Tarifas e remuneração de agentes virtuais de energia no âmbito dos mercados de energia eléctrica
Data
2019-06-21
Autores
Título da revista
ISSN da revista
Título do Volume
Editora
Resumo
Com as mudanças políticas e económicas ocorridas em diversos países no final dos anos 80, os produtores passam a vender energia e serviços directamente aos consumidores. São criados novos mercados de energia eléctrica e novas entidades como o agente virtual de energia (VPP), um novo conceito que permite agregar diferentes entidades, para participarem nos mercados de energia eléctrica, fornecendo novos serviços de geração e promoção de eficiência no consumo aumentando assim, os benefícios dos seus agregados. O trabalho desenvolvido teve como objectivo último dotar os VPPs com capacidade de cobrar o consumo de energia e remunerar de forma estratégica a geração e serviços dos agentes a si agregados, com base em tarifas e contratos estrategicamente definidos de acordo com as suas características, perfis e comportamentos. Foi assim definida a metodologia de Remuneração e Tarifação de Agentes Virtuais de Energia (RemT), uma estrutura de funcionamento para um agente agregador, dividida em diferentes processos de apoio à caracterização e gestão dos seus agregados. São utilizadas metodologias de caracterização de consumidores e perfis típicos de consumo, metodologias de definição de tarifas adaptadas e metodologias de remuneração de consumidores que participam em programas de flexibilização de consumos. Foram utilizadas técnicas de Data Mining para obter um conhecimento específico dos hábitos de consumo de vários consumidores de energia, com base no processo de descoberta de conhecimento em base de dados. Com base neste conhecimento são propostas estruturas tarifárias adaptadas a diferentes perfis típicos de consumo, que consideram o preço da energia formado em mercado, a quantidade de energia renovável (eólica e solar) produzida e a estratégia de actuação do agregador. Para validar estas metodologias foram usadas amostras de dados reais. Foram também propostas novas estruturas de remuneração para os agentes agregados com produção independente e participação em programas de demand response. A análise crítica dos três casos de estudo revelou que a aplicação das novas metodologias de normalização e clustering de dados, possibilitaram um menor erro que os métodos tradicionais, permitindo uma separação dos consumidores quer por quantidade quer por perfil típico de consumo. A metodologia de remuneração permitiu agrupar os consumidores consoante a sua semelhança, identificar os valores base de remuneração tendo em conta os valores à priori de cada consumidor e calcular valores mínimos médios e máximos de remuneração, apropriados para cada grupo de consumidores. Permitiu também definir valores de remuneração dinâmicos, que dependem não só de preços, mas também à variação da geração em cada momento, e ao incentivo à participação em programas de flexibilidade.
With the political and economic changes that took place in many countries in the late 1980s, producers began to sell energy and services directly to consumers. New electric energy markets and new entities such as virtual energy agent (VPP) are created, a new concept that allows to aggregate different entities, to participate in the markets of electric energy, providing new services of generation and promotion of consumption efficiency thus increasing, the benefits of their aggregates. The ultimate objective of the work was to provide the VPPs with the ability to charge energy consumption and to remunerate strategically the generation and services of the aggregate agents, based on tariffs and contracts strategically defined according to their characteristics, profiles and behaviors. It was defined the methodology of Remuneration and Tariffing of Virtual Energy Agents (RemT), an operating structure for an aggregating agent, divided into different processes to support the characterization and management of its aggregates. Methodologies are used to characterize consumers and typical consumption profiles, methodologies to define adapted tariffs and methodologies of remuneration of consumers who participate in programs of flexibility of consumption. Data Mining techniques were used to obtain a specific knowledge of the consumption habits of several energy consumers, based on the knowledge discovery process in database. Based on this knowledge, tariff structures are proposed adapted to different typical consumption profiles, which considers the price of the energy formed in the market, the amount of renewable energy (wind and solar) produced and the aggregator's performance strategy. To validate these methodologies, real data samples were used. New remuneration structures were also proposed for the aggregate agents with independent production and participation in demand response programs. The critical analysis of the three case studies revealed that the application of the new methodologies of normalization and data clustering, allowed a smaller error than the traditional methods, allowing a separation of the consumers either by quantity or by the typical profile of consumption. The remuneration methodology made it possible to group consumers according to their similarity, to identify the basic remuneration values taking into account the a priori values of each consumer and to calculate average minimum and maximum remuneration values appropriate for each consumer group. It also allowed for the definition of dynamic remuneration values, which depend not only on prices, but also on the variation of the generation at each moment, and on the incentive to participate in flexibility programs.
With the political and economic changes that took place in many countries in the late 1980s, producers began to sell energy and services directly to consumers. New electric energy markets and new entities such as virtual energy agent (VPP) are created, a new concept that allows to aggregate different entities, to participate in the markets of electric energy, providing new services of generation and promotion of consumption efficiency thus increasing, the benefits of their aggregates. The ultimate objective of the work was to provide the VPPs with the ability to charge energy consumption and to remunerate strategically the generation and services of the aggregate agents, based on tariffs and contracts strategically defined according to their characteristics, profiles and behaviors. It was defined the methodology of Remuneration and Tariffing of Virtual Energy Agents (RemT), an operating structure for an aggregating agent, divided into different processes to support the characterization and management of its aggregates. Methodologies are used to characterize consumers and typical consumption profiles, methodologies to define adapted tariffs and methodologies of remuneration of consumers who participate in programs of flexibility of consumption. Data Mining techniques were used to obtain a specific knowledge of the consumption habits of several energy consumers, based on the knowledge discovery process in database. Based on this knowledge, tariff structures are proposed adapted to different typical consumption profiles, which considers the price of the energy formed in the market, the amount of renewable energy (wind and solar) produced and the aggregator's performance strategy. To validate these methodologies, real data samples were used. New remuneration structures were also proposed for the aggregate agents with independent production and participation in demand response programs. The critical analysis of the three case studies revealed that the application of the new methodologies of normalization and data clustering, allowed a smaller error than the traditional methods, allowing a separation of the consumers either by quantity or by the typical profile of consumption. The remuneration methodology made it possible to group consumers according to their similarity, to identify the basic remuneration values taking into account the a priori values of each consumer and to calculate average minimum and maximum remuneration values appropriate for each consumer group. It also allowed for the definition of dynamic remuneration values, which depend not only on prices, but also on the variation of the generation at each moment, and on the incentive to participate in flexibility programs.
Descrição
Tese de Doutoramento em Engenharia Electrotécnica – Electrónica e
Computadores
Palavras-chave
Agentes virtuais de energia , mercados eléctricos