Assessment of grid resolution for using Poisson-Nernst-Planck equations in reactive transport modeling
Wed, March 3
E007
Zoom
In reactive transport modeling, multicomponent diffusion systems are often simulated using simple Fick’s laws, which neglect the differences in ion diffusion coefficients. A more refined approach consists of solving the Nernst-Planck equation, which fully considers these differences. However, a local electroneutrality (EN) condition is often assumed instead of solving the full Poisson-Nernst-Planck equations (PNP). The EN condition reduces the number of primary variables to be solved, which might be justified by the accuracy of the numerical algorithms used to solve the transport equations. However, solving the full PNP equations may become necessary in specific modeled systems with spatially highly resolved simulation domains. This work intends to quantify the quality of the reactive transport predictions using the EN condition as a function of grid resolution and ionic strength. By analyzing and comparing the EN and PNP results, we show that the EN condition can be used without loss of accuracy for systems in which the grid discretization is larger than ten times the Debye screening length. Reactive transport simulations satisfy this criterion in most applications.