Integration of Sensor Data With Computational Fluid Dynamics Simulations To Improve Workplace Safety From Airborne Pathogens

In many commercial & residential buildings, natural ventilation is insufficient to provide adequate indoor air quality (IAQ) and/or thermal comfort to all residents. Therefore, building managers (BMs) rely on heating, ventilation, and air conditioning (HVAC) systems to provide clean air & regulate indoor air temperature. As the HVAC system is usually the primary influencer of indoor airflow, it can (if poorly designed) propagate aerosolised droplets & other contaminants (CO2, VOC, and solid particulate matter (PM) of all sizes) throughout an indoor workspace. Therefore, it is imperative to assess & manage the risk of exposure to indoor airborne pathogens whose transmission is facilitated by HVAC systems of particular recent importance is the COVID-19 pandemic. Most indoor workspaces undergo tenancy & occupancy changes throughout their lifecycle, which alongside dynamic daily usage patterns leads to changes in building utilisation, occupant density, and equipment usage. Conventional ventilation systems such as ‘mixing ventilation’ & ‘displacement ventilation’ are ill-suited to adapting to these dynamic demands as these demands and their effects are generally localised. To address this limitation, many studies have used numerical simulations to explore the feasibility of replacing the aforementioned conventional HVAC approaches with flexible systems. These systems are designed to meet the diverse thermal comfort requirements & ventilation demands of different tenants, occupancy levels, and usage patterns. However, the ranges of workspace geometry & staff occupancy patterns investigated by these studies are limited. This thesis evaluates the effectiveness of a relatively new flexible ventilation system termed ‘multi mode adaptive ventilation’ (MAV) at mitigating risk of exposure to airborne COVID 19 droplet aerosol exhaled by infected persons (Delta variant). Risk is quantified using the cumulative infection risk (CIR) concept that is applied atop of computational fluid dynamics (CFD) simulations. Moreover, these simulations are conducted in a steady state manner, which assumes workers remain stationery and exhales droplets at a constant rate in an adiabatic system. The range of cases include different infector densities & occupancy scenarios with the workspace geometry being a realistic open plan layout. Three different MAV air distribution modes (x aligned, y aligned, chequered) are studied to determine their effects on general & localised CIR. These novel results for the MAV HVAC scheme are intended to help improve ventilation system design and evaluate the effectiveness of novel infection risk reduction measures via MAV modes. While steady-state CFD simulations help determine risk of exposure to airborne pathogens in a controlled environment, they cannot account for realtime changes in environmental conditions such as temperature & humidity introduced to varying extents by occupants, equipment, HVAC systems, and human & equipment movement. To address these limitations, the integration of data from wireless sensor networks (WSNs) becomes crucial, as WSNs help elucidate IAQ & pathogen transmission dynamics, in turn helping to develop more effective infection control strategies. However, WSN data represents relatively few discrete points. Interpolation of these data (PM2.5 concentration is used as a safe proxy for COVID 19 aerosol concentration in this thesis) over the region of interest may employ spatial interpolation algorithms such as inverse distance weighted (IDW) & ordinary kriging (OK), which are more flexible distance-weighted schemes than, e.g. simple triangulation & linear interpolation. The novelty of the current approach is that no current studies investigate improvement of predictive accuracy of distance based interpolation schemes by incorporating airflow direction, resulting in the new Flow IDW approach. This approach thus allows for field-gathered HVAC airflow data to underpin predictions of WSN-derived pathogen concentrations to dynamically improve IAQ and improve indoor workplace health & safety.