ESCUELA TÉCNICA SUPERIOR DE INGENIERÍA (ICAI) MÁSTER EN INGENIERÍA INDUSTRIAL

ESCUELA TÉCNICA SUPERIOR DE INGENIERÍA (ICAI) MÁSTER EN INGENIERÍA INDUSTRIAL ESTUDIO DE LA RENTABILIDAD DE INTRODUCIR BATERIAS Y ENERGÍA SOLAR FOTOVOLTAICA EN EL MODELO DE NEGOCIO DE LOS GESTORES DE CARGA: ESTUDIO DEL CASO ESPAÑOL Autor: Pilar Serrano Ojeda Directores: Tomás Gómez San Román y José Pablo Chaves Ávila Madrid Julio 2017

Pilar Serrano Ojeda ESTUDIO DE LA RENTABILIDAD DE INTRODUCIR BATERIAS Y ENERGÍA SOLAR FOTOVOLTAICA EN EL MODELO DE NEGOCIO DE LOS GESTORES DE CARGA: ESTUDIO DEL CASO ESPAÑOL

Scanned by CamScanner

Pilar Serrano Ojeda ESTUDIO DE LA RENTABILIDAD DE INTRODUCIR BATERIAS Y ENERGÍA SOLAR FOTOVOLTAICA EN EL MODELO DE NEGOCIO DE LOS GESTORES DE CARGA: ESTUDIO DEL CASO ESPAÑOL

ESTUDIO DE LA RENTABILIDAD DE INTRODUCIR BATERIAS Y ENERGÍA SOLAR FOTOVOLTAICA EN EL MODELO DE NEGOCIO DE LOS GESTORES DE CARGA: ESTUDIO DEL CASO ESPAÑOL Autor: Serrano Ojeda, Pilar Directores: Gómez San Román, Tomás y Chaves Ávila, José Pablo Entidad Colaboradora: ICAI Universidad Pontificia Comillas. RESUMEN DEL PROYECTO Palabras clave: Vehículo eléctrico, gestor de carga, batería, punto de carga, energía fotovoltaica, tarifas eléctricas 1. Introducción Después de casi un siglo en el que el motor de combustión interna ha dominado el sector del transporte, en los últimos años la industria del vehículo eléctrico ha experimentado un rápido crecimiento. Las ventas de vehículos eléctricos en todo el mundo han experimentado un crecimiento exponencial (Pontes, José. 2017). Sin embargo, la penetración del mercado es distinta según el país. El país con mayor cantidad de vehículos eléctricos es China, seguido por Estados Unidos. Europa está en una buena posición con respecto a las ventas de EV, con Noruega en la primera posición. En cuanto a España, su cuota de mercado apenas representa un 0,33% (Fernández, 2017). A lo largo de esta evolución de la penetración del vehículo eléctrico, se han desarrollado modelos de negocio para cubrir las necesidades de los usuarios. Entre los diferentes servicios que los EV proporcionan, se encuentran servicios de equilibrio y servicios auxiliares. Nuevos agentes, como los gestores de carga o los sistemas de control avanzado, gestionarían la programación de la carga de los vehículos eléctricos teniendo en cuenta la información relativa a las necesidades de movilidad de los usuarios y su horario de conexión a la red. El modelo de negocio para el desarrollo de la infraestructura de recarga pública desarrollado por empresas del mercado consiste en la instalación de puntos de recarga en ubicaciones estratégicas a lo largo de la red. Pero esto no es tan sencillo como puede parecer. Con el fin de proporcionar a los clientes el servicio adecuado, es necesario hacer una estimación de la potencia que se consumirá con el fin de dimensionar los puntos de carga. Esto requiere un pronóstico de la demanda y la potencia que será necesaria. Debido a estas incertidumbres, es difícil tener una previsión exacta. El pronóstico de la planificación de carga es uno de los factores que afecta al dimensionamiento de los gestores de carga ya que es necesario determinar la potencia contratada con la compañía de distribución. Además, los costes energéticos dependerán de los precios de la energía y las tarifas aplicadas. Parte del proyecto está dedicada a explorar las prácticas actuales y las mejoras potenciales de los modelos de negocio para la inserción de vehículos eléctricos en el contexto español. i

2. Objetivos del proyecto Para lograr el objetivo del proyecto, que consiste en analizar la rentabilidad de la introducción de nuevos enfoques en el modelo de negocio del gestor de carga, se siguen los siguientes pasos: a. Revisión de la literatura: Orientada a dos objetivos principales: - Revisión de literatura para recopilar información sobre la situación actual a nivel mundial y en España sobre temas tecnológicos relacionados con los aspectos relacionados con el proyecto. - Cuáles son las posibles características con efecto sobre la rentabilidad de una estación de carga operada por un gestor carga. b. Recolección de datos: Información de los diferentes aspectos técnicos involucrados en el desarrollo del modelo. Los diferentes enfoques son: - Datos de movilidad: Consumo de EVs, requerimientos, tiempo de consumo, frecuencia de consumo... etc. - Costes de energía: precios de la energía ( / kwh) y costes de capacidad ( / kw / año). - Tecnologías: i. Tecnologías actuales y futuras de las baterías y costes asociados. ii. Datos de tecnología solar fotovoltaica c. Modelado: Modelo de optimización para calcular la rentabilidad económica de incorporar una serie de tecnologías en una estación de carga. Este modelo sigue una función objetiva específica para optimizarlo como el enfoque final. d. Casos de estudio: Una vez desarrollado el modelo de optimización, se analizan diferentes casos de estudio. Estos casos se centran en diferentes alternativas de diseño y configuración de las estaciones de carga. i. Caso Base: Estación de carga básica típica ii. iii. iv. Estudio de caso 1: Estación de carga con batería Estudio de caso 2: Estación de carga con batería y venta de energía a la red Estudio de caso 3: Estación de carga con batería y producción de energía solar fotovoltaica v. Estudio de caso 4: Estación de carga con batería y producción de energía solar fotovoltaica considerando la futura reducción de costes de estos componentes. e. Resultados y análisis de sensibilidad: Se analiza la rentabilidad de las soluciones propuestas para el gestor de carga. Se realizan análisis de sensibilidad de los principales supuestos y parámetros. f. Conclusiones y recomendaciones: Conclusiones generales y futuras recomendaciones basadas en los resultados obtenidos a lo largo de los casos de estudio. ii

3. Metodología La metodología del proyecto se compone de dos partes principales, la primera se centra en el ámbito de investigación y la segunda y punto clave del proyecto, centrada en un análisis numérico desarrollado a partir de un modelo de optimización. La primera parte del trabajo de investigación consiste en la investigación de varios aspectos de la industria de la movilidad eléctrica: conocimiento de la dinámica y los procedimientos del gestor de carga en España, y un concepto básico de la tecnología relacionada. La segunda parte del trabajo de investigación consiste en analizar diferentes alternativas de diseño de la estación de carga y los beneficios asociados para el gestor de carga. Esta parte también incluye la búsqueda de datos numéricos sobre factores específicos e inversiones que podrían tener lugar en el modelo de negocio, datos que se utilizarían para la implementación del modelo de optimización posterior. Esta información se obtiene de varias fuentes: artículos del sector, informes de empresas específicas de la industria de la movilidad eléctrica e informes de instituciones del gobierno publicados en el Boletín Oficial del Estado. La segunda parte del proyecto consiste en la implementación de un modelo de optimización que representa el modelo de una estación de carga. Para el desarrollo del modelo, se emplea el Sistema de Modelado Algebraico General (GAMS). Este sistema permite alcanzar una solución óptima basada en una función objetivo y teniendo en cuenta varias restricciones. La función objetivo representa el beneficio potencial que se obtendría en la estación de carga. Aunque se ha desarrollado un primer modelo básico, posteriormente se analizan varios estudios de casos. Para cada uno de ellos, el modelo se modifica dependiendo de las restricciones, variables y parámetros relacionados. La implementación de esta metodología ha permitido comparar el desempeño de cada una de las alternativas para la estación de la carga y el impacto de las distintas opciones de tarifas para la compra y en algunos casos, venta de energía a la red. 4. Resultados Diferentes casos de estudio se han analizado, representando distintas configuraciones y diseños de la estación de carga. Asimismo, dentro de cada uno de estos casos, se han establecido como principales sensibilidades las distintas opciones de precios de energía obtenida de la red: Tarifa 2.1 (Vehículo Eléctrico), Tarifa 3.1 (Media tensión) o tarifa de autoconsumo. De acuerdo con los respectivos casos de estudio, se obtienen distintos resultados. Los diferentes casos que se analizan son los siguientes, con sus respectivas conclusiones principales: a. Caso base: Este es el caso con el menor beneficio Los resultados obtenidos son más favorables en el caso de emplear la tarifa 2.1 b. Caso de estudio 1: Se considera la instalación de una batería en la estación de carga. Se evalúa el potencial para balancear el consumo de energía y la potencia contratada con la consecuente reducción de costes. La introducción de batería supone un aumento en la rentabilidad de la estación de carga para el gestor de carga. Se optimiza la distribución de tarifas de acceso entre períodos, lo que supone una importante reducción de costes. Mayor instalación de cargadores eléctricos y un consecuente mayor beneficio. iii

Mayores beneficios en el caso de introducir la batería en la estación de carga. Mayor capacidad de la batería en el caso de tarifa de Vehículo Eléctrico (tarifa 2.1) debido al perfil de precios de la energía. Mayor costes totales y beneficios con tarifa de Vehículo Eléctrico c. Caso de estudio 2: Además de considerar la instalación de baterías en la estación de carga, surge la opción de vender la energía almacenada en la batería a la red. Con tarifas 3.1 y 2.1 no se obtiene ninguna diferencia en comparación con el caso de no vender energía, ya que el precio al que se vende la energía (precios de mercado) son menores que aquellos a los que se obtiene. La consideración de comprar energía a los precios de mercado hace que la venta de energía sea mucho más rentable que con las otras dos tarifas, ya que los precios de compra y venta de energía son los mismos, lo que hace que haya una posibilidad de obtener beneficios. El perfil de energía descargada a la red (V2G) no sigue una tendencia continua. Los momentos en que la estación de carga vende energía a la red corresponden al punto en el que el precio de la energía es mayor. La cantidad total de energía vendida a la red es mayor en el caso de los precios de mercado, seguidos por la tarifa 2.1. Esto es debido a que, como se ha mencionado anteriormente, el hecho de comprar y vender energía al mismo precio aumenta la posibilidad de obtener beneficios. Costes totales más altos y mayor beneficio final al usar la tarifa 2.1. d. Caso de estudio 3: Estación de carga con energía fotovoltaica. Resultados con respecto al Estudio de Caso 1: Se reduce la potencia contratada, ya que la batería permite obtener una distribución más adecuada entre periodos tarifarios. Disminución de costes totales Aumento de los beneficios La energía obtenida de los paneles solares apenas se utiliza en la carga directa de vehículos eléctricos. La energía de los paneles fotovoltaicos está optimizada a su máximo nivel, principalmente orientado al almacenamiento en la batería. La capacidad fotovoltaica se instala a su nivel máximo independientemente del tipo de tarifa utilizada, confirmando su alto potencial Con la tarifa 2.1 y la tarifa de autoconsumo, la situación óptima es utilizar la cantidad máxima de cargadores (10 unidades), mientras que en el caso de la tarifa 3.1, la cantidad óptima es de siete unidades. e. Caso de estudio 4: Estación de carga con energía fotovoltaica: considerando futuras reducciones en los costes de energía solar y de batería: Misma capacidad fotovoltaica (límite máximo) Mayor capacidad de la batería Se reduce la potencia contratada cuando se reducen los costes. Distribución similar entre periodos. Debido a la reducción de costes de instalación se destinan más ingresos a costes de potencia y energía. Mayor coste de energía y acceso Menores costes totales Mayores beneficios iv

Distribución de energía suministrada similar entre los diferentes periodos La Tabla 1 muestra la evolución de la rentabilidad de los diversos casos de estudio analizados. Al observar la variación al utilizar la Tarifa de Media Tensión (3.1) o la Tarifa de Vehículo Eléctrico (2.1), se observa cómo, a excepción del caso de venta de energía a la red, cada característica añadida en la instalación de la estación de carga aumenta su rentabilidad. También es importante estar al tanto de los beneficios que proporcionan los precios de mercado o las tarifas de autoconsumo en comparación con el caso de la tarifa de Media Tensión (3.1). CASO DE ESTUDIO Caso base No batería Caso de Estudio 1 Estación de carga con batería Caso de Estudio 2 Venta de energía a la red Caso de Estudio 3 Instalación solar fotovoltaica Caso de Estudio 4 Instalación solar fotovoltaica (reducción de costes) Tabla 1. Comparación de los beneficios de los distintos casos de estudio BENEFICIO Tarifa 3.1 Tarifa 2.1 Precios mercado Autoconsumo -1.787 11.656 - - 2.257 58.224-3.578-2.257 58.230 23.456-6.185 65.505-13.595 11.241 - - - 5. Conclusiones El negocio de carga de vehículos eléctricos se enfrenta a un fuerte proceso de desarrollo y existe una alta dependencia entre los fabricantes de vehículos eléctricos y los gestores de carga. La colaboración entre ambos es esencial, y también la elaboración de regulaciones aplicadas a toda la industria. Además, también encontramos la cuestión de la regulación de la figura del gestor de carga. Actualmente no existe un modelo sólido y común establecido para todas las empresas, por lo que la industria es muy sensible a los factores legales y económicos. Partiendo del caso base, que sería la instalación normal de una estación de carga, cada una de las características añadidas proporciona una ventaja adicional para el sistema. El hecho de introducir una batería en el sistema para optimizar la adquisición y venta de energía distribuyéndola en función del tiempo, permite al sistema aumentar sus beneficios y proporcionar un proceso de carga posiblemente mucho más eficiente, tanto para los propietarios de EVs como para para el sistema eléctrico. Los resultados obtenidos en la venta de energía a la red no añaden un beneficio sustancial en el sistema de estación de carga, ni bien aumenta los beneficios. Los resultados al agregar paneles fotovoltaicos muestran que este es un campo interesante de estudio porque la capacidad solar alcanza su nivel máximo de capacidad y supone un aumento considerable en el beneficio. La hipótesis de considerar una reducción de costes en el futuro tanto para baterías como para paneles solares, indica que las estaciones de carga serán más rentable a medida que la industria se desarrolla. Dentro de cada uno de los casos de estudio, se han considerado diferentes tarifas de compra de energía y costes de capacidad para poder analizar los distintos resultados v

y el efecto que tienen en la configuración y la rentabilidad del modelo de negocio del gestor de carga Con respecto a las tarifas de energía, los resultados obtenidos indican que la tarifa de Vehículo Eléctrico (2.1) permite tener una ganancia mayor que en el caso de Media Tensión (3.1) gracias a un mayor uso de la batería. Además, el hecho de tener en cuenta la tarifa de autogeneración (ya que se está autogenerando energía) tiene sus respectivas ventajas, incluyendo un beneficio mayor que en el caso de la tarifa 3.1. Esto se debe a que los precios de la energía y de capacidad son menores. Además, al considerar los precios de mercado para comprar y vender energía, las ventas de energía a la red se vuelven rentables. Por lo tanto, un rediseño de las tarifas de electricidad debe considerarse seriamente para enviar señales económicas de precio eficiente y hacerlas neutrales para saber si la misma tecnología (como las baterías) está conectada en casa, lugar público o en cualquier otro lugar. Como consideración adicional, se ha calculado una tarifa de cobro fija para los consumidores de cara a obtener una base de ingresos y de esta manera analizar cuál sería la configuración óptima para una estación de carga. La tarifa /kwh se ha elegido de acuerdo a la mayor media equivalente de cada uno de los casos, siendo esta la correspondiente a la tarifa 2.1. De acuerdo con esto se obtiene que la configuración más rentable sería aquella en la que se instala una batería y paneles solares fotovoltaicos, usando tarifas de autoconsumo para la compra de energía. Hay muchas oportunidades para el negocio del gestor de carga. Estos avances tendrán un peso creciente en el futuro ya que la tecnología alcanza una base más sólida. Otro factor importante a tener en cuenta es que para el éxito de la industria de la movilidad eléctrica será esencial una gran presencia de este mercado entre los usuarios. Actualmente la cuota de mercado en España es todavía demasiado pequeña para considerar un sistema como el desarrollado en el proyecto. Pero siguiendo las predicciones, es alentador pensar que en algunos años esta industria seguirá un crecimiento exponencial y todas estas suposiciones y posibilidades constituirán una realidad. 6. Referencias PONTES, JOSE. 2017. "EV Sales". Ev-Sales.Blogspot.Com.Es. http://ev-sales.blogspot.com.es/. FERNANDEZ, SERGIO. 2017. "Top 10 De Los Países Que Más Coches Eléctricos Han Vendido En 2016". Forococheselectricos. http://forococheselectricos.com/2017/02/top-10-de-los-paises-que-mas-cocheselectricos-han-vendido-en-2016.html. vi

STUDY OF THE PROFITABILITY OF INTRODUCING BATTERIES AND SOLAR PHOTOVOLTAIC INSTALLATIONS IN THE CHARGING NETWORK OPERATOR S BUSINESS MODEL: SPANISH CASE STUDY Author: Serrano Ojeda, Pilar Supervisors: Gómez San Román, Tomás y Chaves Ávila, José Pablo Collaborative entity: ICAI Universidad Pontificia Comillas. ABSTRACT Key words: Electric vehicle, charging network operator, battery, charging point, photovoltaic energy, electricity tariffs 1. Introduction After nearly a century with the internal combustion engine dominating the transportation sector, in the recent times the electric vehicle industry has experienced a rapid growth. The sales of electric vehicles worldwide follow an exponential tendency (Pontes, Jose. 2017). However, the penetration of the market differs considerably depending on the country. The country with the higher amount of electric vehicles is China, followed by the United States. Europe is also at a good position regarding EV sales, with Norway at the leading position. Regarding Spain, its market share within the mobility industry represents just a 0.33% (Fernandez, 2017). Along this evolution, several business models have been developed in order to cover the necessity demanded by the customer. Among the different services that EVs could provide, balancing and ancillary services have been put forward; new actors such as charging network operators or advanced control systems would manage the charging scheduling of EVs taking into account the information regarding EVs owners mobility needs and their connection schedule to the grid. The actual business model for the development of public charging infrastructure used by several companies in the market consists of the placement of charging points in strategic location along the network. However, this is not as straightforward as it may seem. In order to provide customers with the adequate service, it is necessary to make an estimation of the power that will be consumed at different times of the day in order to dimension the charging points. This requires forecasting the demand and the power that will be needed. Due to these uncertainties, it is difficult to have an exact forecast. The forecast of charging scheduling is one of the factors that affects the dimensioning of the charging installations because it is necessary to determine the contracted power with the distribution company and pay accordingly. In addition, energy costs would depend on the energy prices and tariffs applied. Part of the project is dedicated to explore current practices and potential improvements for business models for electric vehicles aggregation in the Spanish context. vii

2. Project Objectives viii In order to achieve the project final purpose, which is to analyze the possible profitability of introducing new approaches in the charging network operator s business model, the following steps are followed: Previous literature review: Oriented to two main goals: - Literature review to gather information about the current situation worldwide and in Spain about technology concerning the project topic. - Delivering which are the possible features having some effect on the profitability of a charging station operated by a charging network operator. Data collection: Collection of information of different technical aspects involved in the posterior development of the model. The different sectors to focus are: - Mobility data: Consumption of EVs, energy requirements, hours of consumption, frequency of consumption etc. - Energy costs: energy prices ( /kwh) and capacity charges ( /kw/year). - Technologies: Actual and future battery technologies and associated costs. Photovoltaic technology data Modeling: Optimization model to calculate the economic profitability of incorporating a series of technologies into a charging station. This model follows a specific objective function to optimize it as the final approach. Case studies: Once the optimization model is developed, different case studies are analyzed. These cases are focused on different alternatives to design and size the charging installations. These cases are: i. Base case: Typical basic charging station ii. Case study 1: Charging station with battery iii. Case study 2: Charging station with battery selling energy to the grid iv. Case study 3: Charging station with battery and photovoltaic energy production v. Case study 4: Charging station with battery and photovoltaic energy production considering future cost reductions for those components. Results and sensitivity analyses: It is analyzed to what extent the proposed solution is beneficial for the charging network operator. Sensitivity analysis of the main assumptions and parameters are carried out. Conclusions and recommendations: Overall conclusions and future recommendations based on the results obtained along the case study. 3. Methodology The methodology of this project is composed of two main parts. The first one is focused on a research scope, and the second one and key point of the project, centered on a numeric analysis developed by using the developed optimization model. The first part of the research scope consists on a deep research of several aspects of the industry of the electric mobility. This involves having knowledge of the dynamics and procedures of the charging network operator in Spain, and a basic concept of the technology involved. The second part of the research consists of analyzing different

design alternatives for the charging installations and the assessment of the associated profitability for the charging network operator. This part has also involved research based on several sources providing numeric data about specific factors and investments taking place in the business, data that would be used for the implementation of the developed optimization model. This information is obtained from several sources: articles from the sector, reports of specific companies of the electric mobility industry, and reports from institutions such as the government and published in BOE (Boletín Oficial del Estado). The second part of the project consists of the development of an optimization model representing different designs of a charging station. For the development of the model, the General Algebraic Modelling System (GAMS) is used. This choice has been made in order to work with a program that allows reaching an optimal solution of an objective function taking into account several restrictions. The objective function represents the potential profit obtained by the charging station operator. Even though a first basic model is developed, several case studies have been analyzed. For each of them, the model has been adapted changing conveniently the involved restrictions, variables and parameters. The implementation of this methodology has allowed to compare the performance of each of the alternative designs for the charging station and the impact of different tariff options for buying from and, in some cases, selling energy to the electrical grid. 4. Results Different case studies have been analyzed representing configurations and design features of the charging station. In addition, within each of the options, sensibilities according to energy tariffs are considered for the energy bought from the grid: Tariff 2.1 (Electric Vehicle), Tariff 3.1 (Medium Voltage) or Self-generation tariff. According to the respective case studies, several results are obtained from each of the considered situations. The different cases that are analyzed are the following, with their respective main conclusions: a. Base case: Typical charging station with no battery installed This is the case with the lower amount of profits Better results in the case of buying at 2.1 Tariff b. Case study 1: Charging station with a stationary battery. Potential for balancing energy consumption and contracted power along the different time periods with the consequent cost reductions. The introduction of batteries leads to an increase in the profitability of the charging station. The distribution of contracted power among periods is optimized, leading to a significant cost reduction. The fact of introducing a battery in the charging station involves a higher installation of number of electric chargers, and a consequent higher profit. The overall balance of profits is higher in the case of introducing battery in the charging station. Higher battery capacity in the case of Electric Vehicle tariff (2.1 tariff) because of the energy prices profile. Higher total amount of costs and profits with Electric Vehicle tariff ix

c. Case study 2: In addition to considering the installation of batteries in the charging station, it arises the possibility of selling the energy stored in the battery back to the grid. With 3.1 and 2.1 tariff no difference is obtained in comparison with the case of not selling energy, because the selling price (Day-Ahead prices) are lower than the ones at which it is bought. The consideration of buying energy at the Day-Ahead prices makes the energy sale much more profitable than with the other two tariffs, because the prices of buying and selling is the same and therefore there is a chance of getting some revenues. The profile of energy injected back to the grid (V2G) changes with the time. The moments when the charging station sells energy to the grid correspond with periods of high energy prices. The total amount of energy sold to the grid is higher in the case of Day- Ahead prices, followed by tariff 2.1. This is because of what was mentioned before; the fact of being the same the prices of buying and selling gives the system a chance of obtaining more revenues. Higher total costs and higher final profit when using 2.1 tariff, because of the higher usage of the battery and energy prices, and therefore the higher prices charged to the users. d. Case study 3: Charging station with battery and photovoltaic energy production. Results with respect to Case Study 1: Contracted power is reduced as the installation of batteries allow to obtain a more adequate distribution among periods Decrease in total costs due to savings in energy at expensive hours Increase in profits The energy obtained from the installed solar panels is barely used in the direct charge of electric vehicles. The energy from photovoltaic panels is optimized at its maximum level, mostly oriented to the storage in the battery. Photovoltaic capacity is installed at its maximum level independently of the type of tariff employed, confirming its potential. With tariff 2.1 and self-generation tariff, the optimal situation is to use the maximum amount of chargers (10 units), while in the case of tariff 3.1, the optimal amount if of seven units. e. Case study 4 - Charging station with battery and photovoltaic energy production: considering future cost reductions in solar and battery components: Same photovoltaic capacity (maximum limit) Higher contracted power when prices are reduced. Similar distribution among periods. Due to savings on installation costs, expenses can be meant to higher energy and power costs Higher energy and access costs Lower total costs due to the reduce on installation costs Higher profits Similar supplied energy distribution among different hours x

CASE STUDY Base case No battery Case Study 1 Charging station with battery Case Study 2 Charging station selling energy to grid Case Study 3 Solar Photovoltaic installation Case Study 4 Solar Photovoltaic installation (cost reduction) Table 2. Comparison of profits of all the different case studies PROFIT 3.1 Tariff 2.1 Tariff Day-Ahead prices Self- generation -1.787 11.656 - - 2.257 58.224-3.578-2.257 58.230 23.456-6.185 65.505-13.595 11.241 - - - Table 2 shows the profitability of the several case studies. This value is obtained based on the supposition of charging the EV users with the corresponding energy prices of the moment of charging plus a 40% of margin. When looking at the variation of profits when using Medium Voltage tariff (3.1) or Electric Vehicle tariff (2.1), it is observed how, except of the case of selling energy to the grid, each added feature considered in the charging station installation increases its profitability. It is also important to be aware of the benefits that buying at Day- Ahead tariff or Self-generation tariff provide in comparison with the case of Medium Voltage tariff (3.1). 5. Conclusions Electric vehicle charging is facing a strong development involving a close interdependence between electric vehicle manufacturers and charging network operators. The collaboration between both of them will be essential for the elaboration of common terms oriented to build solid legislations. Currently there is not a solid and common understanding of the role of charging network operators. Therefore, the industry would be very sensible to legal and economic implications of that future legislations. Based on the base case, that involves a fast EV charging station, in this project is demonstrated the each of the added features, batteries and solar energy production, would provide an additional advantage to the system. The fact of introducing a battery into the system in order to optimize the acquisition and sale of energy distributing it along the time, allow the system to increase its profits and provide a much more beneficial charging process for both the EVs owners as well as for the whole electricity system. When selling energy back to the grid the battery does not add a substantial benefit in the EV charging, however it increases the profits of the installation. The results when adding photovoltaic panels showed that this is a field interesting to exploit because the solar capacity was installed at its maximum capacity level and led to a considerably rise in the profits. Also the assumption of considering a cost reduction in the future for both PV and batteries will be much more profitable as the industry develops. Within each of the case studies, different tariffs regarding energy and capacity prices have been considered in order to analyze the differences and how they could affect xi

the performance and profitability of the business model of the charging network operator. Regarding charging tariffs, the results obtained indicated that the Electric Vehicle tariff (tariff 2.1) allow to have a higher profit than in the case of Medium Voltage tariff (tariff 3.1) thanks to a higher usage of the battery. In addition, the fact of taking into account self-generation tariff when buying energy has advantages, including a higher profit than in the case of the basic tariff 3.1 due to lower energy costs and capacity costs. Furthermore, when considering the wholesale energy prices to buy and sell energy, energy sales become profitable. Therefore, a redesign of electricity tariffs should be seriously considered to send economic efficient price signals and make them neutral to whether the same technology (such as batteries) is connected at home, public place or elsewhere. As an additional consideration, a fix tariff for EV users has been set in order to obtain a fix revenue base and assess which one is the more profitable configuration of the charging station. The /kwh established is the maximum between all the cases, meaning the one of Tariff 2.1. With this regard, the optimal configuration is the one in which a stationary battery and solar photovoltaic energy source are installed, buying energy a Self-Generation tariff. There are several opportunities for the business of the charging network operator. These advances will have an increasing weight in the future as the technology reaches a higher and more solid basis. Another important fact to take into account is that for the success of the electric mobility industry it would be essential a major presence of this market among customers. Currently the market share in Spain is still too small to consider a system like the one developed in the project. However, following the predictions, it is encouraging to think that in some years this industry will follow an exponential grow and all this assumptions and possibilities would be indeed a reality. 6. References PONTES, JOSE. 2017. "EV Sales". Ev-Sales.Blogspot.Com.Es. http://ev-sales.blogspot.com.es/. FERNANDEZ, SERGIO. 2017. "Top 10 De Los Países Que Más Coches Eléctricos Han Vendido En 2016". Forococheselectricos. http://forococheselectricos.com/2017/02/top-10-de-los-paises-que-mas-cocheselectricos-han-vendido-en-2016.html. xii

INDEX GLOSSARY... 1 1. INTRODUCTION AND OBJECTIVES... 3 2. MOTIVATION... 5 3. TECHNOLOGY DESCRIPTION... 7 3.1 The figure of the charging network operator... 7 3.2 Charging modalities... 8 4. PRESENT OF THE ELECTRIC MOBILITY INDUSTRY... 11 4.1 Global EVs market... 11 4.2 Actual situation in Spain... 14 4.2.1 Electric Vehicles sales evolution in Spain... 14 4.2.2 Charging infrastructure in Spain... 15 4.2.3 Charging network operators in Spain... 16 4.2.4 Public economic incentives in Spain... 21 4.2.5 Similar projects... 24 5. METHODOLOGY... 27 5.1 Assumptions... 27 5.2 Factors involved... 28 5.2.1 Energy demanded by electric vehicles... 28 5.2.2 Cost of contracted capacity... 29 5.2.3 Energy costs... 31 5.2.4 Regulated costs in the case of self-generation... 32 5.2.5 Charging station... 33 5.2.6 Transformer... 34 5.2.7 Batteries... 35 5.2.8 Photovoltaic investments... 37 5.2.9 Sale of energy to the network... 39 5.3 Algorithm... 40 5.3.1 Objective function... 40 5.3.2 Restrictions... 42 6. RESULTS ANALYSIS... 45 6.1 Case study 1: Charging station with battery... 46 6.2 Case study 2: Charging station selling energy to the grid... 54 6.3 Case study 3: Charging station with solar photovoltaic installation... 63 xiii

6.4 Case study 4: Charging station with solar photovoltaic installation future cost reduction... 73 6.5 Results comparative... 78 7. CONCLUSIONS AND FUTURE APPROACHES... 83 7.1 Conclusions... 83 7.2 Future approaches... 86 8. BIBLIOGRAPHY... 87 ANNEX 1... 91 ANNEX 2... 99 xiv

FIGURE INDEX Figure 1. Electric Vehicles sales worldwide... 11 Figure 2. Number of EV per country... 12 Figure 3. EV sales per country... 12 Figure 4. EV sales in Spain per month... 14 Figure 5. EV sales in Spain per year... 15 Figure 6. IBIL normal charging points location in Spain... 17 Figure 7. IBIL fast charging points location in Spain... 17 Figure 8. GIC charging points location in Spain... 18 Figure 9. Endesa charging points location in Spain- Mallorca... 19 Figure 10. Gas Natural Fenosa charging points location in Spain... 20 Figure 11. Fenie Energía charging points location in Spain... 21 Figure 12. Charging station in La Granja, Segovia... 25 Figure 13. Planning of vehicles arriving at the station and energy demanded... 29 Figure 14. Tariff 3.1 periods... 30 Figure 15. Tariff 3.1 and 2.1 Energy price profile... 32 Figure 16. Medium voltage tariff (3.1) and Self-Generation Energy price profile... 33 Figure 17. Transformer s cost... 35 Figure 18. Solar panel costs... 38 Figure 19. Combination of possible solar panel costs... 39 Figure 20. Graphic representation of Case Study 1... 46 Figure 21. Graphic representation of Base Case... 46 Figure 22. Contracted power for tariff 3.1 and 2.1 when no battery is installed. Case Study 1. 47 Figure 23. Contracted power for tariff 3.1 and 2.1 when having battery. Case Study 1... 48 Figure 24. Cost distribution for 3.1 tariff, with and without battery. Case Study 1... 49 Figure 25. Supplied energy procedure, Case Study 1... 50 Figure 26. Battery level VS energy price (3.1 tariff), Case Study 1... 51 Figure 27. Battery level VS energy price (2.1 tariff), Case Study 1... 52 Figure 28. Graphic representation of Case Study 2... 54 Figure 29. Contracted power for tariffs 3.1, 2.1 and Day-Ahead. Case Study 2... 56 Figure 30. Supplied energy procedure, Case Study 2... 58 Figure 31. Battery level VS energy price (2.1 tariff), Case Study 2... 59 Figure 32. Battery level VS energy price (Day-Ahead prices), Case Study 2... 59 Figure 33. Energy withdrawn from the battery to the EVs or to the grid, Tariff 2.1. Case Study 2... 60 Figure 34. Energy withdrawn from the battery to the EVs or to the grid, Day-Ahead prices. Case Study 2... 61 Figure 35. Energy sold from the battery to the grid (B2G). Case Study 2... 61 Figure 36. Graphic representation of Case Study 3... 63 Figure 37. Contracted power for tariffs 3.1, 3.1-self generation and 2.1. Case Study 3... 65 Figure 38. Supplied energy procedure. Case Study 3... 68 Figure 39. Battery storage procedure. Case Study 3... 69 Figure 40. Battery level VS energy price (tariff 2.1). Case Study 3... 70 Figure 41. Battery level VS energy price (tariff 3.1). Case Study 3... 70 xv

Figure 42. Battery level VS energy price (tariff 3.1-self generation). Case Study 3... 71 Figure 43. Battery costs evolution... 73 Figure 44. Contracted power for tariff 3.1, with basic and reduced costs. Case Study 4... 74 Figure 45. Supplied energy procedure. Case Study 4... 76 Figure 46. Battery storage procedure. Case Study 4... 77 xvi

TABLE INDEX Tabla 1. Comparación de los beneficios de los distintos casos de estudio... v Table 2. Comparison of profits of all the different case studies... xi Table 3. Electric Mobility charging figures in Europe and Spain... 7 Table 4. Charging points per country... 13 Table 5. Number of charging points per province in Spain... 15 Table 6. Charging network operators in Spain and number of charging points... 16 Table 7. Endesa electric mobility services... 18 Table 8. MOVEA 2017 plan aids distribution... 22 Table 9. Capacity and autonomy of EV models... 28 Table 10. Tariff 3.0 Periods... 30 Table 11. Regulated Capacity costs for tariff 3.1 A... 31 Table 12. Variable Energy costs for tariff 3.1 (Medium Voltage)... 32 Table 13. Variable Energy costs for tariff 3.1 self-regulation... 32 Table 14. Cost of charging stations... 33 Table 15. Cost of transformers depending on the power... 34 Table 16. Summary of solar panel costs... 38 Table 17. Differences in battery capacity, Case Study 1... 47 Table 18. Costs and profits, Case Study 1... 49 Table 19. Differences in battery capacity. Comparison Case Study 1 and Case Study 2... 55 Table 20. Differences in battery capacity, Case Study 2... 55 Table 21. Costs and profits, Case Study 2... 56 Table 22. Profits for Case Study 1 and Case Study 2... 57 Table 23. Options considered in Case Study 3... 64 Table 24. Differences in battery and solar capacity. Case Study 3... 64 Table 25. Access costs for tariff 3.1 A... 65 Table 26. Costs and profits, Case Study 3... 67 Table 27. Differences in battery and solar capacity. Case Study 4... 74 Table 28. Costs and profits, Case Study 4... 75 Table 29. Comparison of battery capacities of the different case studies... 78 Table 30. Comparison of profits of the different case studies... 79 Table 31. Average /kwh charged to EV users... 80 Table 32. Total costs and profits of the charging station (based on maximum /kwh)... 81 xvii

xviii

i j GLOSSARY Description Units GAMS SETS Half an hour of a period of one year (2016:366 days) Electric vehicle in a specific half hour. p Tariff period: Valley (3), Off-Peak (2) Peak period (1) p PARAMETERS P ij Energy demanded by electric vehicle number j in i. KWh pdem(i,j) g C i Cost of energy from the grid in hour i. /KWh pcener(i) sale C i Price of selling energy to the network /KWh pcsale(i) C cs Cost of charging station pcstation C ch Cost of one charger pccharger access C p Cost of access tariff for each period p /KW - year pcaccess C 1 solar Cost of small-size solar panel per square meter /m 2 pcsolar1 C 2 solar Cost of medium-sized solar panel per square meter /m 2 pcsolar2 C 0 initial fixed cost for medium solar panels. pcsolaricp C rep Cost of replacement of batteries - degradation /kwh pcrep F Fixed cost of the battery pf V Variable cost of the battery /kwh pv D Capacity degradation coefficient pd Δ Discharging efficiency pdelta λ Efficiency of charging from the network; directly to the EV or to the battery plambda ᴪ Transformer efficiency ppsi cos φ Power factor - transformer pcosfi B min Minimum power in battery KWh pbmin P max Maximum power that the battery can supply in an hour KWh ppmax cap b max Maximum size of the battery KWh pmaxbatcap VARIABLES b cap Storage capacity of the battery KWh vbcap g i Power injected to the electric vehicles directly from the grid KWh vg(i) ch i charging from grid to battery KWh vch(i) dh i discharging from battery to electric vehicles KWh vdh(i) g b i Energy sold to the grid from the battery KWh vbatred(i) contrac Contracted capacity. Maximum power for the whole p p period. KW vpcontr(p) b i Battery state at period i KWh vb(ie) C energy Total cost of energy vctenergia C access Total access cost vctacceso C bat Total cost of battery vctbat 1

C infr Total cost of station and chargers - infrastructure vctchar C trans Total cost of transformer vcttrans p trans Power of the transformer KW vptransf N Number of chargers in the station vn p ij Demand connected at moment i kwh vdem(i,j) C sol Cost of installing the solar panel vctsol s cap Size/capacity of solar panels Kw vsolsize s i ch Solar energy used for charging kwh vsolcharge(i) bat s i Solar energy for the battery KWh vsolbat(i) g s i Solar energy sold to the network kwh vsolred(i) BI i Binary variable which indicates if the battery is charging (1) or discharging (0) vbi(i) car j Indicates if the jth car has a station (= 1) or not (= 0) vcar(j) 2

1. INTRODUCTION AND OBJECTIVES After nearly a century with the internal combustion engine dominating the personal transportation sector, in the recent times the electric vehicle has experienced a rapid growth in both developed and developing vehicle markets. The broad-scale adoption of the electric vehicle aims to bring significant changes for society in terms of not only the technologies used for personal transportation, but also moving our economies away from petroleum and reducing the environmental footprint of transportation. Along this evolution, several business models have been developed in order to cover the necessity demanded by the customer. Among the different services that EVs could provide, balancing and ancillary services for the electricity sector have been put forward; new actors such as charging aggregators or advanced control systems would manage the charging scheduling of EVs taking into account the information regarding EVs owners mobility needs and their connection schedule to the grid. Part of the project is dedicated to explore current practices and potential improvements for business models for electric vehicles aggregation in the Spanish context. Those improvements would be mainly based on three different case studies taking into account the introduction of batteries in the charging network operator s business model, considering the usage of renewable sources, such as solar energy, and finally assessing the possibility of selling electricity back to the grid. In fact, the final purpose of the project will be to analyze and compare the profitability of these case studies, which several advantages will be explained in the following sections The project will use and develop computational models to quantify relevant proposals; assessing the profitability of different approaches of business models and analyzing possible sensibilities. 3

2. MOTIVATION The actual business model for the development of public charging infrastructure used by the several companies in the market consists of the placement of charging points in strategic location along the network. But this is not as straightforward as it may seem. In order to provide customers with the adequate service, it is necessary to make an estimation of the power that will be consumed at different times of the day in order to dimension the charging points. This requires to forecast information about the number of vehicles and the power they will need when to charge. Due to the uncertainties about possible variations along the day and the hours, it is difficult to have an exact forecast. Several factors could affect these previsions, for example peak hours during the day, specific days of the month due to holidays or a specific event, market prices, etc. The forecast of charging scheduling is one of the factors that affects the charging dimensioning of the charging network operator because it is necessary to determine the contracted power with the distribution company and pay accordingly. In addition, we should take into account the energy prices. We could find the possibility of making a wrong estimation about the power consumption with the consequent shortage of power to deliver to the customers. Or the opposite, having more power than we needed, and so, additional costs due to overinvestments. These two problems could be dealt with a possible investment on static batteries. What we propose is to investigate the possibility of adding of batteries into the charging network operator s infrastructure. With this measure could be possible to store the surplus of energy and deliver it the next day or when necessary. This way, we would be saving costs from a lower contracted power, and increasing the benefits of the charging network operator. In order to measure the profitability for charging network operator it is necessary to account for battery costs, including charging and discharging efficiency and battery degradation costs in order to make an appropriate estimation about the difference in costs and profits, and demonstrate if it would be profitable or not. Going one step further the possibility of introducing additional renewable sources will be also considered as well as selling electricity to the grid, encouraged by the motivation of contributing to the protection of the environment and the added advantage of increasing the profitability of the business. The overall motivation of the project is oriented to the promotion and facilitation of the usage and expansion of the electric vehicle industry: contributing to the reduction of the environmental impact and reducing the impact on the electric system. 5

3. TECHNOLOGY DESCRIPTION Previous to the development of the model it is important to start from a basic knowledge of the industry and current dynamics taking place. In the following subsections can be found information about the basic concepts that take place along the project. 3.1 The figure of the charging network operator A charging network operator is a trading corporation that being an electric power consumer, has the right of reselling electric energy only with the purpose of recharging electric vehicles. Some of the obligations to be fulfilled by the network operator to its customers are (BOLETIN OFICIAL DEL ESTADO 2011a): i. Report about the origin of the energy supplied. ii. iii. iv. Maintain its facilities in regulatory technical and safety conditions. Be linked to a control center that allows them to interact with the network to participate in the active demand management. Inform the National Energy Commission of the charging points and energy service provided in them. Royal Decree 647/2011, of May 9 (BOLETIN OFICIAL DEL ESTADO 2011a), regulates the activity of the charging network operator system for energy recharging services, and establishes that charging network operators are those corporations that develop the activity of supplying electric power for recharging electric vehicles. The main duties and rights of a charging network operator are: i. Act as market players in the market for electricity production. ii. iii. Access to transmission and distribution networks in the terms provided in the regulations. Invoice and collect the energy delivered in reselling services for electric vehicle charging energy. It is important to highlight that the figure of the charging network operator does not involves the same activities everywhere. In the case of Spain, these companies operate the charging infrastructure and provides the energy demanded to the customers and give them support in the service. This figure is not completely regulated and depending on the area, the business model differs. Table 3. Electric Mobility charging figures in Europe and Spain Most of European countries CPO (Charging Point Operator) + Service Provider Spain Charging network operator ( Gestor de carga ) A shown in Table 3, in Europe there are two figures that together constitute the charging service. In the case of Spain both activities are provided by a unique figure, the charging network operator. 7

3.2 Charging modalities Before talking about different technologies already developed, it is important to know what the different types of charging alternatives are. According to Endesa ("Endesa Vehículo Eléctrico" 2017) there are three main modes of charging: a) Conventional charging: The conventional single-phase load uses the electric current and voltage at the same level as the housing itself (16 amperes and 230 volts). This implies that the electrical power that the charger can deliver for this type of load is approximately 3.7 kw. With this power level, the process of charging the battery takes about 8 hours. This solution is optimal, mainly, to recharge the electric vehicle during the night in a garage of a detached house or community garage. b) Semi-fast charging: Semi-fast charging employs 32 amps of current and 230 VAC of electrical voltage. This implies that the electrical power that can deliver the point for this type of loads is approximately 7.3kW. With this power level, the process of charging the battery takes about 4 hours. This solution is optimal to recharge the electric vehicle during the night in a garage of a detached house or community garage. c) Fast charging: The rapid recharge means that in 30 minutes 80% of the battery can be charged. This solution is what, from the point of view of the customer, resembles its current refueling habits with a combustion vehicle. Fast charging uses a higher electric current and, in addition, delivers the energy in direct current, obtaining an output power of the order of 50kW. Electrical requirements are higher than conventional recharge. This may imply the need to adapt the existing electricity grid. In SAE terminology, 240 volt AC charging is known as Level 2 charging, and 500 volt DC high-current charging is known as DC Fast Charge. Owners can install a level 2 charging station at home, while businesses and local government provide level 2 and DC Fast Charge public charging stations that supply electricity for a fee or free. The International Electrotechnical Commission modes ("Charging Station" 2017) are classified as follows: i. Mode 1: Slow charging from a regular electrical socket (single- or threephase). ii. iii. iv. Mode 2: Slow charging from a regular socket but with some EV specific protection arrangement. Mode 3: Slow or fast charging using a specific EV multi-pin socket with control and protection functions. Mode 4: Fast charging using some special charger technology such as CHAdeMO. 8

When talking about charging station there are several alternatives and a lot has been said when deciding which one would be the best solution (Nieto González 2009). One of the main problems of electric vehicles is the time of charging of batteries. This time is much higher than the one needed to refill the fuel capacity of a conventional one. On the one side, this involves a necessity of changing consumer s mentality regarding the time spent in charging in order to have the maximum autonomy. Another charging modality is the night charging. This way consumers don t need to waste any time in charging. The problem is that this modality normally is only possible in residential parking, so not everyone has access to them. The last charging mode consists in the replacement of the used battery for a new one in the same charging station. The problem of that is the big investment that would suppose to the charging network operator. And also the necessity of having standard batteries for all vehicle models. From these options it can be observed that none of the alternatives could perfectly fit to the current demand. Nevertheless, a lot of advances and different alternatives have been developed. Some of them have been gathered in the next section in order to have some knowledge previous to develop the model of the project. 9

NUMBER OF EV SOLD 210.817 317.895 258.985 548.210 774.384 4. PRESENT OF THE ELECTRIC MOBILITY INDUSTRY This chapter collects information about the electric mobility industry worldwide and in the specific case of the Spanish country. Previous to the development of the goal of the project it is important to be aware of the evolution of the sector in order to depart from the basic concepts and understand the essence of the motivation leading this work. 4.1 Global EVs market The electric vehicle market is a reality that year by year is acquiring more importance, as the sales of electric vehicles worldwide follows an exponential tendency, as shown in Figure 1. Although the data of year 2017 is not complete, it is expected to keep this growing path, as previous years. Figure 1. Electric Vehicles sales worldwide 2013 2014 2015 2016 2017- Q1 YEAR Note Source: Pontes, Jose. 2017. "EV Sales". Going deeper into the evolution of electric vehicle per country, depending on the country, the penetration of the market differs considerably. As shown in Figure 2, the country with the higher amount of electric vehicles is China, followed closely by the United States. Europe is also at a good position regarding EV sales, with Norway at the leading position. Moreover, Japan holds a good position too in the market. It can be observed that in the case of Spain the market is not already developed at all as the number of EV is still too low. 11

34.464 45.492 28.195 36.907 27.694 33.704 23.465 25.154 35.489 24.645 20.154 21.000 8.981 13.454 6.784 10.067 3.015 4.603 114.230 157.181 262.583 333.481 Figure 2. Number of EV per country Spain Canada Sweden 14.314 27.392 30.513 Japan Netherlands Germany France UK Norway 147.000 113.636 74.754 108.065 91.627 135.276 EEUU 570.187 China 645.708-100.000 200.000 300.000 400.000 500.000 600.000 700.000 Note Source: Pontes, Jose. 2017. "EV Sales". In addition to the amount of total EV per country is important to take into account what is the proportion of this number in comparison with the total amount of vehicles matriculated. Regarding this point, the country with a higher market share of electric vehicles is Norway with a 29.1%. In the following positions can be found Netherlands and Sweden with a 6.4% and 3.5% of market share respectively. In the case of Spain this percentage is of the 0.33% approximately. It is evident that it is still too low as to consider it a profitable industry (Fernandez 2017) In Figure 3 is shown the evolution between the years 2015 and 2016 of the sales of electric vehicles per country. Even though most of them are not very high, they all represents a positive evolution in respect to the previous year, except for the case of The Netherlands, that last year 2016 experienced a negative evolution. Figure 3. EV sales per country China EEUU Norway UK France Germany Netherla nds Note Source: Pontes, Jose. 2017. "EV Sales". Japan Sweden Canada Spain 2015 262.583 114.230 34.464 28.195 27.694 23.465 35.489 20.154 8.981 6.784 3.015 2016 333.481 157.181 45.492 36.907 33.704 25.154 24.645 21.000 13.454 10.067 4.603 12

Electric mobility does not only depend on the sales of the vehicles. In order to develop a solid industry, the respective infrastructure to sustain it is also essential. In Table 4 have been gathered information about the number of charging points in some of these countries. In concordance with the extent of the market, China, United Stated and Japan are the countries with the higher development of infrastructure, followed by the European city of Norway. Table 4. Charging points per country Country Charging points China 150.000 United States 33.000 Japan 40.000 Norway 5.000 France 2.500 Spain 2.338 Note Source: Aguilar, Carlos Arturo. 2016. "Países Con Mejor Escenario Para Autos Eléctricos". Motorbit; "Puntos De Recarga En Madrid (España)". 2017. Electromaps.Com. 13

EV Sales 4.2 Actual situation in Spain Once a general overview of the world-wide situation of electric mobility industry, a closer approach is assessed in this section, based on the most relevant drivers and players of the industry applied to the specific situation of Spain. 4.2.1 Electric Vehicles sales evolution in Spain Figure 4 shows the evolution of the electric vehicles sales from January 2017 until May 2017. This number represents the total amount for 100% vehicles gathering passenger cars, industrial vehicles and quads. 500 450 400 350 300 250 200 150 100 50 0 Figure 4. EV sales in Spain per month Note Source: Noya, Carlos. 2017. "Ventas De Coches Eléctricos En España - Forococheselectricos". Forococheselectricos; The remarkable variances between one month and next time may be because of economic incentives provided by the government in specific times of the years, or due to campaigns promoted by the car manufacturers themselves. Having a look at the overall evolution per year in Figure 5 it is observed that this trend follows an exponential evolution, similar to the one of the worldwide sales. 14

1664 1.274 EV SALES 1.958 1.800 3.015 4.603 5000 4500 4000 3500 3000 2500 2000 1500 1000 500 0 Figure 5. EV sales in Spain per year 2012 2013 2014 2015 2016 2017- Q1 YEAR Note Source: Pontes, Jose. 2017. "EV Sales". The market share of electric vehicles in Spain based on the first month of 2017 is of the 0.33%, taking into account a number of 538.710 total amount of matriculated vehicles during this period ("Aniacam.Com" 2017). This percentage is still too low. However, studies are optimistic about the evolution of this industry. In order to success, infrastructure is an essential factor to be developed, as will be explained on the following sections. And this is the reason why the final goal of the project is to focus on the concerning with charging infrastructure. 4.2.2 Charging infrastructure in Spain The lack of charging stations could justify the minimum implementation of the electric vehicle in Spain. A correlation between the number of supply points and of electric cars is evident. According to electromaps ("Puntos De Recarga En Madrid (España)" 2017) there are 2338 charging points throughout the country distributed more or less as shown in Table 5. This gives an approximate idea of the level of infrastructure in the country and how it is distributed. The Community of Madrid and Cataluña have been registered the highest number of electric cars in Spain. Both regions are those that have more charging points: 446 in Barcelona and about 293 in Madrid. Table 5. Number of charging points per province in Spain City Charging points Barcelona 446 Madrid 292 Baleares 180 Tenerife 69 Valencia 69 Sevilla 56 Málaga 53 15

Tarragona 51 Las Palmas 50 Note Source: "Puntos De Recarga En Madrid (España)". 2017. Electromaps.Com. All of these charging points are operated by different charging network operators, that are assessed into more detail in the following subsection. 4.2.3 Charging network operators in Spain The main charging network operators that have a wider operating range according to CNMC are listed in Table 6, with the correspondent number of charging points displayed along the country that have been possible to collect. To see it in more detail, the different charging network operators and its location in Spain, they are listed in a document by CNMC (CNMC 2017). Table 6. Charging network operators in Spain and number of charging points Charging Network Operator Charging points in Spain Fast charging points in Spain IBIL 85 26 GIC 12 Endesa 6 6 Gas Natural Fenosa 27 Fenie Energía 100 Iberdrola edp VIESGO Note Source: "IBIL". 2017. Ibil.Es.; "GIC - Gestión Inteligente De Carga". 2017.; "Endesa Vehículo Eléctrico". 2017.; "Gas Natural Fenosa". 2017. Gas Natural Fenosa.; "Fenie Energía". 2017. Hereafter are described some of the previously network operators listed in order to go deeper into its business model and operating processes, as well as trying to obtain information about its charging mode and corresponding tariffs. IBIL GESTOR DE CARGA DE VEHÍCULO ELÉCTRICO, S.A IBIL is the result of the collaboration agreement between the Basque Energy Board (EVE) and REPSOL, in October 2010. IBIL is the first electric vehicle charging network operator enrolled in the National Energy Commission (CNMC). In addition, it provides an integral service based on 100% renewable energy, facilities, intelligent terminals, and a control infrastructure centre ("IBIL Gestor De Carga De Vehículo Eléctrico" 2017). The services and products provided by IBIL can be divided in: - Electric vehicles in private areas: in garages and in particular companies. - Electric vehicles in public areas: in public car parks, shopping centres, roads, gas stations, etc. 16

Most of the charging points are located in Madrid and País Vasco, see Figure 6. Up to June 2017, IBIL has 59 normal charging points and 26 fast recharge points. Moreover, 2 more normal recharge points are in process of installation in A Coruña and Malaga, and a fast one in Tarragona. Figure 6. IBIL normal charging points location in Spain Note Source:"IBIL". 2017. Ibil.Es. Figure 7. IBIL fast charging points location in Spain Note Source: "IBIL". 2017. Ibil.Es. IBIL users have the advantage of using the IBIL car for the procedure of the charging process. The price for charging is established with a fix cost per kwh, being this one for fast charging: 0.54 /kwh. Moreover, there is a minimum charge of 5, what would be equivalent to 9kWh of energy, sufficient for a 60 kilometer trip more or less (motor.es 2017). Up until today the IBIL recharge card can only be used in IBIL's public points network. Among the charging network operator are working to launch an interoperability or roaming service soon so that all customers can recharge at all recharging points independently of the operator to which they belong. 17

GIC ( Gestión de Carga Inteligente ) GIC is a company forming part of ACS group. Together with the network operator IBIL and promoted by Madrid city council, they have signed an agreement to renew the charging network of the city. From this initiative, 24 new charging points were installed in the city. This points are interoperable, so that any user can make use of all points, regardless of the charging network operator that operates them, IBIL or GIC. In Figure 8 is shown the location of the 12 charging points, all of them located in the city of Madrid. Figure 8. GIC charging points location in Spain Note Source: "GIC - Gestión Inteligente De Carga". 2017. In relation with the costs of the service, when using GIC services, there are two consecutive costs ("GIC" 2017): - Cost for the reservation of the charging point: 1 - Cost for charging: 0.45 /kwh ENDESA The Enel Group s Spanish subsidiary Endesa in one of the principal companies in the European energy sector, committed with the environment and electric mobility. Regarding electric mobility, the products and services offered by the company are the following ("Endesa Vehículo Eléctrico" 2017): Table 7. Endesa electric mobility services Charging Infrastructure Charging point adapting to specific needs. - Private housing - Neighborhood community - Business infrastructure - Charging point installation Electric supply and tariffs - Electrical supply contract 18

- Counselling Note Source: "Endesa Vehículo Eléctrico". 2017. Endesa promoted a project started in Mallorca, with the installation of 6 quick recharge points. These recharge points have been co-financed with ERDF European Funds. This new charger incorporates the three types of connectors that currently exist on the market, allowing to charge all types of electric vehicles, regardless of the manufacturer. In Figure 9 is shown the location of this six charging points Figure 9. Endesa charging points location in Spain- Mallorca Note Source: "Endesa Vehículo Eléctrico". 2017. Initially the registration in the club has a cost of 39 that will be paid by the customer on his first monthly bill of service by direct debit. Subsequently each recharge will have a cost of 6 which will be charged at the end of the month through domiciliation. GAS NATURAL SERVICIOS SDG, S.A. Within its overall mobility strategy, Natural Gas Fenosa bet for the promotion of electric mobility. Gas Natural Fenosa is one of the main private partners of MOVELE and has been directly involved in the installation of charging infrastructure for electric vehicles in Madrid city. As reflected in Figure 10 until the moment there are 27 charging point distributed along the country installed by Gas Natural Fenosa, most of them concentrated in Madrid, and in a lower amount in Barcelona. 19

Figure 10. Gas Natural Fenosa charging points location in Spain Note Source: "Gas Natural Fenosa". 2017. Gas Natural Fenosa. IBERDROLA SERVICIOS ENERGÉTICOS, S.A.U. Iberdrola launched in 2010 a business area intended for the development and commercialization of charging solutions for electric vehicles. By the end of 2015, Iberdrola has installed 224 charging points, of which 19 have been for the refilling of electric vehicles in the BMW headquarters, car dealerships and other corporate centres manufacturer. Iberdrola provides a service that is in charge of everything related with the electric installation of recharging points in houses and community garages. The standard tariff they offer to charge an electric vehicle is from 32 / month with no setup fee. This monthly fee includes ("Productos Y Servicios - Iberdrola" 2017): - Car charging facility with maximum performance. - Access to Aplicacion Recarga Verde so that wherever the customer is, from a computer or from a smartphone, he/she can schedule the load of his/her car, see the charging points available, get information about charges, etc. - Technical Assistance Service for charging point, with 24-hour service and onsite assistance from an authorized installer when necessary. FENIE ENERGÍA Fenie Energía is another charging network operator with an important deployment of charging point in Spain. Through the web page or the mobile application users can make use of the advantages provided by this services. As shown in Figure 11, there are 100 charging point distributed as reflected in the figure, most of them concentrated on the provinces of Valencia and Alicante: 20

Figure 11. Fenie Energía charging points location in Spain Note Source: "Fenie Energía". 2017. Recarga.Fenieenergia.Es. 4.2.4 Public economic incentives in Spain The Ministry of Industry tries to cover this gap with incentive measures, such as forcing dealers to install a charging point worth 1,000 euros in the area where users of electric cars live. However, several more barriers could be found, for example the reduced autonomy of lithium batteries and its high cost. The high cost has been a factor that has penalized electric cars, especially in a context of financial crisis. The difficulty of companies to access credit discouraged them of facing a "strong initial overrun" to supply their car fleets with these vehicles, even though in four or five years the payment could be compensated by fuel savings. Since 2011 there have been grants in Spain for the purchase of electric vehicles of the government. In certain regions there are also some additional assistance, consistent with the State s one. And for certain sectors, such as taxis, there are also some specific subsidies. The intention is to keep developing incentives to boost this technology, both from the industrial and demand promotion points of view. Down below are listed some of this incentives measures implemented in Spain for the lasts years. PLANES ADAPTACION MOVEA 2017 This past June, 2017 the Council of Ministers approved a Royal Decree that regulates the granting of aid for the purchase of vehicles using alternative energies and for the implantation of electric charging points. This initiative is framed within the Strategy of Impulse of the Vehicle with Alternative Energies driven in 2014 and which is valid until 2020 (BOLETIN OFICIAL DEL ESTADO, 2017). MOVEA 2017 maintains without major changes the lines of the past program. The total amount is of 14.26 million euros available to be distributed between different technologies. 21

According to the Minister of Industry, Alvaro Nadal, this amount will be sufficient to cover the demand until the month of October (Noya, 2017). Four months in which with these funds are expected to cover the registration of 2,200 units, taking into account that there is a maximum of 5,500 euros of aid for electric cars. Individuals, self-employed persons, private companies, local entities, autonomous communities and public entities linked or dependent on the General State Administration may be beneficiaries of this aids. In the case of charging points, all of the above may be beneficiaries, except for individuals and individuals. The amount of the aid is fixed with different scales depending on the type of vehicle and the combustible used, as reflected in Table 8: Type of Vehicle Table 8. MOVEA 2017 plan aids distribution Maximum quantity ( ) Minimum quantity ( ) Electric cars, plug-in hybrids or extended range 1.100 12.000 Electric motorcycles 1.000 2.000 Vehicles of fuel cell (hydrogen) 5.500 Electric quads 1.959 2.350 Conventional charging points 1.000 Semi-Fast charging points 2.000 Fast charging points 15.000 LPG vehicles 500 15.000 Natural gas vehicles 1.000 18.000 Note Source: BOLETÍN OFICIAL DEL ESTADO. 2017. Núm. 149, Sec. I. Pág. 51631. The aid for the implementation of charging infrastructure for electric vehicles in public areas can reach an amount of up to 40% of the cost, with a maximum of 1,000 euros per conventional charging point, 2,000 euros per Semi-fast charging point, and 15,000 euros per fast charging point. According to government estimations, the aid will encourage the purchase of 1,800 electric passenger cars and vans and 230 electric motorcycles. The MOVEA 2017 Plan and will be in force until October 15 of this year or until the available funds are exhausted2020 (BOLETIN OFICIAL DEL ESTADO, 2017). PLAN PIVE 8 The "Incentive Program Efficient Vehicle (PIVE-8)" consisted of promoting a reduction of national energy consumption by incentives for the modernization of conventional vehicle (M1) and commercial (N1) models with high energy efficiency, less fuel consumption and CO2 emissions. All of this under the Plan de Ahorro y Eficiencia Energética 2011-2020 ("Programa De Incentivo Al Vehículo Eficiente (PIVE-8) - Energía - Mº De Energía, Turismo Y Agenda Digital" 2017). This measure took place until the 31st December 2015. 22

More specifically, PIVE 8 Plan aimed to promote with incentives the withdraws of 300.000 vehicles with more than 10 years-life in the case of M1 vehicles, and 7 years-life in the case of N1 ones. This would suppose savings of 103 million liters of fuel, with reduced emissions of 114.300 tons of CO2 per year (IDEA 2017). According to The Royal Dree 380/2015 of May 1 (IDEA 2017), this Plan had allocated a budget of 225 million euros to apply to the requests made under the PIVE 8 Plan. PLAN PIMA AIR 3 This plan was initiated the 6th May of 2014. It is the new convening launched by the government, whose beneficiaries are both physical persons and legal persons of private nature and communities of goods seeking to renew a van or commercial vehicle. Also to those who wish to purchase a hybrid or electrical motorcycle or moped. Electric bicycles are also included ("Plan Pima - El Blog De La Sostenibilidad" 2017). The budget amount of the plan was of 5.5 million euros. This quotation was divided into two parts depending on the type of vehicle, allocating a budget of 5 million for commercial vehicles or vans, and 500,000 euros to encourage the purchase of two-wheeled vehicles. 23

4.2.5 Similar projects As mentioned before, in this section several projects have been analyzed as a way of stablishing what are the bases of the current innovative projects taking place. In all these innovative models a common factor is the inclusion of renewable energies oriented to optimize the charging process. SIRVE Project In year 2015, four Aragonese companies (Urbener, Zoilo Ríos SA, Pronimetal and the CIRCE Foundation) allied to create the first charging station in Spain, located in Zaragoza. The project aims to build a charging station for electric vehicles that will integrate renewable electricity generation and energy storage (Durán, 2015). The development of this station is part of the SIRVE (Integrated System for the Recharge of Electric Vehicles) project, which was supported by the Ministry of Economy and Competitiveness, in a commitment to extend the use of the electric vehicle and reduce its impact on the electrical system. The most important and distinguishing feature from other similar projects is its ability to produce renewable electricity and store energy (Movilidad Electrica, 2015). In fact, one of the most revolutionary points is the existence of an energy store inside. The system consists of lithium batteries which allow compensating active and reactive power. This way, a stabilizing effect is achieved in the electrical system, without causing serious imbalances due to the consumption of the car. It also has different charging modes; from the standard one to the fast one. Moreover, it also distinguishes by its quick installation. In only 24 hours can be assembled and its modular design allows the incorporation of consecutive units. After completing the entire research, study and development phase, two pilot installations have been built and have been tested in a real environment for later commercialization ("SIRVE Urbener" 2017). The first SIRVE I was installed in 2014. Its functional elements have been dimensioned for a single parking space. In this installation, the fast mode (50 kw DC) and moderate mode (22 kw AC) take place, oriented to the installation of independent charging points in the public infrastructure. In addition, the slow mode (3.7 kw AC) is intended for individual users usage. The second SIRVE II was installed in 2015. This second installation is also multipoint, with a fast mode (50 KW DC), a moderate (22 KW AC) and three slow mode points (3.7 KW AC). PROYECTO LEVANTE A Valencian project for the assembly of a charging station for the recharging of electric vehicles was among the three finalists to the prize "ZayedFutureEnergy" (Levante 2016). 24

The initiative was developed by Tono Benet, professor of Schools San Jose in Valencia, and has had the participation of teachers and students of higher degree. The project consists of the design and installation of an energy generation plant through the use of renewable photovoltaic and wind technologies that will allow the installation of the charging station. It will provide recharging services for both electric vehicles and bicycles, and will generate the necessary energy to provide the building of Formative Cycles of external lighting and luminosity by using LED technology. In addition, it will allow the loading of mobile phones, tablets and laptops in classrooms throughout the school (Levante 2016). CHARGING STATION IN LA GRANJA, SEGOVIA The 26th of June was inaugurated the charging station of La Granja, province of Segovia. This is supposed to be the first one of a network of free fastcharging points distributed along Spain. Electrolineras Sostenibles is the company in charge of this project (Movilidad Electrica, 2017). Figure 12 shows an image of the charging station. Figure 12. Charging station in La Granja, Segovia Note Source: Movilidad Electrica, 2017. "Inaugurada La Electrolinera Solar De La Granja". The station is totally sustainable. Some of its energy comes from photovoltaic panels that are captured and subsequently introduced to the grid. Another part is supplied by the electricity company through a contract that guarantees that it is 100% coming from renewable energy. All the energy injected into electric cars is of renewable origin. The charging station supports photovoltaic panels capable of generating 50 kw and two super-fast chargers. The stations will have three-phase charging points for those who do not need fast-charging, to avoid the collapse of the station. The charging tariff for electric vehicle users that have not acquired its 25

car through Electrolineras Sostenible will be of 4 approximately (Movilidad Electrica, 2017). The charging corridor is expected to be composed of charging stations on the following provinces: Vizcaya, Burgos, Segovia, Madrid, Toledo, Ciudad Real, Jaén, Córdoba, Sevilla and Cadiz. The stations are still in process of development. As have been covered in the previous examples, each day appear more and more alternatives regarding charging infrastructure, and many opportunities can be found. Some of the features that appear in these projects will be used as a guidance when designing the model of the project. 26

5. METHODOLOGY In this project the final goal is to reach the optimal solution for a charging station in terms of profit, based on specific case studies where the investments of the station differ on each of the cases. The objective function in our case is the supposed profit a charging station obtains based on a series of assumptions and conditions. The assumptions are introduced in the form of parameters, and the conditions are stablished as restrictions. In addition to the objective function that is optimized, there are also a set of variables that define some of the performance features of the process of charging, and therefore determine the optimal conditions for the achievement of a successful business model of the charging network operator. For the development of this model, it is used the General Algebraic Modelling System (GAMS). This choice has been made in order to work with a program that allows to reach an optimal solution based in an objective function and taking into account several restrictions. In the following subsections will be explained the assumptions taken into account, the factors involved or parameters, and the algorithm itself, including equations and restrictions. 5.1 Assumptions Prior to start designing the model, it is essential to have clear the basis by which the model is built. As mentioned before, it is developed a model based on a charging station which mission is to provide with energy refilling to the electric vehicles arriving. First of all, this charging station is supposed to be located in a motorway. One of the reasons is that it is wanted to cover a situation in which the main problem is providing fast charging. If we were talking about a charging station in the middle of the city; in a residential community, or in a parking, it will not be essential to provide fast-charging. However, in the case of trip destined to cover a considerable amount of kilometers, the fact of having a fast-charging station becomes essential for electric vehicle users. This is why the project focus on this concrete situation. For this project a maximum number of chargers in the charging station of 10 units has been established. This quantity is expected to be enough to cover the demand and a reasonable measure in order to obtain sensible results. Regarding the features constituting the charging station performance, they are explained in detail throughout the following sections. 27

5.2 Factors involved To understand the factors involved, it is important to understand what are the basic elements or costs a functional charging station have, because that is the first step in the development of the model. On the one hand, there is the cost of the main infrastructure elements that are common in any kind of charging station. The first one is the charging station and each of the chargers, which do not change with or without battery. Then it comes the price of the transformer, which varies depending on the amount of power needed to obtain from the electric grid. Moreover, in this project it is introduced the cost of the battery that, as will be explained later, is composed of a fixed cost, a variable cost, and an additional cost because of battery degradation. In another case study it is considered the introduction of solar panels as an additional way of obtaining energy, so the corresponding costs are also considered. On the other hand, there are also some costs that vary depending on the amount of power and energy needed, so it is different depending on the circumstances applied. First of all, there is the regulated capacity cost, that depends on the amount of power contracted and is paid annually. In addition, the energy price, that follows a specific profile depending on the day and the hour of supply. Each of these elements are explained more deeply down below, as well as shown the origin from where the corresponding data is obtained. 5.2.1 Energy demanded by electric vehicles The evolution of batteries capacity in electric vehicles is in a constant progress, experimenting an evolution each year. According to a recent article (Ibáñez 2016), a 2016 electric vehicle has 12 times more the autonomy they used to have at the end of nineteenth century, when the early stages of electric mobility took place, and it is 8 times faster. This indicates that in a few years these data would have probably evolves, but for now the focus is on the actual situation of the industry. From an article published in AutoBild (de Haro 2016), some data has been gathered regarding in Table 9, with the most common models of the markets and their correspondent battery power and km of autonomy. These magnitudes of values are the ones that are used in order to represent the power needed by each of the electric vehicles arriving to the charging station Table 9. Capacity and autonomy of EV models Model KW Battery Km autonomy Tesla Model S P100D 100 613 Tesla Model X P100D 100 542 Renault ZOE 41 400 BMW i3 33 190 22 300 Hyundai Ioniq Electric 28 280 Nissan Leaf 30 250 Mercedes-Benz B 250e 28 200 28

Volkswagen e-golf 40 190 Renault Kangoo Z.E. 44 170 Note Source: de Haro, Ignacio. 2016. "Los Diez Coches Eléctricos Con Más Autonomía Del Mercado". Autobild.Es. To perform the most possible realistic model it should be taken into account that an electric vehicle battery is never used at its total capacity. There is always a security margin in order to buffer the charging and discharging. This way, the battery works with cycles between 5% and 95% of its total capacity, extending this way its service life. When the battery has reached almost the 100% of its capacity the recharge power starts decreasing to avoid the overheating of the battery cells. Normally, this decrease takes place when the battery is at the 80% or 90%, getting to the 100% much slower. That is why most manufacturers states that the vehicle can be charged from 0% to 80% in half an hour or less. (Fernandez, 2017) On the basis of all the previous statements, a range of possible demand values in KWh that could take place in the charging station are set. In this case have been stablished values in between 10 KW and 30 KW. The minimum is 10 KW because it is supposed that the vehicles won t arrive completely discharged, but on the contrary, always with some amount of energy in the vehicle battery. Furthermore, when setting the input parameters of demand, it is necessary to take into account that depending on the hour of the day the demand is probably different. For this reason, on the night hours, the number of vehicles arriving are established to be less than the normal ones. Figure 13 shows the process followed to determine the demand. Figure 13. Planning of vehicles arriving at the station and energy demanded Number of vehicles arriving at the station h < 10 & h > 42 Vehicles arriving ϵ [1,4] 10 < h < 42 Vehicles arriving ϵ [5,10] Energy demanded [10 kw, 25 kw] 5.2.2 Cost of contracted capacity The cost of contracted capacity depends on the maximum power withdrawn or injected to the grid in at different periods. In the Spanish case, the corresponding tariff to be applied in the model is Tariff 3.1 A (BOLETÍN OFICIAL DEL ESTADO 2014). The 3.1 A is the network access tariff determined by law for medium voltage (from 1 to 36 KV) with power 29

contracted in all periods equal to or less than 450 kw. This tariff invoice in three periods and a different amount of power can be contracted in each of them. Each period corresponds to a daily time zone where the price of energy and power is different. Those periods are the following: - P1 (Peak period): It is the period in which the power and the energy are more expensive. It is 4 hours a day. - P2 (Off-Peak Period): It is 12 hours a day and the holidays are called P5. - P3 (Valley Period): It is the period in which power and energy are cheaper. It is 8 hours a day. The schedules of the periods also depend on if it is winter or summer, and if the contract is peninsular or insular. All these factors are taken into account when assigning costs. When defining the energy costs by hour, these periods also have an impact that will be mentioned in the next section. The schedule of this tariff has been obtained from the BOE report (BOLETÍN OFICIAL DEL ESTADO 2014a) for the Iberian Peninsula, and is shown in the following table: Table 10. Tariff 3.0 Periods Winter Summer P1: Peak period 18 22 h 11 15 h 8 18 h 8 11 h P2: Off-Peak Period 22 24 h 15 24 h P3: Valley Period 0 8 h 0 8 h Note Source: BOLETÍN OFICIAL DEL ESTADO. 2014A. Núm. 175, Sec. I. Pág. 57163. Figure 14 shows the distribution of periods is shown in a more visual way. Figure 14. Tariff 3.1 periods Note Source: ESIPE (Entidad supervisora independiente de Proyectos Energéticos). 2017. Modalidad De 3 Períodos (Punta, Llano, Valle) - TARIFA 3.0A. TARIFAS DE ACCESO A REDES ELÉCTRICAS - PERÍODOS TARIFARIOS. 30

The costs corresponding to each of these periods has been obtained from BOE of 2014 (BOLETÍN OFICIAL DEL ESTADO 2014b) and it is reflected in Table 11. With the added feature of obtaining energy from photovoltaic panels for the purpose of selling it to the grid, it could be considered as a self-generation installation. Due to this, access tariffs corresponding to self-consumption tariff will be taken into account as an additional case study. These access costs appear as well in the Table 11. Table 11. Regulated Capacity costs for tariff 3.1 A Regulated Capacity costs ( /kw - year) Period 1 Period 2 Period 3 3.1 59,17 36,49 8,36 3.1 Selfgeneration 36,37 7,25 5,04 Note Source: BOLETÍN OFICIAL DEL ESTADO. 2014B. Núm. 28, Sec. I. Pág. 7147. 5.2.3 Energy costs In addition to regulated capacity costs, that accounts for the cost of total contracted power, another term accounting for energy cost per hour is taken into account. The parameters of the energy cost from the energy withdrawn from the network of each hour are obtained from ESIOS web page ("Mercados Y Precios ESIOS Electricidad Datos Transparencia" 2017). For the study model, in order to simulate the situation of one whole year, the prices of 2016 are used as a reference. When analyzing different sensibilities for each of the case studies, the possibilities of performing with two different kind of tariffs are considered. In first place as the project is based on charging station for electric vehicles, tariff 2.1 is taken into account, specially designed for the activity of charging network operator for electric vehicles. Additionally, Medium Voltage tariffs (tariff 3.1) are also considered. This comparison is interesting as it shows how changes in regulations could affect the profitability of the business. Figure 15 reflects the profile of both tariffs (2.1 and 3.1) and the Day-Ahead prices. It can be observed that the dynamics of the evolution of each one depending on the hour is considerably different, as for example in the case of tariff 2.1, the changes between time periods are more drastic than in the case of tariff 3.1. This is because of regulated costs taken place in those time periods. This could lead to a different performance of each of the features taking place in the model of the project. Medium voltage tariff, as reflected in the Figure 15, is the result of adding the energy price in the Day-Ahead market plus the energy component of the access tariff, shown in Table 12 as Variable energy costs. 31

1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106 113 120 127 134 141 148 155 162 169 Energy price ( /kwh) Table 12. Variable Energy costs for tariff 3.1 (Medium Voltage) Variable Energy cost ( /kwh - year) Period 1 Period 2 Period 3 0,014335 0,012754 0,007805 Note Source: BOLETÍN OFICIAL DEL ESTADO. 2014B. Núm. 28, Sec. I. Pág. 7147. 0,16 0,14 0,12 0,1 0,08 0,06 0,04 0,02 0 Figure 15. Tariff 3.1 and 2.1 Energy price profile Time (h/2) 2.1 Tariff 3.1 Tariff Day-Ahead tariff Note Source: "Mercados Y Precios ESIOS Electricidad Datos Transparencia". 2017. Red Eléctrica De España, ESIOS. 5.2.4 Regulated costs in the case of self-generation In the case of considering the charging station as a self-generation installation, another factor should be taken into account: regulated costs. Apart from the cost of contracted capacity, which is fixed for each tariff period depending on the amount of power contracted and the energy costs varying for each hour, the energy cost accounts for regulated charges fixed for each tariff period and that should be added to the energy cost at every hour These charges are used to recover the regulated systems costs: costs of renewables subsidies, institutions, island systems, etc. The periods are the same as explained previously, and the costs to be added appears on Table 13: Table 13. Variable Energy costs for tariff 3.1 self-regulation Variable Energy cost ( /kwh - year) Period 1 Period 2 Period 3 0,016699 0,011411 0,013268 Note Source: BOLETÍN OFICIAL DEL ESTADO. 2015D. Núm. 302, Sec. I. Pág. 117270. Figure 16 reflects the comparison of Medium Voltage tariff (3.1 tariff) and self-consumption or self-generation tariff. It can be observed how the profile maintains a similar direction but there is an offset value added to the selfgeneration one. It is important to take into account that depending on the tariff period, this offset value is higher or lower. 32

1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106 113 120 127 134 141 148 155 162 169 Energy price ( /kwh) Figure 16. Medium voltage tariff (3.1) and Self-Generation Energy price profile 0,1 0,09 0,08 0,07 0,06 0,05 0,04 0,03 0,02 0,01 0 Time (h/2) Self-generation tariff 3.1 Tariff Note Source: "Mercados Y Precios ESIOS Electricidad Datos Transparencia". 2017. Red Eléctrica De España, ESIOS; BOLETÍN OFICIAL DEL ESTADO. 2014B. Núm. 28, Sec. I. Pág. 7147 These regulated costs are considered in the case study of introducing solar panels into the charging installation. 5.2.5 Charging station Charge Station Hardware As it has been explained in the assumptions section, a charging station is modeled based on a fast charging structure, or mode 4. Previous to include in the model the cost of the charging stations taking part in the infrastructure, there are several possible options. In the Level 2 category, which correspond to slow charging, can be found Home chargers, Parking garage chargers or Curb-side chargers. Apart from these, there are also DC Fast Chargers. For each one of these types, the cost can be divided in the following parts: Charging station hardware; Other hardware and material; Electrician and other labor; Mobilization; Permit. Table 14 (Agenbroad and Holland 2014) shows the breakdown of these costs depending on the category: Level 2 Home Table 14. Cost of charging stations Level 2 Parking Garage Level 2 Curb-side DC Fast Charging 415-920 1,380-2,290 1,380-2,750 11,000-32,000 Electrician Materials 45-140 190-470 140-275 275-550 Electrician labor 90-320 1,140-2,700 730-1,380 1,470-2,750 Other materials 45-90 45-140 90-370 Other labor 230-690 2,300-6,880 4,590-13,760 Transformer N/A N/A N/A 9,170-22,930 33

34 Mobilization 45-180 230-460 230-460 550-1,100 Permitting 0-90 45-180 45-180 45-180 Note Source: Agenbroad, Josh, and Ben Holland. 2014. "RMI: What's The True Cost of EV Charging Stations?". Greenbiz. In the case of this project, DC Fast Charging are considered, so the costs involved are the highlighted ones. The transformer costs are considered in a separate section, as they are considered as an independent cost in the study model. The total amount of the charging station is the one obtained in equation 1: Intermediate values have been taken into account in order to obtain an average cost. C cs = 20.000 + 400 + 2.000 + 200 + 7.000 + 1.000 + 100 = 30.700$ = 28.538 (1) C cs = 28.538 (2) The cost of each charger is estimated to be around 800, to which should be added the cost of the installation, which can be specific on the location. An average installation cost is assumed to be around 200/300. The cost of each charger to install is established in 1.200. 5.2.6 Transformer C ch = 1.200 (3) To calculate the transformer cost, a linear function is established by which it depends on the maximum power taken or delivered to the grid, which at the same time determines the contracted power. According to the BOE (BOLETIN OFICIAL DEL ESTADO, 2017), in load mode 4 or fast charging, the supply voltage, that refers to the input voltage of the alternating-to-continuous converter, can be up to 1000 V in three-phase alternating current and 1500 V in direct current. Within this voltage (12 kv U 1 kv), several power values are possible. In order to obtain the costs of the transformer size from the power contracted estimated, an interpolation function has been developed. Table 15. Cost of transformers depending on the power S (KVA) COSTS ( ) Investment O&M 15 39.029 893 25 39.029 893 50 41.100 941 100 41.633 953 160 42.705 977 250 43.815 1.003 400 46.344 1061 630 47.436 1086 1000 48.967 1.121 1250 51.955 1189 Note Source: BOLETÍN OFICIAL DEL ESTADO. 2015C. Núm. 297, Sec. I. Pág. 117270.

Cost ( ) The data shown in Table 15 is transformed into a linear function represented in Figure 17, whose independent variable is the size of the transformer needed (x axis in Fig. 1), that is established to be higher than the contracted power. The dependent variable is the total cost associated to the transformer. Figure 17. Transformer s cost Total costs of transformers y = 9,3417x + 41789 60.000 50.000 40.000 30.000 20.000 10.000 0 0 200 400 600 800 1000 1200 1400 p trans (KVA) p trans p contracted (4) C trans = 9,3417 ptrans cos φ + 41789 (5) 5.2.7 Batteries a. Battery technology chosen for charging stations When choosing a specific type of storage device for the project purpose there are several options with different characteristics regarding capacity, efficiency, duration and costs. All of them have benefits for specific operational requirements or competing technologies. The most common storage technologies are: Compressed Air, Flow Battery, Flywheel, Lead- Acid, Lithium-Ion, Pumped Hydro, Sodium and Zinc. Some of these technologies make economic sense only at large scale, such as compressed, pumped hydro, etc., which is not the scope of this study. An analysis of the different types of batteries have been made in order to choose the most adequate one. The main features considered of the different storage technologies are obtained from a study from Lazard (2015) and described below. 35

- Flow battery A flow battery is a type of rechargeable battery where rechargeability is provided by two chemical components dissolved in liquids contained within the system and separated by a membrane. Its life time is also around 15-20 years. One of the positive points of this batteries is its high and independent capacity of scalable. Moreover, they are not affected by the effect of degradation. However, its efficiency is reduced to its rapid charge/discharge. - Sodium Sodium batteries can be distinguished between high temperature and liquid-electrolyte-flow batteries. They have high power and energy density relative to for example, lead-acid. Its expected useful time is from 5 to 15 years. Despite being a long duration, mature technology, the costs are high and there is a potential risk of flammability due to the need of high temperatures for its operation. - Zinc Zinc batteries cover a wide range of technology applications. They are non-toxic, non-combustible and potentially low-cost due to the abundance of the primary metal. Despite all these positive factors, this technology remains unproven in widespread commercial deployment. Its duration is also 5-15 years. - Lithium-Ion These batteries are commonly used in electronic devices and transportation industries. Lately they are replacing lead-acid batteries in specific applications. The main reason is its high energy density, low self-discharge and high charging efficiency. The reason why it is not totally established is its relatively high cost and the advanced manufacturing capabilities needed to achieve high performance. - Lead-Acid For this study Lead-Acid batteries haven been chosen, which are considered to gather the specific features required. Lead-acid batteries were invented in the 19th century and are the oldest and most common batteries. One of its advantages is its low cost in comparison with some of the others and also the fact that they are adaptable to numerous uses such as electric vehicles, off-grid power systems, uninterruptible power supplies, etc. New technologies related with this type of batteries allow the increase of the efficiency and life time, and also the improvement of partial state-of-charge operability. This is known as Advanced lead-acid battery technology. 36

b. Battery costs involved Once it has been established which kind of battery to use in the model, the different costs involved have been gathered as shown hereafter. As shown in Equation 5, this cost is composed of a fixed and a variable cost. C bat = C F bat + C V bat (5) First of all, there is a fixed cost accounting for installation and transportation costs due at the first step of building the infrastructure. The value for this fixed cost has been set in 4.000. C F bat = 4.000 (6) Then it comes the variable cost, whose composition is shown in Equation (6). i=17568 C V bat = V b cap + C rep D [ch i + dh i ] i=1 As can be observed, the battery s variable cost is composed of one term depending on the battery capacity needed, and a second term referred to degradation. According to several articles (Princeton, 2011), variable cost for Lead Acid batteries are around a range of 100 /kwh and 200 /kwh, so in the model to be optimized it is established on 150 /kwh. The cost term related to degradation depends on the total activity of the battery, in terms of total amount of power charged and discharged, and also on a degradation coefficient D. The cost of replacement C rep and the degradation coefficient have been introduced as parameters according to values obtained from Beer et al analysis (Beer et al., 2012). 5.2.8 Photovoltaic investments As commented before, one possible case study that is analyzed is the introduction of photovoltaic panels in order to obtain renewable energy to contribute to the profitability of the charging station model. The purpose of introducing photovoltaic panels is to either serve the electric vehicles directly with solar energy when possible, or to store this energy in the battery for its subsequent usage. Moreover, as explained later, the possibility of selling photovoltaic energy to the grid is also presented. First of all, the solar operation data is obtained from NREL (National Renewable Energy laboratory) PVwatts calculator ("Pvwatts Calculator", 2017). The chosen location has been Madrid. The parameters introduced in the model are the beam Irradiance corresponding to each hour. The units of this parameter are W/m2. The corresponding costs for solar panels have been obtained from a NREL report (National Renewable Energy Laboratory (NREL) Technical, 2017), as reflects Table 16. (7) 37

Figure 18. Solar panel costs Note Source: National Renewable Energy Laboratory (NREL) Technical. 2017. U.S. Solar Photovoltaic System Cost Benchmark: Q1 2016. Table 16. Summary of solar panel costs Small Size Medium Size Cost [$/W] 2.93 2.13 Cost [ /kw] 2610 1900 Note Source: National Renewable Energy Laboratory (NREL) Technical. 2017. U.S. Solar Photovoltaic System Cost Benchmark: Q1 2016. In Figure 18 the evolution in costs per year can be observed for three sizes of solar panels: Residential PV, Commercial PV and Utility-Scale PV. In the case study of this project Small-Size (Residential) and Medium-Size (Commercial) are considered. The selection of both sizes is used as a function of the total size needed. The inflection point in which medium size becomes more profitable than the small one is considered when the total size needed is superior to 10 Kw. Figure 19 reflects the dynamic used when choosing the kind of solar panels and its correspondent costs. 38