On 8/20/20 2:24 PM, David Eckhardt wrote:
Some excellent points, Jim!
Let me take this concept of reflected power to something we are all more
familiar with: The AC power sockets in our walls. Whether its 120 vrms,
220 vrms, 308 vrms, 408 vrms, or 480 vrms, at 60 Hz, the power source
assumes a purely resistive load. Where else have we run into this
<snip>
PFC correction on loads >70W but smaller than, say, a few kVA, was driven in a lot of cases by the problem with neutral currents on 3phase Y office loads. All those switching power supplies which have a diode feeding a capacitor input filters have high peak currents at light loads. If you have them on three phases, those peaks occur 60 degrees apart - so the load isn't nicely balanced, and the neutral carries 3x the current of the phase legs.
It's the industrial consumers with motors (inductive load when lightly loaded), induction furnaces, and things like "magnetic ballasts" on lighting loads that have the classic lagging power factor which requires correction with capacitor banks or synchronous condensers.
Modern power supplies with "PFC correction" are really doing more about harmonics than actual lagging power factor. In any case, what they do is have a switching regulator that draws "the right amount" of current from the line in phase with the voltage.
On a grid basis, the "impedance matching" issue is mostly about power flow and transients. Utilities set the current phase relative to the voltage to control whether power flows one way or another - in the very short term they do that by switching in capacitors and inductors as needed. In the medium term, they do it by adjusting the generator rotational speed relative to the grid. If you load a generator heavily, it slows down, and the faster generators that are tied to it tend to pick up the load. These relationships are non-linear.
1000 km is 10 milliseconds light time (a big fraction of a cycle at line frequency), and a lot slower in a power line with distributed L and C, so the whole "relative phase management" is very, very complex, and difficult to stabilize. (My father's PhD committee chairman developed a lot of the computer codes to do this in the 60s) This is why they love DC interties - at one end you set constant voltage, and the other end you set constant current, and the system is stable.
So, PFC (and harmonic management) in office and light industrial applications is addressing a different problem than the PFC compensation in heavy industry and power system management.
But yes, an understanding of phasors is really useful for power systems design and operation.
An excellent textbook for this is by Theodore Wildi,"Electrical machines, drives, and power systems" that doesn't get into too much "rotating magnetic field design" for motors and such. Lots of useful stuff on synchronous motors/generators vs induction motors vs DC machines, and also on power systems design.
These PFC circuits are nothing more than matching circuits. They are
designed to provide to the power grid as much as possible a purely
resistive load from a complex power consuming appliance. Since the power
grid is pretty much zero-ohms, non-reactive, any resistive load can be
handled with grace and no need for generating extra power to compensate for
reflections caused by reactive loads to the grid.
Matching loads to sources doesn't apply just to RF energy. It's
everywhere. Bet you weren't aware of the PFC and power grid matching?
Dave - W?LEV
