But as the grid itself has some impedance, if you measure the voltage at the inverter terminals while it is exporting power, and then tell it to stop exporting and measure voltage again (quickly enough that no grid fluctuations occur), the 2nd measurement will be lower than the 1st. So the inverter does have to put out a voltage that is higher than the voltage it sees at its terminals while not exporting power.
It's not a voltage source that depends on a voltage differential to export current. If you connected the inverter directly to the power plant with superconductors so that Vd = 0, it would still export. Vd in the conductors is a result of the exported current, not the cause.
I agree that if you have an isolated system consisting of a grid-following inverter, another perfect voltage source with no source impedance, an interconnect between them that has no impedance (does a superconducting interconnect still have reactance?), and some loads connected to the interconnect, then the voltage on the interconnect (constant along its length as it has no impedance) would not change between the case of the inverter exporting power and the inverter not exporting power.
When there is current across an impedance there will be a voltage drop. I don't think it makes sense to differentiate between cause and effect here. I also don't think it's important to model the inverter as a current source rather than as a dependent voltage source. It may be convenient, but either point of view works.
Whatever way you need to look at it for it to make sense to you, I guess.
A superconductor still cannot have zero impedance, because even the simplest short circuit possible, still has some inductance by virtue of the closed loop. There will inevitably be some voltage difference, whether due to inductance, resistance, or a combination of the two, between the current source and the grid. So Vd will never truly be zero, when combining sources.
The real world characteristics of a superconductor have nothing to do with the thought experiment, which assumes zero impedance.
Even if we assume zero impedance, seems like the inverter will only function as a current source when it is clipping and its thermal programming is to maintain a maximum AC current output. If the inverter is not clipping, or its thermal limits have to do with total power output, rather than current output, then it will function as a constant power source. The difference being how the output current varies (or not) with varying grid voltage.
Of course not. No real world hardware is an ideal anything.
What is the difference in terms of the power source output characteristics? Presumably for linear loads, and to the extent that linear analysis is applicable, there is none. But for non-linear loads, you could see a difference?
PV modules act as current sources over almost all of the voltage range from 0V to Voc, and an inverter cannot itself sink or store energy. It can clip its output by driving the DC MPP over the cliff toward Voc whenever the array would otherwise source more power than the inverter can convert to AC, but when it is operating within its range of rated power it converts whatever the array sends it to AC current at the voltage to which the output is clamped plus the voltage rise that the current generates in the conductors.
OK, but for a given install, when the insolation is fixed, and the inverter is not clipping, if the grid voltage fluctuates, the result will be that the exported AC current goes down as the voltage goes up, while it goes up if the voltage goes down. This is the behavior of a (momentarily fixed) power source, not a current source.
Any clipping is done to the DC side; the inverter output behaves the same whether it is clipping or not.
Um, no, if the inverter is not clipping then its output current will go down.
The spec on an inverter shows both maximum power and maximum current. If the voltage changes either power or current has to change. I have seen inverters spec'ed with an operating AC voltage window and a range of maximum current over that window for max power to remain the same even if it is clipping. Clipping happens on the DC side.
That specification is certainly clear that the thermal maximum is a maximum power. But I could certainly see a spec that just provides max AC current and max AC power, and that if you dig deeper you find that the programming implements max AC current only, and that the max AC power is just a computed figure based on nominal voltage. I seem to recall having read of such an example here, but I may have imagined it.
The clipping is done on the DC side, in response to the limits on the AC side the inverter wishes to implement. If it implements just a maximum AC power limit, then agreed, both with and without clipping it acts as a constant power source (of magnitude depending on the available DC power). But if in part of the voltage range there is implemented a maximum AC current limit, then whenever that limit is controlling the behavior (lower AC voltage, so higher current needed for a given power), the inverter would act as a current source.
Often the spec is somewhere in between, as it were. That is, the max power will exceed the nominal AC voltage times the max current, but won't be as much as the max AC operating voltage times the current. So there is a AC voltage range in which the clipping is based on AC voltage and then a level at which the max AC power spec kicks in.
