Técnicas instrumentales para la monitorización de la calidad de la energía (instrumental techniques for power quality monitoring)

Resumen

Power quality (PQ) in electrotechnical systems refers to a set of characteristics related to the energy transport, and delivery voltage waveform to the end-consumers in order to be properly served in line with the best criteria about the performance of the network. As a first requirement, the not interruption of the power service of the system should be in general as less frequent as possible. Secondly, the use of energy by customers must exhibit an optimum performance. The UNE-EN 50160 standard defines the characteristic from which the electric service of a common power system could be considered secure, continuous and constant. On the other hand, the IEC 61000-4-30 standard summarises the measurement methodologies that must be incorporate into the measuring equipment to be used, regarding their technology and class of precision to ensure that the previously established requirements are met. However, electricity networks are currently in a process of change and adaptation to new technologies and concepts of energy use within the future Smart Grid (SG). This conception is based on the new capabilities of systems production by non-conventional means. Modern instrumentation solutions and expert systems should inform about energy behaviours in real time, and specifically in the field of PQ, inform about its continuity and reliability, dealing with a continuous monitoring of multiple variables. Therefore, standardized methods are not enough to solve these issues and is necessary to introduce new measurement indices that allow a more flexible characterization of the network, including aspects related not only to producers but also consumers. This thesis reflects the state of the art related to PQ analysis solutions, particularly those related to the implementation of new quality indices in the domain of higher-order statistics (HOS). This work represents by itself a compendium of the different research lines related to the Analysis for Power Quality Monitoring issues through HOS. The investigation addresses the detection of PQ problems from two perspectives: the detection of specific events that occur on the network in isolation, as defined by current regulations, and the continuous monitoring detection. The new method is based on the hypothesis that, through HOS studies, an optimal statistical characterization of the network nodes, based on new waveform indicators, can be made. It will be shown that the statistics can not only help understand the events, but also help in a more realistic characterization of a PQ model not only for the conditions of supply at that point and network behaviour but also, load connected under normal operating conditions. Given a specific site or point of connection, the electrical parameters fluctuate due to numerous factors mainly related to the non-linear behaviour or the loads connected and the uncontrolled growth of Distributed Energy Resources (DER). In signal processing for PQ analysis, the second order is the most commonly method implemented in the traditional instrumentation and indicators as RMS and THD, for both voltage and current characterization. The thesis shows that non-Gaussian behaviour could be characterised by cumulants and moments of order, higher than second order. The benefits of using HOS are better explored bridging the gap between signal processing and power. The results obtained indicate that the proposed algorithm based on HOS is robust enough and, with an adequate usage, it is possible to estimate the transitions between the different network states. The main advantage of the method is the ease in terms of extracting features from the time domain, reducing computational complexity, allowing the spatial and temporal scalability of the proposed method and the compression of data within the framework of big data. In addition, the results of this work can be divided into different applications for monitoring both, the reliability and PQ, focused on the detection of events in 2D, bi-dimensional planes of different duration. The method allows considering and incorporating much simpler set points and control limits to the instrumentation, smart meters, and other advanced monitoring solutions. Finally, a summary of the investigation, experiences, and results within the framework of the international research stay in Dresden University of Technology can be found. The document shows how the use of HOS in voltage monitoring solutions is extended to current waveform models. Thus, the new models enabling the characterization of different types of circuit topologies of equipment where their impact on the network could be studied from new perspectives. La calidad de la energía (PQ) se refiere a un conjunto de características que deben cumplir los sistemas electrotécnicos en sus redes de distribución, suministro y entrega de la energía asegurando que esta sea suministrada adecuadamente a los clientes (tensión y corriente) finales, así como que su uso por parte de los clientes garantice un rendimiento óptimo y perturbe en la menor medida posible al sistema en general. El estándar UNE-EN 50160 define los límites para los cuales se considera un suministro constante y estable de electricidad. Por otro lado, el estándar IEC 61000-4-30 resume las metodologías de medición que deben ser incorporadas en los equipos de medida en función de su clase para asegurar que se cumplen los requisitos previamente establecidos. Sin embargo, la necesidad va aún más allá dado que actualmente las redes eléctricas se encuentran en un proceso de cambio y adaptación a nuevas tecnologías y conceptos de uso de la energía dentro de la red inteligente futura (Smart Grid). Esta concepción de lo smart, se basa principalmente en las capacidades de los sistemas de informar en tiempo real sobre los comportamientos de la energía, y concretamente en el campo de la calidad del suministro su continuidad, fiabilidad. En definitiva, se basa en la capacidad de las soluciones de instrumentación, y de los sistemas expertos, de hacer frente a una monitorización de múltiples variables de modo continuo. Por tanto, se hace necesario introducir nuevos métodos e índices de medición que permitan una caracterización más flexible de la red incluyendo aspectos relacionados no sólo con los productores sino también con los consumidores. La presente tesis plasma el estado del arte actual en lo relativo a las soluciones de análisis para la calidad del suministro eléctrico, concretamente las relacionadas con la implementación de nuevos índices de calidad y al dominio de los estadísticos de orden superior (Higher-Order Statistics, HOS). La elaboración del documento representa en sí un compendio de las distintas líneas de investigación relacionadas con la temática análisis de la calidad del suministro (Analysis for Power Quality Monitoring). Esta tesis aborda la detección de eventos de calidad del suministro, desde dos perspectivas, la detección de eventos concretos que se presentan en la red de modo aislado, tal como define la normativa actual, y la detección en continuo. El nuevo método está basado en la hipótesis de que mediante los HOS se pueden optimizar los estudios en el campo de la calidad del suministro mediante la caracterización estadística óptima de los nodos de la red basados en un nuevo indicador de la forma de la onda desde el cual se puede no sólo comprender mejor los problemas de calidad de la energía, sino también ayudar en una caracterización más realista de un modelo de calidad de la energía para las condiciones de suministro en ese punto, y el comportamiento del mismo en base a las cargas conectadas. Los resultados obtenidos indican que el algoritmo propuesto es robusto y estima adecuadamente las transiciones entre los diferentes estados de la red.