Identification and Removal of Performance- and Health-Based Indicator Chemicals in a Mobile, Carbon-Based Direct Potable Reuse Pilot

1. INTRODUCTION
Potable reuse with reverse osmosis (RO) presents brine disposal challenges for inland communities. Carbon-based advanced treatment (CBAT) eliminates these brine disposal challenges. CBAT trains often include combinations of the following processes: ozonation, biologically activate carbon (BAC) filtration, granular activated carbon (GAC), coagulation, microfiltration (MF)/ultrafiltration (UF), UV disinfection, and/or UV advanced oxidation processes (UV/AOP). Much research has been done on chemical removal by the processes used in CBAT. Nevertheless, demonstration projects continue to provide meaningful information about new or overlooked chemicals and their occurrence or removal in wastewater effluent. The concentrations of chemical pollutants vary between wastewater treatment plants (WWTPs) based on industrial point sources or different consumer patterns. WWTPs also have site-specific differences in water quality parameters that impact treatment efficiency [e.g., nitrite consuming the ozone dose or total organic carbon (TOC) competing with trace organics for sorption sites on GAC]. Thus, piloting is necessary at planned reuse sites to (1) demonstrate the safety of reuse to the public and regulators, (2) fine-tune process design, and (3) identify key site-specific health- and performance-based indicators. In 2020, Colorado Springs Utilities (CSU), the Colorado School of Mines (CSM), and Carollo Engineers, Inc. collaborated to design and build a mobile Direct Potable Reuse (DPR) demonstration trailer. The PureWater Colorado demonstration trailer has served multiple purposes: public engagement, research, piloting for Colorado utilities considering reuse, and informing state-level reuse policy. The CBAT process in the trailer includes ozonation, BAC, GAC, MF, UV/H2O2, and chlorination. PureWater Colorado was commissioned in June 2021. Routine sampling is ongoing at multiple points through the multi-barrier, CBAT process. Several evaluations have been conducted as part of this project, but this presentation will focus on the identification and removal of health- and performance-based indicators.
2. METHODS Health-based indicators are defined as specific chemicals that occur at concentrations with human health relevance at the influent to the reuse system-regardless of their difficulty to remove or federal regulatory status (Drewes et al. 2018; Anderson et al. 2010). Health-based indicators specific to the CSU installation were selected using the procedure recommended by a Science Advisory Panel to the state of California to select which trace organics to monitor in potable reuse. Maximum concentrations of each chemical at the pilot influent were divided by Monitoring Threshold Levels (MTLs). These MTLs are based on drinking water guidelines or toxicity information from a variety of sources including the California Notification Levels and the Australian reuse guidelines. Since the focus of the above Science Advisory Panel was on organics, concentrations of metals and inorganics were divided by their USEPA Primary Maximum Contaminant Levels (MCLs). Chemicals with a maximum measured value to MTL ratio (max/MTL) would be necessary to remove to protect the public health, either through treatment or source control. Chemicals with max/MTL greater than 0.1 would be recommended for removal to maintain a desirable margin of safety. Performance-based indicators are defined as specific chemicals that are useful for monitoring whether a treatment process is performing as designed-regardless of their toxicity (NWRI 2019; Thompson and Dickenson 2020). Performance-based indicators specific to CSU were selected using the procedure recommended by a National Water Research Institute (NWRI) Independent Expert Advisory Panel to the state of Colorado. Chemicals were identified that would have 'sweet spot' moderate removal for each process. Ideally, the chemical would still be detectable after the process when the process is functioning properly, but the chemical's removal would drop noticeably during a malfunction. Among chemicals that met these criteria, the one with highest median concentration divided by its Method Reporting Limit (MRL) was selected, since this chemical could be most precisely quantified.
3. RESULTS Three health-based indicators with max/MTL > 1 were identified in the CBAT influent at CSU: the industrial chemical quinoline, the antibiotic amoxicillin, and the perfluoroalkyl substance (PFAS) perfluorooctanoic acid (PFOA) (Table 1). Fortunately, these three chemicals have so far been removed to below detection by the combination of treatment process, though this should continue to be monitored as the adsorptive capacity of the GAC diminishes over time. All measured chemicals were below their MTL or MCL in the CBAT purified water, through nitrate was closest to its limit. This demonstrates the importance of effective, consistent nutrient removal upstream of reuse (Hill et al. 2018). Based on the criteria described in Methods, the pharmaceutical primidone or the pesticide diuron would be suitable performance-based indicators for ozonation at CSU. Primidone would also be the best performance-based indicator for GAC (Figure 1). While no organic chemicals were consistently detected after GAC as of the time of this writing, N-nitrosodimethylamine (NDMA) could be an appropriate performance-based indicator for UV/H2O2, based on (1) its detection after BAF (Figure 1), (2) its known rapid breakthrough in GAC, (3) its resistance to oxidation, and (4) its moderately high removal by UV (Glover et al. 2019; Thompson and Dickenson 2020).
4. CONCLUSIONS AND NEXT STEPS Highest priority health-based indicators and ideal performance-based indicators identified in this study differed from those identified in technical reports or peer-reviewed literature conducted at other locations. For example, Drewes et al. (2018) recommended health-based indicators NDMA, NMOR, and 1,4-dioxane, and performance-based indicators gemfibrozil, sulfamethoxazole, iohexol, and sucralose. This exemplifies the importance of site-specific piloting and indicator selection. Fundamental to the mobile aspect of this project is that when piloting is complete at CSU, the PureWater Colorado trailer will become available for other Colorado utilities considering potable reuse.