Reliability of pathogen control in direct potable reuse: Performance evaluation and QMRA of a full-scale 1 MGD...

To safely progress toward direct potable reuse (DPR), it is essential to ensure that DPR systems can provide public health protection equivalent to or greater than that of conventional drinking water sources. This study collected data over a one-year period from a full-scale DPR demonstration facility, and used both performance distribution functions (PDFs) and quantitative microbial risk assessment (QMRA) to define and evaluate the reliability of the advanced water treatment facility (AWTF). The AWTF's ability to control enterovirus, Giardia, and Cryptosporidium was characterized using online monitoring of surrogates in a treatment train consisting of ozone, biological activated carbon, microfiltration, reverse osmosis, and ultraviolet light with an advanced oxidation process. This process train was selected to improve reliability by providing redundancy, defined as the provision of treatment beyond the minimum needed to meet regulatory requirements. The PDFs demonstrated treatment that consistently exceeded the 12/10/10-log thresholds for virus, Giardia, and Cryptosporidium, as currently required for potable reuse in California (via groundwater recharge and surface water augmentation). Because no critical process failures impacted pathogen removal performance during the yearlong testing, hypothetical failures were incorporated into the analysis to understand the benefit of treatment redundancy on performance. Each unit process was modeled with a single failure per year lasting four different failure durations: 15 min, 60 min, 8 h, and 24 h. QMRA was used to quantify the impact of failures on pathogen risk. The median annual risk of infection for Cryptosporidium was 4.9 × 10^-11 in the absence of failures, and reached a maximum of 1.1 × 10^-5 assuming one 24-h failure per process per year. With the inclusion of free chlorine disinfection as part of the treatment process, enterovirus had a median annual infection risk of 1.5 × 10^-14 (no failures) and a maximum annual value of 2.1 × 10^-5 (assuming one 24-h failure per year). Even with conservative failure assumptions, pathogen risk from this treatment train remains below the risk targets for both the U.S. (10^-4 infections/person/year) and the WHO (approximately 10^-3 infections/person/year, equivalent to 10^-6 DALY/person/year), demonstrating the value of a failure prevention strategy based on treatment redundancy.