Analysis of Capabilities of Machine Learning for Local Energy Communities to Provide Flexibility to the Grid
Energy market design is changing worldwide. Small-scale low carbon electricity generation or so- called Distributed Generation made it possible for neighboring citizens not only to jointly own and operate microgeneration or storage facilities but also be actively involved in the energy market by sel...
| Autor: | |
|---|---|
| Tipo de recurso: | tesis de maestría |
| Fecha de publicación: | 2021 |
| País: | España |
| Institución: | Universitat Politècnica de Catalunya (UPC) |
| Repositorio: | UPCommons. Portal del coneixement obert de la UPC |
| Idioma: | inglés |
| OAI Identifier: | oai:upcommons.upc.edu:2117/368279 |
| Acceso en línea: | https://hdl.handle.net/2117/368279 |
| Access Level: | acceso abierto |
| Palabra clave: | Renewable energy sources Energies renovables Àrees temàtiques de la UPC::Enginyeria elèctrica |
| Sumario: | Energy market design is changing worldwide. Small-scale low carbon electricity generation or so- called Distributed Generation made it possible for neighboring citizens not only to jointly own and operate microgeneration or storage facilities but also be actively involved in the energy market by selling the excess energy and earn a profit. The thesis is investigating the concept of Local Energy Communities from the current regulatory framework and technical point of view mainly assessing capabilities to be flexible on the energy market, meaning delivering or consuming electricity for maintaining the generation-consumption balance and the required grid frequency. Nowadays, thanks to smart meters deployment and sensors' measuring capabilities, the ability to gather data from customers up to the service provider have disrupted the electricity sectors, with the opening of new services and markets. This makes it possible to operate with the energy data more freely and frequently than before. Combining the disruption with new energy data and legislation enabling energy communities operation, this thesis assesses the possibility to make the dispatch and flexibility provision as automatic and “smart” as possible with the help of Artificial Intelligence and Machine Learning techniques. More precisely, the Master’s Thesis is aiming to answer the following questions regarding Local Energy Communities (LECs): 1. What are LECs and what are the positive and negative aspects of LECs' existence? 2. From Local Energy Community to Smart Local Energy Community - Can ML techniques support LEC to automatically dispatch/ feed energy from/to LEC? How can flexibility be used in the context of LEC? 3. What are the market structures and business models for LEC integration? Which energy market players participate in LEC business area? The thesis is organized as follows: The first part of the thesis is a theoretical part, where the literature review was done. After the general definition and legal introduction in Section 1, 2.1, 2.2 and 2.3. Different advantages and disadvantages of LECs are reviewed in Sections 2.3 and 2.4. The following section of this thesis presents a brief review of the literature on Big Data foundations and techniques (Section 4). The second part of the thesis is a practical experiment where the author works with a dataset from real households, performs basic data visualization tasks, and performs machine learning-based generation forecasting to evaluate flexibility. The methodology and results are explained in Sections 5 and 6. The last subsection of the given thesis compares different market models of LEC in different countries (Section 7). Main contributions, conclusions, and future work are discussed in Section 8. A representative list of references is provided at the end of the thesis. |
|---|