Modeling process dynamics in membrane-electrolyte-assemblies of chloralkali electrolyzers considering steric and hydration effects

The dynamic operation of the chloralkali electrolyzer has been considered as one of the solution for balancing the supply-demand fluctuation in the electricity grid. This opportunity becomes a motivation to study the complex transient behaviors of the electrolyzer caused by linked multiscale phenomena. Membrane of the electrolyzer, which is a crucial part of the plant operation and lifetime, is prone to chemical, mechanical and thermal damages. Direct measurement and in situ observation on the phenomena inside the membrane are technically challenging and very limited. Throughout this work, a mathematical model based on the non-equilibrium thermodynamic theory is proposed to model the process dynamic in Membrane-Electrolyte-Assembly (MEA). The model describes the electrolyte in the MEA as a hydrated multicomponent mixture, that consists of five main components, viz. Na⁺, Cl⁻, OH⁻, and water. In general, the formulated model consists of the generalized Poisson-Nernst-Planck equation system coupled with the momentum balance. By considering the appropriate boundary conditions and mechanical equilibrium assumption, the model is then reduced into a 1-D modeling problem consisting of the generalized Poisson-Nernst-Planck system coupled with the momentum balance for the mechanical equilibrium system. The Galerkin finite element method is utilized to discretize the spatial domain, while the time domain is discretized by using the unconditionally stable backward Euler method. The ramp input representing the primary balancing scenario of the demand response is introduced as the dynamic input. The model predicts that chloride ion is the determinant component in the transient mechanism. Different transient mechanism during the ramp up and the ramp down load change are also predicted during the ramp input simulation. The simulations also show that the charged double layer in the vicinity of the solution-membrane interfaces plays an essential role in shaping the transient. The diffusion coefficient and the hydration number significantly influence the permselectivity and the formed double layers, which determines the dynamic behavior of the MEA. The simulation results also show that the limiting current regime in the more hydrated mixture occurs at the higher current density. This trend defines the local overshoots occurring during the transient. The non-linear relation between the mechanical stress acting on the membrane and the current density is presented in a current-pressure curve. Membrane thickness is initially identified as an uncertain property that could influence the process dynamic and lead to abnormal process condition. Four different MEAs with different membrane thickness are modeled and simulated. Overall, the simulations predict that the thinner membrane exhibits shorter transient time, less mechanical stress, less electric potential drop and higher limiting current density. However, this advantageous trend needs to be compromised with the lower permselectivity performed by the thinner membrane. Summarily, this work shows that the developed model is capable to demonstrate the transient mechanism in the MEA, that are governed by the linked multiscale phenomena.

Author dc:creator

• Budiarto, Thomas

Advisor dc:contributor.advisor

• Repke, Jens-Uwe

Licence dc:rights.uri

• https://creativecommons.org/licenses/by-nc/4.0/

Language dc:language.iso

en

Identifier URI

http://dx.doi.org/10.14279/depositonce-11399

OAI identifier oai:identifier

oai:depositonce.tu-berlin.de:11303/12590