Modelling the fate of micropollutants in integrated urban wastewater systems: Extending the applicability to pharmaceuticals

Highlights

• The IUWS_MP model library was extended to simulate the fate pharmaceuticals.

• Different pharmaceuticals were simulated in two integrated urban wastewater systems.

• The uncertainty of simulations overlapped with the domain of measurements.

• Applications shows the potential for reducing risk from pharmaceutical discharges.

Abstract

Pharmaceutical active compounds (PhACs) are a category of micropollutants frequently detected across integrated urban wastewater systems. Existing modelling tools supporting the evaluation of micropollutant fate in such complex systems, such as the IUWS_MP model library (which acronym IUWS stands for Integrated Urban Wastewater System), do not consider fate processes and fractions that are typical for PhACs. This limitation was overcome by extending the existing IUWS_MP model library with new fractions (conjugated metabolites, sequestrated fraction) and processes (consumption-excretion, deconjugation). The performance of the extended library was evaluated for five PhACs (carbamazepine, ibuprofen, diclofenac, paracetamol, furosemide) in two different integrated urban wastewater systems where measurements were available. Despite data uncertainty and the simplicity of the modelling approach, chosen to minimize data requirements, model prediction uncertainty overlapped with the measurements ranges across both systems, stressing the robustness of the proposed modelling approach. Possible applications of the extended IUWS_MP model library are presented, illustrating how this tool can support urban water managers in reducing environmental impacts from PhACs discharges.

Introduction

Micropollutants released from urban areas into surface waters might pose an important environmental risk (Jobling et al., 1998; Johnson et al., 2017; Gosset et al., 2017; Letsinger and Kay, 2019). Micropollutants are typically discharged from integrated urban wastewater systems (consisting of sources, sewer, WWTP and receiving water body) in both dry and wet weather, i.e. from outlets of wastewater treatment plants (WWTPs) (Luo et al., 2014; Margot et al., 2015; Rogowska et al., 2019), separate storm sewer outlets (e.g. Zgheib et al., 2012; Brudler et al., 2019) and combined sewer overflows (e.g. Gasperi et al., 2012; Launay et al., 2016). Micropollutants have been targeted by existing environmental legislations for almost two decades. For example, the European legislation (Water Framework Directive - European Commission, 2000) and following directives (European Commission, 2008, European Commission, 2013) established Environmental Quality Standards (EQS) for 33 (later extended to 45) priority substances, including herbicides, fungicides, surfactants, metals and polycyclic aromatic hydrocarbons.

Cost-effective minimization strategies need to be implemented to ensure good chemical status in natural waters, and mathematical models are widely accepted tools for assessing the effectiveness of different options in reducing micropollutant emissions and in meeting the desired EQS. Integrated urban wastewater systems are characterized by complex and highly dynamic conditions and processes (e.g., source emissions, WWTP operations), which limit the applicability of simple models based on flow analysis. Control strategies explicitly targeting dynamic emissions (e.g. combined sewer overflows treatment) or actively employing dynamic processes (e.g. real-time control of WWTP) require the application of dynamic models.

For this purpose, a dynamic model library (named IUWS_MP, Vezzaro et al., 2014) was specifically developed during the ScorePP project (Source Control Options for Reducing Emissions of Priority Pollutants - 2006–2009). The IUWS_MP model library was designed to support water managers in developing control strategies (e.g. Delli Compagni et al., 2020; Eriksson et al., 2011) by: (i) simulating micropollutants fluxes across the integrated urban wastewater systems, (ii) supporting the development of monitoring campaigns, and (iii) evaluating the compliance of different strategies with EQS and/or with risks for the environment and human health.

In the IUWS_MP model library state-of-the-art models, capable of simulating traditional pollutants are coupled with micropollutant fate processes. The fate of a specific micropollutant is predicted based model parameters such as inherent chemical-physical properties (e.g. sorption, volatilization) and first-order kinetics rates (to describe e.g. biodegradation), which can easily be retrieved from existing databases and scientific literature, with an approach derived from chemical fate assessment that acknowledges a general lack of measurements. This minimizes the need for parameter estimation and allows simulating the fate of a wide range of micropollutants. However, the ranges of parameters found in literature show high variability, stressing the need for propagating such uncertainty to model outputs in order to ensure robust results. So far parameter uncertainty has been investigated in few studies, focusing on single elements of the integrated urban wastewater systems (Vezzaro et al., 2011) or on a single micropollutant (Mannina et al., 2017), while studies investigating multiple micropollutants are still missing.

Recently, a new range of micropollutants has recently come into focus (the EU “watch list”, European Commission, 2015), i.e. pharmaceutically active compounds (PhACs) such as anti-inflammatory substances, antibiotics and hormones. PhACs are ubiquitously consumed chemicals and can have harmful effects on natural ecosystems as a result of their biological activity at low concentrations (Bottoni et al., 2010). PhACs are characterized by different release pathways and behaviours compared to the micropollutants that were originally considered in the IUWS_MP model library. For example, PhACs can be excreted as unchanged molecules (also known as parent substances) and metabolites, and through feces and/or urine. These characteristics lead to high variability in wastewater concentrations (Luo et al., 2014; Margot et al., 2015). Different studies included different PhACs release models in simple mass balance calculations (Letsinger and Kay, 2019; Ashton et al., 2004), or in more sophisticated models (Khan and Ongerth, 2004; Coutu et al., 2016). Dynamic PhAC release patterns have been simulated for sewer systems (Pouzol et al., 2020), but investigations covering all the elements of an integrated urban wastewater system are still missing. Moreover, these models conceptualised the PhAC parent mass excreted through urine and feces as one single flux, although the flux excreted through feces might be considered as sequestered and thereby less bioavailable (i.e. less subject to sorption/desorption equilibria and biodegradation processes) (Gonzalez-Gil et al., 2018). Additionally, PhACs are often ionisable substances and can undergo retransformation processes, such as deconjugation (Plósz et al., 2012). These processes can play an important role for the PhACs fate since the ionisable status (i.e. positive, negative or both) affects PhACs sorption affinity with solid matrices (raw wastewater solids, sludge, sediments), and retransformation of certain metabolites (i.e. glucuronide conjugates) can affect the fate of the parent fraction (Stadler et al., 2012). During the last decades, several monitoring activities have been carried out in integrated urban wastewater systems, targeting a wide range of PhACs (Kasprzyk-Hordern et al., 2009; Zuccato et al., 2010; Patrolecco et al., 2015; Castiglioni et al., 2018). Compared to previous modelling studies (De Keyser et al., 2010; Keller et al., 2007; Schowanek et al., 2001; Vezzaro et al., 2014; Williams et al., 2009), data availability has thus improved, allowing for a more comprehensive evaluation of model performance against measured data.

This work presents: (i) the extension of the IUWS_MP model library with additional release pathways (a consumption-excretion model was developed) and fate processes (deconjugation), allowing for simulation of PhACs, (ii) the quantification of the model results uncertainty deriving from the high variability of PhACs model parameters and; (iii) the verification of the prediction capabilities of the extended IUWS_MP model library against measurements from two different integrated urban wastewater systems.