Study of multi-parameters for the improvement of performance and development of a 1 kW vanadium redox flow battery stack
Date of Issue2018
School of Materials Science and Engineering
The extensive use of intermittent renewable energy supplies from solar and wind in the recent decades resulted in huge demand for large scale energy storage systems. Redox flow batteries (RFB) hold the capacity to meet these demand, of which, the vanadium redox flow battery (VRFB) is the most popular due to its safe chemistry. Compared to other types of secondary batteries, VRFB offers many advantages, such as flexibility to independently design power density and backup time, life-time of decades requiring low maintenance, environmentally-friendly technology, fast response, high efficiency, excellent load levelling and peak shaving capacity, no safety issues with overcharging or deep-discharging and the cheapest levelized cost/cycle/kWh. Despite commercial availability of VRFB, there have been extensive research interest on improving the performance and reducing the cost of the technology. Research works have mainly focused on the electrode, membrane, electrolyte, and stack design. The VRFB performance is affected by various losses such as activation, Ohmic, mass transport, and other losses such as pump loss, self-discharge, shunt current loss and capacity loss. In conventional designs, electrolyte is forced to flow through the cross section of the porous electrode. Therefore, in a large area cell and stack, optimization of the flow distribution through the porous electrode is important for minimizing the Ohmic and mass transport losses, and protecting the electrode material from degradation. In parallel to improving the flow distribution, the reduction of pump power loss due to the pressure drop across the porous electrode also needs to be considered to maintain a high overall energy efficiency. More importantly, there are no suitable methods available to precisely map the flow behavior through the porous media. Hence, this PhD research aimed to develop suitable methods to study (and improve) the flow distribution in conventional VRFB, minimize the various losses and demonstrate all the improvements by the development of a 1 kW flow battery stack system. The 1 kW stack was aimed to operate at a current density of 80 mA cm-2 and deliver a stable energy and voltage efficiency of >80% during long-term testing. To fulfil the objective of understanding and improving the flow distribution through the porous media, segmented cell and pH tracing approach were developed. A single cell with electrode area of 100 cm2 was divided into sixteen segments and was used to study the local voltage and open circuit voltage (OCV) distribution. It was demonstrated for the first time that the in-situ flow behavior can be precisely predicted and quantified using the local OCV or state of charge (SOC) mapping. Similarly, pH tracing test was developed for the easy and fast screening of the flow distribution. This test was found to be useful in comparing the flow frame designs and improving the electrode configurations during the 1 kW stack development. Further, as an approach to improving the flow distribution and reducing parasitic losses, four designs of flow channel on the porous electrode were investigated for the first time. Of the various designs, interdigitated (IDD) channels on the porous electrode was found to be promising due to the improvement of overall energy efficiency up to 2.5% and reduction of pressure drop by ~ 40%, as compared to cell with no flow channels. Next, to fulfil the target of 1 kW stack development, it was necessary to examine and optimize various areas in flow battery that were critical in determining the stack performance. As such, optimization of electrode compression was done by investigating a wide range of compression in order to minimize the Ohmic and mass transport losses. Similarly, a “power drop effect” was reported in operation of flow battery at high current density. This effect was found to be an important performance determining feature, and it is reported for the first time in this dissertation. Further, during flow battery operation, electrolyte species crossover through the membrane causes electrolyte volume imbalance between two tanks and gradual capacity loss. To address this issue, two tanks were made in hydraulic shunt connection with each other and auto rebalancing of electrolyte was made. In addition, the use of a combination of anion exchange membrane (AEM) and cation exchange membrane (CEM) in a stack was investigated to minimize the capacity loss and experimentally verified for the first time that this concept worked effectively. Similarly, electrode activation is necessary to improve the cell kinetics. Therefore, an investigation was carried out to understand the effects of thermal activation on the electrode performance using surface and electrochemical characterization. All the improvements made through multi-parameter tests were demonstrated in a 20 cell-1 kW stack. The stack used novel design of flow frames with integrated flow guides in consideration of uniform electrolyte flow distribution, thermally treated porous electrodes machined with IDD flow channels, and thin BPs from SGL Carbon. The test results of 250 cycles at 80 mA cm-2 showed a stable energy efficiency and voltage efficiency over 80%, pressure drop reduction by ~ 40%, and a stable capacity achieved through auto rebalancing of electrolyte. Hence, the objective of this PhD research was accomplished.