Model Uncertainty and Parameter Sensitivity in Linear Retardance Formulations
William F. McTernan
AbstractPreliminary sensitivity and uncertainty analyses were performed to examine the effects of adsorption parameter and model selection on solute transport. It was found that while sensitivity increased with increasing retardation factor (R), local maximums existed for specific mass transfer coefficients. Comparisons between the concentrations predicted by equilibrium and non-equilibrium models showed that in the time period 0 < T ≤ R, the traditional equilibrium model always under-predicted the concentration simulated by non-equilibrium assumptions. A maximum possible error of -100 percent existed. For the time T > 2R, however, the equilibrium assumption always over-predicted the concentration with a maximum possible error of +25 percent. For the time R < T ≤ 2R, the equilibrium may either under- or over-predict the output concentration, with a possible error ranging from -50 to +25 percent. The discrepancies between equilibrium and non-equilibrium models increased rapidly as the mass transfer coefficient decreased. This indicated that significant errors may result from using equilibrium transport model to simulate solutes undergoing non-equilibrium adsorption, especially for hydrophilic compounds where mass transfer rate are slow.
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 United States License.