The studies which this thesis is based on assessed the stormwater quality in a highway catchment located in Sundsvall, Sweden, and examined the performance of a gross pollutant trap (GPT)-biofilter stormwater treatment train (TT) downstream of the catchment, in terms of removal efficiency, intra-event variability, and environmental risk reduction of organic micropollutants (OMPs) for the receiving water body. Assessing the occurrence and event mean concentrations (EMCs) of all OMPs in the catchment revealed that bisphenol A (BPA), octylphenol (OP), nonylphenol (NP), five carcinogenic and four non-carcinogenic polycyclic aromatic hydrocarbons (PAHs), and C16-C40 fractions of petroleum hydrocarbons (PHCs) all potentially pose an environmental risk to freshwater (EMCs >PNEC: predicted non-effect concentration), while alkylphenolethoxylates, six low- molecular weight PAHs, and lighter fractions of PHCs (C10-C16) do not occur at problematic levels.

In order to assess the impact of the downstream TT in mitigating the risks of the studied OMPs, the performance of the TT compartments (a GPT followed by three filter cells) was analyzed and then compared with each other to identify the importance of each design feature (i.e. pre-treatment GPT, sand-based filter media, vegetation, and chalk amendment). Overall, the TT removed most OMPs from highway runoff effectively. The GTP did not contribute to this treatment, thus, the filter sections were responsible for most of the OMP removal. The results showed that, although the non-vegetated sandfilter (SF) could moderately (<50% removal for phenolic substances) to substantially (50–80% removal for PAHs and PHCs) treat the OMPs, the vegetated biofilters (BF and BFC) considerably improved the removal performance, especially for BPA, OP, and suspended solids (TSS). This observation was explained by additional filtration processes provided by the vegetation topsoil layer, which not only enhanced the particulate/particle-bound OMP physical retention but also physiochemical adsorption of colloidal and soluble substances/fractions (such as BPA and OP).

Further analysis of intra-event concentration (IEC) variations of OMPs and TSS showed that the IECs in the highway stormwater and GPT outflow varied considerably without any particular patterns over the course of the events, but first flush rarely occurred. The IEC variations were attenuated by the SF and BFC cells so that more even pollutant load discharge with no first flush was observed during the filter cells’ outflow events. Yet, the IECs for the SF cell revealed that the IECs often peak at the beginning of the effluent events (within the first 100 m3 out of maximum record of ⁓600 m3) and then decrease and become stabilized towards the end of the event. The early-phase concentration peaks exceeded the PNECs for TSS, five PAHs, BPA, and OP, a fact that was not shown by the EMC-based analysis, thus highlighted the advantage of the IEC analysis.