key: cord-0991073-ycdok8fc
authors: Shutler, J.; Zaraska, K.; Holding, T. M.; Machnik, M.; Uppuluri, K.; Ashton, I.; Migdal, L.; Dahiya, R.
title: Risk of SARS-CoV-2 infection from contaminated water systems
date: 2020-06-20
journal: nan
DOI: 10.1101/2020.06.17.20133504
sha: cf102821df4f59f56a21db6811bdb0e0d72899d7
doc_id: 991073
cord_uid: ycdok8fc

Following the outbreak of severe acute respiratory syndrome coronavirus (SARS-CoV-2) in China, airborne water droplets (aerosols) have been identified as the main transmission route, although other transmission routes are likely to exist. We quantify SARS-CoV-2 virus survivability within water and the risk of infection posed by faecal contaminated water within 39 countries. We identify that the virus can remain stable within water for up to 25 days, and country specific relative risk of infection posed by faecal contaminated water is related to the environment. Faecal contaminated rivers, waterways and water systems within countries with high infection rates can provide infectious doses >100 copies within 100 ml of water. The implications for freshwater systems, the coastal marine environment and virus resurgence are discussed.

The outbreak of the severe acute respiratory syndrome coronavirus (SARS-CoV-2) began 28

in Wuhan province, China in December 2019 and has now spread throughout the world 29 with about 6 million cases confirmed globally within 214 countries and territories. Water 30 aerosols originating from individuals infected by SARS-CoV-2 are considered a major 31 pathway for infection 1 , and the virus has been shown to remain stable in saline solution 2 32 and under varying environmental conditions 3 . Viral shedding in faeces of viable SARS-33

CoV-2 virus is documented (eg 4 ) and SARS-CoV-2 ribonucleic acid (RNA) has been 34 detected in the shed faeces of both symptomatic and asymptomatic children and adults 35 (eg 5 ); with potentially 43% of infections being asymptomatic and unreported 6 . 36

Human viral pathogens that can be transmitted by water that pose moderate to high health 38 significance as defined by the WHO include adenovirus, astrovirus, hepatitis A and E, 39 rotavirus, norovirus and other enteroviruses. The survival of the large family of 40 coronavirus in water systems has been highlighted 7 , and viral loads within untreated 41 wastewater, consistent with population infection rates, have been identified 8 . While 42 evidence for SARS CoV-2 is limited, other human coronaviruses are documented to 43 survive in wastewater effluent 9 , with colder water temperature likely to increase survival 44 considerably 3 . Collectively this evidence suggests that SARS-CoV-2 virus can survive 45 within water and the viral loads within untreated sewage effluent are likely high in countries 46 of high infection rates, a portion of which is viable virus, and therefore water contaminated 47 

. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted June 20, 2020. . and c) countries where relative risk has been calculated with relative risk as a linear scale; grey signifies a country not included.

. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted June 20, 2020. 

. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted June 20, 2020. virus copies, the proportion of infectious viruses in sewage must be known. The presence 117 of infectious virus in stool samples has been demonstrated 4 , but there is a lack of 118 quantitative data on this ratio for SARS-CoV-2 in stool. We instead used literature on the 119 number of infectious adenovirus copies in sewage (eg 16 ) and wastewater discharge into 120 rivers 17 to select high (10 -1 ) medium (10 -2 ) and low (10 -3 ) estimates for the ratio of 121 infectious virus to genome copies to infectious viruses. We note that adenoviruses are 122 known to be particularly resilient, and therefore likely to represent an upper estimate, but 123 also that our selected range is consistent with the 10 -3 value used elsewhere for assessing 124 viral risk in water systems (eg 14 ), including one assessment for SARS CoV-2 transmission 125 risk to wastewater workers 18 . 126

The temperature dependent survivability means that it is likely that the risk posed by 128 wastewater will increase during winter months as the sewage temperature will be lower 129 enabling longer viral survival, but temperature history and age of the sewage will be 130 needed to fully understand any detected viral loads. SARS-CoV-2 infection to, and spread 131 between, domestic cats has occurred due to similarities between human and some animal 132 angiotensin converting enzyme 2 (ACE2) gene 20 . Increased animal foraging can occur 133 downstream from water treatment facilities, relative to upstream, highlighting possible risk 134 of some riparian wildlife infection if feeding occurs after a spill. 135 136

It is possible that SARS-CoV-2 survivability and transport within rivers could impact 138 drinking water supplies in countries where rivers or reservoirs are the primary drinking 139 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 20, 2020. . https://doi.org/10.1101/2020.06.17.20133504 doi: medRxiv preprint water sources and where large populations, with little or no sewage treatment, exist close 140 to the water source, such as within refugee camps or shanty towns. Riverine enteric virus 141 transport and catchment accumulation can occur for common viruses (eg 21 ) and under 142 stratified conditions it would be possible for a river plume to enter a reservoir and 143 subsequently exit through the reservoir outlet without mixing with the main body of water. 144

Filtering of water, followed by ultraviolet disinfection or chlorination are the recommended 145 approaches for virus removal from drinking water sources 22 . Filtering is normally used to 146 remove large particulates. The effective ultraviolet dose for SARS-CoV-2 disinfection 147 appears highly variable and dependent upon the surface to which the virus is attached 23 . 148

The upper dosage value of 1 Joule (J) cm -2 to ensure effective ultraviolet disinfection of 149 SARS-CoV-2 23 is an order of magnitude larger than that typically used (~40 to 90 mJ cm -150 2 ) for low volume domestic drinking water treatment. The World Health Organization 151 (WHO) guidelines state that effective chlorination disinfection occurs at residual chlorine 152 concentrations of ≥0.5 mg L -1 22 , which matches the minimum needed to deactivate 153 SARS-CoV-1 24 . However, the actual chlorine dosage used for water treatment can vary, 154 based on country, region, water origin and infrastructure (eg UK guidelines are 155 concentrations of 0.2 to 0.5 mg L -1 ). Collectively this means that if a drinking water source 156 was to become infected with SARS-CoV-2 the standard virus removal and disinfection 157 approaches of ultraviolet exposure and chlorination may not reduce the virus below 158 detectable limits. Reviewing of regional or countrywide drinking water processing 159 approaches is recommended to reduce the potential for SARS-CoV-2 surviving through 160 drinking water processing systems. Boiling of drinking water will result in the virus being 161 orca and pilot whales 20 . Of particular concern are whales whose throats are exposed to 177 large volumes of water during feeding and who visit coastlines for prey that are known to 178 accumulate around sewage outfalls, such as minke whales feeding on mackerel or orca 179 feeding on chinook salmon. In these instances, the animal could be exposed to a large 180 viral dose, even if the virus is only present within the water in low concentrations. For 181 example, if the riverine viral concentration is low at 1 copie ml -1 , which is undetectable by 182 PCR (detection limit is >100 copies ml -1 ), then a medium sized whale filtering water during 183 feeding could receive repeated doses of 5.65 million copies every second (see methods 184

for calculation). A seafood market is among the suspected sources for the origin of the 185 SARS-CoV-2 virus, so any viral transmission from land to sea may be a circular process. . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 20, 2020. . https://doi.org/10.1101/2020.06.17.20133504 doi: medRxiv preprint

The relative risk of SARS CoV-2 from waste water systems is calculated by using a 217 modified version of equations 1 and 2 from 13 , given as 218

where V ww,c is the per capital daily volume of domestic water usage for country c, and DF c 220 is the dilution factor downloaded from 13 supplemental table 1 and supplemental table 2 We note that measured wastewater viral counts in Paris on the 9 th April were 3.1 × 10 6 235 genome copies L -1 with 82,000 active cases 19 , whereas using our (albeit country specific) 236 method gives the estimate of 1.3 × 10 6 genome copies L -1 , which is within the correct order 237 of magnitude (this calculation used the same number of active cases). 238 239 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 20, 2020. . https://doi.org/10.1101/2020.06.17.20133504 doi: medRxiv preprint

To calculate C faeces we assumed a log-normal distribution and calculated the expected 240 value using the mean and standard deviation from 12 using the standard equation: 241

Where µ is the sample mean and σ is the sample standard deviation of the log normal 243 distribution. 12 state that µ of the distribution is 5.22 log 10 copies ml -1 and σ = 1.86 log 10 244 copies ml -1 which results in an expected C faeces concentration within the sewage effluent of 245 1595.9 million copies ml -1 . V faeces is the mean daily volume of faeces generated per person 246 (0.149 kg, from table 3 of 26 and assuming faeces has a density approximately equal to 247 water 27 . Note we used the 'rich country' value from 26 because the RT-PCR data 12 that 248 we use to estimate C faeces was measured from samples collected in Germany. The four orders of magnitude, and as such we selected high (10 -1 ), medium (10 -2 ) and low (10 -257 3 ) estimates (which equate to 10%, 1% and 0.1% proportion of viable versus within the 258 total viral genome counts). 259

The expected number of copies of infectious virus resulting from a sewage spill into a river, 261 lake or coastal region for a given country can therefore be calculated as 262

C spill,c was estimated for May 3 2020 21 countries that contain large inland water bodies 264 and were known to rely upon reservoirs for drinking water 28 . Long-term statistical mean 265 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 20, 2020. . https://doi.org/10.1101/2020.06.17.20133504 doi: medRxiv preprint water temperature, needed to calculate virus survivability, was calculated from a climate 266 quality global lake temperature dataset (see below). Temperature values for each country 267

were the countrywide mean lake temperature within a rectangular box matching a 268 simplified country outline. The dilution factors reported in 13 can vary by several orders of 269 magnitude and were deemed to provide the major source of uncertainty in the calculation. The concentration of SARS-CoV-2 virus needed for infection is not known. 30 provides 10 3 284 copies for influenza. The Infectious dose for SARS-CoV-2 is likely significantly lower 285 because 31 ranks influenza as "very high infective dose" and SARS-CoV-2 as "low". We 286 therefore use a value of 100 copies as a concentration that could result in infection. is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 20, 2020. . https://doi.org/10.1101/2020.06.17.20133504 doi: medRxiv preprint number of active cases (±20%), mass of faeces generated per capita per day (0.095 kg, 292 see table 3 of 26 , mean number of viral genome copies in faeces (3.54×10 12 ) and density of 293 faeces was not included in the uncertainty analysis. This resulted in a combined 294 uncertainty budget of ±68% copies ml -1 . It is important to note that this value does not 295 include uncertainty in the dilution factors or the ratio of viral genome copies to infectious 296 virus. Instead, the C inf calculation was repeated for high, medium and low values of these 297 parameters. 298 299

As reported in 3 , the virus concentration in water follows an exponential decay, with its 301 half-life decreasing with decreasing temperature and the pH control of half life is very small 302 over the pH range of 3-10 (which encompasses the range found in natural freshwater and 303 marine systems). Based on the in vitro data presented in 3 , the following empirical model 304 was derived to describe virus concentration reduction factor due to the temperature-305 dependent die-off: 306 = 10 !.05!°! !!.32 (5) 307 = ! 10 !!" (6) 308

Where C 0 is initial virus concentration (copies ml -1 ), n(t) is virus concentration after time t 309 (days) and r is 24 hour survival factor due to temperature T driven die off. This model fit to 310 the in vitro data gives a root mean square difference (RMSD) of ±1% for water at 4°C 311 which increases to ±7.5% at 22°C. When considering temperature controlled survival in 312 the waste water system, C ww,c becomes the value used for the initial viral load C 0 following 313 a sewage effluent spill. As noted in 5,12 , the viral load follows a heavy-tailed distribution 314 with the majority of patients shedding around 10 5 copies ml -1 ) but some having viral loads 315 as high as 10 12 copies ml -1 . This results in the super-spreader problem where a tiny 316

proportion of the infected population can become responsible for contributing a majority of 317 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted June 20, 2020. . https://doi.org/10.1101/2020.06.17.20133504 doi: medRxiv preprint viral load in the wastewater. For a large infected population, this approach allows robust 318 statistical modeling of viral load. However, in case of smaller communities with low number 319 of infections, the actual viral load could be severely underestimated if a super-spreader is 320 present within the population. 321 322

The example volume flow rate through the mouth of a medium sized Bowhead whale 324 whilst feeding was provided by 32 ) A flow rate of 5.65 m 3 s -1 is given for a 15 m whale 325 (mouth pressure of -1768 Pa at a 4 km h -1 foraging speed, assuming an oral opening of 326 5.09 m 2 with an opening radius = 1.27 m). Assuming a low viral concentration of 1 copies 327 per ml -1 , which equates to 1000 copies l -1 . 5.65 m 3 s -1 equates to 5650 L s -1 . The dosage 328 per second as the whale swims during feeding is given by 1000 (copies L -1 ) × 5650 (L s -1 ) 329 = 5.65 million copies s -1 . 330

Severe acute 333 respiratory syndrome coronavirus 2 (SARS-CoV-2) and coronavirus disease-2019 334 (COVID-19): The epidemic and the challenges

Evaluation of saline, phosphate buffered saline and minimum 337 essential medium as potential alternatives to viral transport media for SARS-CoV-2 338 testing

Stability of SARS-CoV-2 in different environmental conditions

Infectious SARS-CoV-2 in Feces of Patient with Severe COVID-19

Virological assessment of hospitalized patients with COVID-2019. 344

Suppression of COVID-19 outbreak in the municipality of

First detection of SARS-CoV-2 in untreated wastewaters in Italy

First confirmed detection of SARS-CoV-2 in untreated wastewater 350 in Australia: A proof of concept for the wastewater surveillance of COVID-19 in the 351 community

Potential fecal transmission of SARS-CoV-2: Current evidence and 356 implications for public health

High levels of sewage contamination released from urban areas 359 after storm events: A quantitative survey with sewage specific bacterial indicators

An analysis of SARS-CoV-2 viral load by patient age

Worldwide estimation 364 of river concentrations of any chemical originating from sewage-treatment plants 365 using dilution factors

Discharge-based 368 QMRA for estimation of public health risks from exposure to stormwater-borne 369 pathogens in recreational waters in the United States

Coronavirus 372 in water environments: Occurrence, persistence and concentration methods -A 373 scoping review

An improved 375 infectivity assay combining cell culture with real-time PCR for rapid quantification of 376 human adenoviruses 41 and semi-quantification of human adenovirus in sewage

Quantification of human adenovirus and norovirus in river water in 379 the north-east of France

QMRA of SARS-CoV-2 for workers in wastewater treatment 382 plants

Evaluation of lockdown impact on SARS-CoV-2 dynamics through 384 viral genome quantification in Paris wastewaters

Broad Host Range of SARS-CoV-2 Predicted by Comparative and 387 Structural Analysis of ACE2 in Vertebrates

Evidence of viral dissemination and seasonality in a Mediterranean 390 river catchment: Implications for water pollution management

Water, sanitation, hygiene, and waste water management for the COVID-19 393 virus: interim guidance

Rapid 395 evidence summary on SARS-CoV-2 survivorship and disinfection, and a reusable 396 PPE protocol using a double-hit process. medRxiv (2020)

Study on the resistance of severe acute respiratory syndrome-399 associated coronavirus

Freshwater Clams As Bioconcentrators of Avian Influenza 402 Virus in Water. Vector-Borne Zoonotic Dis

The Characterization of Feces and 405 Urine: A Review of the Literature to Inform Advanced Treatment Technology

Review of synthetic human faeces 409 and faecal sludge for sanitation and wastewater research

World water: resources, usage and the role of man made reservoirs

Lake Surface Water Temperature (LSWT) 414 v4.0 (1995-2016)

Influenza Virus Aerosols in the 417 Air and Their Infectiousness

Emerging threats from zoonotic coronaviruses-from SARS 419 and MERS to 2019-nCoV

Models of hydrodynamic flow in the bowhead whale filter feeding 422 apparatus