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

Abstract

The COVID-19 pandemic has raised concern of viral spread within buildings. Although near-field transmission and infectious spread within individual rooms are well studied, the impact of aerosolized spread of SARS-CoV-2 via air handling systems within multiroom buildings remains unexplored. This study evaluates the concentrations and probabilities of infection for both building interior and exterior exposure sources using a well-mixed model in a multiroom building served by a central air handling system (without packaged terminal air conditioning). In particular, we compare the influence of filtration, air change rates, and the fraction of outdoor air. When the air supplied to the rooms comprises both outdoor air and recirculated air, we find filtration lowers the concentration and probability of infection the most in connected rooms. We find that increasing the air change rate removes virus from the source room faster but also increases the rate of exposure in connected rooms. Therefore, slower air change rates reduce infectivity in connected rooms at shorter durations. We further find that increasing the fraction of virus-free outdoor air is helpful, unless outdoor air is infective in which case pathogen exposure inside persists for hours after a short-term release. Increasing the outdoor air to 33% or the filter to MERV-13 decreases the infectivity in the connected rooms by 19% or 93% respectively, relative to a MERV-8 filter with 9% outdoor air based on 100 quanta/h of 5 μm droplets, a breathing rate of 0.48 m³/h, and the building dimensions and air handling system considered.

Keywords: ACH, air changes per hour; AHU, air handling unit; Airborne transmission; COVID-19; Indoor air quality; MERV, minimum efficiency reporting value; Multizone buildings; Well-mixed; Wells-Riley.

Graphical abstract

Fig. 1

Essential elements of a central air handling system (a) in a generic small building (b) with their process flow representation. The two scenarios considered include a virus containing particle source that was either internal (Scenario 1) or external (Scenario 2) to the building (from a source room or outdoor air, respectively). The virus spreads to the connected rooms via a centrally connected plenum and air handling unit (AHU).

Fig. 2

Concentrations scaled on the total virus concentration shed, C_total, (a and c) and cumulative probabilities of infection (b and d) versus time for the source (a and b) and connected (c and d) rooms across the four conditions of no filter and MERV-8, MERV-11, and MERV-13 filters (Scenario 1 Cases 1–4).

Fig. 3

Concentrations scaled on the total virus concentration shed, C_total, (a and c) and cumulative probabilities of infection (b and d) versus time for the source (a and b) and connected (c and d) rooms for air change rates of 1.8, 3, 6, and 12 ACH (Scenario 1 Cases 1 and 5–7).

Fig. 4

Concentrations scaled on the total virus concentration shed, C_total, (a and c) and cumulative probabilities of infection (b and d) versus time for the source (a and b) and connected (c and d) rooms for virus-free outdoor air fractions of 0, 6, 9, and 33% (Scenario 1 Cases 1 and 8–10).

Fig. 5

Concentrations scaled on the total outdoor air concentration of virus, C_total,OA, (a, c and e) and cumulative probabilities of infection (b, d, and f) versus time for (a and b) no filter and MERV-8, MERV-11, and MERV-13 filters (Scenario 2 Cases 1–4); (c and d) air change rates of 1.8, 3, 6, and 12 ACH (Scenario 2 Cases 1 and 5–7); outdoor air fractions of 0, 6, 9, and 33% (Scenario 2 Cases 1 and 8–10). For an outdoor virus source, all rooms are connected rooms.

Fig. 6

Concentrations scaled on the total virus concentration, C_total, shed (a and c) and cumulative probabilities of infection (b and d) versus time for the source (a and b) and connected (c and d) rooms for 3, 10 and 30 rooms (Scenario 1 Case 1 varying f_Q and corresponding supply flow rates through the AHU).

Fig. 7

Cumulative probabilities of infection versus time for the source (a) and connected (b) rooms for q = 10, 30, 100, and 300 quanta/h (Scenario 1 Case 1).

Fig. A.1

Concentrations scaled on the total virus concentration shed, C_total, for the baseline condition with both viral decay and settling included (solid), without viral decay but with settling (short dash), and with viral decay but without settling (long dash).