Incorporating nonlinear isotherms into robust multilayer sorptive barrier design

Sorptive barrier technologies have emerged as useful tools for addressing a wide range of remediation problems. When simulating barrier performance, numerous isotherm expressions are available for relating aqueous and sorbed concentrations. However, isotherm selection is non-trivial because alternative expressions may yield comparable fits to experimental data. Additionally, concentration data collected for parameter fitting is often outside the range of concentrations relevant to simulating barrier performance. This incompatibility necessitates extrapolation of isotherm behavior during simulation–optimization. Consequently, equally plausible isotherms may predict significantly different barrier performance. Numerical experiments involving organic contaminants were performed to examine the influence of isotherm selection and extrapolation on optimal barrier design. Ten isotherms were calibrated to existing experimental data and evaluated using information-theoretic selection criteria. When incorporated into simulation–optimization, extrapolation effects were clearly evident and optimal designs varied according to the chosen isotherm. To ensure robust barrier design in the presence of such variability, a simple methodology is proposed that utilizes a piecewise-minimum isotherm concept. By favoring plausible isotherms that predict the least amount of sorption, the methodology encourages conservative barrier design while respecting available data.