Abstract
This study examines the impact of altruistic and aggressive driver behavior on the effectiveness of an evacuation for a section of downtown Atlanta. The study area includes 37 signalized intersections, seven ramps, and 48 parking lots that vary by size, type (lot versus garage), peak volume, and number of ingress and egress points. A detailed microscopic model of the study area was created in VISSIM. Different scenarios examined the impacts of driver behavior on parking lot discharge rates and the loading rates from side streets on primary evacuation routes. A new methodology was created to accurately represent parking lot discharge rates. This study is also unique in that it assumes a "worst case scenario" that occurs with no advance notice during the morning peak period, when vehicles must transition from inbound to outbound routes. Simulation results indicate that while overall network clearance times are similar across scenarios, the distribution of delay on individual routes and across parking lots differ markedly. More equitable solutions (defined as the allocation of delay from parking lots and side streets to main evacuation routes) were observed with altruistic driver behavior.