Optimal design of vanadium flow batteries

A number of effects impact on the performance a Vanadium Flow Battery (VFB) multicell stack, such as shunt currents, and hydraulic losses. The formers are caused by the different cell electric potentials which drive the vanadium charged species to move along the hydraulic flow paths inside the cells and stack, resulting in electric currents (dubbed “shunt”) and the related Joule losses [1]. The latter are due to the friction losses in the porous electrodes, in the stack hydraulic paths and in the external hydraulic piping. In addition, losses are also due to the hydraulic ancillary devices, to the cell electrochemical overpotentials and to self-discharge effects resulting from vanadium species crossover through the membranes [2]. Shunt-currents and hydraulic losses can be kept low with a proper hydraulic circuit design, even if they call for opposite choices. In fact, the trade-off between shunt current losses and pumping losses is the crucial issue in designing VRFB hydraulic circuits [3]. The evaluation of shunt current is quite difficult from experimental point of view [4]. Indeed, they can be computed numerically by solving an equivalent electrical circuit in which each cell was represented as an ideal voltage source E0,n in series with an internal resistance Ri,n, which represents the cell overpotential voltage drop, both parameters being SOC dependent. The electrolyte flow segments of the stack, i.e. manifolds, flow channels and felts, can be modelled as resistors [1]. In this work the authors provide a shunt current evaluation by calculating the electrical resistance of the n-th segment with length lk,n and cross section Ak,n as Rk,n,± = lk,n/σk±,n Ak,n, where k stands for manifold (m) and cell flow channels and felts (c). In particular this work is focuses on the evaluation of Rc by means FEM analysis by varying the cell geometry. The final goal is a multiphysic optimization of a multicell VFB stack by reducing shunt currents losses and pressure loss, i.e. by optimizing the cell area and cell number. This work presents the preliminary results of such optimization analysis for VFBs stacks, performed at the Electrochemical Energy Storage and Conversion Laboratory (EESCoLab) at the University of Padua (Italy). In Fig. 1 is presented a preliminary evaluation of the electrical resistance of a flow channel geometry Rc of a large-scale cell, that is the starting point for the evaluation of the optimal size of the VFB stack for a desired power.