Emission-driven surface water quality modeling to prioritize sewage treatment plants for abatement of pharmaceutical emissions

Human pharmaceuticals and its derivatives form a group of bioactive chemicals that are abundant in the water cycle and are found in relatively low concentrations in surface waters and drinking water. Although there is no immediate concern regarding human health based on individual chemicals, concerns exist on mixture effects and adverse effects on fish are observed. Major points of entry into the aquatic environment are sewage treatment plants (STPs). In the Netherlands there is much discussion on if, how and where additional water treatment technologies at STPs would be desirable. Also other abatement options are discussed to reduce emissions of human pharmaceuticals to surface waters, in particular those having a susceptible function. To explore where further abatement of emissions would optimally relieve pressure upon susceptible functions, a consumption-based emission estimation method and a meta-model, derived from the WFD explorer which captures the complex Dutch surface water network, were combined. From this, surface water concentrations for carbamazepine and ibuprofen in a dry and a wet season were modeled. Two types of susceptible functions, i.e. drinking water abstractions and Natura2000 areas, were identified and coupled to modeled concentrations in the water units. For drinking water, besides surface water intakes and river bank filtration, ground water abstractions influenced by surface waters were included. To select the areas where further abatement has the highest priority, STPs were ranked upon their impact on the susceptible functions. Relative contribution to the concentration in susceptible functions was compared between Dutch STPs and to foreign sources; for the latter it was found that the relative contribution from Dutch STPs compared to foreign sources is rather heterogeneous over the country. Besides, different impact factor calculations suggest that a relatively low number of STPs have a relatively high impact on susceptible functions. Impact factors are for a part determined by normative aspects and thus steered by interests and can serve as a backbone in the discussion between stakeholders. This study showed the potentials and bottlenecks of water quality modelling and STP selection strategy. Further development of the concept and refinement of assumptions are required for a more reliable STP prioritization. Moreover, the model can be used as a supplement to other models; modeling on the national scale may serve as “umbrella” for regional models and as a piece of the puzzle, e.g. on the European scale