key: cord-0759155-75qq7e3t
authors: Hart, Olga E.; Halden, Rolf U.
title: Computational analysis of SARS-CoV-2/COVID-19 surveillance by wastewater-based epidemiology locally and globally: Feasibility, economy, opportunities and challenges
date: 2020-04-22
journal: Sci Total Environ
DOI: 10.1016/j.scitotenv.2020.138875
sha: 365139f7c4fa3b5b5719377854b6d36df6083b07
doc_id: 759155
cord_uid: 75qq7e3t

Abstract With the economic and practical limits of medical screening for SARS-CoV-2/COVID-19 coming sharply into focus worldwide, scientists are turning now to wastewater-based epidemiology (WBE) as a potential tool for assessing and managing the pandemic. We employed computational analysis and modeling to examine the feasibility, economy, opportunities and challenges of enumerating active coronavirus infections locally and globally using WBE. Depending on local conditions, detection in community wastewater of one symptomatic/asymptomatic infected case per 100 to 2,000,000 non-infected people is theoretically feasible, with some practical successes now being reported from around the world. Computer simulations for past, present and emerging epidemic hotspots (e.g., Wuhan, Milan, Madrid, New York City, Teheran, Seattle, Detroit and New Orleans) identified temperature, average in-sewer travel time and per-capita water use as key variables. WBE surveillance of populations is shown to be orders of magnitude cheaper and faster than clinical screening, yet cannot fully replace it. Cost savings worldwide for one-time national surveillance campaigns are estimated to be in the million to billion US dollar range (US$), depending on a nation's population size and number of testing rounds conducted. For resource poor regions and nations, WBE may represent the only viable means of effective surveillance. Important limitations of WBE rest with its inability to identify individuals and to pinpoint their specific locations. Not compensating for temperature effects renders WBE data vulnerable to severe under-/over-estimation of infected cases. Effective surveillance may be envisioned as a two-step process in which WBE serves to identify and enumerate infected cases, where after clinical testing then serves to identify infected individuals in WBE-revealed hotspots. Data provided here demonstrate this approach to save money, be broadly applicable worldwide, and potentially aid in precision management of the pandemic, thereby helping to accelerate the global economic recovery that billions of people rely upon for their livelihoods.

Months into the pandemic of coronavirus disease 2019 , it remains a global challenge to identify the presence and spread of the SARS-CoV-2 biohazard. The prolonged incubation time and virus shedding from asymptomatic infected cases have allowed the virus to spread quickly and to avoid medical detection and containment. Current estimates of SARS-CoV-2 occurrence are heavily biased toward regions where medical screening of individuals is under way. In contrast, resource-poor regions remain under-tested and disease occurrence underreported. While some countries are now trying to test every individual (e.g., Iceland) to obtain population-wide data, this approach is impractical, slow, and cost-prohibitive for most nations around the world.

Wastewater-based epidemiology (WBE) has been identified as a population-wide infectious disease surveillance tool featuring a proven track record for polio and hepatitis A (Ashgar et al., 2014; Hellmér et al., 2014) , and holds considerable promise for population-wide surveillance of the COVID-19 pandemic. When first proposed for tracking of SARS-CoV-2, the prevailing scientific opinion was that the virus may be shed into wastewater at insufficiently high rates, and that both the virus itself and its RNA may be too labile to facilitate detection in wastewater. Recent reports of coronavirus shedding in human stool (Gao et al., 2020; Holshue et al., 2020; Jiehao et al., 2020; Tang et al., 2020; Wölfel et al., 2020; Zhang et al., 2020a; 2020c; 2020d; 2020e) and three preliminary reports of successful SARS-CoV-2 detection in municipal wastewater from the Netherlands the United States and Australia have dispelled some of these concerns (Ahmed 2020, Lodder 2020, Medema et al. 2020 , Wu et al. 2020 ). Yet, considerable uncertainty remains as to what information may be gleaned from monitoring for SARS-CoV-2 RNA in wastewater and whether a WBE assay, once perfected and shown to be J o u r n a l P r e -p r o o f 5 reproducible across different laboratories, will be sensitive enough to inform public health responses. Better population-wide data could aid in reducing the economic damage and social burden placed on populations dealing with stay-at-home ordinances, furlough and involuntary unemployment. At present, public health interventions are implemented with a broad brush; potentially excluding communities that would benefit from them and putting a burden on areas where the virus may currently not pose a threat, thereby rendering hardship-inducing containment measures not only ineffective but also economically and socially disruptive.

The present study combined consideration of economy, scaling, practicability, and data analysis, with a particular focus on wastewater temperature as an underappreciated source of data bias in WBE when performing population-wide screening for SARS-CoV-2.

The SARS-CoV-2 load to municipal wastewater was estimated using excretion rates in human stool recently reported by Zhang et al., (2020e) and Wölfel et al. (2020) , and assuming a fecal load in the range of 100 -400 g feces/day/person, and a fecal density of 1.06 g/mL (Brown and Butler, 1996) .

The degradation over time of a biomarker of interest present in wastewater can be expected to follow exponential decay, described by the formula as: where ( ) is the quantity that still remains and has not yet decayed after a time (i.e., the amount measured by the sampling campaign); 0 is the initial quantity of the substance that was excreted and discharged into the wastewater collection system; 1/2 is the half-life of the biomarker, and is the time elapsed between the time of excretion (time = 0) and time of observation/sample collection (time = ).

The adjusted biomarker half-lives were based on the calculated wastewater temperature, a series of initial biomarker half-lives reported at ambient temperatures, and the Arrhenius Equation as shown in Equation 2:

where 1 is the initial decay rate, equal to the negative natural log of two divided by the initial reported half-life (Laidler 1984) . When solving for the half-live, Equation 3 is obtained:

(2)× 10

( 2− 1 10℃ ⁄ )

Where 1 2 ,1 is the initial half-life, 1 is the temperature at which initial half-life was derived, 1 2 ,2 is the half-life at seasonally-and spatially-adjusted wastewater temperature calculated in this study, 2 is the calculated temperature to which initial half-life is adjusted to, and 10 is a factor of temperature-dependent of rate change, ranging between 2 and 3 for most biologic systems (Hart and Halden, 2020a) .

This study focused on the City of Tempe, Arizona, USA. The city covers an area of 104 square kilometers and is land-locked by neighboring cities comprising the metropolitan Phoenix people per square kilometer. Land use is predominantly residential, with some industrial and commercial activity.

Data related to the physical layout of the wastewater collection system was obtained from the City of Tempe Water Utilities Department. Wastewater loading was estimated using historic wastewater meter data to derive a per capita wastewater loading rate. Population density estimates were based on Maricopa Association of Governor's Traffic Analysis Zones, with residential wastewater loads assigned proportionally to population density and maintenance access hole (node) count. Industrial wastewater loads were assigned to the collection system node nearest to the industrial facility, with average flow rates and diurnal curves based on meter data. A previously employed US EPA SWMM model was set up to simulate a 72-hour period representing typical weekday conditions subject to dry weather flows only. No leakage or infiltration were incorporated into the hydraulic model.

Hydraulic modeling to calculate in-sewer travel time of wastewater in the collection system, volumetric wastewater flow rates, and velocities was performed using the U.S. 

The per-assay costs of clinical and WBE screening may vary greatly among geospatial regions and around the world due to differing labor costs, safety requirements, existing infrastructure, etc. In order to obtain a more robust and geographically scalable cost estimate, the 

Where Cost P,clinical is the total cost in USD for reagents needed to test a population of the size P,

and Cost P,WBE is the total cost in USD for reagents required to screen community wastewater produced by population P in N wastewater treatment plants (N WWTPS ).

J o u r n a l P r e -p r o o f 9

The SARS-CoV-2 load to municipal wastewater is estimated to be bracketed by the lower and upper bounds estimate of 56.6 million to 11.3 billion viral genomes per infected person per day. This mass load translates into concentrations of 0.15 to 141.5 million viral genomes per liter of wastewater generated in North America and Europe. This is based on the presence of between a reported 600,000 (Zhang et al., 2020e) to 30,000,000 (Wölfel et al., 2020 ) viral genomes of SARS-CoV-2 per mL of fecal material, and assuming a fecal load of 100 -400 g feces/day/person with a density of 1.06 g/mL (Brown and Butler, 1996) . A further refinement of the lower and upper bounds estimate is desirable and will be informed by future studies providing more comprehensive information on virus shedding by symptomatic and asymptomatic infected individuals.

Assuming a detection limit of 10 coronavirus RNA genomes per mL sewage, and further assuming absence of additional stormwater, commercial, and industrial flow inputs to the sewer system, successful detection of SARS-CoV-2 by qRT PCR in fully homogenized wastewater will require at worst as much as 0.88% of the population in a monitored sewershed to be infected (1 in 114 individuals) and at best, as few as 0.00005% (1 infected case in about 2 million noninfected individuals). This implies that the practical limit of detection of SARS-CoV-2 in community wastewater is well within the useful range and potentially superior to the alternative approach of randomly testing of 100 to 2 million people to establish presence or absence of symptomatic or asymptomatic cases in a population of interest. The relationship between virus concentration, water use and SARS-CoV-2 detectability is shown in Figure 1 (Table 1) . These values are in the same order of magnitude as the typical hydraulic retention times of municipal wastewaters in most sewerage systems around the world. Thus, in most situations in conditions where wastewater flow is at a temperature of 20°C, at least 25% of the virus load should still remain even in situations where the average in-sewer travel time is long (e.g., 10 h) and the virus stability is relatively low (t 0.5 = 4.8 h).

In the absence of seasonal temperature fluctuation, a biomarker discharged into the wastewater collection system at a constant rate would degrade at a rate proportional to its residence time in the system and arrive available for observation at some downstream sampling location at a seasonally invariant rate. However, in actuality, wastewater temperature varies seasonally, and its modulation over the course of a year differs around the world (Hart and Halden, 2020a) . Seasonal changes in air and soil temperature affect the transfer of heat between wastewater and the surrounding environment ( Figure 2, panel A) . Thus, temperature-adjusted degradation of such a biomarker will not be constant over a year, and instead may vary significantly as shown here for a use case, the city of Tempe, AZ, USA (Figure 2, panel B) . As a result, the same hypothetical constant loading deposited into the sewer upstream will result in different masses available for observation at the downstream monitoring location (Figure 2 , panel C). All else held constant, the degree to which the same seasonality in wastewater temperature will affect downstream observations will increase with increasing in-sewer travel timesi.e., it will be magnified for outfalls serving larger sewersheds, and be more subtle in locations serving smaller sewersheds. We recently showed that failing to account for the role that rendering temperature a marginal factor overall. However, as WBE becomes more common and is being practiced long-term across the changing seasons and differing climate regions, its potentially pronounced effects will have to be taken into consideration in order to obtain robust data and to inform selection of an appropriate public health response.

In prior work, we introduced a deterministic model for computing wastewater temperatures at any location globally (Hart and Halden, 2020a) . Here, we applied this model to examine annual changes in wastewater in locations representing past, present and emerging hot zones of the COVID-19 pandemic; for illustrative purposes, we also include the study use case, the city of Tempe, that thus far has reported relatively fewer cases and a low frequency of detection of SARS-CoV-2 in wastewater (unpublished data). 

The coming onset of warm weather (April 2020) and increasing wastewater temperatures 

Despite the relative longevity of the genetic material of SARS-CoV-2 in wastewater and a low estimated limit of detection by qRT PCR, seasonal changes in wastewater temperature result in a changing population focus for WBE observations over the course of a year. Using the City of Tempe as a case study, the pronounced difference in the composition of a sewershed outfall's sample makeup during the winter and summer months is further illustrated in Figure 5 .

Here contracting and expanding effective area observed had been described previously for chemical markers by our team (Hart and Halden, 2020a,b) .

Even in countries like Germany, where the testing capacity is highest in Europe at about 100,000 clinical assays per day (NYT, 2020), almost 3 months of non-stop testing would be required to assess the infection status of the entire 83 million population just once. This implies that comprehensive screening of a nation's population using individual test kits is not only expensive (approximately $1.25 billion USD for Germany for assay reagents only) but also impractically slow (3 months, during which major changes can occur in the prevalence of the virus in the Arizona State University, this shared resource Halden et al. 2015; Venkatesan and Halden, 2014) was created over the course of a dozen years primarily with discretionary funds from philanthropists and support from partnering municipalities (Gushgari et al. 2018 (Gushgari et al. , 2019 Chen et al. 2019; Driver et al, 2020) . Among the latter, the city of Tempe, AZ stands out as an innovator in WBE data communication, by immediately releasing obtained wastewater data on locally consumed licit and illicit opioids via an online dashboard to inform the public, emergency response teams, and policy makers (ASU-Tempe, 2020).

As is true for any modeling study, output data are only as good as the input information provided. For the present work, a number of areas would benefit from additional experiments and the parameterization of phenomena important for understanding virus detectability.

Improvements would be welcome in lowering the detection limit and in increasing the precision of quantitative data, possibly achievable via use of digital PCR (Majumdar et al. 2015) . Whereas processing of larger sample volumes and more effective concentration of virus particles and viral RNA may help to lower method detection limits, there also is a desire to keep required samples sizes low enough to facilitate economic shipping between sampling location and the analytical laboratory.

Finally, it is important to note that the simulations conducted here are exclusively directed at the detection of the occurrence of the virus and the number of virus particles per unit wastewater.

The data provided here should not be interpreted as providing information on the presence of intact, infective virus particles. Data on the infectivity of sewage-borne SARS-CoV-2 is urgently needed but difficult to obtain, due to the temporary closure of many research laboratories and the J o u r n a l P r e -p r o o f 18 need for biosafety level certification in excess of Level 2 to conduct this type of work. Such work also should address a study on the decay and detectability of SARS-CoV-2 as a function of temperature.

This computational modeling study and cost analysis served to identify WBE as a rapid, ; B). Assuming shedding of a steady virome load into the sewer (100M genomes/d), the detectable signal at a city's monitoring location (here Tempe, AZ, USA) will change significantly over the course of a year (C). This is due to temperature-induced variability in the time-to-depletion of the virus below the assumed detection limit of 10 genomes/mL (D). . Tempe, AZ undergoes extreme seasonal temperature changes which are predicted to impact SARS-CoV-2 detectability as illustrated by computer simulations. The city has been monitoring wastewater for opioids since May 2018, first in 3 and currently in 5 areas, and recently started monitoring for COVID-19 with the intent of adding this new data source to an existing public health online dashboard to which wastewater analysis data get posted in real-time (https://arcg.is/ey0Ha).

J o u r n a l P r e -p r o o f 

Environmental surveillance for polioviruses in the global polio eradication initiative

World's first online dashboard for immediate public release of information collected by wastewater-based epidemiology

Theoretical evaluation of using wastewater-based epidemiology to assess the nutritional status of human populations

Gross solids transport in small diameter sewers

Alcohol and nicotine consumption trends in three U.S. communities determined by wastewater-based epidemiology

Inactivation of an Enveloped Surrogate Virus in Human Sewage

Survival of surrogate coronaviruses in water

Longitudinal study of alcohol, nicotine, and caffeine consumption on a public university campus via wastewater-based epidemiology

Monitoring of regional drug abuse through wastewater-based epidemiology-A critical review

2019 Novel coronavirus infection and gastrointestinal tract

Long-term Tracking of Opioid Consumption in Two United States Cities Using Wastewater-based Epidemiology

Tracking Narcotics Consumption at a Southwestern U.S. University Campus by Wastewater-based Epidemiology

Survival of coronaviruses in water and wastewater

Tracking harmful chemicals and pathogens using the Human Health Observatory at ASU

Modeling wastewater temperature and attenuation of sewage-borne biomarkers globally

Simulated 2017 nationwide sampling at 13,940 major U.S. sewage treatment plants to assess seasonal population bias in wastewater-based epidemiology

NetSTORM-A Computer Program for Rainfall-Runoff Simulation and Precipitation Analysis

Detection of pathogenic viruses in sewage provided early warnings of hepatitis A virus and norovirus outbreaks

First case of 2019 novel coronavirus in the United States

A case series of children with 2019 novel coronavirus infection: clinical and epidemiological features

Measuring spatial and temporal trends of nicotine and alcohol consumption in Australia using wastewater-based epidemiology

The development of the Arrhenius equation

SARS-CoV-2 in wastewater: potential health risk, but also data source

Digital PCR Modeling for Maximal Sensitivity, Dynamic Range and Measurement Precision

Can a Paper-Based Device Trace COVID-19 Sources with Wastewater-Based Epidemiology?

Presence of SARS-Coronavirus-2 in sewage

Storm Water Management Model User's Manual Version 5.1. Office of Research and Development. Water Supply and Water Resources Division. US Environmental Protection Agency

Global, regional, and country level need for data on wastewater generation, treatment, and use

SCORE -Sewage Analysis CORe group Europe COST Action ES1307

Detection of Novel Coronavirus by RT-PCR in Stool Specimen from Asymptomatic Child

Wastewater Treatment Plants as Chemical Observatories to Forecast Ecological and Human Health Risks of Manmade Chemicals

United States National Sewage Sludge Repository at Arizona State University -A New Resource and Research Tool for Environmental Scientists, Engineers, and Epidemiologists

Virological assessment of hospitalized patients with COVID-2019. Nature 1-10

Fecal specimen diagnosis 2019 novel coronavirus-infected pneumonia

Virus shedding patterns in nasopharyngeal and fecal specimens of COVID-19 patients

Molecular and serological investigation of 2019-nCoV infected patients: implication of multiple shedding routes

Isolation of 2019-nCoV from a Stool Specimen of a LaboratoryConfirmed Case of the Coronavirus Disease

Virus shedding patterns in nasopharyngeal and fecal specimens of COVID-19 patients. medRxiv, in press

The authors wish to thank Mayor Mark Michell, Rosa Inchausti, Wydale Holmes, Phillip Brown, Stephanie Deitrick, Andrea Glass, Darrell Duty, Greg Ruiz, and the many other friendly personnel and staff members of the City of Tempe for their assistance with the WBE case study.We also thank the institutional support of Arizona State University and our sponsors that have and continue to make possible the ongoing free-of-charge, robotic screening of both clinical samples and wastewater process samples in our laboratories at the ASU Biodesign Institute in Tempe, AZ.

☐ The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.☒The authors declare the following financial interests/personal relationships which may be considered as potential competing interests:Dr. Halden is co-founder of AquaVitas, LLC, an Arizona State University (ASU) startup company that works in the intellectual space touched upon by this study. Dr. Halden further is founder of ASU Foundation's OneWaterOneHealth, a nonprofit project providing wastewater-based health assessments to underserved US communities.