| Part of the problem we are having in communication is that of | engineering jargon. | | If pf is defined as cos(theta) where theta is the phase difference | between voltage and current phasors, then assigning positive pf to | inductive loads and negative pf to capacitive loads is mathematically | untenable.
I think what makes this hard is the notion that the "normal" case is defined as being 1 instead of defined (somehow) as being 0. One could construct a formula to make it be 0 for the non-reactive.
The problem is, it really is a circular situation. That is, the phase angle can be represented precisely as a vector that can be at any position in a circle. So really, PF is merely the real component of a complex value. So we need an additional imaginary value to account for the reactance and define the phase angle completely (in two dimensions). But for most scenarios it seems the PF is all that is needed.
I've always wondered what would happen on a shared neutral circuit if an almost pure inductive load was on one side, and an almost pure capacitive load was on the other side. Normally with resistive loads, or at least with loads of like reactance, the neutral would have only the imbalance current, which would be zero with identical loads. But in the case of inductive on one side and capacitive on the other, of equal current, the neutral is going to be getting double current (assuming single phase).