key: cord-0887445-g42jx2lc
authors: Katal, A.; Albettar, M.; Wang, L.
title: City Reduced Probability of Infection (CityRPI) for Indoor Airborne Transmission of SARS-CoV-2 and Urban Building Energy Impacts
date: 2021-01-20
journal: nan
DOI: 10.1101/2021.01.19.21250046
sha: ace52fc319cebd7323eb894c30b5ce232891f644
doc_id: 887445
cord_uid: g42jx2lc

Airborne transmission of aerosols produced by asymptomatic individuals is a large portion of the SARS-CoV-2 spread indoors. Outdoor air ventilation rate, air filtration, room occupancy, exposure time, and mask-wearing are among the key parameters that affect its airborne transmission in indoor spaces. In this work, we developed a new web-based platform, City Reduced Probability of Infection - CityRPI, to calculate the indoor airborne transmission of COVID-19 in various buildings of a city scale. An archetype library of twenty-nine building types is developed based on several standards and references. Among the mitigation strategies recommended to reduce infection risk, some could result in significant energy impacts on buildings. To study the combined effects of energy consumption and reduced infection probability, we integrated CityRPI with City Building Energy Model. We applied the integrated model to Montreal City and studied the impact of six mitigation measures on the infection risk and peak energy demand in winter. It shows that the same strategy could perform quite differently, depending on building types and properties. In the winter season, increasing the outdoor air ventilation rate may cause massive building energy consumption. All strategies are shown to reduce the infection risk but wearing a mask and reducing exposure time are the most effective strategies in many buildings, with around 60% reduction. Doubling the outdoor air ventilation rate is not as effective as other strategies to reduce the risk with less than 35% reduction. It also significantly increases building peak heating demand with 10-60%.

direct and indirect contact transmission with the infected people, respiratory transmission by 1 large droplets within 1 m distance from an infected person, and transmission by airborne 2 aerosols [3] [4] [5] . Human expiratory activities such as coughing, sneezing, speaking, singing, and 3 even breathing release particles in a wide range of sizes, with most of them between 2 and 4 100

[6]. The small respirable particles < 5 − 10 can remain airborne and can spread at 5 long distances. The largest droplets fall next to a source, within a distance of 1-2 m, as a result 6 of gravitational force [7] . The latest research findings show that aerosols' airborne transmission 7 is a large portion of the spread of COVID-19 disease, especially in indoor spaces with poor 8 ventilation conditions, large gathering, and long-duration exposure to high concentrations of 9 aerosols [7] [8] [9] [10] [11] . Several studies show that improving the ventilation condition, wearing a face 10 mask, avoid overcrowding, and shortening the event time (exposure time) can significantly 11 reduce the airborne infection risk in indoor environments [12] [13] [14] . Buonanno rate to ensure a probability of less than 1% infection for different exposure times using the 20 Wells-Riley model [18] . They modeled some typical scenarios and concluded that the minimum 21 required ventilation rate can be reduced to a quarter by wearing the mask, which can be 22 The existing studies, e.g., Jiménez et al.'s estimator, evaluated the mitigation measures of 26 reducing airborne transmission risk in limited buildings or specific cases, e.g., choir, classroom, 27 subway, supermarket, and stadium. In contrast, the effectiveness of mitigation measures can 28 vary with building type because of different ventilation conditions, occupants' density, event 29 time, dominant age of occupants, etc. Therefore, it is essential to cover different building types, 30

preferably at an urban scale for multiple cities, and investigate the impact of different strategies 31 in reducing airborne infection risk based on the building usages. Each city or state also has its 32 unique COVID-19 prevalence with daily variations; e.g., the top five most infected state 33 . CC-BY-NC-ND 4.0 International license It is made available under a perpetuity.

is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted January 20, 2021. ; https://doi.org/10.1101/2021.01. 19.21250046 doi: medRxiv preprint to use the service without prior technical knowledge from his/her mobile devices or computers. 1

All the archetype library data are based on publicly available data. All the private input 2 parameters are not collected and discarded at the client-side after an active web session. 3

Moreover, many risk mitigation strategies create significant building energy impacts: e.g., 4

increasing outdoor ventilation rates to 130% or 200% of the baseline during the winter of 2020 5 may increase the building's total energy consumption and peak demand. It could overload the 6 existing system and result in a power outage at an urban or community scale. Therefore, while 7 reducing the probabilities of indoor infections is essential, it is equally important to evaluate 8 each mitigation strategy's energy impact at both a building and an urban scale to be 9

implemented. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The following theory was based on the original work of the COVID-19 Aerosole Transmission 5

Estimator" by Jiminez et al. (2020) while this study extended the work to various building types 6 at a whole-city scale for North America, investigating various mitigation strategies and their 7 energy impacts, which have not been addressed previously. The estimator is based on five 8 assumptions: i) one infector in a space with a constant SARS-CoV-2 quanta generation rate, ii) 9 zero initial quanta in the space, iii) latent period of the disease is longer than the time duration 10 of the event, iv) indoor environment is well-mixed, and v) the infectious quanta removal is a 11 first-order process. Two types of P.I. are calculated: conditional P.I. and absolute P.I. The is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint 

where is the number of quanta inhaled by a susceptible. Consider a susceptible in the room 3 who inhales at rate B ( 3 ℎ ⁄ ) and is present for T hours. The expected number of quanta inhaled 4 is calculated by Eq. 2: 5

is the time-average quanta concentration ( 3 ⁄ ); is the fraction of people in the room 6 who wears the mask, and is the inhalation mask efficiency. By solving the well-mixed 7 material balance equation for the room (Eq. 3), the is calculated using Eq. 4. 8

Eq. 3

is the volume of the room ( 3 ); is the first-order loss rate coefficient for quanta (ℎ −1 ); 9 and is the net quanta emission rate (ℎ −1 ). is calculated based on the number of infected and activity levels (resting, light activity, and heavy exercise). 15

The infectious quanta could be removed from room air by first-order processes reflecting 17 several mechanisms: outdoor air ventilation 1 , filtration 2 , deposition on surfaces 3 , and 18 airborne inactivation 4 . 19

Eq. 6

First, infectious quanta are removed with a first-order rate constant 1 equal to the air change 20 of outdoor air per hour (ℎ −1 ) supplied to the room by the HVAC system or opening the 21 . CC-BY-NC-ND 4.0 International license It is made available under a perpetuity.

is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted January 20, 2021. ; https://doi.org/10.1101/2021.01.19.21250046 doi: medRxiv preprint windows. Second, the in-room air filtration using portable air purifiers and/or duct filters in 1 HVAC systems can remove infectious quanta with a rate constant 2 . Third, infectious quanta 2 are removed by gravitational settling with a first-order rate constant 3 . The deposition rate is 3 calculated by [15] and is equal to 0.24 ℎ −1 . Finally, infectious quanta are inactivated with a 4 first-order rate constant 4 . The quanta inactivation was evaluated based on the SARSCoV-2 

is the disease prevalence in the community that depends on the state of the pandemic in 12 the region of study and the period of the disease. and are the number of daily new cases 13 and the population of the studied region, respectively.

is the fraction of unreported cases. 14 A study on ten diverse geographical sites in the U.S. shows that the estimated number of 15 infections was much greater (6 to 24 times) than the number of reported cases in all sites [35] . 16 is the duration of the infectious period of SARS-CoV-2 [36] . and are the total 17 number and number of susceptible people in the room, respectively.

is the fraction of 18 the population that has had the disease and has some immunity against it. It can be estimated 19 using the total recovered cases in the region of study [35, 37] . is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted January 20, 2021. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted January 20, 2021. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted January 20, 2021.

The susceptible person's breath rate in the room is estimated using US EPA Exposure Factors 1 Handbook (Chapter 6) (U.S. Environmental Protection Agency, 2011). The breath rate depends 2 on age and activity level. A typical activity type and average age are defined for each building 3 type to assign the buildings' breath rate. The quanta emission rate by an infective person in the 4 room depends on the activity level. We used the recommended quanta emission rate by 5 Buonanno et al. (2020a Buonanno et al. ( , 2020b and standard activity levels to estimate each building type's 6 quanta emission rate. The ceiling height and stay time are obtained based on the standard values 7 of different building types. Therefore, in this work, we assumed that all buildings are designed 8 and operating based on the available standards to calculate the indoor airborne infection risk 9 and energy consumption. Table 2 shows some key input parameters of the archetype library 10 assigned to each building type. 11

Using local weather data for building energy simulation instead of nearby weather station data 13 such as airports can significantly affect its accuracy [27, 28] . In this work, mesoscale is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted January 20, 2021. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted January 20, 2021. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted January 20, 2021. 

The accuracy of the CityBEM for the calculation of building energy performance is already 7 investigated by comparing the annual and hourly electricity consumption with corresponding 8 measurement data [27, 28] . In this work, for the validation of indoor air temperature results 9 calculated by CityRPI, Montreal City is simulated from June 24-29, 2020. We compared the 10 calculated indoor air temperature of two school buildings with measurement data provided by is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted January 20, 2021. ; https://doi.org/10.1101/2021.01.19.21250046 doi: medRxiv preprint system, or the A.C. systems were not operating at the study time. Therefore, the indoor air 1 temperature is directly affected by the cooling load components, and an accurate calculation of 2 indoor air temperature could indicate the accuracy of building energy load estimation. Figs. 5a 3 and 5b show the schools' aerial view map and weather data (outdoor air temperature and solar 4 radiation) used for the simulation, respectively. According to Fig. 5b , the studied days were 5 quite hot and sunny. In the absence of A.C. systems in the buildings, indoor air temperature is 6 much higher than cooling-setpoint temperature. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted January 20, 2021. ; https://doi.org/10.1101/2021.01.19.21250046 doi: medRxiv preprint simplified single-zone model used for the whole building. The RMSE between the average 1 temperature of classrooms and simulation results of schools 1 and 2 are 1.8 ℃ and 1.6 ℃, 2 respectively. 3

According to Equations 1-9, the indoor aerosol infection risk is a function of several parameters. 5

Some of them, such as the number of occupants, ventilation rate, stay time, and breathing rate, 6 depend on the building usage type. Therefore, infection risk and the effectiveness of mitigation 7 measures can change by building type. In this section, we modeled the Montreal City by 8

CityRPI and studied the impact of six mitigation strategies on buildings' indoor aerosol 9 infection risk. Increasing the outdoor air ventilation rate by opening the windows or more 10 outdoor air from the HVAC intakes can improve the room ventilation condition and reduce 11 infection risk. But it can also significantly increase building energy consumption in the winter. 12 Therefore, it is essential to find the most effective strategy considering both the reduced 13 probability of infection and the building's energy consumption. For this purpose, we conducted 14 the simulation over Montreal from February 12-21, 2020, which was the coldest period of 15 winter 2019, and the effectiveness of three mitigation strategies for improving the indoor 16 ventilation condition (more outdoor air, upgrading duct filter, portable air cleaner) are 17 compared. 18

The simulation area that covers Montreal City is shown in Fig. 2 (Table 2) . 26

In this section, first, we study the conditional and absolute P.I. of buildings at baseline 28 conditions (e.g., minimum design ventilation rate by default). We compare the infection risk 29 between different building types to find the most vulnerable buildings against the airborne 30 infection risk of COVID-19. We also study the daily variation of to investigate its change 31 with the prevalence rate of the disease. Then, to reduce the infection risk in buildings, we 32 . CC-BY-NC-ND 4.0 International license It is made available under a perpetuity.

is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted January 20, 2021. ; https://doi.org/10.1101/2021.01.19.21250046 doi: medRxiv preprint designed several scenarios using different mitigation measures. We applied the mitigation 1 measures to all buildings in the city to find the most effective strategies for each building type. 2 Some strategies also affect building energy consumption. Therefore, we calculated the 3 increased peak demand for all buildings and identified the recommended strategy for risk 4 reduction and energy consumption. 5 Fig. 6 The historical prevalence rate of COVID-19 in Montreal and of seven 6 selected building types. 7

To study P.I. variation with the city's prevalence rate, we calculated all buildings' historical P.I. 9

from the start of the pandemic in Montreal until January 6, 2021. The first wave of the COVID-10 with the reopening of schools and daycares around January 1. The second wave of pandemic 13 begun around the middle of September and the number of daily new cases is still rising. The 14

Quebec government reimposed a partial lockdown on December 25 that will be continued until 15 is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted January 20, 2021. ; https://doi.org/10.1101/2021.01.19.21250046 doi: medRxiv preprint 1 infected among 25 persons averaged over the whole population. It includes both symptomatic 2 and asymptomatic cases. The historical of seven building types are plotted and compared 3 in Fig. 6 . We calculated the of all 29 building types, but to compare the variation of 4 relative with time, the result is plotted for seven buildings with the most usage and 5 importance during the pandemic. We did not plot the of buildings because it is not a 6 function of the prevalence rate and is constant with time. According to Fig. 6 , even though the 7 varies with the prevalence rate but the relative infection risk of buildings is the same; for 8 example, daycare remains with the high among selected buildings. 9

To compare the P.I. between all twenty-nine building types, we calculated the conditional and 10 absolute P.I. of buildings on the worst day (May 3, 2020) with the largest daily new cases and 11 prevalence rate. The number of daily new cases, total number of confirmed cases, total number 12 of deaths, and the total number of recovered patients were 1652, 16251, 1365, and 9123. The 13 prevalence rate of the disease was around 4.03%. Fig. 6 shows the weekly prevalence rate, and 14 this specific day is not plotted in this figure, but as it can be seen, the number of daily new cases 15 is still increasing during the second wave and arrives in close to the worst case. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted January 20, 2021. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted January 20, 2021. ; https://doi.org/10.1101/2021.01.19.21250046 doi: medRxiv preprint

The number of buildings per type is another important parameter when comparing different 1 buildings. One building type can have a large , but if there are few buildings of this type 2 in a city or urban area, it may not affect the city-scale risk. To consider its impact, we calculated 3 the weighted probability of infection, which is the probability of infection multiplied by 4 building type fraction, × , , ⁄ , which , is the number of buildings with a 5 specific type and , is the total number of studied buildings. Most of the city buildings 6 are residential buildings, but because our focus is on public buildings, residential buildings are 7 removed from the analysis. Fig. 7b shows the weighted conditional and absolute P.I. The school 8 buildings offer the largest weighted because of large and building fraction. 9

Shopping centers, industrial buildings, offices, and daycares are the next vulnerable buildings. 10

Regarding the weighted , which is "person to person" transmission without considering 11 the prevalence rate, office, industrial building, school, shopping center, sports center, and 12 daycare are the most vulnerable buildings. The office building was the 6 th vulnerable building 13 considering only the , however, it is the most vulnerable building at the city scale because 14 of the large number of office buildings in the city. Therefore, to reduce the city's airborne 15 infection risk, it is essential to focus on the buildings with the largest weighted P.I. 16

As mentioned previously, airborne infection risk depends on building properties. The in 17 is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted January 20, 2021. ;

Before analyzing the impact of scenarios on all buildings, it is crucial to study the RPI change 1 with the prevalence rate. For this purpose, we calculated the classroom daily RPI from the 2 beginning of the pandemic until January 6, 2021 (Fig. 8) . We selected the school building 3 because it is the most vulnerable building based on the Fig. 7b . Results show that all strategies' 4 effectiveness is almost constant with time and is independent of the prevalence rate. For a 5 classroom, wearing a face mask by all occupants is the most effective strategy with 64% RPI. 6

The next effective strategies are upgrading the duct filter to the MERV-13 filter, using a 7 portable air cleaner with 480 CFM flow rate, half stay time, and half occupancy with 50%-58% 8 RPI. Doubling outdoor air ventilation rate seems the least effective strategy with only 31% RPI. 9

It should be noted that the duct filter RPI is the median of all school buildings in the city because 10 the HVAC system airflow rate depends on the design condition and building properties. A study 11 by Curtius et al. (2020) on using portable air cleaners in the classrooms show that a total air 12 exchange rate of 5.7 h -1 reduces the airborne transmission by 80%, the equivalent air exchange 13 rate of air cleaner modeled in this work is around 2.7 h -1, and the reduced infection risk is 53%. 14 By using more air cleaners in the room, we can get more reduction. The conditional P.I. is not 15 plotted in this figure because it is independent of the prevalence rate. 16 is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted January 20, 2021. ; https://doi.org/10.1101/2021.01.19.21250046 doi: medRxiv preprint According to Fig. 8 , the relative effectiveness, RPI, of all strategies seems independent of the 1 prevalence rate. Therefore, to study the impact of mitigation measures on all building infection 2 risk reduction, we calculated the RPI on the worst pandemic day. Energy impact is another 3 important factor that must be considered. Some of the strategies can increase building energy 4 consumption while reducing the infection risk. For example, more outdoor air ventilation rate 5 in the winter needs more pre-heating and thus more energy. Upgrading to duct filters with 6 higher efficiency may increase fan power consumption, and a portable air cleaner also adds 7 extra electricity usage. The number of daily cases in Montreal is rising during the winter; 8 therefore, we calculated the increased building peak energy demand for the coldest period of 9 winter 2019 by assuming the winter of 2020 would have similar weather. 10 is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted January 20, 2021. ;

The distribution of buildings by usage type is plotted in Fig. 10a . We plotted the of all 1 buildings in the City downtown for double outdoor air ventilation rate and wearing a face mask 2 in Figs. 10b and 10c. This region includes diverse building types and covers almost all building 3 types studied in this work. The RPI range for double outdoor ventilation rate is between 0.1%-4 35% but for wearing a face mask is always larger than 60%. Our analysis shows that four times 5 outdoor air ventilation rate could provide a similar level reduction of infection risk as wearing 6 a face mask in many cases, consistent with the result presented by Dai and Zhao (2020). 7

The increased peak energy demand for double outdoor air ventilation varies between 10%-60% 8 depending on the building type. At the same time, it is less than 2% for upgrading duct filters 9 (MERV 13) or using a portable air cleaner, and other strategies do not have an energy penalty. 10

The results are plotted in Figs. 10d and 10e. The energy penalty using double outdoor air 11 ventilation rate is between 2%-70% for all buildings while it is less than 2% for upgrading the 12 duct filter. Considering both reduced infection risk and increased peak demand, we may 13 conclude that double outdoor air ventilation rate is the least effective strategy. To get the same 14 RPI as other strategies such as wearing a face mask or half stay time, we need to further increase 15 the outdoor air ventilation rate while unavoidably increasing the heat demand. As a result, this 16 could build excessive pressures on the electricity grid and increase the chance of power outages 17 in the city. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted January 20, 2021. The calculation of infection risk involves many input parameters, and the results could be 28 subject to the uncertainties of these parameters. As future work, a sensitivity analysis of the 29 infection risk model becomes necessary to quantify these uncertainties' impacts on the results. 30

It will help to determine which parameters are dominant and improve the accuracy of the 31 . CC-BY-NC-ND 4.0 International license It is made available under a perpetuity.

is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted January 20, 2021. especially for an event with many occupants and for a high prevalence rate in a city because 2 the conditional probability is based on the scenario that there is only one infector present in a 3 room. With more people and higher prevalence, the chance of more than one infectors in the 4 same room could become high, which could be considered in a future study. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted January 20, 2021. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted January 20, 2021. ; https://doi.org/10.1101/2021.01.19.21250046 doi: medRxiv preprint

COVID-19) events as they happen

WHO, Coronavirus disease (COVID-19)

WHO, Q&A on coronaviruses (COVID-19), Who. (2020)

Infection Control: Severe acute respiratory syndrome coronavirus 2

Basic facts

Covid-19 airborne transmission and its prevention: 20 Waiting for evidence or applying the precautionary principle?

Consideration of the Aerosol 23 Transmission for COVID-19 and Public Health

Aerosol transmission of SARS-CoV-2? 2 Evidence, prevention and control

Transmission of SARS-CoV-2 by 6 inhalation of respiratory aerosol in the Skagit Valley Chorale superspreading event

Identifying airborne transmission 9 as the dominant route for the spread of COVID-19

How can airborne transmission of COVID-19 indoors be 17 minimised?

Review of indoor aerosol generation, transport, 19 and control in the context of COVID-19

Estimation of airborne viral emission: Quanta 22 emission rate of SARS-CoV-2 for infection risk assessment

Testing mobile air purifiers in a school classroom: 25 Reducing the airborne transmission risk for SARS-CoV-2

Integrating IAQ control strategies to reduce the risk of asymptomatic SARS

CoV-2 infections in classrooms and open plan offices

Mathematical models for assessing the role of airflow 8 on the risk of airborne infection in hospital wards

The efficacy of social distance and ventilation effectiveness in 11 preventing COVID-19 transmission

COVID-19 Airborne Transmission Tool Available | CIRES

Ventilation control of indoor transmission of airborne 16 diseases in an Urban community

Ventilation for Acceptable 21 Indoor Air Quality

Toward understanding the risk of secondary 1 airborne infection: Emission of respirable pathogens

Using a Mathematical Model to Evaluate the Efficacy of

Quantitative assessment of the risk of airborne 6 transmission of SARS-CoV-2 infection: Prospective and retrospective applications

Aerosol and surface stability of SARS-CoV-2 as compared with

SARS-CoV-1

2020_COVID-19_Aerosol_Transmission_Estimator -Google Sheets

Seroprevalence of

Antibodies to SARS-CoV-2 in 10 Sites in the United States

GitHub -microsoft/CanadianBuildingFootprints: Computer generated 1 building footprints for Canada

Computer generated building footprints for the United States: 3 microsoft/USBuildingFootprints

Google Earth 5 Engine: Planetary-scale geospatial analysis for everyone, Remote Sens. Environ

Unités d'évaluation foncière -Jeu de données

Exposure Factors Handbook: 2011 Edition

Environment Canada, HRDPS data in GRIB2 format

Scattered Data Approximation

GitHub -nytimes/covid-19-data: An ongoing repository of data on 14 coronavirus cases and deaths in the

Esri Canada, COVID-19 Canada

This study developed the CityRPI model to calculate the airborne transmission of COVID-19 2 in a city. We developed an archetype library based on standards and references to estimate 3 different parameters for calculating building infection risk. We modeled Montreal City and 4 calculated the infection risk of all buildings in the city, and evaluated the impact of six 5 mitigation strategies on reducing infection risk. Some strategies impact the energy consumption 6 of a building, especially in the winter. We integrated CityRPI with the CityBEM model to 7 calculate all building's peak energy demand. The main conclusions are as follows: 3. The mitigation study shows that wearing a face mask and half stay time are the most effective 18 strategies for most buildings. Double outdoor air ventilation rate is the least effective strategy 19 in many buildings, and the corresponding RPI is much lower than other strategies. 20 4. Double outdoor ventilation rate also significantly increases building peak heating demand in 21 the winter. Therefore, this strategy is not as effective as other strategies considering the 22