Assessing the effectiveness of shelter-in-place as an emergency response to large-scale outdoor chemical releases

Publication Type




Large-scale outdoor chemical releases can cause severe harm to people in nearby communities. Sheltering in buildings may be used as a temporary measure to reduce health risk from exposure to the toxic materials. Shelter-in-place (SIP) is relatively straightforward to implement because most people are already in buildings most of the time, and so exercising the emergency response simply means closing windows and doors, and turning off ventilation fans. However, air leakage variability in the building stock can lead to considerable differences in the effectiveness of buildings in protecting occupants against outdoor releases. The effectiveness of SIP for the community can also vary for different release conditions. This dissertation identifies and assesses the key factors that affect community-scale SIP effectiveness. Large-scale airborne toxic chemical releases are simulated to assess the potential acute health effects for the exposed population. Modeling of the distribution of indoor concentrations is accomplished through detailed analysis of the air leakage of residential and non-residential buildings and simulation of their air infiltration rates. The expected outcome for a population that shelter indoors is quantified by a communitybased metric that captures the variability among buildings. Sensitivity of SIP effectiveness to model parameters is evaluated under different release scenarios by comparing changes in the casualty reduction estimates. Aside from the physical, biological, and chemical factors that influence SIP effectiveness – such as the building air exchange rate, the degree of nonlinearity of the dose-response relationship, and the extent of chemical sorption onto indoor surfaces – human factors, such as community response time in emergencies are also important factors that govern whether SIP can provide adequate protection for an exposed population. After the plume has dispersed, SIP should be terminated by means of exiting or deliberately ventilating the buildings. In most situations, however, it is found that a short delay in terminating SIP would not significantly degrade the overall effectiveness of the strategy. On the other hand, a potentially large enhancement of SIP effectiveness can be realized by reducing the time delay for SIP initiation. The understanding gained from these analyses can guide decisions in emergency response and pre-event planning.


Department of Civil and Environmental Engineering

Year of Publication



Research Areas