Analysis And Control Of Low Frequency Oscillation In Power Systems

With the growing scale and complexity of power system, the power system dynamic stability issue became a critical problem. The inter-area low frequency oscillations caused by the interconnection of weakly coupled power systems threaten the security and stability of the interconnected power systems badly. It is an urgent task to investigate the mechanism and influencing factors of the inter-area low frequency oscillations, and design effective measures to damp the inter-area oscillations.When analyzing the oscillation phenomenon in certain ideal and uniform multi-machine system, it is indicated that the oscillation frequencies are equivalent to the frequencies of two separate systems, where one system is a simple system with two generators while the other is the single-machine infinite bus system. By the example of a typical 3-generator system, the influences of system structure, operating conditions and parameters on the low oscillation frequency are studied. System frequencies might be comparable under certain parameter conditions. When the frequency difference of two system oscillation modes is small and the amplitudes of these two modes in some electric variable are comparable, this electric variable will present differential frequency oscillation with relatively low frequency. Furthermore, if the frequency difference is small enough, the phenomena of ultra low frequency oscillation will result.The validity of two non-linear analysis methods for low frequency oscillation in power systems, i.e. normal form method and modal series method, are analyzed and compared from many aspects in this dissertation. An index El is defined to analyze the validity of the second-order participation factors of normal form method. Then the second-order participation factors of normal form method are qualitatively compared with that of modal series method. To evaluate the solution precision of different low frequency oscillation analysis methods, an error index err describing the closeness between the analysis result and the nonlinear simulation result is defined. The concept of valid region is then derived, which refers to the region in parameter space satisfying certain error index. The validity of normal form method, modal series method and linear modal method is compared through valid regions and the validity under different system stress condition. Besides, a nonlinearity index for normal form method is defined based on its original counterpart. The rationality of proposed index is proofed through comparing the real response and linear response of system modes.The most suitable location for power system stabilizer(PSS) to damp low frequency oscillations is investigated in this dissertation. When the phase of some generator's speed-participation factor is in the second or third quadrant, the damping can be worsened when this generator is equipped with PSS even if the generator has a large speed participation factor. So the phase of speed-participation factor has a prominent influence on the damping control. A composite participation index that takes into account both the input and control effect of PSS controllers is proposed to identify the best PSS location, which can compromise the requirements of multiple system operating conditions. Hence it is a more reasonable and effective method than the conventional participation factor method.The wide area feedback signal of power system stabilizer(PSS) is also discussed in this dissertation. A residue index is proposed to select the feedback signal for damping controllers and its validity is verified by the value and phase of real residues. This residue index is very convenient for it just needs the left and right eigenvectors of characteristic matrix. Through analyzing the influence of system operating condition on participation factors and residues, it is indicated that the change direction of generator speed residue is just the same as its participation factor. This gives some clues to select the wide area feedback signal for damping controller. By the example of a typical 2-area 4-generator power system, the validity and robustness of local signal, inner-area composite signals, inter-area composite signals used as feedback signal is compared.