Aportaciones a la monitorización y predicción del recurso solar para la integración de fuentes renovables en sistemas eléctricos basadas en internet de las cosas

  1. Paredes Parra, José Miguel
Supervised by:
  1. Ángel Molina García Director
  2. María del Carmen Bueso Sánchez Director

Defence university: Universidad Politécnica de Cartagena

Fecha de defensa: 16 December 2022

Committee:
  1. J. V. Salcedo Chair
  2. Antonio Mateo Aroca Secretary
  3. Mariano Alarcón García Committee member

Type: Thesis

Abstract

This doctoral dissertation has been presented in the form of thesis by publication. The need to achieve a decarbonized economy to fight climate change has prompted several international commitments that oblige Spain and Europe to progressively reduce their greenhouse gas emissions until reaching climate neutrality in 2050. This goal can only be reached with a new energy model based on a massive integration of renewable energies. Among these technologies, photovoltaic solar energy is called to become one of the pillars to decarbonize the economy decarbonization thanks to its own technological maturity and the significant cost reduction, it has experienced in recent years. Within the different types of photovoltaic installations, one of the fastest growing is that of self-consumption, which are installations located near the points of electricity demand managed by the consumers themselves who also become providers of energy and services to the network, becoming the so-called prosumers.This massive integration of new distributed generation plants poses a challenge for the management of the electricity grid, which has traditionally responded to a unidirectional power flow model. In this unidirectional model, the power flows from the large generation units to the consumption points, adjusting the generation to achieve a power balance that allows effective regulation and control of the voltage and frequency values within the admissible ranges. The new decentralized distributed model requires new tools and management strategies to deal with the variability that renewable energies present due to their very nature and the large number and the geographical dispersion of installations that require greater flexibility to ensure a constant supply to cover demand and meet new challenges in the form of bidirectional flow of power These new management strategies are supported by the great advances made by information and communication technologies (ICTs) that have been gradually incorporated by the network operators in order to collect various variables related to generation units and consumer´s behavior in order to optimize distribution and consumption. In the case of photovoltaics, until a few years ago, the cost and complexity of monitoring systems for photovoltaic installations limited their use to large-capacity photovoltaic plants (from 1 MW, both for economic and regulatory reasons), but the appearance and rapid evolution in the market of the so-called Internet of Things, IoT, In this thesis, which has a practical approach, new contributions have been done to the three layers that make up an IoT system (perception, communication, and application) to propose a new management and communication prototype for selfconsumption photovoltaic installations based on open standards and low-cost IoT solutions. In the perception layer, several prototypes of monitoring systems have been developed and evaluated in accordance with the requirements of the EC–61724 standard, which describes the general guidelines for monitoring and analyzing the performance of photovoltaic power plants. Within the communications layer, work has been done on the integration and evaluation of new low-cost, high coverage, and low energy demand communications systems (LPWAN) to exchange data. Finally, in the applications layer, the analysis of short-term power generation prediction models for these facilities has been addressed to provide reliability and stability to the network, studying both different sources of irradiance data for the application of these models such as the influence of the fundamental parameters of the communications network in their results.