Network-aware approach for energy storage planning and control in the network with high penetration of renewables Khashayar Mahani ⇑ , Farbod Farzan, Mohsen A. Jafari Rutgers, The State University of New Jersey, USA highlights  Approximate solution for energy storage (ES) sizing and operation planning.  Approximate solution significantly reduces the complexity of problem.  Approximate solution for complex network with multiple applications of ESs.  A rule-based control scheme for the near real-time operation of complex ES network.  The control schema has been developed by mining the I/O statistical relationship. article info Article history: Received 6 December 2016 Received in revised form 23 March 2017 Accepted 24 March 2017 Keywords: Network-aware planning and control Energy storage network Data-driven control Optimal planning Community level micro-grid abstract In this paper, we consider multiple energy storage nodes distributed over a power distribution network, and are purposed for multiple applications. The research problems of interests are to optimally locate these nodes over the distribution network and to create day-ahead plans according to planned applica- tions. The two problems are formulated as stochastic optimization problems, and hourly and time- aggregated approximate solutions are presented. The approximation identifies time periods where load and generation patterns demonstrate low variability, and marks the whole period as a single time zone, thus significantly reducing the number of decision variables and the overall problem size. We show that aggregate and hourly planning solutions are close. The planning problem can handle any number of stor- age nodes with general topology and load connections, and deterministic or stochastic capacities. In this paper, we focus on network of static energy storages with deterministic capacity. Finally, we build a novel rule based control scheme for the near real time operation of the storage network by mining the statis- tical relationship between input and optimal charge and discharge patterns. Ó 2017 Elsevier Ltd. All rights reserved. 1. Introduction Energy storage (ES) has the potential to offer a new means of added flexibility on the electricity distribution systems. This flexi- bility can be used in a number of ways, including adding value towards asset management, power quality and reliability. An important factor in evaluating the feasibility of ES technology is the application(s) for which the storage is used for [1]. ES can pro- vide local level services such as, peak shaving and renewable inte- gration [2,3], and network level services, such as voltage and frequency control [4]. It can also be utilized for loss minimization and deferral of network infrastructure upgrades. With the use of energy storage in a distribution networks for multiple applications, however, comes the challenge of determining how best to control these storage units under load and system state uncertainties. For example, with increasing number of Electrical Vehicles (EVs) the uncertainty in the electricity demand rises due to EV charging demand [5–7]. But, on the other hand, Vehicle-to-Grid (V2G) tech- nology, while mitigating some of this uncertainly, can add system dynamics complexities to the network [8–10]. Han et al. [11] and Wong et al. [12] provide control algorithms to maximize EV owner’s profit, which comes from selling power to grid and participating in the frequency regulation market. They formulate the problem as a discrete-time Markov decision process and solve it by introducing an online learning algorithm which iterates every hours based on available information. Koutsopoulos et al. [13] study the optimal energy storage control problem by tak- ing the point of view of a utility operator and focuses on arbitrage application of energy storage. The authors show that the model can be extended to account for a renewable source that feeds the stor- age device. The same problem was considered in [14], where the http://dx.doi.org/10.1016/j.apenergy.2017.03.118 0306-2619/Ó 2017 Elsevier Ltd. All rights reserved. ⇑ Corresponding author. E-mail address: km723@scarletmail.rutgers.edu (K. Mahani). Applied Energy 195 (2017) 974–990 Contents lists available at ScienceDirect Applied Energy journal homepage: www.elsevier.com/locate/apenergy