Enhanced Newton-Raphson algorithm for normal, controlled and optimal powerflow solutions using column exchange techniques

A column exchange approach for powerflow (PF) problems is presented, in which various functional Jacobian matrices are obtained from a new basic formulation. Transformations between these matrices are described, which allow simple and flexible algorithms to be programmed. An enhanced version of the partitioned Newton-Raphson algorithm for normal, controlled and optimal PF yielded by the new approach is also presented. The new approach is shown to be much more powerful than the standard method. In addition to the calculation of normal PF solutions, the new algorithm can conveniently and efficiently model simple control activities, such as generator VAr limits, load-tap-changing transformers and variable phase shifting transformers. Complex control activities, such as control of remote variables or nonstandard system controls, as well as the calculation of sensitivity vectors for both objective functions and system constraints in optimal PF (OFF) can also be handled efficiently. Even when generator VAr limits, LTC transformers and transformer tap limits are included, the new algorithm still exhibits quadratic convergence. Computational efficiency is much greater than the traditional Newton-Raphson implementation (about seven times faster for practical UK systems) and is comparable with the fast decoupled loadflow (FDLF). The algorithm is also very reliable for systems with either meshed or radial configurations, including ill-conditioned networks