Indigenous microbial surrogates in wastewater used to understand public health risk expressed in the Disability-Adjusted Life Year (DALY) metric

Large scale water recycling schemes are increasing in importance. Climate change and urbanisation are placing pressure on the continued ability to meet regional water demand through conventional water sources such as lakes and aquifers¹. Similarly, the global challenges posed by inadequate water supply, sanitation, and hygiene highlight the need to increase safe water recycling practices².

Various forms of water recycling are taking place in Australia³. The largest scale applications involve domestic wastewater contributed by the public to the sewer. When treated to a high standard, recycled water can be safely used by industry and domestic users. There are many Australian schemes where recycled water is supplied through a network that is entirely separate to the drinking water supply (Figure 1), such as the Rouse Hill scheme in north-west Sydney⁴. Such schemes offset the use of the conventional water sources for non-drinking uses with relatively low human exposure, including garden watering and toilet flushing. There are also schemes that are used to safely augment a drinking water supply, such as the Beenyup scheme in Perth, which injects highly treated water into the water table^5,6. Doing so replenishes the groundwater that is later treated for drinking.

Figure 1.  Recycled water used for non-drinking purposes in the Greater Sydney region of Australia.

In the Australian Guidelines for Water Recycling, EPHC-NRMMC-AHMC³ set out a framework for the effective and reliable management of safety of water recycling schemes. The approach is formalised in a 12-element recycled water safety plan. The preparation and implementation of a water safety plan is a key regulatory requirement of water supply authorities across Australia. Similar requirements apply in many international jurisdictions.

In formulating the water safety plan, much attention is paid to the risk posed by enteric pathogens in the untreated source water. This risk defines the types of treatment processes required and how well these processes must perform to protect public health. In Australia, the level of pathogens in recycled water deemed to pose a negligible risk is based on a specified public health measurement (referred to as a ‘health-based target’). The target is an upper limit of 10⁻⁶ DALY person per person per year, and it was adopted by the National Health and Medical Research Council over a decade ago³.

When considering the application of the DALY metric to recycled water exposure scenarios, a method is required for translating pathogen levels into public health risk. Quantitative microbial risk assessment (QMRA) is the recognised process to do this^7,8. It models the nature of exposure, associated health effects, and other relevant factors across a theoretical reference population (Figure 2). Results are then compared to a targeted tolerable (negligible) reference level.

Figure 2.  Quantitative microbial risk assessment process.

QMRA has been formally embedded in water recycling guidelines in Australia since 2006³. It has a much longer history of being used to estimate health risks in drinking water supplies^9,10. However, novel developments in the approach to implementation are not often presented in literature. The next section of this paper provides a case study of adaptive implementation of the guidelines in the Greater Sydney region of Australia. The approach represents the further development of the method discussed in Microbiology Australia by Cox et al.⁴.

The case study focuses on a water recycling scheme that supplies to commercial and local government entities. The water is used for purposes categorised as ‘municipal irrigation’, as defined by EPHC–NRMMC–AHMC³. The frequency and volume of public exposure associated with this use are reflected in the required level of pathogen reduction.

The level of pathogen reduction by the wastewater treatment processes was estimated based on monitoring indigenous surrogate organisms representing three major pathogen groups (Table 1) across the treatment process during typical operating conditions. The surrogates are understood to have similar removal characteristics to their represented pathogens but are generally of higher concentration in domestic wastewater^3,11,12. The use of surrogates was warranted due to the levels of pathogens in wastewater being too low to fully measure the high level of reduction required. This reflects the stringency of public health requirements and the analytical limitations of detecting pathogens in low concentrations in treated water. The more common approach of challenge-testing the treatment processes was not feasible due to the potential to disrupt the co-located wastewater disposal processes. Similarly, testing during commissioning did not occur due to the original purpose of the plant being for environmental discharge rather than for water recycling.

Table 1.  Surrogate organisms monitored to understand the fate of each major pathogen group.

At the studied water recycling plant, located in Greater Sydney, the microbial surrogates were monitored at three phases: before treatment, following the biological digestion process, and following the chlorination process. A total of 27 samples were analysed over two campaigns, from April 2015 to May 2016 (n = 14) and from April to June 2019 (n = 13). The level of reduction, expressed in terms of the ‘logarithmic reduction value’ (LRV), was calculated for each treatment process and each pathogen group as log₁₀(C_in/C_out) (Table 2). C_in was the level of surrogate before a treatment process (influent) and C_out was the level after (effluent). C_in and C_out were paired by the date of sampling. Only the results from the latter monitoring campaign were used to assess protozoal removal due to the recent improvement of treatment performance.

Table 2.  Summary statistics of the daily log-reduction.

Three major steps were involved in determining whether the health-based target was met (Table 3). First, an LRV for each major pathogen group was estimated for the water recycling plant based on the surrogate monitoring results. Because the lower range of performance was of interest, the fifth percentile statistic of the LRV results was used in the subsequent assessment stages (Table 2). Use of the fifth percentile is conservative in comparison to the more common approach of using an average LRV.

Table 3.  Aggregated assessment of pathogen reduction requirements.

The second step was the incorporation of default LRVs associated with end-use limitations, as guided by EPHC-NRMMC-AHMC³ (Table 3). These limitations seek to reduce exposure of the public to recycled water. They are codified as management practices in the water safety plan.

The final step involved the comparison of the aggregated LRVs to the default LRV requirements (Table 3). The default LRV requirements were taken from the draft revised Australian Guidelines for Water Recycling¹³, rather than the current version, for two reasons. First, the revised guidance recognises the decreased incidence and burden of rotavirus infection following the recent implementation of broad community vaccination^13–15. Second, newer literature relevant to the selection of dose-response models and other assessment assumptions are incorporated¹³. The health-based target was considered met if the aggregated LRV for each pathogen type was greater than or equal to the relevant LRV requirement.

Using the microbial surrogate data and the above method, the recycling scheme was shown to meet the public health requirements for waterborne enteric pathogens (Table 3). The key benefit of the surrogate monitoring approach is its ability to overcome the described limitations inherent to operational water recycling schemes when assessing public health requirements. The approach complements existing methods for the assurance of safe water recycling. Overall, it is crucial that QMRA implementation is subject to ongoing critical evaluation, adapts to new evidence, and incorporates conservative assumptions where appropriate.

The authors declare no conflicts of interest.

This research did not receive any specific funding.