Modelling and control of vanadium redox flow battery for profile based charging applications

Batteries are the crucial link in accelerating the penetration of renewables in distributed grid systems. With multiple sources (most of which are intermittent), there is a need to optimize the usage of these energy storage systems. It is vital to have a charging algorithm to intelligently charge the battery, considering the losses and state of health. Several battery charging techniques exist for most of the mainstream batteries such as lithium ion and lead acid batteries. These charging algorithms are developed for specific battery chemistries to optimize the charging operation by controlling the charging current. There has been no specific charging methodology proposed, considering the uniqueness of flow batteries. In this paper a flow battery model is developed to accommodate the implementation of charging protocols by controlling the current and flow rate in real time. The aim is to demonstrate a methodology where different user requirements can be programmed for efficient operation of the battery to perform a given task. The control is established by an intuitive fuzzy logic controller for changing the input variables based on the intrinsic losses, pump power loss and the capacity decay due to the aging of the battery. The need for such a controller to handle different scenarios of operation arises from the vast applicability of large scale energy storage. A simple example of such a system is proposed with demonstrating three operator selectable modes namely Autonomous (A), Energy bias (E) and Time bias (T). In each case, a specific current value and flow rate value is selected by the controller and is finetuned during the charging process. The values are tuned for specific modes of operation and different aspects of the battery performance are studied for these modes. This work demonstrates a method to leverage the use of a physical model based system in a high level user interface for controlling flow batteries. (C) 2017 Elsevier Ltd. All rights reserved.