Pharmaceuticals are inherently biologically active substances and ubiquitous water contaminants that have shown detrimental effects on aquatic organisms at low concentrations. The presence of pharmaceuticals in rivers is starting to be regulated by European and worldwide environmental legislations. Therefore, countries are starting to plan and implement measures (Wastewater treatment plants (WWTP) upgrades and source control) to reduce pharmaceutical concentrations in rivers. Decision-makers use models to predict the fate, removal and transport of pharmaceuticals in rivers and to evaluate the effectiveness of measures for the reduction of pharmaceutical concentrations at catchment scale. However, there is still large uncertainty around the processes driving the fate, removal and transport of pharmaceuticals in rivers which compromises decision-making. Moreover, the cost of implementing WWTP upgrades at catchment or national level can be daunting, hence the development of tools that optimize the upgrading costs are indeed required. In addition, there is little scientific information on the effectiveness of source control measures for the reduction of pharmaceutical concentrations at catchment scale.
Thus, the aim of this thesis is to provide decision-makers with modelling tools for the evaluation of measures (WWTP upgrades and source control) to reduce pharmaceutical concentrations in rivers. The modelling tools include uncertainty in the whole decision-making process.
The first section describes the development and calibration of a Microcontaminant Fate and Transport model for the estimation of pharmaceutical concentrations in rivers including uncertainty. The model was successfully calibrated and the uncertainty in the concentrations decreased after using Bayesian inference and measurements of diclofenac concentrations in WWTPs and rivers.
The second section deals with the influence that the model uncertainty has on the selection of WWTP upgrades designed to decrease pharmaceutical concentrations (i.e. diclofenac) in rivers. For this purpose, we evaluated different scenarios of model uncertainty and WWTP diclofenac removal efficiencies using the model developed in the first section. We concluded that the installation of tertiary treatments results in apparent reductions of diclofenac concentrations regardless of the uncertainty. However, apparent reductions after upgrading secondary treatments require lower uncertainty.
The third section shed light on the relationship between proposed Environmental Quality Standards (EQS) for pharmaceuticals (i.e. diclofenac) and the optimal cost of the WWTP upgrades at catchment level. For this purpose, we optimized the number of WWTPs requiring an upgrade for different EQS and uncertainty levels using multi-objective genetic algorithms and the model calibrated in the first section. We used minimization of costs and total EQS exceedance as the objective functions. We found that there is a non-linear relationship between EQS and the costs and, hence there is an optimal EQS that balances costs and ecosystem protection.
The fourth section illustrates the effect that source control measures (i.e. substitution of diclofenac by naproxen) have on the required WWTP upgrades for the reduction of pharmaceuticals in rivers. For this purpose, we optimized the number of WWTP upgrades for different levels in the consumption of diclofenac and naproxen, different EQS and uncertainty levels. We found that apparent reductions in the number of WWTP upgrades are achieved only when more than half of the diclofenac consumed is substituted by naproxen. However, we conclude that any substitution between pharmaceuticals requires a model-based evaluation because the substitution may be harmful for the environment under specific scenarios of EQS.
Finally, we discussed the factors that influence the selection of measures for the reduction of pharmaceuticals: uncertainty in the estimates of pharmaceutical concentrations, EQS setting, hydrological conditions and consumption of pharmaceuticals. Therefore, we recommend decision-makers to follow adaptive management of pharmaceuticals at catchment level in response to the changing factors that influence the selection of measures.