Per-unit system
The per-unit system is the trick that lets power engineers compare a 765 kV transmission line, an 11 kV generator, and a 400 V motor on the same page without drowning in zeros. You pick a 'base' value for each quantity — say 100 MVA and the local rated voltage — and then express everything as a fraction of that base. A voltage of 1.0 pu means 'exactly nominal'; 0.95 pu means '5% low'. Suddenly every voltage on the grid, no matter the actual kilovolts, hovers reassuringly around 1.0.
Its real magic appears at transformers. Convert impedances to per-unit on a common power base and the transformer's turns ratio vanishes from the maths — the high side and low side share the same per-unit impedance, so you can analyse a multi-voltage network as if it were one seamless circuit. This is why utility short-circuit studies, load-flow programs, and protection settings all speak in per-unit: it strips away the bookkeeping of voltage levels and exposes the engineering underneath.
A classic exam trap: per-unit impedance changes when you change the base. To move from a device's nameplate base to the study's common base, scale by (V_old/V_new)² × (S_new/S_old). Forget this and a fault current comes out wildly wrong.