Comparison of ALOHA, Gaussian Plume, and OpenFOAM CFD for Urban Atmospheric Dispersion of Ammonia: A Companion Analysis to the Colorado Springs Chlorine Baseline

This companion study extends the Colorado Springs Urban Dispersion Study by substituting ammonia (NH3) for chlorine as the release chemical while retaining the identical 21-receptor Option A geometry, meteorological conditions, and three-model comparison framework established in the baseline chlorine analysis (Peters, 2026a). Ammonia is lighter than air (molecular weight 17.03 g/mol; gas-phase density approximately 0.59 times that of air), in contrast to chlorine's dense-gas behavior (density approximately 2.5 times that of air). Three models are evaluated at t = 660 s: ALOHA 5.4.7, an analytical Gaussian plume model, and a previously generated OpenFOAM passive-tracer CFD baseline. Spearman rank correlations show strong agreement between the Gaussian plume and OpenFOAM (ρ = .896, n = 14, p < .001) and between ALOHA and the Gaussian plume (ρ = .836, n = 21, p < .001). The ALOHA-to-Gaussian concentration enhancement factor at centerline receptors is 3.1–3.4× for ammonia, markedly lower than the 5.4–6.0× factor observed for chlorine, consistent with ammonia's buoyant rather than dense-gas transport behavior. The OpenFOAM peak shifts from receptor 4 (100 m, centerline) in ALOHA and Gaussian predictions to receptor 11 (300 m, centerline), indicating that urban channeling alters the along-wind concentration distribution. All inputs, outputs, and analysis scripts are archived in a SHA-256-verified artifact bundle (DOI: 10.5281/zenodo.18637518).