Synchronization of isolated microgrids with a communication infrastructure using energy storage systems Jairo Giraldo a,⇑ , Eduardo Mojica-Nava b , Nicanor Quijano a a Departamento de Ingeniería Eléctrica y Electrónica, Universidad de los Andes, Bogotá, Colombia b Electrical and Electronics Department, National University of Colombia, Bogotá, Colombia article info Article history: Received 21 December 2013 Received in revised form 7 May 2014 Accepted 27 May 2014 Available online 25 June 2014 Keywords: Communication infrastructure Energy storage Frequency control Microgrids Multi-agent systems Networked control systems abstract There exists an increasing interest in networked systems due to the wide number of applications in dis- tributed and decentralized control of large scale systems, such as smart grids. We address the problem of distributed frequency synchronization of several isolated microgrids, each one described by a linear-time continuous system, composed by different types of generators, whose outputs are measured and sent through a communication infrastructure. We assume that each microgrid possesses renewable resources with storage capabilities that helps to improve stability of the network when small damping ratio is con- sidered. Thus, using the smart grid communication infrastructure and the data flow through the network, we propose a cooperative control strategy based on the consensus algorithm that simultaneously man- ages the turbine governor input and the amount of energy that the storage devices have to absorb/inject from/into the grid. Nevertheless, physical constraints need to be included, which can be modeled using saturation non-linearities, and conditions to assure synchronization even with saturation are obtained based on multi-agent systems. Additionally, we consider that sensor measurements are sampled and we extend the results of frequency synchronization with saturation to the case of control discretization and sampling-period independence is demonstrated using passivity concepts. Ó 2014 Elsevier Ltd. All rights reserved. Introduction The power network is a large-scale and complex system that involves a wide number of elements (e.g., generators, loads, control devices) that are interconnected. This makes control in power net- works a challenging research field that has been addressed in the last century, where the main objective is to preserve stability and avoid unnecessary and damaging oscillations [1]. Neverthe- less, in the last few years the inclusion of small and medium volt- age generators and the increasing penetration of renewable resources have introduced new advantages due to the possibility of producing clean, sustainable, and low cost energy [2]. However, new challenges have emerged because of the uncertainties induced by the renewable resources, and the high amount of information that needs to be processed [3]. Furthermore, the future power network (or smart grid) needs high speed, reliable, and secure data communication networks to manage the high amount of informa- tion coming from sensors all over the power network (from generation to user level) and take smart decisions [4]. On the other hand, considering the smart grid as only one large- scale system increases the difficulty of analysis and design due to the high amount of variables that need to be considered. In this respect, microgrids emerge as a solution to these problems. Microgrids (MGs) are smaller systems, usually of medium or small voltage, that include distributed generators (e.g., small hydro tur- bines, diesel generators, solar panels) and storage devices (e.g., energy capacitors, batteries, flywheels). Microgrids could increase the efficiency of power systems facilitating monitoring and control because of its reduced size and the possibility of using hierarchical control [5]. Therefore, we can see the power network as the inter- connection of several microgrids, each one with control and energy generation capabilities, and a communication infrastructure that transmits data between them. In general, an MG can operate as a grid-connected system or as an island, where the latter takes place by unplanned events like faults in the network or by planned actions like maintenance requirements [6]. One of the main goals in control of a power system is frequency synchronization, where each node needs to work at the same fre- quency and voltage in order to avoid failures and malfunctioning. The importance of synchronization lies in the fact that if a genera- tor works with a different speed (frequency) or voltage than the power system due to failures or sudden changes in loads, http://dx.doi.org/10.1016/j.ijepes.2014.05.042 0142-0615/Ó 2014 Elsevier Ltd. All rights reserved. ⇑ Corresponding author. Tel.: +57 3167277994. E-mail address: ja.giraldo908@uniandes.edu.co (J. Giraldo). Electrical Power and Energy Systems 63 (2014) 71–82 Contents lists available at ScienceDirect Electrical Power and Energy Systems journal homepage: www.elsevier.com/locate/ijepes