Process-based multimedia models are frequently used to simulate the long-term impacts of pollutants and to evaluate potential remediation actions that can be put in place to improve or manage polluted marine environments. Many such models are detailed enough to encapsulate the different scales and processes relevant for various contaminants, yet still are tractable enough for analysis through established methods for uncertainty assessment. Inclusion and quantification of the uncertainty associated with local efficacy of remediation actions is of importance when the desired outcome in terms of human health concerns or environmental classification shows a nonlinear relationship with remediation effort. We present an updated fugacity-based environmental fate model set up to simulate the historical fate of polychlorinated dibenzo-p-dioxins and dibenzo-furans (PCDD/Fs) in the Grenland fjords, in Norway. The model is parameterized using Bayesian inference and is then used to simulate the effect of capping parts of the polluted sediments with active carbon. Great care is taken in quantifying the uncertainty regarding the efficacy of the activated carbon cap to reduce the leaching of contaminants from the sediments. The model predicts that by capping selected parts of the fjord, biota will be classified as moderately polluted approximately a decade earlier than a natural remediation scenario. Our approach also illustrates the importance of incorporating uncertainty in local remediation efforts, as the biotic concentrations scale nonlinearly with remediation effort.