Model reduction, stability, and coordinated tuning of supplementary controllers in electric power systems

The first part of this thesis deals with the analysis and system order reduction of synchronous and induction machine models using the method of integral manifolds. The classical steady-state stability of a synchronous machine equipped with an excitation system is investigated when connected to an infinite bus. Improved second-order swing and voltage models incorporating the effects of an automatic voltage regulator (AVR) or the combined effects of an automatic voltage regulator plus a power system stabilizer (AVR/PSS) are extracted from the original detailed model. These models give insight into the effects of critical parameters on the overall steady-state stability of the system. A new steady-state stability result is obtained using the voltage model. The reduced models also provide a tuning strategy for the excitation system parameters so that, for a wide range of loading conditions, there is an increase in steady-state stability limits without reducing the already poorly-damped electromechanical oscillations. Similar results show that the additional damping from the PSS is proportional to the product of a washout gain times a lead time constant. On the other hand, the PSS does not have much effect on the voltage modes obtained using only the AVR.; In a second tutorial example, synchronous and induction machine model order reduction is studied using the method of integral manifolds. It is found that the classical quasi-steady-state order reduction based on the neglect of stator transients gives erroneous small-signal stability results for small synchronous machines and for both small and large induction machines operating in the generator mode. A method based on the theory of higher-order integral manifold corrections allows the almost exact reconstruction of the original stability regions without the need for reintroducing the neglected stator transients. This method can easily be programmed in conventional small-signal stability programs to obtain more accurate stability results for power systems for which the classical quasi-steady-state order reduction does not give robust regions of stability.; As a second part of this research, a method for the coordinated tuning of power system controllers using linear programming is developed. Here, the concept of eigenvalue sensitivity is used to symbolically formulate the real part variation of the system dominant eigenvalue as an analytical function of controller parameter increments. Since no special assumptions have been made in the derivation of this algorithm, the proposed method can be used for the simultaneous tuning of diverse supplementary controllers such as PSS, FACTS and HVDC controllers.; The proposed tuning method is first applied to several cases of a one-machine-infinite-bus system where the combination of a PSS and a FACTS device controller is used to improve the damping of small-signal oscillations. The proposed tuning method is extended to the simultaneous tuning of multiple controllers in a multi-machine power system. The program is tested in several cases of a multimachine power system where the tuning of multiple controllers, including PSS and FACTS controllers, need to be coordinated. In each case, the system is initially given a set of PSS and FACTS controller parameters and the tuning program automatically drives the PSS and FACTS controller parameters to appropriate settings providing enough system damping for the electromechanical modes of oscillation. The effectiveness of the tuning method is also validated with digital simulations of the original nonlinear power systems.