Background formulae for an a.c. line
- Instantaneous voltage and current: $$v = V_0 \sin \omega t, \;\; i = I_0 \sin (\omega t - \phi)$$
- r.m.s. values: $$V = V_0/\sqrt 2, \; I = I_0/\sqrt 2$$
- Average (real) power delivered: $$P = VI \cos \phi$$. Here $$\cos\phi$$ is called the power factor.
When electricity is sent from a generating station (source) to a consumer (load), the two relevant quantities are
- the load current $$I$$ which the station must supply so that the consumer receives the desired real power $$P$$,
- the line (ohmic) loss in the transmission resistance $$R$$: $$P_{\text{loss}} = I^{2} R$$.
(a) Why a low power factor implies large power loss
Suppose the consumer needs a fixed real power $$P$$ and the r.m.s. transmission voltage is fixed at $$V$$ (this is normally the case for a given line rating).
From the power relation $$P = VI \cos \phi$$, the required line current is
$$I = \dfrac{P}{V \cos \phi}.$$
This current has to pass through the (unavoidable) resistance $$R$$ of the transmission line. The ohmic loss is therefore
$$P_{\text{loss}} = I^{2} R = \left( \dfrac{P}{V \cos \phi} \right)^{\!2} R = \dfrac{P^{2} R}{V^{2} \cos^{2} \phi}.$$
Hence, for fixed $$P,\,V,\,R$$,
$$P_{\text{loss}} \propto \dfrac{1}{\cos^{2} \phi}.$$
If the power factor is low (i.e. $$\cos \phi \ll 1$$), the denominator $$\cos^{2}\phi$$ becomes small, the required current $$I$$ becomes large, and the resistive loss $$P_{\text{loss}}$$ increases quadratically. Therefore, a low power factor leads directly to a large power loss during transmission.
(b) How a capacitor improves the power factor
Most industrial and commercial loads (induction motors, transformers, fluorescent lamps, etc.) are dominantly inductive. An inductive reactance $$X_L = \omega L$$ makes the current lag behind the voltage by an angle $$\phi_L$$, so the power factor $$\cos \phi_L$$ is less than unity.
Attach a capacitor of reactance $$X_C = \dfrac{1}{\omega C}$$ either in series with the load (for simple single-device circuits) or, more usually, in parallel (shunt) across the load (power-factor correction banks used in distribution networks):
- The capacitor current leads the voltage by $$90^{\circ}$$, exactly opposite to the lag introduced by the inductor.
- Because the two reactive currents are out of phase, they partially cancel each other.
Quantitatively, if the inductive and capacitive reactances appear in series with a resistance $$R$$, the net reactance becomes
$$X = X_L - X_C = \omega L - \dfrac{1}{\omega C}.$$
By choosing $$C$$ so that $$X_C \approx X_L$$, we can make $$X \to 0$$. The impedance $$Z$$ then approaches the pure resistance $$R$$:
$$Z = \sqrt{R^{2}+X^{2}} \longrightarrow R, \qquad \phi = \tan^{-1}\!\left(\dfrac{X}{R}\right) \longrightarrow 0.$$
Therefore $$\cos \phi \longrightarrow 1$$ and the power factor is said to be improved (brought as close to unity as economically feasible). With a higher power factor:
- The load draws a smaller current for the same real power.
- Transmission losses $$P_{\text{loss}} = I^{2}R$$ fall accordingly.
Capacitor banks are inexpensive, have negligible energy dissipation, and supply only the necessary leading reactive power. Hence they are the standard devices for power-factor correction in a.c. power systems.