Splitting one inverter output to two separate services

[GoldDigger]

However, in the case I brought up, the 240V load is using all of the PV system's output. Current is in phase with voltage on the meter for the customer with the load...

Nice diagram!

But rethink it again. The customer with the load does not pay anything (if the load current exactly equals the PV current.) This is the normal situation where the PV power (or half the PV power in this case) just balances out the load. The meter reads zero because the load is being supplied by the PV.

On the other side take a close look at your voltage and current directions and I think you will see that the meter there is in fact running backwards, so that customer gets credit for his 1/2 of the PV power.

If neither customer has a load, each of them gets meter credit for backfeeding 1/2 of the PV output. If both customers have loads equal to the PV current, both meters read zero. And the load is being run "for free" just as we would expect.

A little more explanation: Since you are showing a 240 volt load which is drawing the full PV current (assuming that in your diagram only the red wires have non-zero current), the load is using the full power equivalent of the PV output, not just one half of it. So the return current on the right conductor of the load customer, without the matching current on the left hand wire, will result in 1/2 of the load power being metered as consumed.
On the non-load customer side, there will be current coming back through the corresponding phase conductor without a matching current in the other phase, so again 1/2 of the load power amount will be counted, but it will be counted as supplied to the grid, not as drawn from the grid.
One meter reads zero, the other reads a credit of PV/2. All exactly as it should be.
Again, the only potential problem is if the meters are built in such a way that a net unbalanced current flow (counting both phases and the neutral which is not shown) will not be metered properly.

Just look at one thing to start off: The direction of the current in the right-hand phase line in the two meters is opposite. So whatever one meter reads as delivered, the other will read as received.

 

[ggunn]

I still think the AHJ would take one look at this and stamp "REJECTED" on the application.

 

[jaggedben]

I admit to not having a firm grasp of how meters work, so here is my question:

In Smart$'s diagram, how are the voltage readings at the meters affected?

Doesn't the meter normally spin backwards because the voltage goes 180 degrees out of phase? Seems to me that voltage reading at meter for the customer with no load could be zero, with only one leg 180 degrees out of phase. So the meter wouldn't be moving?

I also admit to being fairly confused on the subject at this point. :lol:

I still think the AHJ would take one look at this and stamp "REJECTED" on the application.

We're all agreed on that. We just can't resist going down the rabbit-hole of curiosity on this one, as far as to what would actually happen.

 

[ggunn]

We're all agreed on that. We just can't resist going down the rabbit-hole of curiosity on this one, as far as to what would actually happen.

The dilithium crystals kinna handle it, Cap'n! She's gonna blow!

 

[GoldDigger]

Doesn't the meter normally spin backwards because the voltage goes 180 degrees out of phase? Seems to me that voltage reading at meter for the customer with no load could be zero, with only one leg 180 degrees out of phase. So the meter wouldn't be moving?

I also admit to being fairly confused on the subject at this point. :lol:

Look at the question you posed and ask yourself "180 degrees out of phase with respect to what?". It cannot be 180 degrees out of phase with the voltage on the other side of the meter, since the line and load sides are connected by a very low resistance path.
So it has to be the voltage being 180 degrees (or at least more than 90 and less than 270 degrees) out of phase with respect to the current.

Adding in the PV and the wiring shown in the diagram does not do anything to the absolute phase of the voltage. It can only affect the amount and direction of the current flowing. If, during the positive half cycle with respect to ground/neutral on any individual wire, the current is flowing toward the utility that is the direction the power is flowing and that is the direction the bidirectional meter will count in. At any moment the left line wire is on the opposite side of the half cycle from the right line wire. So both sides of a line-to-line load current loop will cause the meter to count upward. If the return current follows some other path instead of going back through the meter, the meter speed would be 1/2 of what it is with both sides of the loop going through the meter.

 

[Smart $]

Nice diagram!

But rethink it again. The customer with the load does not pay anything (if the load current exactly equals the PV current.) This is the normal situation where the PV power (or half the PV power in this case) just balances out the load. The meter reads zero because the load is being supplied by the PV.

On the other side take a close look at your voltage and current directions and I think you will see that the meter there is in fact running backwards, so that customer gets credit for his 1/2 of the PV power.

If neither customer has a load, each of them gets meter credit for backfeeding 1/2 of the PV output. If both customers have loads equal to the PV current, both meters read zero. And the load is being run "for free" just as we would expect.

A little more explanation: Since you are showing a 240 volt load which is drawing the full PV current (assuming that in your diagram only the red wires have non-zero current), the load is using the full power equivalent of the PV output, not just one half of it. So the return current on the right conductor of the load customer, without the matching current on the left hand wire, will result in 1/2 of the load power being metered as consumed.
On the non-load customer side, there will be current coming back through the corresponding phase conductor without a matching current in the other phase, so again 1/2 of the load power amount will be counted, but it will be counted as supplied to the grid, not as drawn from the grid.
One meter reads zero, the other reads a credit of PV/2. All exactly as it should be.
Again, the only potential problem is if the meters are built in such a way that a net unbalanced current flow (counting both phases and the neutral which is not shown) will not be metered properly.

Just look at one thing to start off: The direction of the current in the right-hand phase line in the two meters is opposite. So whatever one meter reads as delivered, the other will read as received.

But you are looking at it from an overall net metering, not per POCO customer (i.e. each meter a different party, not same party for both meters). If the PV system is supposed to be beneficial to both on an as needed basis, the customer with the load has to pay POCO for 1/2 the power delivered by the PV system, while the other no-load customer gets credit for 1/2 PV power delivered. Think of it from the perspective that you are the customer with the load...

 

[GoldDigger]

But you are looking at it from an overall net metering, not per POCO customer (i.e. each meter a different party, not same party for both meters). If the PV system is supposed to be beneficial to both on an as needed basis, the customer with the load has to pay POCO for 1/2 the power delivered by the PV system, while the other no-load customer gets credit for 1/2 PV power delivered. Think of it from the perspective that you are the customer with the load...

Oh, I do. But there is no equitable way to split the PV production between the two on an "as needed" basis. And with net metering the only fair as well as feasible way to split it would be 50-50, which this meter circuit does.

There is a real problem that there is no good way to sell back to POCO and still keep track of who should get the credit. By your argument, if neither customer is using any power, it should go back to POCO for free, since neither of them "need" it at the moment.

A grid tied system is in effect using POCO as a battery to store power for future use until needed (even if the need was in the past, think about that for a moment!) The GTI will always be producing whatever the panels allow whether there are any loads or not.
If this were an off grid system, you could have the panel output split equally to two different battery banks, one for each customer, or you could get sort of the result you want by having them share a battery bank. But in the ways of human nature, that would result in each of them trying to use more power to get their 'fair share'.

If they are on good terms, the split does not matter. If the relationship is strained, only 50/50 should be acceptable to both sides.

 

[Smart $]

... But in the ways of human nature, that would result in each of them trying to use more power to get their 'fair share'.

...

See it as you will. Unless you know the arrangement, it is hard to say what is fair. I agree 50/50 is a fair arrangement... but measurement isn't always limited to tangible means.

 

[GoldDigger]

See it as you will. Unless you know the arrangement, it is hard to say what is fair. I agree 50/50 is a fair arrangement... but measurement isn't always limited to tangible means.

Quite true. And between the two meter readings and the output information from the GTI itself, the two have all the numbers that they need to divide it up any way that they want. POCO just will not help by giving each of them exactly what they want know from their own meter alone.