Appropriate understanding of power system dynamics is in a big correlation with succesful operation and planning of electric power system (EPS). One of the transients is also an electromechanical disruption, which is caused by fluctuation in real power. The understanding of electromechanical wave propagation is necessary, if one intends to detect the arrival of an electromechanical wave on specific location. The accurate monitoring of wave propagation is the basis for calculation of fault location.
The electromechanical wave preparation speed is not constant in EPS. It depends both on electrical and mechanical parameters. By using the model for spreading of electromechanical disturbance (wave), the effects of electrical and mechanical parameters on wave propagation speed were analysed. The analysis was carried out within Matlab software environment.
Generators inertia constant H and line reactance X have the highest impact on wave propagation speed. The increasing of specific generators inertia constant decreases wave preparation speed, while decreasing of inertia constant increases wave preparation speed. Varying transmission line reactance has a similar effect. The electromechanical wave propagation is spread slower on transmission lines with higher reactance value.
Different methods for measuring the time of arrival on a specific location were also studied. It can be concluded that using the wave peak value moment is appropriate only when the dimensions of waves all across the system are equal. Only slight improvement of results was noticed by using wave bifurcation point. The most accurate approach is measuring the wave time of arrival with setting up the constant threshold on a proper low value. If the threshold is set too low, extra problems might appear while measuring in real EPS. Namely, it often occurs that constant threshold is exceeded in a steady state already. So pre-disturbance situation should be more reasonably referred to as a quasi-stationary state, as in case constant threshold is set too low, the measured time of arrival is completely wrong.
Finally, a new method for measuring wave time of arrival with setting up several constant thresholds is presented. Using this approach the precision of measuring can be increased and several other problems connected with setting the constant threshold too low can be avoided. In addition, the approach is simple to use, even compared with measuring the bifurcation point. More accurate measurements will increase the accuracy of applications that can calculate the location of a failure on the basis of measured times of arrival.
|
