Real and reactive power optimization by approximate sub-optimum techniques

Abstract

Guided by the experience that real power is more sensitive to nodal phase angles and reactive power to voltage magnitudes, a new formulation of power flow op timization is developed which exploits this decoupling and consequently reduces problem size, computer memory and solution time. The decomposition of the variables and the load flow equations results in the formation of two subproblems- one corresponding to the reactive power equations (Q-Optimization) and the other to the real po wer equations (P- Optimization). The two subproblems are alternatively solved until the desired accuracy is at tained. The method has considerable potantiel for real time applications and the solution of only one problem, if desired. Approximate models are preferred for both mod ules. The errors induced by ignored set of power flow equations, which are the major limitations common to all approximate models, are overcome by the proposed formu lations. Q-optimization subproblem is devoted to the minimi zation of the real power transmission losses (RPTL) and the improvement of voltage profiles by adjusting the con trol variables of the module, i.e. generator voltage mag nitudes, VAR injection values of the shunt capacitor/reac tors and tap values of the tap changing transformers, p-optimization module is used to minimize the hourly fuel costs (HFC) of the thermal generating units. It is also possible to eliminate an unfavourable increase in the RPTL throughout the latter module. The proposed formulations are programed and the ef ficiency is tested by solving several test systems. The results are discussed and compared with those reported in the literature.