Fast decoupled load flow: hypothesis, derivations, and testing

A framework is presented that allows systematic studies on the hypothesis underlying a variety of versions of the fast decoupled load flow method, and their derivation. Decoupling is not seen as merely zeroing coupling submatrices of a full Newton Jacobian matrix. Instead, it is treated as an intelligent two-step procedure that solves the full Newton iteration equations without extra approximations. A new derivation of the standard fast decoupled method is presented, and the mechanisms governing the good performance of the general-purpose version of the method are clarified. Test results and illustrative examples supporting the basic theory are reported. Testing includes studies on very difficult real-life systems, e.g. a 1138 bus transmission system and a 308 bus low-voltage underground distribution system. >