Crosscutting Multiscale Modeling of Electrolysis Cells for Accelerating Materials R&D

Abstract

Numerical models are tools that can be used to a) elucidate fundamental phenomena that are difficult to observe/resolve experimentally, and b) observe device performance as a function of materials properties and operating conditions. The latter use is particularly valuable for informing the design of cell components and accelerating their implementation at practical scales. This presentation will discuss multiple modeling approaches for electrochemical cells and the corresponding computational methods used to solve them. We will describe how high-fidelity models can be used to analyze cell components by assessing their performance, lifetime, and reliability.In our approach, a general framework for electrochemical modeling is constructed using principles of conservation of mass, momentum, charge, energy, and species. Then, application-specific, user-defined source terms are implemented which generally account for chemical and electrochemical reactions. These source terms are translated to all the conservation equations mentioned above, which expands the functionality of the computational fluid dynamics software and allows the basic modeling approach to be very flexible. Through the specific mathematical descriptions for the molecular conversion processes, such as H2 production or CO2 electrolysis, the modeling methods can be readily adapted to various electrochemical cell architectures as a general approach for component and device design or performance optimization.After validating to lab-scale experimental data, the models can be used to perform sensitivity studies on the material properties to predict the most impactful avenues for developing a system component. In addition, existing component designs can be compared to determine which is better suited to a particular application. Ranges of operating conditions can be explored in silico to observe their effect on performance at a much faster rate than can be investigated experimentally. Examples of using this approach in the past include: Analyzing the performance of various thin foil LGDL designs for low temperature proton exchange membrane electrolysis cellsStudying the effect of operating conditions on the leakage current phenomena in proton conducting solid oxide electrolysis cellsInvestigating the role of the thickness of the catholyte buffer layer in low temperature CO2 electrolysis cells These diverse applications demonstrate the usefulness of the generalized modeling framework.This presentation will demonstrate how computational modeling can link experimental materials research with the final component or product design and help close the gap between innovation and implementation. Our approach facilitates materials R&D and advances computational tools to support industry and other high technology readiness level efforts. These insights are used to inform the device design and obtain performance, degradation, and life cycle predictions, which are then used to accelerate material implementation into a finalized product.