M.Ing. / This study investigates the limits to which distortion can be compensated in electrical distribution networks. Many factors hinder the "ideal" compensation of the line current to a perfect sinusoidal, in phase, scaled replica of the supply voltage. The different power theories that can be used to derive the reference compensation current, is one of the fundamental limits. The differences and correspondences between the FBD and Czarnecki theories are investigated in detail. Furthermore, it is shown by simulation and practical results, that the effectivity of compensation by the instantaneous power theory is load dependent. The compensation strategy and topology also impose limits on the effectivity of compensation. The position and strategy of a compensator determine whether the consumer is able to isolate his distortion from the rest of the network, or isolate himself from the distortion of other consumers. Distortion frequencies and system impedances are chosen specifically to visually show the effect of the different topologies and strategies, by means of simulation. One of the most important limits brought about by new technology, is the lagging of the reference compensation current, due to the use of signal processing in determining the reference signals. The effect of this lagging reference is clearly shown by means of simulation and practical compensation systems. An effectivity index for this phenomenon is defined for steady state systems. The dynamic limitations of this lagging reference compensation current is investigated thoroughly. A method is derived by which the maximum fault, due to the sampling time and signal processing time, can be calculated when a high f-region occurs in the line current. The dynamic response of the compensation system is also limited by the ' di-ability of the converter. Therefore, a method is developed to calculate the individual fault contribution of the lagging reference as well as the converter. This method enables designers to determine the parameters of the signal processing system and the converter in the planning phase. The use of a maximum sampling frequency is stressed. In the experimental work, a 10kVA, PWM-switched IGBT inverter is used as a parallel voltage fed compensator. It compensates for two non-linear loads : a three phase diode rectifier with an inductive load, and a three phase diode rectifier with a capacitive load. The practical compensation according to the instantaneous power theory of these two loads, confirms that the result of this compensation is load dependent. The effect of the lagging reference compensation current in a practical system is also shown.
| Identifer | oai:union.ndltd.org:netd.ac.za/oai:union.ndltd.org:uj/uj:9493 |
| Date | 16 August 2012 |
| Creators | Le Roux, Wiehan |
| Source Sets | South African National ETD Portal |
| Detected Language | English |
| Type | Thesis |
Page generated in 0.0022 seconds