key: cord-0938413-raht37b8
authors: Coyle, Jayme P.; Derk, Raymond C.; Lindsley, William G.; Boots, Theresa; Blachere, Francoise M.; Reynolds, Jeffrey S.; McKinney, Walter G.; Sinsel, Erik W.; Lemons, Angela R.; Beezhold, Donald H.; Noti, John D.
title: Reduction of exposure to simulated respiratory aerosols using ventilation, physical distancing, and universal masking
date: 2022-02-22
journal: Indoor Air
DOI: 10.1111/ina.12987
sha: cbe5cebbfe0a65fbb78639dc38d68f1e5bb1dcbd
doc_id: 938413
cord_uid: raht37b8

To limit community spread of SARS‐CoV‐2, CDC recommends universal masking indoors, maintaining 1.8 m of physical distancing, adequate ventilation, and avoiding crowded indoor spaces. Several studies have examined the independent influence of each control strategy in mitigating transmission in isolation, yet controls are often implemented concomitantly within an indoor environment. To address the influence of physical distancing, universal masking, and ventilation on very fine respiratory droplets and aerosol particle exposure, a simulator that coughed and exhaled aerosols (the source) and a second breathing simulator (the recipient) were placed in an exposure chamber. When controlling for the other two mitigation strategies, universal masking with 3‐ply cotton masks reduced exposure to 0.3–3 µm coughed and exhaled aerosol particles by >77% compared to unmasked tests, whereas physical distancing (0.9 or 1.8 m) significantly changed exposure to cough but not exhaled aerosols. The effectiveness of ventilation depended upon the respiratory activity, that is, coughing or breathing, as well as the duration of exposure time. Our results demonstrate that a layered mitigation strategy approach of administrative and engineering controls can reduce personal inhalation exposure to potentially infectious very fine respiratory droplets and aerosol particles within an indoor environment.

masks, maintaining physical distances, and avoiding crowded indoor and outdoor spaces, among other strategies. 10, 11 Universal masking reduces respiratory aerosol exposure through source control, that is, limiting the release of infectious very fine respiratory droplets and aerosol particles into the ambient environment at the point of generation, while face masks also provide some protection from aerosols for the mask wearer. 12 While the generalized effectiveness of masking for source control has been established, [13] [14] [15] its effectiveness is neither absolute nor uniform in practice. Variations in filtration efficiency, air flow resistance, user compliance, and mask fit can limit the effectiveness of masks as source control and protection for the wearer. 16, 17 Despite these limitations, comparison of COVID-19 cases among states employing mask mandates demonstrate an association between universal masking and reduced incidence rates [18] [19] [20] as well as community transmission of COVID-19. 21, 22 Physical distancing reduces infectious material transfer via respiratory-derived droplets and aerosol particles. Routine respiratory actions, such as breathing and normal speech, produce micron and submicron scale particles that can remain airborne for minutes to hours. 23 By comparison, coughing, loud speech, and singing can project aerosols and droplets over greater distances, thus potentially increasing the probability for pathogen transmission. For example, droplets and aerosols produced by coughing may travel up to 8 m. 24 Analysis of a super spreader event among a cohort of choir members, all of whom were unmasked and within 1.8 m of physical distance during practice, estimated the SARS-CoV-2 attack rate as between 53.3% and 86.7%. 25 Correlative analyses support an association between physical distancing policies and reduction in COVID-19 incidence, 26, 27 and agree with case-control studies. 28 Engineering controls, such as room ventilation, are an effective and reliable strategy to ensure good air quality while mitigating infection transmission in the indoor environment. 9, 29 While evidence clearly shows increasing ventilation rates as an effective measure in exposure mitigation, 30, 31 air flow patterns can influence the dispersion of potentially infectious respiratory aerosols and personal exposure, 32 particularly in confined spaces. 33 The overall effectiveness of ventilation can be difficult to generalize since ventilation is unique to each room and operates alongside other exposure mitigating strategies, such as masking and physical distancing. As such, the current investigation examines the combined effect of physical distancing, universal masking, and ventilation on exposure to simulated very fine respiratory droplets and aerosol particles generated during breathing and coughing within a controlled indoor environment. The results of this investigation quantitatively examine the contribution of the matrix of controls employed on respiratory infection mitigation strategies within the indoor environment.

The testing environment consisted of an environment chamber measuring 3.15 m × 3.15 m × 2.26 m (gross internal volume of 23.8 m 3 , Figure 1 ).

An internal re-circulating high-efficiency particulate air (HEPA) filtration system (Flow Sciences, Inc.) was used to reduce background aerosol/ particle concentrations to near-zero prior to each experiment. The HEPA system consisted of a 10.8 cm return duct positioned along the left wall 55.9 cm from the ground leading to the central motor/filter unit and a supply duct positioned along the right wall at a height of 2.19 m from the floor; no external fresh air was introduced into the environmental chamber during experimentation. The HEPA system utilized for experimentation was configured for dilution ventilation for our experiments in order to maximize the removal of aerosol particles from the test chamber. Six Grimm 1.108 optical particle counters (OPCs; GRIMM Aerosol Technik Ainring GmbH & Co. KG) were positioned at a height of 152 cm throughout the chamber. The OPCs measured particle concentrations in channels ranging from 0.3 to 3.0 µm at a frequency of 1 Hz, except for one OPC sampler at 0.167 Hz. Four OPCs were affixed to telescopic stands 152 cm above the floor and referred to as "area samplers." One OPC was positioned 3.2 cm next to the mouth central axis and anteriorly planar to the mouth opening of the recipient simulator (see below) and fit behind a mask affixed to the simulator; this position is denoted as "at the mouth of the breather" for presentation purposes. The remaining OPC was positioned 8.9 cm next to the mouth central axis and anteriorly planar to the mouth opening of the recipient simulator to allow for measurement in the personal breathing zone outside of a mask affixed to the simulator.

All OPCs were controlled and data logged using a custom program in LabVIEW v. 2009 (National Instruments).

In addition to particle removal, the HEPA system provided ventilation, with a variable transformer (Staco Energy Products, Co.) used F I G U R E 1 Experimental setup and simulators. Diagram of environmental chamber setup showing positions of the aerosol source simulator (red), recipient simulators (blue; position adjustable between 0.9 and 1.8 m), and OPCs (green dots) for area measurements (S1-4) and personal breathing zone measurements at the mouth (M) and beside the head (B) of the recipient. The room air supply (at the ceiling) and return (near the floor) for the HEPA system are each shown with a circle and "X." The HEPA filter and blower unit are demarcated by the red square containing an "X" to set the HEPA system flow rate. Air exchange rates were determined via single-point measurement of the linear air flow at the return duct using a Model 5725 VelociCalc rotating vane anemometer (TSI, Inc.) equipped with a tapered air cone (TSI, Inc.). The return duct was straightened for a length of >10 diameters from the return opening to minimize turbulent flow during anemometer readings for air changes per hour (ACH) derivation. The HEPA system was set to 0 ACH, 4 ACH (0.255 m 3 /s flow), 6 ACH (0.382 m 3 /s), and 12 ACH (0.765 m 3 /s); calculations assumed zero leakage into the chamber. Effective air filtration rates were derived empirically. Briefly, the chamber was saturated with particles using a stand-alone TSI Model 8026 generator until the 0.3-0.4 μm particle size channel reached 10 5 particles per liter under constant mixing using a household fan. The particles for effective air changes per hour were generated using a 1% solution of NaCl in distilled water formulated from 100 mg tablets provided with the TSI Model 8026 generator as per manufacturer's instructions. After a 15-min mixing period, the HEPA filtration system was set to the desired ACH based on anemometer measurements. Particle concentrations were measured for 20 min using five of the six OPCs to derive particle exponential decay curves spatially throughout the chamber.

Theoretical particle exponential decay curves were modeled from the three smallest size bins (0.3-0.4 µm, 0.4-0.5 µm, and 0.5-0.65 µm) assuming negligible loss to chamber surfaces and aerosol agglomeration using MATLAB v. 9.6 (Mathworks). The slope of the modeled particle decay was assumed to be first order as per equation:

Where: C t is the particle concentration at time t (#/cm 3 ).

C i is the initial particle concentration at time zero (#/cm 3 ).

e is Euler's number, approximated to 2.71828.

λ is the slope of particle concentration change over the time (#/ cm 3 /s).

t is time (s).

Empirical concentrations of particles measured by the five area OPCs were then fitted via log-linear regression and the resultant decay coefficient (λ) derived to estimate the effective OPC-specific ACH.

The source simulator had a head form with pliable skin (Hanson Robotics) as described in previous work. 14 For these tests, a single cough and two versions of simulated breathing were examined. The followed by a single 4.2 L rapid exhalation at a peak flow rate of 11 L/ min 33 ; the simulator did not breathe following the cough. For breathing tests, the simulator breathing rate was 12 breaths/min with a tidal volume of 1.25 L and ventilation rate of 15 L/min. The breathing parameters correspond to the ISO standard for females performing light work. 34 For the breathing modality, the nebulizer was cycled 10 s on and 50 s off continuously throughout the test duration.

Tests were conducted for a duration of 15 min, except for a limited subset of testing conditions which were conducted for 60 min. As an additional examination of the time dependency of ventilation in reducing recipient exposure, additional tests were conducted using a modified aerosol generation cadence during the breathing action.

During these tests, the nebulizer generated aerosol continuously for the first 3 min of the test, after which the nebulizer was turned off, and are henceforth designated short-term aerosol generation tests.

To simulate source aerosol exposure to a recipient, a breathing simulator (Warwick Technologies Ltd.) with a pliable skin head form 

For experimental trials with masking conditions, a 3-ply cotton mask (Hanes Defender, HanesBrand, Inc.) was fitted to the respective simulator followed by fit factor assessment using the PortaCount Pro+ (TSI, Inc.) in the N95 mode (measuring negatively charged particles 55 nm in diameter) 15 as per manufacturer's instructions. A daily quality assurance test was conducted using the 3M 1860 N95 respirator (Saint Paul, MN).

To test the effect of layering aerosol mitigation strategies of universal masking, physical distancing, and ventilation, experiments consisting of a matrix of the three variables were conducted (Table 1) . For masking, the combinations of no masking (neither simulator wore a mask) and universal masking (both simulators wore a 3-ply cotton mask) were examined. For physical distancing, given the limitation of the distance due to the size of the environmental chamber, 0.9 and 1.8 m distances were examined. For ventilation, four ACH rates were selected: 0, 4, 6, and 12.

After mask fitting and distance configuration, the environmental chamber was sealed, and the HEPA filtration system run at maximal rate to minimize background airborne particles. Thereafter, the HEPA filtration system was either turned off (0 ACH) or set to the desired air exchange rate (4-12 ACH) and allowed to run for 15 min, during which time all OPCs were initialized to begin particle concentration data collection and the recipient simulator activated to begin

breathing. After the air exchange stabilized, the source simulator was initiated to cough or breathe, and aerosol concentrations were measured for 15 min. The chamber was allowed to cool to 22°C between experiments to reduce the inter-test temperature variability.

Three independent experimental replicates were conducted for each unique experimental condition without condition randomization.

The background aerosol concentration was determined based on the mean particle concentration during the 3 min prior to cough or exhalation. The bin-specific particle counts per cubic meter of air were converted to volume based on the mean bin diameter (assuming spherical particles) and then to mass concentration by multiplying by the density of KCl (1.984 g/cm 3 ). The total mass concentration was calculated by summing the bin-specific mass concentrations for all size bins. The mean mass concentration was calculated as the average mass concentration over the test duration and served as the exposure metric in these simulations. OPC data were processed using the R statistical Unstandardized regression coefficients are presented in addition to back-transformed coefficients expressed as percent reduction in the outcome variable (mean mass concentration). Statistical significance was set at p < 0.05.

Across all experiments, the mean chamber temperature was 24.1 ± 1.1°C with a relative humidity of 26.0% ± 2.4%. The temperature change during all experiments rose 0.4°C ± 0.2°C and the relative humidity change was 0.2°C ± 0.2°C. Particle clearance by the ventilation system followed first-order exponential decays, with overall clearance rates 74.1% ± 4.4% of decay rates estimated by anemometer readings (Range: 73.1%-76.7%; Figure 2A ). Particle decay rates throughout the chamber, as measured by the five OPCs, were largely homogeneous ( Figure S1 ). The experimental decay rates after single coughs were 76.1% ± 1.5% of theoretical values (Range:

74.4%-77.3%). These experimental decay rate magnitudes and variances were comparable to those obtained from particle decay testing, which suggests that the ventilation system promoted adequate air mixing to disperse cough aerosols through the chamber volume.

Therefore, we presume similar air mixing within the chamber during ventilation studies for the other two modalities tested.

Chamber aerosol concentrations during simulated respiratory events are shown in Figure 2B 

The time-concentration curves at the 1.8 m physical distance are shown in Figure 3A ; analogous results for the 0.9 m physical distance are presented in Figure S3 . The reduction in mass concentration was likely due to preferential filtration of aerosols >1 µm in diameter by the 3-ply cotton masks fitted to the source and recipient simulators (Figure 4 ). The differences in exposure reduction among the aerosol generation modalities were likely due to specific changes in aerosol spatiotemporal dispersion when the source was masked. Aerosol plumes generated during both breathing modalities and a single cough escape through face seal leaks. 38 The plumes would then be deflected behind and/or to the side of the source and thus effectively farther from the recipient compared with the experiments with no masks. Without chamber mixing, as observed with no ventilation, the cough aerosol deflected by the mask took longer to disperse throughout the chamber compared to without a mask as was observed in Figure 3A . for coughing and 44.3% ± 14.0% for breathing). 15 Lastly, since the current investigation utilized static breathing simulators, the results do not account for the potential contribution of anthropogenic movement and individual behavior to aerosol particle exposure. 39 We have previously observed the presence of the exhalation from a breathing receiver can influence aerosol particle exposure within the experimental configuration contained in this work. 12 Table 2 ), while universal masking expectedly reduced mean mass concentration significantly.

When condensing the total aerosol generation period to the initial 3 min in the short-term aerosol generation tests, increasing ACH became a significant predictor in exposure reduction (5.2%; 95% CI:

3.8%-6.5%; p < 0.001; Table S2 ). The time-concentration curves of the short-term aerosol generation tests demonstrated the log-linear decay similarly to time-concentration profiles observed from a single cough, albeit shifted to the right to reflect the longer aerosol generation period ( Figure S4 ). This result demonstrates that attainment of a dynamic equilibrium with continuous aerosol input or removal of aerosols produced by an intense, short-term generation event through increasing ventilation can result in significant exposure reduction for a recipient. We did not examine the extended exposure duration for a single cough over 60 min, though we expect increasing ventilation will remain a significant predictor of mean mass concen- the direction of initial smoke dispersion shifted toward the recipient, likely due to the pressure drop produced from the ventilation system supply stream above and behind the recipient ( Figure S5 ). The 

The current investigation has several noteworthy limitations that must be considered. First, the mass concentration of aerosol generated during the experimental modalities, particularly breathing, was higher than those produced from human exhalations. 23 The higher concentrations combined with the wide dynamic range of the OPC allowed for stable and reproducible measurements while assuring attainment of quantitative limits of detection among all tests. Second, the simulators lack generation of body heat, do not generate a thermal exhalation plume, and exhale or cough dry salt particles, all of which affect aerosol size, aerosol dispersion, and inhalation exposure. [45] [46] [47] Given the confines of the environmental chamber, the internal ventilation setup, and the high aerosol concentrations, we would not expect substantial differences in mean mass exposure given the small volume of the chamber. Therefore, limits must be placed on the interpretation of the results within a larger indoor environment, especially considering the dispersion potential of an exhalatory thermal plume and the strong influence of ventilation supply air flow observed. Third, the range of human respiratory aerosols can be smaller and larger than the measured range of this investigation (0.3-3.0 µm). 35, 37 For droplets, the effect of physical distancing may be higher than those suggested by the observed results.

Fourth, the study investigated the exposure reduction of a single 3-ply cotton mask. The authors recognize the limitation of having tested a single mask, since the effectiveness of exposure reduction by other masks could be either higher or lower, depending on the mask. Nonetheless, the analytics of the study allow for reasonable expectation of exposure reduction of the other predictor variables provided the aerosol behavior does not significantly deviate from this study with another type of mask.

The current investigation highlights the contribution of three common engineering and administrative controls recommended for limiting SARS-CoV-2 exposure within an indoor environment: ventilation, physical distancing, and universal masking. When controlling for the other two mitigation strategies, universal masking with a 3-ply cotton mask contributed to the plurality of the observed reduction in aerosol mass exposure irrespective of aerosol generation modality.

This reduction was due, in part, to preferential reduction of particles 

The authors wish to acknowledge the facilities, maintenance, and security personnel of NIOSH Morgantown for their hard work and dedication, which were integral to the completion of this work.

Aerosol transmission of infectious disease

It's about time: a comparison of Canadian and American time-activity patterns

Indoor transmission of SARS-CoV-2

Estimation of the asymptomatic ratio of novel coronavirus infections (COVID-19)

Airborne transmission of SARS-CoV-2 in indoor environments: a comprehensive review

COVID-19 outbreak associated with air conditioning in restaurant

Probable airborne transmission of SARS-CoV-2 in a poorly ventilated restaurant

Evidence supporting transmission of severe acute respiratory syndrome coronavirus 2 while presymptomatic or asymptomatic

A paradigm shift to combat indoor respiratory infection

CDC calls on Americans to wear masks to prevent COVID-19 spread

Keep a Safe Distance to Slow the Spread

Efficacy of universal masking for source control and personal protection from simulated cough and exhaled aerosols in a room

Respiratory virus shedding in exhaled breath and efficacy of face masks

Efficacy of face masks, neck gaiters and face shields for reducing the expulsion of simulated cough-generated aerosols

A comparison of performance metrics for cloth face masks as source control devices for simulated cough and exhalation aerosols

Comparison of performance of three different types of respiratory protection devices

Surgical mask filter and fit performance

Community use of face masks and COVID-19: evidence from a natural experiment of state mandates in the US

Trends in county-level COVID-19 incidence in counties with and without a mask mandate -Kansas

Association of state-issued mask mandates and allowing on-premises restaurant dining with County-level COVID-19 case and death growth rates -United States

Face masks, public policies and slowing the spread of COVID-19: evidence from Canada

Causal impact of masks, policies, behavior on early covid-19 pandemic in the U

Aerosol emission and superemission during human speech increase with voice loudness

Turbulent gas clouds and respiratory pathogen emissions: potential implications for reducing transmission of COVID-19

High SARS-CoV-2 attack rate following exposure at a choir practice -Skagit county

Enacting national social distancing policies corresponds with dramatic reduction in COVID19 infection rates

Examining the effect of social distancing on the compound growth rate of COVID-19 at the county level (United States) using statistical analyses and a random forest machine learning model

Social distancing decreases an individual's likelihood of contracting COVID-19

Role of mechanical ventilation in the airborne transmission of infectious agents in buildings

Influence of room ventilation settings on aerosol clearance and distribution

Small droplet aerosols in poorly ventilated spaces and SARS-CoV-2 transmission

Mathematical models for assessing the role of airflow on the risk of airborne infection in hospital wards

Efficacy of an ambulance ventilation system in reducing EMS worker exposure to airborne particles from a patient cough aerosol simulator

ISO. Respiratory protective devices -Human factors -Part 1: Metabolic rates and respiratory flow rates. ISO/TS 16976-1:2015. International Organization for Standardization

Size distribution and sites of origin of droplets expelled from the human respiratory tract during expiratory activities

Modality of human expired aerosol size distributions

Size distribution of exhaled particles in the range from 0.01 to 2.0m

Visualizing the effectiveness of face masks in obstructing respiratory jets

Close contact behavior in indoor environment and transmission of respiratory infection

Effect of human movement on airborne disease transmission in an airplane cabin: study using numerical modeling and quantitative risk analysis

Flow and contaminant transport in an airliner cabin induced by a moving body: model experiments and CFD predictions

Investigation of potential aerosol transmission and infectivity of SARS-CoV-2 through central ventilation systems

Factors involved in the aerosol transmission of infection and control of ventilation in healthcare premises

Role of ventilation in airborne transmission of infectious agents in the built environment -a multidisciplinary systematic review

How human thermal plume influences nearhuman transport of respiratory droplets and airborne particles: a review

Physico-chemical characteristics of evaporating respiratory fluid droplets

Fluid dynamics of respiratory infectious diseases

A systematic approach to estimating the effectiveness of multi-scale IAQ strategies for reducing the risk of airborne infection of SARS-CoV-2

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.

Jayme P. Coyle involved in conceptualization, investigation, methodology, data curation, visualization, wrote the manuscript, and edited the manuscript. Raymond C. Derk involved in conceptualization, investigation, methodology, visualization, wrote the manuscript, and