Studies Of Effects Of System Operating Condition Variation On Oscillation Modes And Damping Torques

To further investigate the computational error in the probabilistic eigenvalue analysis, the multi-machine system dynamic model is used in this thesis to quantitatively study the varying characteristic of electromachenic oscillation modes and damping torques under system multi-operating conditions. Obtained result is compared with that of probabilistic eigenvalue analysis. Effects of system parameters on the system small signal stability are also analyzed on a practical system.The conventional stability analysis is usually based on one or several deterministic system operating conditions. However, in power system operation, there are many fluctuations and random factors, as a result, the electromachenic oscillation modes and electric torques will vary in some degree. The quantitatively investigation for the varying characteristics will provide the base for the further PSS settings and the probabilistic eigenvalue analysis involving multi-operating conditions.In modal analysis, the main concern is the participation degree of generators to oscillation modes. In this thesis, with the zero gain PSSs on all machines, the participation degrees of machines to the electromechanical modes are determined by computing the enginvalue sensitivities. For the 720 operating conditions of an eight-machine system, the computation is repeated 720 times.The order that each generator takes participation in oscillation modes is studied. For convenience, generators are classified as strong, moderate and weak related. Results show that the participation degree of generators to electromechanical oscillation modes is changed more or less under different operating conditions; the inter-area mode with more generators involved becomes more complicated; however, with the movement of operating condition, the participation orders of stronger participated generators in corresponding modes are less alterative in. From these, the result obtained from the probabilistic approach is validated.From the generalized multimachine representation, the electrical torque provided directly by PSS can be analytically represented. With the 720 operating conditions of the eight-machine system as samples, the damping torques provided by PSSs are calculated under all operating conditions. From the results, each PSS always providessignificant positive damping to the corresponding main oscillation mode under considered system operating samples, though the value of the damping torque is varied with the change of generator power; under some samples, a PSS may provide much small negative damping torques to other modes. These also indicate that the probabilistic technique is effective.Eigenvalue sensitivities quantitatively indicate the magnitude and trendency of parameter effects, and also give the direction for improving system stability. Based on the typical operating mode of a practical system, eigenvalue sensitivities with respect to system operating parameters and network parameters are computed. Effects of generator powers, PV voltages, line impedances, transformer taps on the system small signal stability are studied. These will provide the selection for the system stability improvement.