Study on architecture design of electroactive sites on Vanadium Redox Flow Battery (V-RFB)

. Numerous researches have been conducted to look for better design of cell architecture of redox flow battery. This effort is to improve the performance of the battery with respect to further improves of mass transport and flow distribution of electroactive electrolytes within the cell. This paper evaluates pressure drop and flow distribution of the electroactive electrolyte in three different electrode configurations of vanadium redox flow battery (V-RFB) cell, namely square-, rhombus- and circular-cell designs. The fluid flow of the above-mentioned three electrode design configurations are evaluated under three different cases i.e. no flow (plain) field, parallel flow field and serpentine flow field using numerically designed three-dimensional model in Computational Fluid Dynamics (CFD) software. The cell exhibits different characteristics under different cases, which the circular cell design shows promising results for test-rig development with low pressure drop and better flow distribution of electroactive electrolytes within the cell. Suggestion for further work is highlighted.


Introduction
Energy storage is recognised as an important technology to work with variable and intermittent in nature of renewable power sources. Therefore, energy storage has been introduced to retain and sustain the required operational and system reliability of the power sources [1] [2] Through various energy storage technologies, vanadium redox flow battery (V-RFB) shown in Fig.1 [3] offers a promise because of its unique features by providing effective and simple operation, capability of decoupling high power and energy, may operate at room temperature, fast response and recharging contenders. Furthermore it comprises an excellent chemical stability that shows an extremely long-round-trip cycle, have long discharge times, good in modular design, exhibits for highly reversible redox kinetics, enable the technology for large-scale applications with theoretically reasonable and controlled maintenance cost compared to conventional battery [1] [4][5][6][7]  As the name applies, flow battery is an electrochemical cell that produces electrical output through electrochemical reactions of two electrolytes that flow within the cell. Most of the researches in the past focused on improving poor kinetic reactions [8], conductivity [9], redox couples [10] electrodes [11] and battery characterisation [12]. A more recent works by Houser [13] et al. and Kumar [14] et al. were focusing on improving mass transport issues in redox flow battery. Mass flow is play an important role in battery to perform well [15] [16]. The basic idea of study is to overcome the mass transport polarization issues and outcomes by introducing new design channel that may distribute the electrolyte well with reasonable pumping energy required. Recently, researchers systematically discuss about the electrode porous (electroactive sites) that acts as key in V-RFB systems [11][17] [18].
In the conventional flow field of porous electrode, the electrolyte is directly pumped through the electroactive sites electrodes. This may results in unevenly distributed flow within the cell because the electrolyte flow through random pores of the electroactive sites part. This may causes pressure drop arising together with flow direction and this lead to a mass transfer losses and deterioration problems when the electrolyte flow into the electrode just from one lateral side. Previous researchers [19] [20] have explored on creating new and appropriate flow channel for control the electrolyte flow and directly gives a positive feedback on improvement of V-RFB performance. There are literatures on several experimental and numerical investigation of the V-RFB design system. Xu et al. [19] [21] proposed experimentally and numerical investigation on architecture of redox flow battery. The study focused on evaluation of vanadium redox flow battery performance based on flow of electrolytes within the cell, focusing on with and without flow field electrode configurations supported by experimental data and numerically designed three-dimensional model of V-RFB. The study concluded that energy efficiency of V-RFB is much higher by 5% when electroactive sites of flow field was applied as compared to energy efficiency without flow field.
Conversely, Wang et al. [22] studied on parallel flow design configurations concluded that energy efficiency of V-RFB could be further be enhanced by 5% in staggered arrangement as compared to symmetrical arrangement. Wang et al. claimed that design of flow field for V-RFBs is an important areas that needs more attention for further investigation and it was concluded that limited work in this area of research has been conducted.  On the other hand, Bortolin et al. [23] and Khor et al. [8] introduced a methodology included several performance indicators by using computational fluid dynamic simulation on determining the flow distribution in a particular geometry of V-RFB cell. Ressel et al. [24] studied the influence of different flow cell geometry design by introduced a tubular cell design with a flow by electrode configuration on the performance and system efficiency of V-RFB cell.
Based on the brief review, it is important to note that both flow field electrode configurations and architecture design of the cell play important roles for future advancement of flow battery technology.
Nevertheless, apart from Khor et al. which worked on rhombus shape electrode geometry design, all other highlighted studies were very much focused on squareshape based cell geometry design.
This study attempts to extend the works in both flow field electrode configurations and cell design geometry; looking into perspective of effect of flow distribution within three different cell design i.e. square shape, rhombus shape and circular shape electrode geometry design. The first part of studies were focused on identifying the best shape to be implemented in V-RFB cell design with respect to flow distribution within the cell based on no flow field electrode configurations. Extended works were focused on application of the best shape of electrode geometry design in further two electrode configurations to identify the lowest pressure drop which could resulted in better performance of the V-RFB.

Simulation details
Numerical simulations were carried out and solved using the ANSYS 16.2 package with a combination of the selfwritten source terms in console zone. Three dimensional models were built in fluid flow (fluent) under analysis systems for handling the flow in the porous carbon electrode. The fluid model used is sulfuric acid solution diluted with vanadium oxides. The fluid was considered to be incompressible and isothermal, with density and dynamic viscosity were 1334.9 Kgm -3 and 0.0024 Pa respectively.
The model domains were developed based on a 100 cm 2 (10 cm x 10 cm) active sites of electrode with a thickness of 3 mm with 0.00424 m cross section area of the channel. The details operating parameter in geometry and setup parameter are shown in Table 1. This dynamic model were simulated with different velocity inlet condition that has been calculated analytically and a specified uniform flow velocity at the inlet and 0 Pa pressure at the outlet was set in the boundary condition. No slip condition on the walls and no flux condition were applied to the domain surfaces exceptional for inlet outlet flow channels. In all simulations, the fluent was set with double precision and pressure-based solver. The porous media model was evaluated with a pressurevelocity coupling algorithm SIMPLE (Semi Implicit Method for Pressure-Linked Equation) in solution methods. The CFD simulation were conducted at different flow rates, flow with/without distribution channel and with different electrode compartment design in order to evaluated the best configuration for transport phenomena and performance in the V-RFB.

Pressure drops in cell
Pressure drop is a crucial issue in the system for determining the smooth continuous flow circulation of the liquids inside the cell stack. At variance of flow pattern design in the cell electrode (active area) may give a different flow distribution and pressure drop. In this study, two conditions were set; at lower flow rate, the flow was consider laminar flow whereas higher flow rate was considered as turbulent flow. This consideration of flow was supported by the calculation by referring Reynolds number formulae. The Reynolds number were determined respectively are shown in Equation 1. In this paper, various flow pattern design were considered including without distribution channel (direct feeding) in a porous media electrode and indirect feeding channel; a parallel and serpentine pattern. (2) The permeability is dependent on the porosity of the electroactive sites of carbon electrode. Flow crosssectional area (A) changes with different cell geometry (active sites).

Computational domains and boundary conditions
A symmetry boundary condition were elucidated to set the operational point of V-RFB cell that consist of inlet velocity magnitude, gauge-pressure at outlet region, symmetry planes and wall condition surface .The no-slip boundary condition was applied on the electroactive sites of the electrode excluded the inlet and outlet regions. Next, on the solution method setup part, pressurevelocity coupling was selected and performed by the SIMPLE algorithm and the second-order upwind scheme was used in a computational domain. Finally, numerical analysis were simulated through continues running by maintaining a steady state conditions of the operation of the battery cell and mass-weighted values applied also monitored.

Result and Discussions
With the aim to investigate the influence of the cell's electrode geometry design on flow distribution within the cell, flow rates at 5 cm 3 s -1 to 30 cm 3 s -1 were fed into three proposed 100 cm 2 square shape, rhombus shape and circular shape electrode geometry design. Fig. 3 indicates the shapes plays significant parts affecting the flow distribution within the three proposed electrode geometry designs when applied through no flow field (plain) electrode configuration at flow rate of 30 cm 3 s -1 . The result shows that circular shape design gives the best uniformity electrolyte distribution as compared to square shape and rhombus shape electrode geometry design. Other results for different flow rates will be published in forthcoming extended paper. Moreover, pumping power is one of the impact parameter for determining a better RFB system. It should be noted that, lower pressure drop indicates a low pumping consumption. In this study, pressure drop has been calculated to determine which proposed geometrical cell design consume a lesser pumping power. In this case, area-weighted average of static pressure has been computed for each geometrical designs. Fig. 4 shows the pressure drop for each geometrical designs at various flow rate of 5 cm 3 s -1 until 30 cm 3 s -1 . It should be noted that, the flow rate increases in accordance to applied flow rate, indicating of uniform distribution of electrolytes within the cell with the expense of higher pumping losses; higher pumping power required is proportional to higher pumping losses which could affecting in overall efficiency of V-RFB system. The results shows that the square flow frame has the highest pressure drop (17.3 KPa) at flow rates 30 cm 3 s -1 , followed by rhombus cell (16.5 KPa) and circular shape is (12.2 KPa). Furthermore, with circular shape cell design represent best result for both flow distribution and pressure drop at various flow rates, the study is extended to determine which flow field configurations would be the best to be applied on 100 cm 2 circular shape cell design. Fig.5 represents the pressure drops for no flow field porous medium model and two different flow pattern at the flow rates 5 cm 3 s -1 until 30 cm 3 s -1 . The result of the numerical simulation indicates that the parallel flow field shows the lowest pressure drop at 253 Pa whereas the no flow field and the serpentine distribution channel has represented a 422 Pa and 2469 Pa pressure drop. This indicates that serpentine flow channel has a highest pressure drops and no flow field design is two times higher than parallel design. By comparing the parallel and no flow field designs, it should be noted that at lowest flow rates, by adding flow field at bipolar plates it can significantly reduce a pressure drop. Nevertheless, when the flow rate is increased, for example of 30 cm 3 s -1 , the pressure drop for the porous electrode and two types of flow patterns increase accordingly as shown in Fig.5. For the first two fields, the pressure drop increases from 253 to 12432 Pa and from 2469 Pa to 45701 Pa respectively whereas without flow field electrode it increases from 422 Pa to 17375 Pa. Details data of effect on pressure drops for respective square shape, rhombus shape and circular shape cell geometry design with respect to flow rates 5 cm 3 s -1 until 30 cm 3 s -1 will be discussed in forthcoming paper.

Conclusions
This study demonstrates that the circular shape cell geometry design has most uniform electrolyte distribution within the cell at applied 30 cm 3 s -1 flow rates as well as lowest average pressure drop as compared to square and rhombus cell design counterparts, whereas parallel flow field configurations represents the best options for 100 cm 2 circular shape cell geometry design of V-RFB at 5 cm 3 s -1 -30 cm 3 s -1 flow rates. Further works on effect of cell architectures to the performance of V-RFB with respect to coulombic, voltage and energy efficiency will be discussed in forthcoming paper.