Date of Award

1989

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Electrical and Computer Engineering

First Advisor

Owen T. Tan

Abstract

The rapidly escalating costs of generation and transmission facilities as well as energy production have focussed attention on the need to optimize system capacity release and energy losses. The problem is generally solved by controlling the voltage or reactive power which also has a pivotal role within the distribution automatization schemes. A review of reactive power compensators and methodologies in designing and controlling distribution capacitors is given. In the presence of voltage source harmonics, a local load reactive power compensation method, cost-constrained power factor optimization using an LC compensator, is discussed. A new methodology based on a piecewise method for designing switchable capacitors installed on large distribution systems is described. Considered is the Var compensation problem concerning the optimal sizing and tap settings of capacitors installed on a radial distribution system such that for a given set of conforming load profiles, the energy and maximum power losses are minimized while the capacitor cost is taken into account. The algorithm is based on tearing the system into smaller subsystems, optimizing the individual subsystems and coordinating the subsystem solutions to yield the overall system optimization. Studies on a test system show the piecewise method to have satisfactory results as well as the significance of using appropriate load models. The extent of benefit of shunt compensator application can be further maximized by providing an appropriate integrated control strategy for the compensators. An expert system using a two-stage artificial neural control network is proposed to control in real time the multitap capacitors installed on a distribution system for a nonconforming load profile such that the system losses are minimized. The first stage of this control network has to predict the load profile from a set of prevailing input data obtained from direct measurements at certain buses as well as from the current tap positions of the capacitors. The second stage of the control network will select the optimal capacitor tap positions based on the load profile obtained in the first stage. The implementation of the control method to the test system shows the expert system to be computationally very efficient while having satisfactory results.

Pages

107

Share

COinS