Analytical upscaled models that can describe the depletion of dense nonaqueous phase liquids (DNAPLs) and the associated mass discharge are a practical alternative to computationally demanding and data-intensive multiphase numerical simulators. A major shortcoming of most existing upscaled models is that they cannot reproduce the nonmonotonic, multistage effluent concentrations often observed in experiments and numerical simulations. Upscaled models that can produce multistage concentrations either require calibration, which increases the cost of applying them in the field, or use dual-domain conceptual models that may not apply for spatially complex source zones. In this study, a new upscaled model is presented that can describe the nonmonotonic, multistage average concentrations emanating from complex DNAPL source zones. This is achieved by explicitly considering the temporal evolution of three source zone parameters, namely source zone projected area, the average of local-scale DNAPL saturations, and the average of local-scale aqueous relative permeability, without using empirical parameters. The model is evaluated for two real and twelve hypothetical centimeter-scale complex source zones. The proposed model captures the temporal variations in concentrations better than an empirical model and a dual-domain ganglia- to-pool ratio model. The results provide evidence that effluent concentrations downgradient of DNAPL source zones are controlled by the evolution of the aforementioned macroscopic parameters. This knowledge can be useful for the interpretation of field observations of effluent concentrations downstream of DNAPL source zones, and for the development of predictive upscaled models. Advances in DNAPL characterization techniques are needed to quantify these macroscopic parameters that can be used to guide DNAPL remediation efforts.
Kokkinaki, A., C. J. Werth, and B. E. Sleep (2014), Comparison of upscaled models for multistage mass discharge from DNAPL source zones, Water Resour. Res., 50, 3187–3205, http://dx.doi.org/10.1002/2013WR014663.