If I may add to the above discussion.
In general, the reactive power associated with any circuit VI Sin (Theta). The symbol for reactive power is Q and its unit of measure is ' Volt-Amps-reactive' . The Q is derived from the quadrature (90 degrees) relationship between the various powers.
Therefore if the average power is zero and the energy supplied is returned within one cycle, why is reactive power of any significance ?
At every instant of time along the power curve(ie a sine-wave)that the curve is above the positive axis, energy must be supplied to the inductor, even tho it will be returned during the negative portion of the cycle.
This power requirement during the positive portion of the cycle requires that the generating plant provide this energy during that inteval.
The effect of reactive elements such as the inductor can be to raise the power requirement of the generating plant even though the reactive power is not dissapated but simply 'borrowed'. The increased power demand during these intervals is a cost factor that must be passed on to the industrial consumer. In fact most larger users of electrical energy pay for the apparent power demand rather than the watts dissipated since the VA used are sensitive to the reactive power requirement. The closer the power factor of the plant is to one, the more efficient is the plants operation since it is limiting its use of 'borrowed' power.
The net flow of power to the pure(ideal) Inductor is zero over a full cycle, and no energy is lost in the transaction. Therefore any energy lost in this transaction is not due to the inductor rather than to the accompanying resistance associated with the inductance.
As far as transformers go, The magnetization current in the transformer is not sinusoidal. The higher frequency components in the Mag current are due to magnetic saturation in the transformer core.
Once peak flux reaches the saturation point in the core a small increase in peak flux requires a very large increase in the peak magnetization current.
The core-loss current is nonlinear because of the nonlinear effects of hysteresis. the core loses are modeled as a resistance rather than an inductor, because of Hysteresis and eddy current losses.
So the total EXCITATION current is Iex = I(hysteresis + eddy current) + Im (the magnetization current). These are added as phasors and Im can be found by using the(as mentioned) square root sum of the squares and
the angle of the Im is the Arctan(X/R) or the (reactive component/real component)
Hope this helps
woc
