Risk Calculator: Understanding Indoor COVID Infections Deutsch | English ✓ Respiratory transmitted viruses such as SARS-CoV-2 spread as aerosols, especially indoors. This raises two questions: “What is the likelihood that someone is infectious?” “What is the probability of infection in the specific room?”.The risk calculator makes this assessment visible and comparable on the basis of current incidence, air exchange, duration, activity, and number of people – and helps to estimate how risky situations actually are. Please enable JavaScript. Terms, briefly explained Air exchange (ACH): How often the entire room air volume is exchanged or equivalently filtered per hour. Mechanical ventilation, cross-ventilation, and HEPA filters add up in their effective removal. (ACH = Air changes per hour) HEPA filter: Clean recirculated air and provide “filter ACH.” Depending on the device/size, 2–5 ACH per device are realistic. Aerosol: Tiny airborne particles that form virus-laden respiratory clouds and can remain distributed indoors for minutes to hours. Unlike large droplets, they do not immediately fall to the ground; their buildup is determined by air exchange, room size, and activity, and reduced through ventilation or HEPA filtration. CO₂: Not a virus, but a proxy for ventilation. Sustained < 800 ppm often indicates good air exchange – but says nothing about the viral load of those present. Quanta: A model unit for the average infectious dose. More activity ⇒ more quanta. Prevalence: Estimated proportion of infectious individuals, derived here from incidence and a typical infectious phase. Incidence: Number of newly reported infections per defined population size (usually per 100,000 people) within a specified time period (e.g., 7 days). It serves as a basis for estimating the current probability of infection in the population. Preloaded room: The room was occupied shortly before the stay; as a result, the viral concentration starts > 0. In the calculator, this can be selected via the option “Potentially preloaded room”. Assumption: The preloading time is chosen dynamically based on the block’s exposure time and capped at 1 hour (t_pre = min(duration, 1 h)). Ventilation/HEPA and natural decay apply as usual. Daily risk vs. long-term risk – what is the difference? A single event may be acceptable, while repeated occurrence builds up a significant overall risk. The daily risk describes the probability of infection at one occasion. The long-term risk sums up many comparable occasions over weeks – such as school or office routines. Especially when including incidence forecasts, it becomes possible to estimate how high the infection risk is over the course of a wave. What is calculated in the background The calculator combines two components: Presence risk: From the 7-day incidence (alternatively: incidence forecast), prevalence is estimated – i.e. the share of infectious individuals in the population. From this follows the probability that in a group at least one person is infectious. Infection risk in the room: Using a simplified box/Wells-Riley approach, the number of infectious quanta inhaled during the stay is calculated. Key factors are air changes per hour (ACH), room volume, duration, activity (quiet breathing vs. speaking/singing/exercise), masks, and HEPA filtration. Ventilation/filters increase the removal of virus-laden particles, masks reduce both emission and inhalation, and natural losses (sedimentation/inactivation) are considered as a constant additional decay. Together, this yields the probability of infection, both once and cumulatively. Preloaded room (optional): In small, frequently re-used rooms (e.g., treatment rooms, changing rooms, dense meetings, course rooms), the calculation can start from an already elevated room concentration. In the calculator this is available as “Potentially preloaded room”. The preloading duration is set dynamically to the block’s duration and capped at 1 hour (t_pre = min(duration, 1 h)). Why aerosols are complex – and which simplifications the risk calculator uses Aerosols are not distributed uniformly: proximity to a speaking person (near field), air currents, thermal effects, open doors/windows, or changing occupancy create inhomogeneities. To remain practical, the calculator applies simplifying assumptions: The room is considered well-mixed. Local peaks (e.g. directly in front of a singing person) are thus underestimated; as a decision aid, the model still works reliably.Quanta emission and breathing rate are averaged over time and scaled according to the chosen activity.Removal/decay (ventilation, filtration, natural losses) is assumed to be constant. Short-term fluctuations are smoothed out in the model calculation.Masks act as an effective reduction in both directions (source/recipient) – in reality there are ranges, but conservative values are applied here. The model thus provides trend statements: Which levers reduce risk the most? It does not replace a tailored airflow assessment for special situations. Method (concise) Framework. Well-mixed room using Wells–Riley with the Gammaitoni–Nucci build-up term. The inhaled dose for one infectious source is: D = (q · p)/(V · λ) × [ t − (1 − exp(−λ · t))/λ ] where q = quanta emission (quanta/h, activity-dependent), p = breathing rate (m³/h), V = room volume (m³), t = exposure time (h), λ = total removal [h⁻¹] = ventilation (ACH) + filtration + decay/deposition. Optional: Preloaded room. If the infectious source was present tpre before you entered, the dose gets an additional start term: Dpre = (q · p)/(V · λ) × [ (1 − exp(−λ · tpre)) · (1 − exp(−λ · t)) / λ ] In the calculator: tpre = min(t, 1 h). Infection probability (≥ 1 infector present). P = 1 − exp(−D) Prevalence mode (unknown infectors). With n other people and pprev estimated from the 7-day incidence I₇ via p_prev ≈ (I7/100000) · T_inf/7 with T_inf ≈ 5 days: P_block = 1 − [ (1 − p_prev) + p_prev · exp(−D) ]^n Daily / cumulative. P_day = 1 − ∏(1 − P_block), P_cum(d) = 1 − ∏_{τ ≤ d} (1 − P_day(τ)) Masks & HEPA. Masks scale D (source/receiver). HEPA increases effective ACH. A small capped chain-effect approximates secondary spread in fixed groups. Assumptions/limits. Well-mixed air, constant rates, no near-field jets, independent contacts; q dominates uncertainty. References: Wells–Riley; Gammaitoni–Nucci (1997). Implementation follows the standard Poisson dose–response used in the aerosol literature. Estimating air exchange with a CO₂ monitor A CO₂ monitor shows how quickly stale air is replaced by fresh air. If CO₂ levels drop by half after ventilation (or in an empty room) in about 40–45 minutes, that corresponds roughly to ≈ 1 air change per hour; in 20–22 minutes about ≈ 2 ACH; in 10–11 minutes about ≈ 4 ACH. HEPA filters reduce viral load but do not affect CO₂ levels – they increase the equivalent air exchange rate for aerosols but do not replace fresh air supply. How to use the calculator First, the setting is defined: incidence, room volume, duration, number of people, and activity. Alternatively, a preconfigured scenario can be used. Partial scenarios can also be defined to represent complex situations. The assumed quanta per activity are SARS-CoV-2-specific. The output is provided in different scenarios (e.g. 0/1/3/6 ACH, with/without HEPA, with masks). For long-term risk, an incidence forecast can be included (if available) or an average incidence over the next 3 months selected. Weekends can also be excluded and/or time periods without infection risk indicated (e.g. school holidays or vacation). The colored traffic-light dots next to the results help to interpret the risk level of each scenario. Preloaded room: For suitable partial scenarios (e.g., treatment rooms, changing rooms, dense meetings, course rooms with room changes) you can enable “Potentially preloaded room”. The preloading time is applied dynamically based on the block duration and capped at 1 hour (t_pre = min(duration, 1 h)), and many templates preselect it where appropriate. If “Assume at least one infectious person is present” is enabled, preloading is conservatively applied automatically. Limitations & interpretation The result is a statistical estimate and not a substitute for individual advice. People differ in susceptibility, behavior, and immune status; especially near-field situations or superspreading events can create much higher local risks. Forecasts remain forecasts. The values are intended as a decision aid, not as a guarantee – also because risks are not evenly distributed and some situations require more caution. More detailed analyses: WHO ARIA For more complex room and usage profiles, the WHO tool ARIA (Airborne Risk Indoor Assessment) is suitable. In addition to the well-mixed approach, it accounts for time-varying ventilation, filter stages (MERV/HEPA), differentiated mask efficiencies, activity and occupancy profiles, as well as uncertainties in quanta emission. ARIA complements the calculator shown here, allows finer scenarios and sensitivity analyses – but requires more input data/assumptions and does not calculate cumulative risks. Transparency & data If available, the long-term module uses a simplified and idealized incidence projection for Germany based on an assumed immunity–infection equilibrium. Forecast data may change at any time and are merely an estimate of future infection dynamics.Alternatively, a constant average incidence can be applied. The curve shows the cumulative risk; the light gray area sketches the 7-day average of the daily probability that at least one infectious person is present. Conclusion Indoor environments remain the central place for SARS-CoV-2 transmission. The calculator helps to weigh risks – for single events (e.g. doctor visits, school enrollment, celebrations) as well as over longer periods and infection waves – in a conscious way.