WBT Sampling Strategy

An essential component of this study will be the evaluation and new implementation of methods to collect wastewater samples that maximize both the probability of environmental viral detection and the ability for correlation with both hospital patient data and human community surveillance data. Our methods evaluation is critical because of at least three competing processes. The first is that there are several sources of water to the sewage system and not all these sources are suspected to have SARS-CoV-2. Wastewater (sanitary sewage carried by a network of pipes from a building) includes blackwater (from toilets and urinals) and greywater (from dishwashing, laundering, hand washing, bathing, showering). Sanitary sewer systems should not be connected to stormwater, although it is recognized that even in properly designed systems, infiltration and inflow occur depending upon weather and groundwater conditions. Given the different sources of water within a building and within the sewage collection network, the relative proportion of blackwater, which would carry the bulk of the SARS-CoV-2 RNA, will vary depending upon the activities and time of day within the building that would also generate greywater and the inherent infiltration and inflow to the system. To account for this variation in wastewater quality, samples will be normalized by physical, chemical, and microbiological markers that correlate with blackwater (e.g., fecal indicator bacteria). Normalization of sewage contaminants against other more readily measured physical-chemical parameters has been shown to work well for studies focused on using WBE to document illicit drug use (7).

The second competing process is that the signal (the viral RNA) decays upon release from an infected individual and as it travels through the sewage network. Upon fecal or urinary release of SARS-CoV-2 RNA from an infected individual, the viral particle is subjected to environmental processes that may accelerate its decay (20, 24, 30). The persistence of SARS-CoV-2 in wastewater is unknown, especially under natural conditions. Studies based upon other viruses have found that survival depends upon temperature, humidity, pH, and salinity (37, 44), wherein general, a higher temperature is correlated with shorter survival due to denaturation of proteins and extracellular enzymes (18, 23). In this proposal, we will be controlling for the basic physical and chemical parameters (e.g., temperature, etc.) to document degradation rates using encapsulated viral controls with unique mutations (Zeptometrix), and we will integrate this information with sewage travel times to develop a model that estimates the number of human infection cases that contribute to a location in the sewage system.

The third competing process is associated with watershed scale, with watershed defined as the network of buildings contributing to a point in the collection network. In general, the variability is larger for smaller watersheds in comparison to larger watersheds due to the averaging effects of aggregating contributions. The variability influences how wastewater samples are collected. Sample collection approaches can be broadly separated into two categories: (a) grab samples which are collected from the wastewater system at a moment in time, and (b) composite samples which are a mixture of grab samples collected at the same location but at different points in time. Composite samples are usually used to average out variations in sample composition that would be expected over time. The traditional approach has been to collect 24, one-hour composite samples over the period of 1 day to average the variability in wastewater quality. However, it is recognized that the SARS-CoV-2 signal degrades during the 24-hour holding period needed for compositing. As such within SF-RAD, we will be targeting three watershed scales: the county scale (CDWWTP), the community scale (UM campus sewage networks), and the building scale (individual buildings within UM). SF-RAD will evaluate the competing processes (proportion of blackwater contribution, viral signal decay, and watershed scale) to identify optimum strategies for collecting samples in terms of when grab samples should be used and when composites (and their time averaging period) should be used to maximize the SARS-CoV-2 signal.

Hypotheses: We hypothesize that the optimum method chosen to collect samples (grab versus composite, and how to composite the samples) for tracking human surveillance data will be dependent upon the holding time and population contributing to the sewage sample. We further hypothesize that because of holding time effects, correlations with human health surveillance data will be improved by normalizing SARS-CoV-2 measures in sewage against readily-measurable physical-chemical (e.g., conductivity, turbidity, temperature) and microbiological parameters (fecal indicator microbes by culture and qPCR). The outcome of this work will create practical knowledge of optimal time for the collection of grab samples and when/how composite samples should be processed to optimize the correlation between the viral signal and human health surveillance results.