Localización y valoración de fuentes de distorsión armónica y/o desequilibrio en redes eléctricas

Resumen

At present, electric power distribution is characterized by a wide variety of loads at its points of production, with, in most cases, electronic power converters comprising singlephase or three-phase loads. As power networks develop towards decentralized generation systems, more units, largely from renewable sources, are being directly connected to the distribution network through electronic converters. Irregular patterns of consumption alongside this distributed generation, contribute to an increase in harmonic distortion and current unbalance that together lead to an increase in network failure and equipment malfunction. At the same time, the very elements that have allowed the development and modernization of electric power distribution networks are those that display greater sensitivity to the disturbances (that they cause) in the network, endangering the system. Consequently, identification of harmonic distortion and/or unbalanced sources in the system is essential, both to demand responsibility from the originator/s (the electrical supply system and/or customer installation), and to take the necessary measures to solve the problem. The different approaches to tackling this problem can be classified into two groups. The first approach concerns measurements taken from a localized section (single-point), while the second concerns multi-point distributed measurement methods (that is, measurements taken simultaneously at different sections on the distribution network). Single-point measurements offer the advantages of low cost and ease of implementation, but in some conditions the information recorded with regard to the state of network disturbances can be inaccurate. The multi-point methods enables the recording of accurate information regarding the state of distortion and unbalance of the complete power system, and is consequently considered capable of definitively resolving the issue. In practice, however, approaches of this nature have to date proved difficult to perform and require the use of expensive and more complex instrumentation from both a hardware and software perspective. What is required is the definition of new indices and corresponding methods for evaluating of the levels of periodic disturbances in the measurement section. Currently, advances in digital signal processing techniques and the interconnection of equipment allowing real-time communication over communication networks make it feasible to address the practical development of a distributed measurement system aimed at quality power in electrical networks. However, it is necessary to define a distributed index that computes the information received from the different locations. These indices are formed from the combination of three indices that have been extensively used to assess distortion and unbalance by single-point methods. The combination of these presupposes greater success in the identification of periodic disturbances. Tests in different conditions have demonstrated that they cannot always clearly identify the source of disturbance, for which reason new proposals would be welcome. This thesis proposes a new multi-point distributed measurement index for the identification of the sources of harmonic distortion and/or unbalance, based in power terms on the IEEE Std. 1459 framework. The Std. 1459 presents a model of apparent power decomposition terms in conditions of distortion and asymmetry. However, this model lacks a practical assessment of the unbalance caused by the load. The new index algorithm is defined using a new method of apparent power decomposition which separates it into four components. The new index introduces two significant developments. First, it is defined as a function of power terms in the Std. 1459 framework, although it retains a fourth perturbation term. And second, the power terms used represent an evolution of the Std terms that most clearly separate distortion and unbalance. The index is applied and validated by means of an experimental platform connected to a real, low-voltage network. The thesis is organized as follows: in chapter I we study the problem of harmonic distortion and imbalance in electrical power systems in a conceptual way, in addition to a classification of the methods of locating sources of disturbance. In Chapter II a study of the methods based on measurements carried out at a single point in the network is carried out. Chapter III has a double objective, on the one hand the most representative multipoint distributed indices of the technical literature are presented, and on the other, the new proposal for the index of distributed measures for the identification of harmonic distortion source and/or unbalance is presented in the network, based on terms of power within the framework of the IEEE 1459 standard. In Chapter IV, in order to contrast and validate the new proposal, a comparative analysis of the multipoint distributed indices is carried out through an experimental platform . Finally, chapter V develops the conclusions.