Oklahoma City Contaminant Dispersion: Concentration Data Processing and Analysis for a Scaled Puff Release Experiment

Understanding and simulating the dispersion of contaminants within the atmospheric boundary layer can provide important information for emergency response infrastructure and personnel. Field tests, laboratory experiments, and CFD models have been used to study scalar transport within the urban environment. One such example is the Joint Urban 2003 study held in Oklahoma City at full-scale, for which a body of related experimental and numerical studies have been conducted. With the motivation of providing a high fidelity, three-dimensional data set for comparison with JU2003 and subsequent studies, magnetic resonance techniques were leveraged to obtain velocity and concentration measurements for a novel puff release contaminant dispersion study. The study involved a 1:2206 scaled model of downtown Oklahoma City as it was in 2003, which was then placed in a water channel with fully turbulent flow (Re=36,000). The puff injection cycle was created using a solenoid valve to oscillate between injection and redirection of the contaminant fluid, and an MRI system was used to take scans at 12 time-specific measurement phases throughout the cycle. The present work details processing methods applied to the nearly 650 million magnetic resonance concentration (MRC) data points obtained from the study. Measurements included several distinct scan types, including a Standard scan with contaminant in the injection and water in the main flow, a Background scan with water in both the injection and main flows, and a Reference scan with contaminant in both flows. Processing entailed the calculation of a concentration field from raw MRI signal data. Background subtraction was used to remove signal variation inherent to the measurement system, and normalization with respect to the Reference scan provided data ranging from zero to unity. Additionally, high molarity scans involving four and ten times the Standard injection concentration were taken during experimentation. Uncertainty was reduced through the scaling and combination of the high molarity scans, specifically in the low concentration regions of the flow field. Processing methods are followed by a preliminary investigation of the results, which highlights noteworthy elements of scalar transport within the data set and the need for further investigation of the complex flow field. Several regions within the city model were subject to residual contamination across the 12 measurement phases, and complex regions can be observed in regions directly upstream of buildings, as well as their wakes. The study ultimately demonstrates the applicability of magnetic resonance techniques to puff release and dynamic experimental conditions, as well as a method for working with data from phase-locked measurements.