20 minutes ago, plonkster said:

Not necessarily. In a perfectly balanced 3-phase system there is no current on the neutral wire, which means all the current arriving from inverter 1 actually goes through inverter 2 and then eventually through inverter 3.

Not that I've set up a 3 phase inverter system, but I'd imagine that it conforms to this  standard arrangement.

In which case I have to disagree.

 

All phase currents return via the fourth wire neutral and circulate through their own  phase inverter respectively.

They do behave independently in this arrangement. 

If the loads are matched then the neutral current will be zero. If they are not matched the residual current of the vector sum will flow in the neutral.

This in turn divides into each separate phase contribution at the starpoint to complete the circuit.

Each inverter only circulates its own contribution.

15 minutes ago, phil.g00 said:

Each inverter only circulates its own contribution.

You may be right. I'd have to think about it some more.

My first thought: To make the 350VAC-400VAC you get between phases, would take the work of at least two inverters (those adjacent to that phase), one making 230V RMS at that point, and the other one making 230 * Cos(2pi/3) at that point, so if there is any three phase loads on there will involve all three. Of course, as long as the three phase loads are smaller than the smallest of the three, I can still see how you might get away with such a lopsided setup.

Maybe the manufacturers just don't do it because there's too much added complexity for little reward.

Edited November 1, 20187 yr by plonkster

A 3 phase load could be thought of as 3 single phases loads from the inverters perspective when connected in star.

Three separate circuits that share a common fourth wire. It's just not that intuitive that in a balanced system the vector sum of the return currents in the neutral is zero.

With an ammeter it looks like the current goes from source to load and doesn't come back, but it does.

Otherwise, Mr Kirchoff would be most annoyed.

20 minutes ago, phil.g00 said:

when connected in star

And when not connected in star? :-)

When connecting a bunch of single phase loads to each leg, then of course you are right... it might as well be three separate single-phase setups that just share some wiring. I am presently still unsure about what happens when the neutral is unused or only lightly used because you are well balanced.

Edited November 1, 20187 yr by plonkster

Thanks guys,

I've read the manual more intensely and confirmed the following:

1) If in 3-phase sync mode, all 3 will shut down if any one shuts down due to failure or overload (it makes sense). In case of failure, the remaining inverters can be temporarily re-configured to run 'stand-alone' until repaired/replaced. (obviously no 3-phase motors can be used during this time). The dead phase can be bridged to any of the remaining inverters as long as the load remains withing max limits. (the bridging has to be done on the dead side of the inverters-to-loads contactors to allow generator to supply 3-phase)

2) Found this pic showing different numbers of inverters on different phases showing 3-Parrallel  on P1 and 2-Parallel on P2 and 1 alone on P3.

It would suggest that the inverters won't mind unbalanced loads between the phases.

 

12 minutes ago, Trober said:

It would suggest that the inverters won't mind unbalanced loads between the phases.

As long as the largest three phase load don't overload the smallest leg of course.

In Delta the phase contribution of each adjoining two inverters  influence the line  current. The line current being the the current in the wires to the load.

Obviously there is no neutral so that excludes single phase loads without introducing a virtual neutral.  Electrical regulations and common sense will dictate that the system can't be floating from earth so the most common practice would be to earth a phase to add a ground to the system.

I can see not reason to connect inverters in delta, only headaches.

But if you're still stuck on the neutral of a star connection  you'll really battle with that concept.

Current cannot flow without a closed circuit. Ergo it doesn't go from A to B and stay there, it comes back from B to A in the Neutral.

The vector sum of 1Amp at 0 degrees + 1 Amp at 120 degrees and 1 Amp at 240 degrees is zero amps. And that's what you'll measure in the neutral, but don't think of it as zero, it actually 3 balanced vector quantities.

 Once that neutral wire gets back to the source starpoint you will measure that 1 amp in each phase will magically appear out of that neutral wire ( that supposedly carried no current) to each inverter.

Can you run a system without a neutral wire?  Yes, you can, should you No.

There is always a neutral even when there is no neutral wire. The neutral is just a electrical place in the system that represents electrical balance. Picture this, the pivot point or the balance point of a three-legged seesaw that is the neutral, it exists whether you want the seesaw to balance or you don't.

Now because the phase loads are constantly changing that pivot point would also be constantly changing, but you could always make that seesaw balance somewhere.

The neutral wire fixes this balance where we want it and to a reference voltage level that we desire, namely at earth potential and stops it moving around.

Neutral current in the fourth wire is a consequence of this.

Otherwise for sufficiently unbalanced loads  high voltages can develop  between earth and the load starpoint unless you allow current to flow with a fourth wire. That would result in hazards and stress insulation, and some weird power factors.

Basically, don't do it.

 

@Trober

1) Yes, there are work-arounds, but not as easy as not having to do them.

2) That's surprising,  I think this application maybe very inverter manufacturer specific.  I don't think Victron inverters are capable of this , so I wouldn't take this capability for granted across the board.

1) Agreed

2) I think since the inverter sinewave is microprocessor generated, its an easy task for it to sync to another reference (in this case, the P2 inverter syncs (with 120 degree shift) to the P1 inverter and so on). For the 3-phase sync mode, there are additional connections (cables) required linking all 3 inverters for this purpose. (as well as the 3-phase unison shutdown scenario on any phase failure )

I am not doubting it can be done, just that it may not be implemented by some manufacturers.

 

@Trober If you are going to use motors, specially borehole, i wouldnt use a High Frequency inverter.

Low Frequency are more capable against overvoltage or surge power.

I'd concur, up until now this discussion hasn't considered the starting current of motors, which can be a multiple of running load.

However, that is a issue that affects 1 ph motors and 3 ph motors alike.

Thanks Javi and phil,

The Axpert spec. states: "Surge Capacity 2 X Rated Power for 5 seconds"

sorry previous message not complete....

The Axpert spec. states: "Surge Capacity 2 X Rated Power for 5 seconds"  which is 8kW @ 230 V meaning about 34 Amp max surge capability. (ignoring Power Factor)

It remains to be seen if this accurate and if it will actually start motors without issues. (my guess.... probably not)

The 3-phase borehole pump is targeted to get a VFD which (I understand) will "soft start" the motor thereby eliminating starting currents.

Perhaps this is a solution for the other motors as well?

The 3-phase diesel gen (15kVA) has no problem starting these motors but its got a heavy flywheel that stores lots of energy for peak demand surges.

 

Be careful to install a VFD for a borehole pump.

Firstly, submersible pump refrigerates itself with the water, you cant Up down Frequency a lot.

The wave of Voltage in a VFD is rectangular and harmonics use to appear. If the lenght from VFD to motor is Big, may be you should install a LC filter, which is not cheap.

Algo regenerated currents may cause overvoltage and then you should install a brake resistor.

 

40 minutes ago, Trober said:

sorry previous message not complete....

The Axpert spec. states: "Surge Capacity 2 X Rated Power for 5 seconds"  which is 8kW @ 230 V meaning about 34 Amp max surge capability. (ignoring Power Factor)

It remains to be seen if this accurate and if it will actually start motors without issues. (my guess.... probably not)

The 3-phase borehole pump is targeted to get a VFD which (I understand) will "soft start" the motor thereby eliminating starting currents.

Perhaps this is a solution for the other motors as well?

The 3-phase diesel gen (15kVA) has no problem starting these motors but its got a heavy flywheel that stores lots of energy for peak demand surges.

 

Problem at start use to be current in motors. The active power surge from an Axpert could be enough, but the problem is that starting cos phi use to be round 0.4. Then, the useful parameter in a  inverter is the surge current or crest factor.

Edited November 2, 20187 yr by Javi Martínez

37 minutes ago, Trober said:

The Axpert spec. states: "Surge Capacity 2 X Rated Power for 5 seconds"  which is 8kW @ 230 V meaning about 34 Amp max surge capability. (ignoring Power Factor) 

It remains to be seen if this accurate and if it will actually start motors without issues. (my guess.... probably not) 

The Victron Multiplus inverters can do twice their nominal power for around 1 second. If I compare the big toroidal transformers in the Multi to the smaller E-core one in the Axpert, I would be very surprised if the Axpert can do this for the stated 5 seconds. Well... maybe it won't switch off, but it's going to show one hell of a voltage sag while it does it :-)

@Javi Martínez is right about a VFD, there are some things to keep in mind, like minimum running speeds or the minimum torque required to even start/keep it moving. I'm pretty sure this is not a new problem to the makers of such drives. Regarding harmonics and all that stuff, having just done a huge amount of research into RCDs and seeing the literature mention such nuisance problems often where VFDs are involved, it would seem that there certainly is somewhat of an extra cost associated with such  devices :-)

Nowadays, for standard applications borehole pump, the best is PMSM motor. It has the best efficiency all over the spectrum of frequencies.

 

Very good info thanks.More homework for me....

7 hours ago, Trober said:

Very good info thanks.More homework for me....

You can see here the specs of a Low Frequency Inverter. 

https://files.sma.de/dl/17632/SI_OFF_ON_6H_8H-DEN1642-V21web.pdf

 

Continuous current is 20 A, 230 V, but peak current is 120 A!

Edited November 2, 20187 yr by Javi Martínez
Add