AEROSOL HOUSE LAB · EXPLORATORY MODEL

What changes airborne infection risk?

Explore a COVID-like indoor scenario. Change the time spent together, ventilation, masks and assumed infectious emission. Compare shared room air with an exploratory near-source air zone.

Conditional model estimate. The percentage assumes an infectious person is present and the chosen quanta emission is correct. Emission, immunity and mixing are uncertain. Use it to explore scenarios. It is an unvalidated educational model and cannot estimate an individual's personal probability from distance alone.

Choose a stage above. The dashboard changes in place—no long-page scrolling required.

01 / Design the scenario

Set the scenario

All values are independent of the house dashboard. The initial air is assumed free of infectious quanta. Sources and occupants stay for the full exposure.

Start with an illustrative scenario

Each preset uses assumed values. Change one control and watch the result respond.

Does the infectious source wear a face mask?

“With source face mask” applies an illustrative 50% reduction to particles and infectious quanta emitted by each source. The visible mask, airborne particles and risk estimate update together.

Particle size and emission assumptions

Explore one representative size at a time. These example inputs describe particles released by each source person before source masking. The model applies the same selected source-mask removal and air-purifier settings to that size.

02 / Follow the infectious air

Estimated airborne infection probability

60 minutes

Drag or play to see accumulated risk and the chart marker change. Playback covers the selected exposure duration.

03 / Inspect the room

See the people in the room

Orange people are assumed infectious. Blue people can breathe the shared air. More dots indicate more modelled particles per cubic metre. Watch the source mask appear or disappear, the door open and the air purifier switch on as you change settings.

Digital room meter

CO₂ starts at 420 ppm and assumes 18 litres per hour per person, with air mixed throughout the room. Masks and particle filtration have no direct effect in this CO₂ calculation. Movement and conversation illustrate the scenario.

🟠 Assumed infectious🔵 Susceptible person

People stay blue because this model estimates probability, rather than identifying who becomes infected. The particle estimates use the assumed emission, diameter and density. Mass assumes spherical particles of one size. Dots show relative number concentration and are capped at 48 per zone. Quanta describe infectious dose separately. The picture shows up to 12 people from each group and up to 48 dots per air zone. The illustrated door represents higher outdoor air exchange in the model.

The visual reflects the modelled probability at the selected time. It does not depict a counted infectious dose or a distance boundary.

Compare a change

Save a result, then change ventilation, duration or masks to compare the estimates.

Risk accumulates over the exposure under the selected assumptions.

How to interpret the estimate

The transient Wells–Riley approach integrates inhaled quanta and uses P = 1 − exp(−dose). A quantum is a statistical infectious dose unit. PM₂.₅, CO₂ and the house model's arbitrary respiratory tracer do not directly measure infectious quanta.

Equations, mixing and removal

Whole room model. dC/dt = Iq(1−ηs)/V − λC, dose = p(1−ηr)∫Cdt, where λ = ACH + CADR/V + deposition/inactivation. The initial concentration is zero and emissions continue throughout the exposure. The concentration integral is solved analytically.

Two-zone exploration. Vn dCn/dt = Iq(1−ηs) − β(Cn−Cf) − λVnCn. Vf dCf/dt = β(Cn−Cf) − λVfCf. Vn + Vf = V. Exchange conserves transported quanta. Ventilation, cleaning and other losses are applied in proportion to zone volumes. Numerical midpoint integration uses a step limited by the exchange/removal rates.

The breathing rate applies to the receiver. Source and receiver masks multiply emission and inhaled dose respectively. Assumed complete removal at 100% is an ideal mathematical limit. The number of susceptible people changes expected infections, while emission depends on the number of infectious people.

Close proximity depends on respiratory jets, direction, activity and local air movement. This two-zone scenario is a sensitivity exercise with assumed mixing parameters. It excludes direct droplet or surface transmission, individual susceptibility, immunity, variant calibration, source heterogeneity and infectious-dose uncertainty. The displayed emission sensitivity range is not a confidence interval.

For a real building, measured airflow, source estimates and independently evaluated infection data would be needed to assess model performance.

UKHSA guidance on ventilation and close contact ↗

Buonanno, Stabile & Morawska (2020), quanta emissions and infection risk ↗

Sze To & Chao (2010), Wells–Riley and dose-response approaches ↗