[A-HW19] Tracer Hydrology: Advances in Measurement and Modelling

[AHW19-P03] Sensitivity-Based Parameter Estimation Framework for Dual Permeability Model for Stratified Porous Medium

*Aman Chandel¹, Deepak Swami¹, Nitin Joshi² (1.School of Civil and Environmental Engineering, Indian Institute of Technology Mandi HP India, 2.Department of Civil Engineering, Indian Institute of Technology Jammu India)

Keywords:Dual Permeability, HYDRUS, Parameter Estimation, Sensitivity Analysis, Tracer transport

The advancement in computational capabilities and a deeper comprehension of tracer transport physics have led to the emergence of sophisticated models such as multi-porosity multi-permeability models. These advanced models effectively capture multiple inflection points in breakthrough curves (BTC). However, with increased model complexity comes a higher number of parameters requiring estimation, necessitating an inverse modeling framework. To streamline the optimization process, a novel approach is introduced, which reduces dimensionality by identifying critical parameters and assigning constant values to less significant ones. This innovative framework integrates Global Sensitivity Analysis (GSA) and Parameter ESTimation (PEST) with the Dual Permeability (DP) HYDRUS module tailored for stratified porous media. To validate its efficacy, a tracer transport experiment through a stratified medium is conducted, specifically focusing on breakthrough curves exhibiting multiple peaks and elongated tailing. The analysis concentrates on capturing four key inflection points: a rising limb, two peaks, and a tailing limb. Sensitivity analysis reveals that, during the early stages of initial breakthrough, fracture domain parameters and immobile porosity account for 64% of total sensitivity. In the rising limb, matrix parameters like porosity and permeability, along with immobile region porosity, contribute significantly, representing about 48% of total sensitivity. The elongated tailing limb sees the dominance of matrix parameters (porosity and permeability), contributing 91%, with immobile porosity sensitivity at 4%. Consequently, parameter estimation based solely on BTC fitting tends to prioritize peak concentration, potentially overlooking accurate estimates for other crucial parameters such as mass recovered, residence time, and outlet variance. In contrast, the developed parameter estimation framework, based on cumulative mass recovery, efficiently captures significant inflection points, peak concentration time, and long tailing behavior. This framework enhances mass recovery estimation at the outlet compared to BTC-based methods.