M.Ing. / Electrical quality of supply affects every consumer, from the homeowner to large industry and mining. Customer awareness of poor power quality is increasing due to the more sensitive nature of modern equipment. The positioning of a new plant is therefore very important and cannot only be determined by the availability of resources in the area. To determine the quality of the electricity in a certain network, years of historical data is required. To get enough information regarding a distributions network performance would require voltage dip measurements at a few places in the network for at least three years. For this reason a new method are used in this study. By using the existing information in a network, like: • The amount of lightning strikes per year • The amount of electrocuted birds • The type of protection used • The amount of vandalism incidents on a line per year • The tower footing resistance • The fault level in the network The fuzzy logic principle can be used to determine how susceptible a network is to voltage dips, by determining the favourable and unfavourable conditions for a number of membership functions. Conventional set theory (Boolean) states that an element is either a member of a set or not. Fuzzy logic is an extension of conventional set theory enabling an element to belong to a set degree. This study will look at all the important factors influencing voltage dips, and to use these factors in a fuzzy logic program to determine the voltage dip risk factor for new customers setting up plant in a specific network. The voltage dip risk factor will then determine if it is cost effective for a customer to take on the additional cost in dip proofing his plant.

The increase in energy costs, restructuring of electricity supply industry, quality of supply standards and environmental responsibility, initiated an investigation to utilize available carbon monoxide for cogeneration at an industrial site. A literature study indicated that most of the elements involved in establishing a cogeneration plant, were investigated individually, but had not been evaluated as a system with an indication of their relationships. The main objective of this study was to create a methodology to evaluate the impact of load scheduling, cogeneration and electrical tariff structures on the energy cost of an industrial operation. A modelling methodology was developed to evaluate the requirements for each of the elements, which were identified as the following processes: Plant and stockpiles Gas and fuel Power generation technology Electricity tariffs Financial evaluation Each of the processes' input requirements were evaluated in terms of being sufficient in providing either useful information or a model from which information could be manipulated. This methodology was then applied to a titanium slag producer with electric arc furnaces and excess carbon monoxide, which was burnt and treated as waste. / Dissertation (M Eng (Electrical Engineering))--University of Pretoria, 2006. / Electrical, Electronic and Computer Engineering / unrestricted

An investigation has been undertaken to ascertain how readily a power system lends itself to statistical modelling. A nonlinear state variable model has been derived in terms of measurable states. This model is linear in its coefficients which are evaluated by the least squares fitting technique of regression analysis. The statistical model's performance is evaluated by comparison of its predicted system responses with those predicted by Park's formulation, and with those produced by a laboratory power system model. / Applied Science, Faculty of / Electrical and Computer Engineering, Department of / Graduate

The current practice in modelling unknown large power systems is reviewed in Chapter 1, and the inadequate representation by constant voltage and constant frequency is discussed. To determine the unknown large power system dynamics, estimation must be used. A complete model for estimation, including known and unknown power systems, is derived in Chapter 2 and the mathematical formulation in Chapter 3. Chapter 4 presents the estimation algorithm, data and results. It is found that the estimated unknown system parameter values are unique, independent of various operating conditions. Conclusions are drawn in Chapter 5. / Applied Science, Faculty of / Electrical and Computer Engineering, Department of / Graduate

No description available.

No description available.

No description available.

No description available.

No description available.

No description available.