Per-unit scaling in electric power systems

The authors propose a re-examination and formalization of per-unit scaling as applied to electric power system analysis. There are some desirable characteristics that the optimum per-unit scaling procedure should possess: the system should be compatible with older traditional systems, insofar as such compatibility does not violate the scaling axioms, and assigned bases should be selected so that scaled values are normally near unity. Procedures are suggested for base assignments. The principles of the scaling method are demonstrated with numerical examples