![]() | Only 14 pages are availabe for public view |
Abstract Power system networks are subjected to harmonics injection due to the presence of nonlinear loads and all kinds of static power converters ( converters, inverters, ...etc.). Harmonic flow analysis is the first step to be performed in order to obtain the existing harmonic distortion levels at different locations in any network. If the harmonic standards are violated, then there should be a method for controlling harmonics and reduce them to the acceptable limits. This work presents a harmonic flow analysis and different methods for controlling harmonics at different voltage levels of electrical networks. Some methods for controlling harmonics are based on filtering techniques and others are based on system loading control. Different practical examples are used to evaluate the performance of different proposed methods. An example of a factory, supplied from subtransmission network, using arc furnaces is studied. Harmonic measurements and flow analysis showed that capacitor banks installed for power factor correction lead to resonance and high harmonic levels. The capacitor banks suffered from blowing their fuses. Two alternative counter measures are suggested for overcoming this problem. The first one inserting series inductor with the capacitor bank. Another method for harmonic control in EHV networks is proposed. The method uses a certain combination of shunt and series passive filters to control harmonic current flow in a practical extra high voltage network. Harmonic flow analysis for this power network showed that considerable reduction in the background voltage harmonic level, at all network buses, is achieved. A new thyristor controlled filter is proposed for harmonic flow control at the customer location. The filter consists of a shunt passive filter controlled by a set of back-to-back thyristors or a triac. This filter overcomes the known disadvantages associated with the use of passive and active filters. The proposed filter is used to reduce the harmonic current distortion resulting from some typical nonlinear loads. The nonlinear loads under study are a high power 6-pulse converter and a thyristor controlled reactor. The results showed that considerable reduction in the total harmonic current distortion is achieved for each of these applications. Neural network is used to determine filter parameters that correspond to minimum total harmonic current distortion for each of the filter applications. |