GENERALISED POWER COMPONENTS DEFINITIONS FOR SINGLE AND THREE-PHASE ELECTRICAL POWER SYSTEMS UNDER NON-SINUSOIDAL AND NONLINEAR CONDITIONS

There is a need for generalised definitions of electrical powers to provide a simultaneous common base for measurement, compensation, power quality and identification of source of distortion. The major problem area today is definitions of powers in the presence of harmonics and nonlinear loads in the electrical power system. In such a scenario, there is a problem to accurately measure especially reactive (nonactive) power. This is important for accurate energy billing. Another important area is the mitigation equipment used to remove unwanted polluting quantities from the power system. Definitions of powers have an important role to play in providing the correct information for the optimal design and performance of such equipment. Evaluation of the quality of the power system to enable appropriate allocation costs to those causing deterioration in the power quality also cannot be discounted. To enable this cost allocation, there is a need to identify the polluters and the definitions should indicate degradation in power quality as well as identify the source of this degradation. Finally, it would be very useful if the definitions could also be used to perform a general analysis of the power system. This thesis commenced with investigation of the problem with an in-depth study of the existing definitions, and what other researchers have indicated about this problem, from the definitions perspective. The issues identified with current definitions are that some definitions do not possess the attributes that are related to source-load properties, and others are based on mathematical consideration and lack physical meaning. One issue in measurement of nonactive power is its nature of having zero average value. Another contributing factor is that the presence of source impedance is neglected in definitions. The use of RMS quantities to determine powers, especially instantaneous powers, in the presence of multi-frequency voltages and currents also contributes to the problem. Additionally, RMS based definitions are based on heating effect while not all sourceload relationships are totally of a heating nature. The RMS based definitions also do not satisfy the energy conservation principle. Another issue is that though harmonic currents are used, current definitions still utilise the RMS value of the voltage wave thus losing harmonic information. The solution is to decompose, as accurately as possible, the total instantaneous power into active and nonactive components utilising DC, fundamental and harmonics of voltage and current as well as being based on the power system properties. To enable this, the load model must closely represent the reality. This thesis presents the new instantaneous power definitions to achieve this. In addition to the fundamental, five sub-components for each of the active and nonactive parts are defined. The definitions are based on both the voltage and current DC, fundamental and harmonic components thus retaining harmonic information. Thus these definitions are not only mathematically based but also have a direct relationship with the load. The definitions do not make the assumption of zero source impedance. With good knowledge of the time profile of active and nonactive power components, an accurate time-domain measurement of the active and nonactive power is achieved. The components of powers introduced in the proposed definitions can be utilised to gauge power quality, to identify the source of distortion and to achieve optimal compensation. Based on the new instantaneous power definitions, the definitions for average values of the powers are also proposed. The recognition of positive going and negative going parts of the nonactive power waveform in defining the average nonactive power alleviates the problem of the “zero average nature” of nonactive power. It also retains energy information and satisfies the principle of energy conservation. The new definitions are evaluated for linear and non-linear loads in the presence of harmonics using benchmark case studies. Evaluation results demonstrate good performance of the proposed definitions. The practical applications of the definitions are explored with a number of examples from the areas of measurement of power and energy, compensation, detection of source of distortion and power quality. An application example showing the capability of the definitions in general analysis of a system is also presented. Good and useful results are obtained for all these examples. The proposed definitions are implemented on prototype systems with digital signal processors to demonstrate their practical usability. The proposed definitions are shown to be consistent with the traditional definitions under the conventional sinusoidal conditions, and their relationships to the commonly used existing definitions are also revealed.