Keywords
MII-E (H62-electrotecnica), energía fotovoltaica, pequeño generador, modelo predictivo, corto plazo, mercado diario, photovoltaic energy, small producer, predictive model, short-term, spot market
Estudio y elaboración de estrategias de precios para pequeños generadores fotovoltaicos que acudan a mercado.
Abstract
Está habiendo una fuerte penetración de las energías verdes cuyo recurso renovable está disperso por todo el territorio nacional, conllevando a la aparición de pequeños generadores.
Con motivo del cumplimiento de los objetivos del 2020 el estado español asignó un régimen retributivo especifico a 8.037 MW de producción verde en el pasado año 2017. Esta financiación quería asegurar la rentabilidad de estos nuevos agentes.
Pese a ello, los generadores siguen prefiriendo acudir a un mercado a plazos vendiendo su energía a través de contratos bilaterales en vez de acudir al mercado diario.
Este proyecto ha querido proporcionar una herramienta a pequeños generadores fotovoltaicos para que descubran la rentabilidad que podrían obtener si acudieran directamente al mercado spot.
La estrategia que deberán seguir será ofertar siempre a su coste variable debido a su poco poder de mercado. Sin embargo, ya que el mercado diario ibérico es un mercado marginalista, serán retribuidos con la oferta que haya realizado el último generador casado.
Se ha realizado un modelo predictivo del precio horario de la electricidad en el corto plazo. Las variables de entradas usadas han sido variables de fácil acceso, certeras y públicas. Se han escogido como variables influyentes tras verificarlas con el método ANOVA: estación del año, mes del año, tipo de día, día de la semana, hora del día, perfil de producción horaria, precio marginal del día anterior, energía negociada en el día anterior y las temperaturas máximas y mínimas de los 10 núcleos poblacionales más importantes de España.
Como modelo predictivo se ha optado por un modelo de regresión no lineal basado en una red neuronal con 160 capas internas.
El análisis económico muestra que una instalación tipo de 4 MW tendrá una TIR entorno a un 10%. Los análisis de sensibilidad muestran que la variable exógena más influyente es el precio medio de la electricidad.
A change on the paradigm of the electricity generation mix is taken place. The green energies are strongly penetrating the market, whose resource is dispersed all along the national territory, leading to the creation of small producers. With the intention of fulfilling the objectives of the 2020 the Spanish government has assigned a special retributive regime to 8.037 MW of green production on the past year 2017. This finance wanted to ensure the profitability of these new agents. Nevertheless, the producers still prefer to sell their energy on the forward marked through bilateral contracts than selling it on the spot market, This project has tried to provide a tool to the small photovoltaic producers so they can discover the profitability the could achieve if they would go to the spot market. The strategy they should follow is bidding at their variable cost, due to its lack of market power. Nonetheless, as the Iberian spot market is a marginal market, they will be remunerated with the clearing offer price. A predictive model of the electricity hourly price for the next day has been developed. The income used variables have been: season, month, type of day (working, Saturday or Sunday/public holiday), day of the week, hour of the day, hourly production profile for a photovoltaic power plan according to the edification technical code, marginal hourly price of the day before, energy negotiated on the spot market the day before and the maximum and minimum temperatures of the ten most inhabited cities on Spain. As a predictive model it has been opted for a nonlinear regression model based on a neural network with 160 layers. The economic analysis proves that a facility of 4MW would have an IRR around 10%. The sensibility analysis show that the exogenous variable which affects the most the IRR is the mean price of electricity.
A change on the paradigm of the electricity generation mix is taken place. The green energies are strongly penetrating the market, whose resource is dispersed all along the national territory, leading to the creation of small producers. With the intention of fulfilling the objectives of the 2020 the Spanish government has assigned a special retributive regime to 8.037 MW of green production on the past year 2017. This finance wanted to ensure the profitability of these new agents. Nevertheless, the producers still prefer to sell their energy on the forward marked through bilateral contracts than selling it on the spot market, This project has tried to provide a tool to the small photovoltaic producers so they can discover the profitability the could achieve if they would go to the spot market. The strategy they should follow is bidding at their variable cost, due to its lack of market power. Nonetheless, as the Iberian spot market is a marginal market, they will be remunerated with the clearing offer price. A predictive model of the electricity hourly price for the next day has been developed. The income used variables have been: season, month, type of day (working, Saturday or Sunday/public holiday), day of the week, hour of the day, hourly production profile for a photovoltaic power plan according to the edification technical code, marginal hourly price of the day before, energy negotiated on the spot market the day before and the maximum and minimum temperatures of the ten most inhabited cities on Spain. As a predictive model it has been opted for a nonlinear regression model based on a neural network with 160 layers. The economic analysis proves that a facility of 4MW would have an IRR around 10%. The sensibility analysis show that the exogenous variable which affects the most the IRR is the mean price of electricity.
Información detallada
Materias, derechos, colecciones e identificadores