This document discusses innovative dispersion modeling practices to achieve reasonable conservatism in regulatory modeling demonstrations. It presents a case study evaluating the Emissions and Meteorological Variability Processor (EMVAP) and approaches to establish background concentrations. The case study models SO2 concentrations from a power plant using 1) constant emissions, 2) variable emissions, and 3) EMVAP. EMVAP provides more realistic concentrations while accounting for emission variability. Using the 50th percentile monitored background concentration when combining with modeled values provides statistical conservatism compared to using high percentile values.