key: cord-0188290-oi4cphh9
authors: Raghav, Vrishank; Tan, Zu Puayen; Bhatt, Surya P.
title: Dispersion of speech aerosols in the context of physical distancing recommendations
date: 2020-07-08
journal: nan
DOI: nan
sha: c94e60978072dced25511c2dc208250bfd7d2847
doc_id: 188290
cord_uid: oi4cphh9

High-speed particle image velocimetry (PIV) was used to quantify the dispersion of aerosol-laden gas clouds generated during phonetic vocalization by a human subject at different sound intensity levels. The measured PIV data was used to quantify the initial penetration depth. Using classical pulsed jet scaling laws propagation distances were computed for time periods beyond the measured duration. Our results indicate that the penetration distance was comparable between loud intensity speech (for example during singing, classroom lectures, parties etc.) and moderate intensity cough. Based on theoretical aerosol propagation distance and time, the 6 feet physical distancing recommendations are likely sufficient to avoid incidental exposure by the initial penetration of the aerosol cloud, but insufficient for prolonged exposure to slow propagating aerosol clouds.

Aerosol propagation during vocalization of specific phones or cough (interrupted jets) occurs in two stages (7) : rapid initial penetration followed by slow propagation. Penetration depth was quantified from the measured PIV data as the maximum horizontal distance traveled by the plume from the subject before reaching ambient velocity. Propagation beyond initial penetration was theoretically projected using classical scaling laws (7) for a transient non-buoyant puff (interrupted jet) ( ) = 1 + 2 . Here, ( ) is the propagation distance as function of the propagation time, = 0.25 for transient jets and puffs (7), 1 and 2 are constants that depends on the diameter of the mouth ( ≈ 1.5

) and the average velocity ( 0 ) at the mouth exit. Measured PIV data were used to determine 1 and 2 , from which propagation distances were computed for time periods beyond the measured duration.

Loud cough was associated with the highest initial penetration depth (up to 5.5 feet, Figure 1A -B and Animation). Among individual phones, the plosives (/ti/ and /ta/) had the highest initial penetration (up to 4 feet) followed by both the fricatives (/si/ and /sa/) and nasal sounds (up to 2.5 feet). The initial penetration depth of the phrase 'Stay Healthy' (comprised of most of the individual phones) was 2.9 feet. Among the plosives and cough, higher sound intensity levels were associated with higher penetration ( Figure 1B) . The penetration depth of loud intensity plosives (4.1 feet) was comparable to normal (3.3 feet) and moderate (4.4 feet) cough. Based on theoretical projections of propagation time, the subsequent dispersion of the aerosol cloud was estimated to reach as far as 7-13 feet for plosives and 12-16 feet for cough within 2 minutes ( Figure 1C ).

Discussion: For the first time, this work quantifies the penetration distance of aerosol clouds generated during regular speech at different sound intensity levels. Our results indicate that the penetration distance was comparable between loud intensity speech (for example during singing, classroom lectures, parties etc.) and moderate intensity cough. This is particularly noteworthy because prior studies have shown that the size and quantity of aerosol generation are equivalent between some speech phones and coughing (8) . Furthermore, based on theoretical aerosol propagation distance and time, we characterize potential exposure risk burden for contacts; around 2 minutes of exposure to an asymptomatic person as far as 7-13 feet apart during regular speech is likely to place contacts at risk of inhaling infectious aerosol. As such, physical distancing recommendations are likely sufficient to avoid incidental exposure by the initial penetration of the aerosol cloud, but insufficient for prolonged exposure to slow propagating aerosol clouds. While the concentration of virus particles in the aerosol cloud is bound to dilute with propagation distance and time, it is important to acknowledge the high virulence of SARS-CoV-2 (9) and the unknown infectious dose (10). Figure 1 

Our study has some limitations. Although our findings were generated by evaluation of a single subject, prior studies have reported minimal variance by age, gender, and BMI in aerosol generation during speech (11) . Additionally, the propagation time and distance of the aerosol cloud beyond instantaneous penetration were estimated based on a theoretical model for a non-buoyant puff with no ambient air currents. This estimate represents an "average risk" scenario that is likely to be influenced (increase or decrease) by the directionality and magnitude of the prevailing indoor air currents.

In conclusion, given the undetermined infectious dose of SARS-CoV-2 and the expected cumulative aerosol dose in practical scenarios, such as several minutes of conversation, it is prudent to treat the six-foot recommendation as a conservative limit for short encounters. As wearing a mask can significantly alleviate aerosol dispersion (12) , contact with an asymptomatic individual with SARS-CoV-2 for extended periods is likely unsafe, at any practical distancing limit, without a mask.

Aerodynamic analysis of SARS-CoV-2 in two Wuhan hospitals

Transmission of 2019-nCoV infection from an asymptomatic contact in Germany

Public Health Recommendations for Community-Related Exposure

Particle image velocimetry: a practical guide

Violent expiratory events: on coughing and sneezing

Recognition of aerosol transmission of infectious agents: a commentary

Self-preserving properties of unsteady round nonbuoyant turbulent starting jets and puffs in still fluids

Characterization of expiration air jets and droplet size distributions immediately at the mouth opening

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

COVID-19: A Risk Assessment Perspective

Aerosol emission and superemission during human speech increase with voice loudness

A schlieren optical study of the human cough with and without wearing masks for aerosol infection control