Individually Grid-tie on 3-phase

[Renier82]

Good morning everyone. I have searched around on the forum, couldn't find exactly what I want to ask, so please forgive me if it has been asked before.

I have 3 phase coming into a premises. Dbs with 1 phase appliances are split onto each phase. I'm curious if one could use single phase Infinisolar inverters for each phase, connect certain critical equipment to the inverter, and feeding whatever PV produces in excess of consumption back into the grid (hopefully to some extend to substitute consumption by appliances not powered by inverter)

The reason I would like to explore feasibility of this route is mostly due to the limitation of pylontech's one can connect in parallel

 

[phil.g00]

I am going to answer this in a inverter agnostic way.

It depends on this:

1. Do you have, or can you guarantee you never have 3 phase appliances?

2. Do your inverters still supply AC in the event of a power failure.

I think you can see where I am going with this.

Using them as individual single phase systems, ( even from the same battery) while the grid is dictating phase and frequency is OK.

Using them as individual single phase systems, to single-phase equipment while the grid is down is OK.

However, if the grid was down, each individual system would act to set its own phase and frequency, which would not work for 3 phase equipment and interconnected inverters.

There is nothing to maintain the phases relative to one another.

That is why, there is a comms channel between inverters set up in 3 phase, so that the phase and frequency relationship between the 3 phases is maintained.

 

[Coulomb]

14 hours ago, phil.g00 said:

That is why, there is a comms channel between inverters set up in 3 phase, so that the phase and frequency relationship between the 3 phases is maintained.

Right. But three individual Infinis with "parallel" cards do just that. I assume that the master sets the timing and phase for the two slaves.

[phil.g00]

3 hours ago, Coulomb said:

Right. But three individual Infinis with "parallel" cards do just that. I assume that the master sets the timing and phase for the two slaves.

Yes, then there will be a comms channel that gets them talking to each other. 

[___]

22 hours ago, phil.g00 said:

However, if the grid was down, each individual system would act to set its own phase and frequency, which would not work for 3 phase equipment and interconnected inverters.

American split-phase inverters do something interesting. When the grid goes out, they simply tie the two phases together. All the single-phase stuff continues to work, and any 240V appliances now has 0V across its terminals and doesn't work.

You could do the same thing for three phase as long as none of your 3-phase appliances have a neutral connection (ie only 3-wire connections, no 4-wire...not counting earth).

[phil.g00]

1 hour ago, plonkster said:

You could do the same thing for three phase as long as none of your 3-phase appliances have a neutral connection (ie only 3-wire connections, no 4-wire...not counting earth).

Nope, it wont work, all the inverters have to march to the same drum.,

Sorry, @plonksterI have to correct this because normally people on the forum take your word as gospel, (and 99.9% of the time rightly so). But in this instance it may be an expensive lesson.

Electrical vectors can be hard to get your head around, so I'll try and explain:

Even Delta connections ( no 4th wire) rely on 120 degree difference between phases, the slightest difference in frequencies of any of the individual phase outputs will ruin that 3 phase relationship.

Just because there isn't a fourth wire doesn't mean there isn't a system neutral. In the diagram Line voltage is Va-b, but it still relies on the relative phase relationship between VA & VB to be constant, for Va-b to remain constant.

 

Line voltages will not remain constant at SQRT3*Vph  or 1.732*230Vac = 398V, because the angle between the voltages will change unless the  voltage vectors spin at exactly the same rate. And the rate they spin at is the frequency. The vectors will not spin at an identical frequency from 3 individual systems, This means the vectors will overtake one another.

The phase to phase voltage between any 2 phases on a 230Vac 3 phase system will then vary between 0V ( when 2 phases are in phase) and 460V( when 2 phases are 180 degrees out of phase) and everywhere in between as they spin at different speeds around the system neutral.

This varying voltage will be a feature when measuring L1-L2, L2-L3 & L3-L1.

No 3 phase appliance will tolerate this.

 

 

Edited December 3, 2019 by phil.g00

[___]

1 hour ago, phil.g00 said:

Nope, it wont work, all the inverters have to march to the same drum.,

I think you misunderstood me, and probably justifiably so cause I wasn't very clear. I didn't mean the (3-phase) appliance would work. I meant this way of making sure all the single phase stuff works (by essentially reconfiguring everything into a large single-phase thing) can also be done on a 3-phase network. There would be zero potential difference between the phases, so for any 3-wire 3-phase appliance it's just like being disconnected. Exactly the same as with the American split-phase system hack I described.

Edited December 3, 2019 by plonkster