SYMPOSIUM DE GENIE ELECTRIQUE(SGE’14) : EF-EPF-MGE 2014, 8-10 JUILLET 2014, ENS CACHAN, FRANCE Optimization methodologies for the power management and sizing of a microgrid with storage Rémy Rigo-Mariani, Bruno Sareni, Xavier Roboam Université de Toulouse, LAPLACE, UMR CNRS-INP-UPS, ENSEEIHT 2rue Camichel, 31071 Toulouse Cedex 07 ABSTRACT–In this paper, we investigate a design approach aiming at simultaneously integrating the power management and the sizing of a small microgrid with storage. We particularly underline the complexity of the resulting optimization problem and how it can be solved using suitable optimization methods in compliance with relevant models of the microgrid. We specifically show the reduction of the computational time allowing the microgrid simulation over long time durations in the optimization process in order to take seasonal variations into account. Keywords — Smart grid, sizing, optimal dispatching, linear programming, dynamic programming, evolutionary algorithms, efficient global optimization, kriging interpolation. RESUME –Dans cet article, nous nous intéressons à une démarche de conception optimale intégrant la planification des flux énergétiques et le dimensionnement des éléments d’un micro-réseau avec stockage. Nous montrons plus particulièrement comment l’adéquation entre les méthodes d’optimisation utilisées et le modèle du micro-réseau employé peut permettre la réduction significative des temps de calcul et la détermination d’une configuration optimale du micro-réseau, valable sur des horizons temporels intégrant les alternances saisonnières. Mots clés — Smart grid, dimensionnement, plannification optimale,programmation linéaire, programmation dynamique, algorithmes évolutionaires, krigeage. 1. INTRODUCTION With the development of decentralized power stations based on renewable energy sources, the distribution networks has strongly evolved to a more meshed model [1]. It can be considered as an association of various "microgrids" both consumer and producer that have to be run independently while granting the global balance between load and generation. Smarter operations now become possible with developments of energy storage technologies and evolving price policies [2]. Those operations would aim at reducing the electrical bill taking account of consumption and production forecasts as well as the different fares and possible constraints imposed by the power supplier [3].This paper deals with a microgrid devoted to a set of industrial buildings and factories with a typical subscribed power of 156 kW (Fig. 1a). It includes photovoltaic (PV) production and a storage unit composed of high speed flywheels(FW). The strategy chosen to manage the overall system is based on a daily off-line optimal scheduling of power flows for the day ahead. Then, in real time, an on-lineprocedure adapts the same power flows in order to correct errors between forecasts and actual measurements [4]. Such control strategy based on the on-line adaptation of off-line optimal references has been extensively studied in the literature (e.g. [5-7] and is not the subject of this work. Our study mainly focuses on the microgrid design investigating the coupling between the power management (i.e. off-line control) and the sizing of the microgrid components (i.e. PV production and storage). Consequently, finding an optimal configuration of the microgrid results in a two level optimization problem including the optimal sizing of the microgrid and the optimal power flow dispatching over a long period of time. In the following sections, we will address this issue and its complexity with regard to energy cost optimization and computational time. To face this problem, we will show how it can be solved using suitable optimization methods in compliance with relevant models of the microgrid. The rest of the paper is organized as follows: in the second section, the power flow model of the microgrid is presented. In section 3, several power dispatching strategies ensuring the minimization of the energy cost are compared. In particular, a fast optimization approach based on Linear Programming (LP) and on a linear model of the microgrid is introduced in order to reduce the computational time of the power flow dispatching. In section 3, a second optimization level is presented. It consists in determining the optimal sizing of the microgrid with regard to the energy cost computed over a complete year in order to take seasonal variations into account. Finally, conclusions are drawn in section 4. 2. MODEL OF THE MICROGRID 2.1. Power Flow Model of the Microgrid The power flow model of the microgrid is given in Fig. 1b. All the microgrid components are connected though a common DC bus. Voltages and currents are not represented and only active power flows are considered.In the rest of the paper the instantaneous values are denoted as P i (t) while the profiles over the periods of simulation are written in vectors P i . Due to the grid policy, three constraints have to be fulfilled at each time step t:  P 1 (t)  0: the power flowing through the consumption meter is strictly mono-directional  P 10 (t)  0: the power flowing through the production meter is strictly mono-directional