, Power). Can be controlled by adjusting ESS setpoint. So, actually, deviation does not matter. Is it liniar? I personally do not care, as we talk about this small amount of power.)
Manually added this. Could haven been a requirement, but as I could see my meter on others screens, I’ve believed was enough. But not… No clue. It did work for me.
I’ve asked “AIround” and followed “ad literam” what was mentioned .Settings.pdf (964.5 KB)
Kept batteries optimize WITH battery life. (My chose, as I have 16K and will not be needed full capacity every day, even in winter time.).
I hope will work for others too.
PS. In the beginning, all 3 phases will be shown, but 2 and 3 with “-”. They will disappear at page reload.
